Add documentation for widget

Added documentation for current widget usage, currently supported arguments, using custom CSS/JS

.github/workflows/ci.yml

@@ -1,27 +1,31 @@
 name: CI
 
-on: [push]
+on:
+  push:
+    branches:
+      - master
+  pull_request:
 
 jobs:
   Macos:
     runs-on: macos-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@v1
-      - name: Setup python 3.5
-        uses: actions/setup-python@v1
+        uses: actions/checkout@v2
+      - name: Setup python 3.10
+        uses: actions/setup-python@v2
         with:
-          python-version: '3.5'
+          python-version: '3.10'
       - name: Install packages
-        uses: mstksg/get-package@v1
-        with:
-          brew: gcovr pkg-config ninja
+        run: |
+          brew update
+          brew install gcovr pkg-config ninja
       - name: Install python modules
         run: pip3 install meson==0.49.2 pytest
       - name: Install deps
         shell: bash
         run: |
-          ARCHIVE_NAME=deps2_osx_native_dyn_kiwix-lib.tar.xz
+          ARCHIVE_NAME=deps2_osx_native_dyn_libkiwix.tar.xz
           wget -O- http://tmp.kiwix.org/ci/${ARCHIVE_NAME} | tar -xJ -C $HOME
       - name: Compile
         shell: bash
@@ -37,23 +41,14 @@ jobs:
           export LD_LIBRARY_PATH=$HOME/BUILD_native_dyn/INSTALL/lib:$HOME/BUILD_native_dyn/INSTALL/lib64
           cd build
           meson test --verbose
-          ninja coverage
         env:
           SKIP_BIG_MEMORY_TEST: 1
-      - name: Publish coverage
-        shell: bash
-        run: |
-          curl https://codecov.io/bash -o codecov.sh
-          bash codecov.sh -n osx_native_dyn -Z
-          rm codecov.sh
-        env:
-          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
 
   Linux:
     strategy:
       fail-fast: false
       matrix:
-        target:
+        name:
           - native_static
           - native_dyn
           - android_arm
@@ -61,34 +56,36 @@ jobs:
           - win32_static
           - win32_dyn
         include:
-          - target: native_static
-            image_variant: xenial
+          - name: native_static
+            target: native_static
+            image_variant: bionic
             lib_postfix: '/x86_64-linux-gnu'
-          - target: native_dyn
-            image_variant: xenial
+          - name: native_dyn
+            target: native_dyn
+            image_variant: bionic
             lib_postfix: '/x86_64-linux-gnu'
-          - target: android_arm
-            image_variant: xenial
-            lib_postfix: '/x86_64-linux-gnu'
-          - target: android_arm64
-            image_variant: xenial
-            lib_postfix: '/x86_64-linux-gnu'
-          - target: win32_static
-            image_variant: f31
+          - name: android_arm
+            target: android_arm
+            image_variant: bionic
+            lib_postfix: '/arm-linux-androideabi'
+          - name: android_arm64
+            target: android_arm64
+            image_variant: bionic
+            lib_postfix: '/aarch64-linux-android'
+          - name: win32_static
+            target: win32_static
+            image_variant: f35
             lib_postfix: '64'
-          - target: win32_dyn
-            image_variant: f31
+          - name: win32_dyn
+            target: win32_dyn
+            image_variant: f35
             lib_postfix: '64'
     env:
       HOME: /home/runner
     runs-on: ubuntu-latest
     container:
-      image: "kiwix/kiwix-build_ci:${{matrix.image_variant}}-26"
+      image: "kiwix/kiwix-build_ci:${{matrix.image_variant}}-31"
     steps:
-    - name: Extract branch name
-      shell: bash
-      run: echo "##[set-output name=branch;]$(echo ${GITHUB_REF#refs/heads/})"
-      id: extract_branch
     - name: Checkout code
       shell: python
       run: |
@@ -98,13 +95,13 @@ jobs:
           'git', 'clone',
           'https://github.com/${{github.repository}}',
           '--depth=1',
-          '--branch', '${{steps.extract_branch.outputs.branch}}'
+          '--branch', '${{ github.head_ref || github.ref_name }}'
         ]
         check_call(command, cwd=environ['HOME'])
     - name: Install deps
       shell: bash
       run: |
-        ARCHIVE_NAME=deps2_${OS_NAME}_${{matrix.target}}_kiwix-lib.tar.xz
+        ARCHIVE_NAME=deps2_${OS_NAME}_${{matrix.target}}_libkiwix.tar.xz
         wget -O- http://tmp.kiwix.org/ci/${ARCHIVE_NAME} | tar -xJ -C /home/runner
     - name: Compile
       shell: bash
@@ -121,9 +118,9 @@ jobs:
           MESON_OPTION="$MESON_OPTION --cross-file $HOME/BUILD_${{matrix.target}}/meson_cross_file.txt"
         fi
         if [[ "${{matrix.target}}" =~ android_.* ]]; then
-          MESON_OPTION="$MESON_OPTION -Dandroid=true"
+          MESON_OPTION="$MESON_OPTION -Dstatic-linkage=true"
         fi
-        cd $HOME/kiwix-lib
+        cd $HOME/libkiwix
         meson . build ${MESON_OPTION}
         cd build
         ninja
@@ -134,7 +131,7 @@ jobs:
       if: startsWith(matrix.target, 'native_')
       shell: bash
       run: |
-        cd $HOME/kiwix-lib/build
+        cd $HOME/libkiwix/build
         meson test --verbose
         ninja coverage
       env:
@@ -143,7 +140,7 @@ jobs:
     - name: Publish coverage
       shell: bash
       run: |
-        cd $HOME/kiwix-lib
+        cd $HOME/libkiwix
         curl https://codecov.io/bash -o codecov.sh
         bash codecov.sh -n "${OS_NAME}_${{matrix.target}}" -Z
         rm codecov.sh

.github/workflows/package.yml

@@ -0,0 +1,91 @@
+name: Packages
+on: [push, pull_request]
+
+jobs:
+  build-deb:
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        distro:
+          - ubuntu-jammy
+          - ubuntu-impish
+          - ubuntu-focal
+          - ubuntu-bionic
+    steps:
+      - uses: actions/checkout@v2
+
+      # Determine which PPA we should upload to
+      - name: PPA
+        id: ppa
+        run: |
+          if [[ $REF == refs/tags* ]]
+          then
+            echo "::set-output name=ppa::kiwixteam/release"
+          else
+            echo "::set-output name=ppa::kiwixteam/dev"
+          fi
+        env:
+          REF: ${{ github.ref }}
+
+      - uses: legoktm/gh-action-auto-dch@master
+        with:
+          fullname: Kiwix builder
+          email: release+launchpad@kiwix.org
+          distro: ${{ matrix.distro }}
+
+      - uses: legoktm/gh-action-build-deb@ubuntu-jammy
+        if: matrix.distro == 'ubuntu-jammy'
+        name: Build package for ubuntu-jammy
+        id: build-ubuntu-jammy
+        with:
+          args: --no-sign
+          ppa: ${{ steps.ppa.outputs.ppa }}
+
+      - uses: legoktm/gh-action-build-deb@ubuntu-impish
+        if: matrix.distro == 'ubuntu-impish'
+        name: Build package for ubuntu-impish
+        id: build-ubuntu-impish
+        with:
+          args: --no-sign
+          ppa: ${{ steps.ppa.outputs.ppa }}
+
+      - uses: legoktm/gh-action-build-deb@ubuntu-focal
+        if: matrix.distro == 'ubuntu-focal'
+        name: Build package for ubuntu-focal
+        id: build-ubuntu-focal
+        with:
+          args: --no-sign
+          ppa: ${{ steps.ppa.outputs.ppa }}
+
+      - uses: legoktm/gh-action-build-deb@ubuntu-bionic
+        if: matrix.distro == 'ubuntu-bionic'
+        name: Build package for ubuntu-bionic
+        id: build-ubuntu-bionic
+        with:
+          args: --no-sign
+          ppa: ${{ steps.ppa.outputs.ppa }}
+
+      - uses: actions/upload-artifact@v2
+        with:
+          name: Packages for ${{ matrix.distro }}
+          path: output
+
+      - uses: legoktm/gh-action-dput@master
+        name: Upload dev package
+        # Only upload on pushes to master
+        if: github.event_name == 'push' && github.event.ref == 'refs/heads/master' && startswith(matrix.distro, 'ubuntu-')
+        with:
+          gpg_key: ${{ secrets.LAUNCHPAD_GPG }}
+          repository: ppa:kiwixteam/dev
+          packages: output/*_source.changes
+
+      - uses: legoktm/gh-action-dput@master
+        name: Upload release package
+        # Only upload on pushes to master or tag
+        if: github.event_name == 'push' && startsWith(github.event.ref, 'refs/tags') && startswith(matrix.distro, 'ubuntu-')
+        with:
+          gpg_key: ${{ secrets.LAUNCHPAD_GPG }}
+          repository: ppa:kiwixteam/release
+          packages: output/*_source.changes
+

.gitignore

@@ -2,3 +2,8 @@
 *.swp
 subprojects/googletest-release*
 *.class
+build/
+.vscode/
+builddir/
+.cache/
+.clangd/

ChangeLog

@@ -1,3 +1,189 @@
+libkiwix 11.0.0
+===============
+
+ * [server] Add support for internationalization (@veloman-yunkan #679)
+ * [server] Use gzip compression instead of deflat (mgautierfr #757)
+ * [server] Version the static resources. This allow better invalidating
+   browser cache when resources are changed (@veloman-yunkan #712)
+ * [server|front] Use integer to query the host for page length (@juuz #772)
+ * [server] Improve multizim search API:
+   - Improvement of the cache system
+   - Better API to select on which books to search in.
+   - SysAdmin is now able to limit the number of book we search in for a multizim search
+ * [server] Introduce a opensearch API for multizim fulltext search
+ * [wrapper] Remove java wrapper
+ * Testing:
+   - Testing of search result pages content (@veloman-yunkan #765)
+   - Better testing structure of xml search result (@veloman-yunkan #780)
+
+libkiwix 10.1.1
+===============
+
+ * Correctly detect the number of article for older zims (<=6) (@mgautier #743)
+ * [server] Fix fulltext search (@mgautierfr #724)
+ * [server][internal] New way to build Error message (@veloman-yunkan #732 #738 #744)
+ * Fix CI (@mgautierfr #736)
+
+libkiwix 10.1.0
+===============
+
+ This release is an important one as it fixes a Xss vulnerability introduced
+ in libkiwix 10.0.0
+
+ * [SECURITY] Fix a Xss attack vulnerability (introduced in 10.0.0) (@juuz0 #721)
+ * [server] Add a option to set a limit on the number of connexion per IP (@kelson42 #700)
+ * [server] Do not display a lang tag in the UI if the book has no language (@juuz0 #706)
+ * [server] Add the book title associated to a search results (@thavelick #705, @mgautierfr #718)
+ * Add `dc:issued` to opds output stream (@veloman-yunkan #715)
+ * Add handling of several languages not provided by ICU (@juuz0 #701)
+ * [server] Add a caching system for search and suggestion (@maneeshpm #620)
+ * Fix cross-compilation (@kelson42 #703)
+ * Add unit-testing of suggestions and error pages (@veloman-yunkan #709 #710 #727)
+ * Better testing system of html response (@veloman-yunkan #725)
+
+libkiwix 10.0.1
+===============
+
+ * [server] The catalog search interpret `count=0` as no limit.
+   This was the case for a long time. This was changed unintentionally
+   (@veloman-yunkan #686)
+ * [server] Correctly generere a human friendly title in the server frontend.
+   (@juuz0 #687, @kelson42 #689)
+ * [server] Fix download button if there is no url do download from.
+   (@juuz0 #691)
+ * Add non-minified isotope.pkdg.js
+   Needed for debian packaging as we need the source and minified version is
+   not the source (@legoktm #693)
+ * [server] Add a tooltip with the full language for the lang tag.
+ * CI fixes (@kelson42 @legoktm)
+
+libkiwix 10.0.0
+===============
+
+ This release is huge release.
+ The project has been renamed to libkiwix, it is more coherent with the library name.
+
+ * Server front page :
+  - Use js in the front page to display the available book,
+     using the OPDS stream as source. The front page is now populated only with
+     the visible books and user can search for books. (@MananJethwany #530, #541, #534)
+     (@kelson42 #628)
+  - Revamp css (@MananJethwany #559)
+  - Correctly Convert 3iso language code to 2iso (@juuz0 #672)
+
+ * Server suggestions search :
+  - Add pagination for suggestion search (@maneeshpm #591)
+  - Fix suggestion system (@MananJethwany #498)
+  - Provide the kind and path (when adapted) to the suggestion answer (@MananJethwany #464)
+  - The displayed suggestion have now highligth on the searched terms (@maneeshpm #505)
+  - Properly handle html encoding of suggestions (@veloman-yunkan #458)
+
+ * Server improvements :
+  - Remove meta endpoints (@mgautier #669)
+  - Add raw endpoints to get the raw content of a zim (@mgautierfr #646)
+  - Add details on 404 error pages (@soumyankar #490)
+  - Fix headbar insertion when `<head>` tag has attributes (@kelson42 #440)
+  - Better headbar insertion (after charset definition) (@kelson42 #442)
+
+ * New OPDS Stream v2 :
+  - Add a list of categories (@veloman-yunkan)
+  - Support for partial entries (@veloman-yunkan #602)
+  - Support multiple icons size in the OPDS stream (@veloman-yunkan #577 #630)
+  - Add language endpoint to catalog (@veloman-yunkan #553)
+  - Add illustration API to get the illustration of a book (@mgautierfr #645)
+  - OPDS search can now filter books by category (@veloman-yunkan #459)
+
+ * Library improvements :
+  - Allow the libray to be live reloaded when the library.xml changes (@veloman-yunkan #636)
+  - Properly handle removing of book from the library (@veloman-yunkan #485)
+  - Use xapian to search for books in the library (@veloman-yunkan #460, #488)
+
+ * Added methods/functions :
+  - Fix `fileExist` and introduce `fileReadable` (@juuz0 #668)
+  - Add `getVersions` and `printVersions` functions (@kelson42 #665)
+  - Add `getNetworkInterfaces()` and `getBestPublicIP()` functions (@juuz0 #622)
+  - Add `get_zimid()` method to the search result (@maneeshpm #510)
+
+ * Various improvements :
+  - Better secret value for aria2c rpc (@juuz0 #666)
+  - Avoid duplicated Archive/Reader in the Searcher (@veloman-yunkan #648)
+  - Add basic documentation (@mgautierfr #640)
+  - Do not use Reader internally (@maneeshpm #536 #576)
+  - Remove dependency headers from our public headers (@mgautierfr #574)
+  - Downloader now don't write metalink on the filesystem (@kelson42 #502)
+  - Support opening a zim file using a fd (@veloman-yukan #429)
+  - Use C++11 std::thread instead of pthread (@mgautierfr #445)
+  - [READER] Do not crash if zim file has no `Counter` metadata (@mgautierfr #449)
+  - Ensure libzim dependency is compiled with xapian (@mgautierfr #434)
+  - Support video and audio mimetype in `getMediaCount` (@kelson42 #439)
+  - Better parsing of the counterMap (@kelson42 #437)
+  - Adapt libkiwix to libzim 7.0.0 (@mgautierfr #428)
+  - Remove deprecated methods (@mgautierfr)
+  - CI: Build package for Ubuntu Hirsute, Impish and Jammy (@legoktm #431 #568) and remove Groovy
+  - Fix compilation for FreeBSD (@swills g#432)
+  - Many fixes and improvement (@MananJethwany, @maneeshpm, @veloman-yunkan, @mgautierfr)
+
+kiwix-lib 9.4.1
+===============
+
+ * Fix `M/Counter` parsing.
+ * [SERVER] Adjust body padding-top for taskbar
+ * Fix potential crash when stoping a server not started.
+ * Various fix in build system and the CI.
+
+kiwix-lib 9.4.0
+===============
+
+ * [SERVER] Make the headers handling case insensitive.
+ * [SERVER] Make server answer 204 http status code for empty search
+ * [PACKAGING] Made CI build deb packages.
+ * [SERVER] Add a way to prevent taskbar and external link bloquer at article
+   level.
+ * Fix meson file to be compatible with meson 0.45
+ * [SERVER] Update search requests to use pageStart/pageLength instead of
+   pageStart/pageEnd arguments.
+ * [SERVER] Set a fixed favicon size in the main page.
+ * [SERVER] Refactor the response system code to better handling future new
+   libzim api.
+ * Fix segmentation fault around exchange with aria2 process making
+   kiwix-desktop crash at exit.
+
+kiwix-lib 9.3.1
+===============
+
+ * Fix handling of samba path on windows.
+ * Do not include `kiwix_config.h` in public header.
+ * Fix compilation with libmicrohttpd v0.97.1
+ * Increase default test timeout to 160seconds/test.
+ * Add automatic debian packaging.
+ * Use non-minified version of jquery-ui.js
+ * Pass `-latomic` compile option for sh4 architecture.
+ * Make mesion install `kiwix-compile-resources` man page.
+
+kiwix-lib 9.3.0
+===============
+
+ * Add a thread safe method to search suggestions.
+   Old methods are now deprecated.
+
+kiwix-lib 9.2.3
+===============
+
+ * Add test on byte-range
+ * Fix compilation on bionic and windows.
+ * Allow building using debian packaged kainjow-mustache
+ * Pass `-latomic` compile option for architectures that need it
+
+kiwix-lib 9.2.2
+===============
+
+ * Fix handling on empty content in byte range management (wrong assert)
+
+kiwix-lib 9.2.1
+===============
+
+ * Fix support of byte range request.
+
 kiwix-lib 9.2
 =============
 

README.md

@@ -1,15 +1,16 @@
-Kiwix library
-=============
+Libkiwix
+========
 
-The Kiwix library provides the [Kiwix](https://kiwix.org) software
-suite core. It contains the code shared by all Kiwix ports (Windows,
+The Libkiwix provides the [Kiwix](https://kiwix.org) software suite
+core. It contains the code shared by all Kiwix ports (Windows,
 GNU/Linux, macOS, Android, iOS, ...).
 
-[![Download](https://api.bintray.com/packages/kiwix/kiwix/kiwixlib/images/download.svg)](https://bintray.com/kiwix/kiwix/kiwixlib/_latestVersion)
-[![AUR version](https://img.shields.io/aur/version/kiwix-lib)](https://aur.archlinux.org/packages/kiwix-lib/)
-[![Build Status](https://github.com/kiwix/kiwix-lib/workflows/CI/badge.svg?query=branch%3Amaster)](https://github.com/kiwix/kiwix-lib/actions?query=branch%3Amaster)
-[![CodeFactor](https://www.codefactor.io/repository/github/kiwix/kiwix-lib/badge)](https://www.codefactor.io/repository/github/kiwix/kiwix-lib)
-[![Codecov](https://codecov.io/gh/kiwix/kiwix-lib/branch/master/graph/badge.svg)](https://codecov.io/gh/kiwix/kiwix-lib)
+[![Release](https://img.shields.io/github/v/tag/kiwix/libkiwix?label=release&sort=semver)](https://download.kiwix.org/release/libkiwix/)
+[![Repositories](https://img.shields.io/repology/repositories/libkiwix?label=repositories)](https://github.com/kiwix/libkiwix/wiki/Repology)
+[![Build Status](https://github.com/kiwix/libkiwix/workflows/CI/badge.svg?query=branch%3Amaster)](https://github.com/kiwix/libkiwix/actions?query=branch%3Amaster)
+[![Doc](https://readthedocs.org/projects/libkiwix/badge/?style=flat)](https://libkiwix.readthedocs.org/en/latest/?badge=latest)
+[![CodeFactor](https://www.codefactor.io/repository/github/kiwix/libkiwix/badge)](https://www.codefactor.io/repository/github/kiwix/libkiwix)
+[![Codecov](https://codecov.io/gh/kiwix/libkiwix/branch/master/graph/badge.svg)](https://codecov.io/gh/kiwix/libkiwix)
 [![License: GPL v3](https://img.shields.io/badge/License-GPLv3-blue.svg)](https://www.gnu.org/licenses/gpl-3.0)
 
 Disclaimer
@@ -17,31 +18,38 @@ Disclaimer
 
 This document assumes you have a little knowledge about software
 compilation. If you experience difficulties with the dependencies or
-with the Kiwix libary compilation itself, we recommend to have a look
-to [kiwix-build](https://github.com/kiwix/kiwix-build).
+with the Libkiwix compilation itself, we recommend to have a look to
+[kiwix-build](https://github.com/kiwix/kiwix-build).
 
 Preamble
 --------
 
-Although the Kiwix library can be (cross-)compiled on/for many
-sytems, the following documentation explains how to do it on POSIX
-ones. It is primarly thought for GNU/Linux systems and has been tested
-on recent releases of Ubuntu and Fedora.
+Although the Libkiwix can be (cross-)compiled on/for many sytems, the
+following documentation explains how to do it on POSIX ones. It is
+primarly thought for GNU/Linux systems and has been tested on recent
+releases of Ubuntu and Fedora.
 
 Dependencies
 ------------
 
-The Kiwix library relies on many third parts software libraries. They
-are prerequisites to the Kiwix library compilation. Following
-libraries need to be available:
-
+The Libkiwix relies on many third party software libraries. They are
+prerequisites to the Libkiwix compilation. Following libraries need to
+be available:
 * [ICU](https://site.icu-project.org/) (package `libicu-dev` on Ubuntu)
 * [ZIM](https://openzim.org/) (package `libzim-dev` on Ubuntu)
 * [Pugixml](https://pugixml.org/) (package `libpugixml-dev` on Ubuntu)
-* [Aria2](https://aria2.github.io/) (package `aria2` on Ubuntu)
 * [Mustache](https://github.com/kainjow/Mustache) (Just copy the
 header `mustache.hpp` somewhere it can be found by the compiler and/or
-set CPPFLAGS with correct `-I` option). Use Mustache version 3 only.
+set CPPFLAGS with correct `-I` option). Use Mustache version 4.1 or above.
+* [Libcurl](https://curl.se/libcurl) (`libcurl4-gnutls-dev`, `libcurl4-nss-dev` or `libcurl4-openssl-dev` on Ubuntu)
+* [Microhttpd](https://www.gnu.org/software/libmicrohttpd) (package `libmicrohttpd-dev` on Ubuntu)
+* [Zlib](https://zlib.net/) (package `zlib1g-dev` on Ubuntu)
+
+To test the code:
+* [Google Test](https://github.com/google/googletest) (package `googletest` on Ubuntu)
+
+The following dependency needs to be available at runtime:
+* [Aria2](https://aria2.github.io/) (package `aria2` on Ubuntu)
 
 These dependencies may or may not be packaged by your operating
 system. They may also be packaged but only in an older version. The
@@ -50,13 +58,13 @@ In the worse case, you will have to download and compile bleeding edge
 version by hand.
 
 If you want to install these dependencies locally, then use the
-`kiwix-lib` directory as install prefix.
+`libkiwix` directory as install prefix.
 
 Environment
 -------------
 
-The Kiwix library builds using [Meson](https://mesonbuild.com/) version
-0.43 or higher. Meson relies itself on Ninja, pkg-config and few other
+The Libkiwix builds using [Meson](https://mesonbuild.com/) version
+0.45 or higher. Meson relies itself on Ninja, pkg-config and few other
 compilation tools.
 
 Install first the few common compilation tools:
@@ -71,7 +79,7 @@ section.
 Compilation
 -----------
 
-Once all dependencies are installed, you can compile the Kiwix library
+Once all dependencies are installed, you can compile the Libkiwix
 with:
 ```bash
 meson . build
@@ -79,12 +87,20 @@ ninja -C build
 ```
 
 By default, it will compile dynamic linked libraries. All binary files
-will be created in the "build" directory created automatically by
+will be created in the `build` directory created automatically by
 Meson. If you want statically linked libraries, you can add
 `--default-library=static` option to the Meson command.
 
 Depending of you system, `ninja` may be called `ninja-build`.
 
+The android wrapper uses deprecated methods of libkiwix so it cannot be compiled
+with `werror=true` (the default). So you must pass `-Dwerror=false` to meson:
+
+```bash
+meson . build -Dwrapper=android -Dwerror=false
+ninja -C build
+```
+
 Testing
 -------
 
@@ -97,7 +113,7 @@ meson test
 Installation
 ------------
 
-If you want to install the Kiwix library and the headers you just have
+If you want to install the Libkiwix and the headers you just have
 compiled on your system, here we go:
 ```bash
 ninja -C build install
@@ -140,6 +156,49 @@ cp ninja ../bin
 cd ..
 ```
 
+Custom Index Page
+-----------------
+
+to use custom welcome page mention `customIndexPage` argument in `kiwix::internalServer()` or use `kiwix::server->setCustomIndexTemplate()`.
+(note - while using custom html file please mention all external links as absolute path.)
+
+to create a HTML template with custom JS you need to have a look at various OPDS based endpoints as mentioned [here](https://wiki.kiwix.org/wiki/OPDS) to load books.
+
+To use JS provided by kiwix-serve you can use the following template to start with ->
+
+```
+<!DOCTYPE html>
+<html xmlns="http://www.w3.org/1999/xhtml">
+  <head>
+    <meta charset="UTF-8" />
+    <meta name="viewport" content="width=device-width,initial-scale=1" />
+    <title><-- Custom Tittle --></title>
+    <script src="{{root}}/skin/isotope.pkgd.min.js" defer></script>
+    <script src="{{root}}/skin/iso6391To3.js"></script>
+    <script type="text/javascript" src="{{root}}/skin/index.js" defer></script>
+  </head>
+  <body>
+  </body>
+</html>
+```
+
+- To get books listed using `index.js` add - `<div class="book__list"></div>` under body tag.
+- To get number of books listed add - `<h3 class="kiwixHomeBody__results"></h3>` under body tag.
+- To add language select box add - `<select id="languageFilter"></select>` under body tag.
+- To add language select box add - `<select id="categoryFilter"></select>` under body tag.
+- To add search box for books use following form -
+    ```
+        <form id='kiwixSearchForm'>
+        <input type="text" name="q" placeholder="Search" id="searchFilter" class='kiwixSearch filter'>
+        <input type="submit" class="kiwixButton" value="Search"/>
+        </form>
+    ```
+
+
+If you compile manually Libmicrohttpd, you might need to compile it
+without GNU TLS, a bug here will empeach further compilation
+otherwise.
+
 If the compilation still fails, you might need to get a more recent
 version of a dependency than the one packaged by your Linux
 distribution. Try then with a source tarball distributed by the

android-kiwix-lib-publisher/kiwixLibAndroid/build.gradle

@@ -26,10 +26,10 @@ task writePom {
     project {
         groupId 'org.kiwix.kiwixlib'
         artifactId 'kiwixlib'
-        version '9.2' + (System.env.KIWIXLIB_BUILDVERSION == null ? '' : '-'+System.env.KIWIXLIB_BUILDVERSION)
+        version '10.1.1' + (System.env.KIWIXLIB_BUILDVERSION == null ? '' : '-'+System.env.KIWIXLIB_BUILDVERSION)
         packaging 'aar'
         name 'kiwixlib'
-        url 'https://github.com/kiwix/kiwix-lib'
+        url 'https://github.com/kiwix/libkiwix'
         licenses {
             license {
                 name 'GPLv3'
@@ -44,9 +44,9 @@ task writePom {
           }
       }
       scm {
-          connection 'https://github.com/kiwix/kiwix-lib.git'
-          developerConnection 'https://github.com/kiwix/kiwix-lib.git'
-          url 'https://github.com/kiwix/kiwix-lib'
+          connection 'https://github.com/kiwix/libkiwix.git'
+          developerConnection 'https://github.com/kiwix/libkiwix.git'
+          url 'https://github.com/kiwix/libkiwix'
       }
     }
   }.withXml {

debian/changelog

@@ -0,0 +1,5 @@
+libkiwix (0.0.0) unstable; urgency=medium
+
+  * Initial release
+
+ -- Kunal Mehta <legoktm@debian.org>  Wed, 08 Jul 2020 18:12:32 -0700

debian/control

@@ -0,0 +1,46 @@
+Source: libkiwix
+Priority: optional
+Maintainer: Kiwix team <kiwix@kiwix.org>
+Build-Depends: debhelper-compat (= 13),
+ meson,
+ pkg-config,
+ libzim-dev (>= 7.2.0~),
+ libcurl4-gnutls-dev,
+ libicu-dev,
+ libgtest-dev,
+ libkainjow-mustache-dev,
+ liblzma-dev,
+ libmicrohttpd-dev,
+ libpugixml-dev,
+ zlib1g-dev
+Standards-Version: 4.5.0
+Section: libs
+Homepage: https://github.com/kiwix/libkiwix
+Rules-Requires-Root: no
+
+Package: libkiwix-dev
+Section: libdevel
+Architecture: any
+Multi-Arch: same
+Depends: libkiwix10 (= ${binary:Version}), ${misc:Depends}, python3,
+ libzim-dev (>= 7.2.0~),
+ libicu-dev,
+ libpugixml-dev,
+ libcurl4-gnutls-dev,
+ libmicrohttpd-dev
+Description: library of common code for Kiwix (development)
+ Kiwix is an offline Wikipedia reader. libkiwix provides the
+ software core for Kiwix, and contains the code shared by all
+ Kiwix ports (Windows, Linux, OSX, Android, etc.).
+ .
+ This package contains development files.
+
+Package: libkiwix10
+Architecture: any
+Multi-Arch: same
+Depends: ${shlibs:Depends}, ${misc:Depends}, aria2
+Conflicts: libkiwix0, libkiwix3, libkiwix9
+Description: library of common code for Kiwix
+ Kiwix is an offline Wikipedia reader. libkiwix provides the
+ software core for Kiwix, and contains the code shared by all
+ Kiwix ports (Windows, Linux, OSX, Android, etc.).

debian/copyright

@@ -0,0 +1 @@
+See COPYING in the repository root.

debian/libkiwix-dev.install

@@ -0,0 +1,4 @@
+usr/include
+usr/lib/*/libkiwix.so
+usr/lib/*/pkgconfig
+usr/bin

debian/libkiwix-dev.manpages

@@ -0,0 +1,2 @@
+usr/share/man/man1/kiwix-compile-resources.1*
+usr/share/man/man1/kiwix-compile-i18n.1*

debian/libkiwix10.install

@@ -0,0 +1 @@
+usr/lib/*/libkiwix.so.*

debian/rules

@@ -0,0 +1,8 @@
+#!/usr/bin/make -f
+export DEB_BUILD_MAINT_OPTIONS = hardening=+all
+
+%:
+	dh $@ --buildsystem=meson
+
+override_dh_auto_test:
+	dh_auto_test -- -t 3

debian/source/format

@@ -0,0 +1 @@
+3.0 (native)

docs/.gitignore

@@ -0,0 +1,2 @@
+api
+xml

docs/conf.py

@@ -0,0 +1,72 @@
+# Configuration file for the Sphinx documentation builder.
+#
+# This file only contains a selection of the most common options. For a full
+# list see the documentation:
+# https://www.sphinx-doc.org/en/master/usage/configuration.html
+
+# -- Path setup --------------------------------------------------------------
+
+# If extensions (or modules to document with autodoc) are in another directory,
+# add these directories to sys.path here. If the directory is relative to the
+# documentation root, use os.path.abspath to make it absolute, like shown here.
+#
+import os
+# import sys
+# sys.path.insert(0, os.path.abspath('.'))
+
+
+# -- Project information -----------------------------------------------------
+
+project = 'libkiwix'
+copyright = '2022, libkiwix-team'
+author = 'libkiwix-team'
+
+
+# -- General configuration ---------------------------------------------------
+
+on_rtd = os.environ.get('READTHEDOCS', None) == 'True'
+
+# Add any Sphinx extension module names here, as strings. They can be
+# extensions coming with Sphinx (named 'sphinx.ext.*') or your custom
+# ones.
+extensions = [
+    'breathe',
+    'exhale'
+]
+
+# Add any paths that contain templates here, relative to this directory.
+templates_path = ['_templates']
+
+# List of patterns, relative to source directory, that match files and
+# directories to ignore when looking for source files.
+# This pattern also affects html_static_path and html_extra_path.
+exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
+
+
+if not on_rtd:
+    html_theme = 'sphinx_rtd_theme'
+
+# Add any paths that contain custom static files (such as style sheets) here,
+# relative to this directory. They are copied after the builtin static files,
+# so a file named "default.css" will overwrite the builtin "default.css".
+html_static_path = ['_static']
+
+breathe_projects = {
+    "libkiwix": "./xml"
+}
+breathe_default_project = 'libkiwix'
+
+exhale_args = {
+    "containmentFolder":   "./api",
+    "rootFileName":        "ref_api.rst",
+    "rootFileTitle":       "Reference API",
+    "doxygenStripFromPath":"..",
+    "treeViewIsBootstrap": True,
+    "createTreeView" : True,
+    "exhaleExecutesDoxygen": True,
+    "exhaleDoxygenStdin":    "INPUT = ../include"
+}
+
+primary_domain = 'cpp'
+
+highlight_language = 'cpp'

docs/index.rst

@@ -0,0 +1,15 @@
+.. libkiwix documentation master file, created by
+   sphinx-quickstart on Fri Jul 24 15:40:50 2020.
+   You can adapt this file completely to your liking, but it should at least
+   contain the root `toctree` directive.
+
+Welcome to libkiwix's documentation!
+==================================
+
+.. toctree::
+   :maxdepth: 2
+   :caption: Contents:
+
+   usage
+   api/ref_api
+   widget

docs/meson.build

@@ -0,0 +1,7 @@
+
+sphinx = find_program('sphinx-build', native:true)
+
+sphinx_target = run_target('doc',
+    command: [sphinx, '-bhtml',
+              meson.current_source_dir(),
+              meson.current_build_dir()])

docs/requirements.txt

@@ -0,0 +1,2 @@
+breathe
+exhale

docs/usage.rst

@@ -0,0 +1,15 @@
+Libkiwix programming
+====================
+
+Introduction
+------------
+
+libkiwix is written in C++. To use the library, you need the include files of libkiwix have
+to link against libzim.
+
+Errors are handled with exceptions. When something goes wrong, libkiwix throws an error,
+which is always derived from std::exception.
+
+All classes are defined in the namespace kiwix.
+
+libkiwix is a set of tools to manage zim files and provide some common functionnality.

docs/widget.rst

@@ -0,0 +1,82 @@
+Kiwix serve widget
+====================
+
+Introduction
+------------
+
+The kiwix-serve widget provides an easy to embed way to show the `kiwix-serve` homepage.
+
+Usage
+-----
+
+To use the widget, simply add an iframe with its `src` attribute set to the `widget` endpoint.
+Example HTML Page ::
+
+    <!DOCTYPE html>
+    <html lang="en">
+    <head>
+    <title>Widget Test</title>
+    </head>
+    <body>
+    <iframe src="http://192.168.18.8:8080/widget?disabledesc&disablefilter&disabledownload" width=1000 height=1000></iframe>
+    </body>
+    </html>
+
+This creates an iframe with the kiwix-serve homepage contents.
+
+Arguments are explained below.
+
+Possible Arguments
+-------------------
+
+Currently, the following arguments are supported.
+
+disabledesc (value = N/A)
+    Disables the description part of a tile.
+
+disablefilter (value = N/A)
+    Disables the search filters: language, category, tag and search function.
+
+disableclick (value = N/A)
+    Disables clicking the book to open it for reading.
+
+disabledownload (value = N/A)
+    Disables the download button (if avaialable at all) on the tile.
+
+
+Custom CSS and JS
+-----------------
+
+You can add your custom CSS rules and Javascript code to the widget.
+
+To do that, use the following code as template::
+
+    <iframe id="receiver" src="http://192.168.18.8:8080/widget?disabledesc=&disablefilter=&disabledownload=" width="1000" height="1000">
+    <p>Your browser does not support iframes.</p>
+    </iframe>
+
+    <script>
+    window.onload = function() {
+    var receiver = document.getElementById('receiver').contentWindow;
+    function sendMessage() {
+        let msg = {
+        css: `
+        .book__header {
+            color:red;
+        }`,
+        js: `
+        function widgetTest() {
+        console.log("Testing widget");
+        }
+        widgetTest();
+        `
+        }
+        receiver.postMessage(msg, 'http://192.168.18.8:8080/widget');
+    }
+    sendMessage();
+    }
+    </script>
+
+
+The CSS/JS fields are optional, you may send both or only one.
+

format_code.sh

@@ -7,6 +7,7 @@ files=(
 "include/common/otherTools.h"
 "include/common/regexTools.h"
 "include/common/networkTools.h"
+"include/common/archiveTools.h"
 "include/manager.h"
 "include/reader.h"
 "include/kiwix.h"
@@ -22,6 +23,7 @@ files=(
 "src/common/pathTools.cpp"
 "src/common/regexTools.cpp"
 "src/common/otherTools.cpp"
+"src/common/archiveTools.cpp"
 "src/common/networkTools.cpp"
 "src/common/stringTools.cpp"
 "src/xapianSearcher.cpp"

include/book.h

@@ -21,28 +21,54 @@
 #define KIWIX_BOOK_H
 
 #include <string>
+#include <vector>
+#include <memory>
+#include <mutex>
+#include "common.h"
 
 namespace pugi {
 class xml_node;
 }
 
+namespace zim {
+class Archive;
+}
+
 namespace kiwix
 {
 
 class OPDSDumper;
-class Reader;
 
 /**
  * A class to store information about a book (a zim file)
  */
 class Book
 {
- public:
+ public: // types
+  class Illustration
+  {
+    friend class Book;
+   public:
+    uint16_t width = 48;
+    uint16_t height = 48;
+    std::string mimeType;
+    std::string url;
+
+    const std::string& getData() const;
+
+   private:
+    mutable std::string data;
+    mutable std::mutex mutex;
+  };
+
+  typedef std::vector<std::shared_ptr<const Illustration>> Illustrations;
+
+ public: // functions
   Book();
   ~Book();
 
   bool update(const Book& other);
-  void update(const Reader& reader);
+  void update(const zim::Archive& archive);
   void updateFromXml(const pugi::xml_node& node, const std::string& baseDir);
   void updateFromOpds(const pugi::xml_node& node, const std::string& urlHost);
   std::string getHumanReadableIdFromPath() const;
@@ -59,6 +85,7 @@ class Book
   const std::string& getDate() const { return m_date; }
   const std::string& getUrl() const { return m_url; }
   const std::string& getName() const { return m_name; }
+  std::string getCategory() const;
   const std::string& getTags() const { return m_tags; }
   std::string getTagStr(const std::string& tagName) const;
   bool getTagBool(const std::string& tagName) const;
@@ -67,9 +94,13 @@ class Book
   const uint64_t& getArticleCount() const { return m_articleCount; }
   const uint64_t& getMediaCount() const { return m_mediaCount; }
   const uint64_t& getSize() const { return m_size; }
-  const std::string& getFavicon() const;
-  const std::string& getFaviconUrl() const { return m_faviconUrl; }
-  const std::string& getFaviconMimeType() const { return m_faviconMimeType; }
+  DEPRECATED const std::string& getFavicon() const;
+  DEPRECATED const std::string& getFaviconUrl() const;
+  DEPRECATED const std::string& getFaviconMimeType() const;
+
+  Illustrations getIllustrations() const;
+  std::shared_ptr<const Illustration> getIllustration(unsigned int size) const;
+
   const std::string& getDownloadId() const { return m_downloadId; }
 
   void setReadOnly(bool readOnly) { m_readOnly = readOnly; }
@@ -90,17 +121,20 @@ class Book
   void setArticleCount(uint64_t articleCount) { m_articleCount = articleCount; }
   void setMediaCount(uint64_t mediaCount) { m_mediaCount = mediaCount; }
   void setSize(uint64_t size) { m_size = size; }
-  void setFavicon(const std::string& favicon) { m_favicon = favicon; }
-  void setFaviconMimeType(const std::string& faviconMimeType) { m_faviconMimeType = faviconMimeType; }
   void setDownloadId(const std::string& downloadId) { m_downloadId = downloadId; }
 
- protected:
+ private: // functions
+  std::string getCategoryFromTags() const;
+  const Illustration& getDefaultIllustration() const;
+
+ protected: // data
   std::string m_id;
   std::string m_downloadId;
   std::string m_path;
   bool m_pathValid = false;
   std::string m_title;
   std::string m_description;
+  std::string m_category;
   std::string m_language;
   std::string m_creator;
   std::string m_publisher;
@@ -114,9 +148,11 @@ class Book
   uint64_t m_mediaCount = 0;
   bool m_readOnly = false;
   uint64_t m_size = 0;
-  mutable std::string m_favicon;
-  std::string m_faviconUrl;
-  std::string m_faviconMimeType;
+  Illustrations m_illustrations;
+
+  // Used as the return value of getDefaultIllustration() when no default
+  // illustration is found in the book
+  static const Illustration missingDefaultIllustration;
 };
 
 }

include/downloader.h

@@ -23,7 +23,6 @@
 #include <string>
 #include <vector>
 #include <map>
-#include <pthread.h>
 #include <memory>
 #include <stdexcept>
 

include/entry.h

@@ -1,192 +0,0 @@
-/*
- * Copyright 2018 Matthieu Gautier <mgautier@kymeria.fr>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#ifndef KIWIX_ENTRY_H
-#define KIWIX_ENTRY_H
-
-#include <stdio.h>
-#include <zim/article.h>
-#include <exception>
-#include <string>
-
-#include "common.h"
-
-using namespace std;
-
-namespace kiwix
-{
-
-
-class NoEntry : public std::exception {};
-
-/**
- * A entry represent an.. entry in a zim file.
- */
-class Entry
-{
-  public:
-    /**
-     * Default constructor.
-     *
-     * Construct an invalid entry.
-     */
-    Entry() = default;
-
-    /**
-     * Construct an entry making reference to an zim article.
-     *
-     * @param article a zim::Article object
-     */
-    Entry(zim::Article article);
-    virtual ~Entry() = default;
-
-    /**
-     * Get the path of the entry.
-     *
-     * The path is the "key" of an entry.
-     *
-     * @return the path of the entry.
-     */
-    std::string getPath() const;
-
-    /**
-     * Get the title of the entry.
-     *
-     * @return the title of the entry.
-     */
-    std::string getTitle() const;
-
-    /**
-     * Get the content of the entry.
-     *
-     * The string is a copy of the content.
-     * If you don't want to do a copy, use get_blob.
-     *
-     * @return the content of the entry.
-     */
-    std::string getContent() const;
-
-    /**
-     * Get the blob of the entry.
-     *
-     * A blob make reference to the content without copying it.
-     *
-     * @param offset The starting offset of the blob.
-     * @return the blob of the entry.
-     */
-    zim::Blob   getBlob(offset_type offset = 0) const;
-
-    /**
-     * Get the blob of the entry.
-     *
-     * A blob make reference to the content without copying it.
-     *
-     * @param offset The starting offset of the blob.
-     * @param size The size of the blob.
-     * @return the blob of the entry.
-     */
-    zim::Blob   getBlob(offset_type offset, size_type size) const;
-
-    /**
-     * Get the info for direct access to the content of the entry.
-     *
-     * Some entry (ie binary ones) have their content plain stored
-     * in the zim file. Knowing the offset where the content is stored
-     * an user can directly read the content in the zim file bypassing the
-     * kiwix-lib/libzim.
-     *
-     * @return A pair specifying where to read the content.
-     *         The string is the real file to read (may be different that .zim
-     *         file if zim is cut).
-     *         The offset is the offset to read in the file.
-     *         Return <"",0> if is not possible to read directly.
-     */
-    std::pair<std::string, offset_type> getDirectAccessInfo() const;
-
-    /**
-     * Get the size of the entry.
-     *
-     * @return the size of the entry.
-     */
-    size_type   getSize() const;
-
-    /**
-     * Get the mime_type of the entry.
-     *
-     * @return the mime_type of the entry.
-     */
-    std::string getMimetype() const;
-
-
-    /**
-     * Get if the entry is a redirect entry.
-     *
-     * @return True if the entry is a redirect.
-     */
-    bool isRedirect() const;
-
-    /**
-     * Get if the entry is a link target entry.
-     *
-     * @return True if the entry is a link target.
-     */
-    bool isLinkTarget() const;
-
-    /**
-     * Get if the entry is a deleted entry.
-     *
-     * @return True if the entry is a deleted entry.
-     */
-    bool isDeleted() const;
-
-    /**
-     * Get the entry pointed by this entry.
-     *
-     * @return the entry pointed.
-     * @throw NoEntry if the entry is not a redirected entry.
-     */
-    Entry getRedirectEntry() const;
-
-    /**
-     * Get the final entry pointed by this entry.
-     *
-     * Follow the redirection until a "not redirecting" entry is found.
-     * If the entry is not a redirected entry, return the entry itself.
-     *
-     * @return the final entry.
-     */
-    Entry getFinalEntry() const;
-
-    /**
-     * Convert the entry to a boolean value.
-     *
-     * @return True if the entry is valid.
-     */
-    explicit operator bool() const { return good(); }
-
-  private:
-    zim::Article article;
-    mutable zim::Article final_article;
-
-    bool good() const { return article.good(); }
-};
-
-}
-
-#endif // KIWIX_ENTRY_H

include/kiwix.h

@@ -22,4 +22,4 @@
 
 #include "library.h"
 
-#endif
+#endif

include/library.h

@@ -24,6 +24,9 @@
 #include <vector>
 #include <map>
 #include <memory>
+#include <mutex>
+#include <zim/archive.h>
+#include <zim/search.h>
 
 #include "book.h"
 #include "bookmark.h"
@@ -35,6 +38,7 @@ namespace kiwix
 {
 
 class OPDSDumper;
+class Library;
 
 enum supportedListSortBy { UNSORTED, TITLE, SIZE, DATE, CREATOR, PUBLISHER };
 enum supportedListMode {
@@ -48,18 +52,23 @@ enum supportedListMode {
 };
 
 class Filter {
-  private:
+  public: // types
+    using Tags = std::vector<std::string>;
+
+  private: // data
     uint64_t activeFilters;
-    std::vector<std::string> _acceptTags;
-    std::vector<std::string> _rejectTags;
+    Tags _acceptTags;
+    Tags _rejectTags;
+    std::string _category;
     std::string _lang;
     std::string _publisher;
     std::string _creator;
     size_t _maxSize;
     std::string _query;
+    bool _queryIsPartial;
     std::string _name;
 
-  public:
+  public: // functions
     Filter();
     ~Filter() = default;
 
@@ -93,33 +102,87 @@ class Filter {
     /**
      * Set the filter to only accept book with corresponding tag.
      */
-    Filter& acceptTags(std::vector<std::string> tags);
-    Filter& rejectTags(std::vector<std::string> tags);
+    Filter& acceptTags(const Tags& tags);
+    Filter& rejectTags(const Tags& tags);
 
+    Filter& category(std::string category);
     Filter& lang(std::string lang);
     Filter& publisher(std::string publisher);
     Filter& creator(std::string creator);
     Filter& maxSize(size_t size);
-    Filter& query(std::string query);
+    Filter& query(std::string query, bool partial=true);
     Filter& name(std::string name);
 
+    bool hasQuery() const;
+    const std::string& getQuery() const { return _query; }
+    bool queryIsPartial() const { return _queryIsPartial; }
+
+    bool hasName() const;
+    const std::string& getName() const { return _name; }
+
+    bool hasCategory() const;
+    const std::string& getCategory() const { return _category; }
+
+    bool hasLang() const;
+    const std::string& getLang() const { return _lang; }
+
+    bool hasPublisher() const;
+    const std::string& getPublisher() const { return _publisher; }
+
+    bool hasCreator() const;
+    const std::string& getCreator() const { return _creator; }
+
+    const Tags& getAcceptTags() const { return _acceptTags; }
+    const Tags& getRejectTags() const { return _rejectTags; }
+
+private: // functions
+    friend class Library;
+
     bool accept(const Book& book) const;
 };
 
 
+class ZimSearcher : public zim::Searcher
+{
+  public:
+    explicit ZimSearcher(zim::Searcher&& searcher)
+      : zim::Searcher(searcher)
+    {}
+
+    std::unique_lock<std::mutex> getLock() {
+      return std::unique_lock<std::mutex>(m_mutex);
+    }
+    virtual ~ZimSearcher() = default;
+  private:
+    std::mutex m_mutex;
+};
+
 /**
  * A Library store several books.
  */
 class Library
 {
-  std::map<std::string, kiwix::Book> m_books;
-  std::map<std::string, std::shared_ptr<Reader>> m_readers;
-  std::vector<kiwix::Bookmark> m_bookmarks;
+  // all data fields must be added in LibraryBase
+  mutable std::mutex m_mutex;
+
+ public:
+  typedef uint64_t Revision;
+  typedef std::vector<std::string> BookIdCollection;
+  typedef std::map<std::string, int> AttributeCounts;
+  typedef std::set<std::string> BookIdSet;
 
  public:
   Library();
   ~Library();
 
+  /**
+   * Library is not a copiable object. However it can be moved.
+   */
+  Library(const Library& ) = delete;
+  Library(Library&& );
+  void operator=(const Library& ) = delete;
+  Library& operator=(Library&& );
+
   /**
    * Add a book to the library.
    *
@@ -132,6 +195,11 @@ class Library
    */
   bool addBook(const Book& book);
 
+  /**
+   * A self-explanatory alias for addBook()
+   */
+  bool addOrUpdateBook(const Book& book) { return addBook(book); }
+
   /**
    * Add a bookmark to the library.
    *
@@ -148,9 +216,18 @@ class Library
    */
   bool removeBookmark(const std::string& zimId, const std::string& url);
 
-  Book& getBookById(const std::string& id);
-  Book& getBookByPath(const std::string& path);
-  std::shared_ptr<Reader> getReaderById(const std::string& id);
+  // XXX: This is a non-thread-safe operation
+  const Book& getBookById(const std::string& id) const;
+  // XXX: This is a non-thread-safe operation
+  const Book& getBookByPath(const std::string& path) const;
+
+  Book getBookByIdThreadSafe(const std::string& id) const;
+
+  std::shared_ptr<zim::Archive> getArchiveById(const std::string& id);
+  std::shared_ptr<ZimSearcher> getSearcherById(const std::string& id) {
+    return getSearcherByIds(BookIdSet{id});
+  }
+  std::shared_ptr<ZimSearcher> getSearcherByIds(const BookIdSet& ids);
 
   /**
    * Remove a book from the library.
@@ -166,7 +243,7 @@ class Library
    * @param path the path of the file to write to.
    * @return True if the library has been correctly saved.
    */
-  bool writeToFile(const std::string& path);
+  bool writeToFile(const std::string& path) const;
 
   /**
    * Write the library bookmarks to a file.
@@ -174,7 +251,7 @@ class Library
    * @param path the path of the file to write to.
    * @return True if the library has been correctly saved.
    */
-  bool writeBookmarksToFile(const std::string& path);
+  bool writeBookmarksToFile(const std::string& path) const;
 
   /**
    * Get the number of book in the library.
@@ -183,53 +260,56 @@ class Library
    * @param remoteBooks If we must count remote books (books with an url)
    * @return The number of books.
    */
-  unsigned int getBookCount(const bool localBooks, const bool remoteBooks);
+  unsigned int getBookCount(const bool localBooks, const bool remoteBooks) const;
 
   /**
-   * Get all langagues of the books in the library.
+   * Get all languagues of the books in the library.
    *
    * @return A list of languages.
    */
-  std::vector<std::string> getBooksLanguages();
+  std::vector<std::string> getBooksLanguages() const;
+
+  /**
+   * Get all languagues of the books in the library with counts.
+   *
+   * @return A list of languages with the count of books in each language.
+   */
+  AttributeCounts getBooksLanguagesWithCounts() const;
+
+  /**
+   * Get all categories of the books in the library.
+   *
+   * @return A list of categories.
+   */
+  std::vector<std::string> getBooksCategories() const;
 
   /**
    * Get all book creators of the books in the library.
    *
    * @return A list of book creators.
    */
-  std::vector<std::string> getBooksCreators();
+  std::vector<std::string> getBooksCreators() const;
 
   /**
    * Get all book publishers of the books in the library.
    *
    * @return A list of book publishers.
    */
-  std::vector<std::string> getBooksPublishers();
+  std::vector<std::string> getBooksPublishers() const;
 
   /**
    * Get all bookmarks.
    *
    * @return A list of bookmarks
    */
-  const std::vector<kiwix::Bookmark> getBookmarks(bool onlyValidBookmarks = true);
+  const std::vector<kiwix::Bookmark> getBookmarks(bool onlyValidBookmarks = true) const;
 
   /**
    * Get all book ids of the books in the library.
    *
    * @return A list of book ids.
    */
-  std::vector<std::string> getBooksIds();
-
-  /**
-   * Filter the library and generate a new one with the keep elements.
-   *
-   * This is equivalent to `listBookIds(ALL, UNSORTED, search)`.
-   *
-   * @param search List only books with search in the title or description.
-   * @return The list of bookIds corresponding to the query.
-   */
-  DEPRECATED std::vector<std::string> filter(const std::string& search);
-
+  BookIdCollection getBooksIds() const;
 
   /**
    * Filter the library and return the id of the keep elements.
@@ -237,7 +317,7 @@ class Library
    * @param filter The filter to use.
    * @return The list of bookIds corresponding to the filter.
    */
-  std::vector<std::string> filter(const Filter& filter);
+  BookIdCollection filter(const Filter& filter) const;
 
 
   /**
@@ -247,43 +327,44 @@ class Library
    * @param comparator how to sort the books
    * @return The sorted list of books
    */
-  void sort(std::vector<std::string>& bookIds, supportedListSortBy sortBy, bool ascending);
+  void sort(BookIdCollection& bookIds, supportedListSortBy sortBy, bool ascending) const;
 
   /**
-   * List books in the library.
+   * Return the current revision of the library.
    *
-   * @param mode The mode of listing :
-   *             - LOCAL    : list only local books (with a path).
-   *             - REMOTE   : list only remote books (with an url).
-   *             - VALID    : list only valid books (without a path or with a
-   *                          path pointing to a valid zim file).
-   *             - NOLOCAL  : list only books without valid path.
-   *             - NOREMOTE : list only books without url.
-   *             - NOVALID  : list only books not valid.
-   *             - ALL : Do not do any filter (LOCAL or REMOTE)
-   *             - Flags can be combined.
-   * @param sortBy Attribute to sort by the book list.
-   * @param search List only books with search in the title, description.
-   * @param language List only books in this language.
-   * @param creator List only books of this creator.
-   * @param publisher List only books of this publisher.
-   * @param maxSize Do not list book bigger than maxSize.
-   *                Set to 0 to cancel this filter.
-   * @return The list of bookIds corresponding to the query.
+   * The revision of the library is updated (incremented by one) only by
+   * the addBook() operation.
+   *
+   * @return Current revision of the library.
    */
-  DEPRECATED std::vector<std::string> listBooksIds(
-    int supportedListMode = ALL,
-    supportedListSortBy sortBy = UNSORTED,
-    const std::string& search = "",
-    const std::string& language = "",
-    const std::string& creator = "",
-    const std::string& publisher = "",
-    const std::vector<std::string>& tags = {},
-    size_t maxSize = 0);
+  Revision getRevision() const;
+
+  /**
+   * Remove books that have not been updated since the specified revision.
+   *
+   * @param  rev the library revision to use
+   * @return Count of books that were removed by this operation.
+   */
+  uint32_t removeBooksNotUpdatedSince(Revision rev);
 
   friend class OPDSDumper;
   friend class libXMLDumper;
+
+private: // types
+  typedef const std::string& (Book::*BookStrPropMemFn)() const;
+  struct Impl;
+
+private: // functions
+  AttributeCounts getBookAttributeCounts(BookStrPropMemFn p) const;
+  std::vector<std::string> getBookPropValueSet(BookStrPropMemFn p) const;
+  BookIdCollection filterViaBookDB(const Filter& filter) const;
+  void updateBookDB(const Book& book);
+  void dropCache(const std::string& bookId);
+
+private: //data
+  std::unique_ptr<Impl> mp_impl;
 };
+
 }
 
 #endif

include/libxml_dumper.h

@@ -38,7 +38,7 @@ class LibXMLDumper
 {
  public:
   LibXMLDumper() = default;
-  LibXMLDumper(Library* library);
+  LibXMLDumper(const Library* library);
   ~LibXMLDumper();
 
   /**
@@ -69,10 +69,10 @@ class LibXMLDumper
    *
    * @param library The library to dump.
    */
-  void setLibrary(Library* library) { this->library = library; }
+  void setLibrary(const Library* library) { this->library = library; }
 
  protected:
-  kiwix::Library* library;
+  const kiwix::Library* library;
   std::string baseDir;
  private:
   void handleBook(Book book, pugi::xml_node root_node);

include/manager.h

@@ -22,10 +22,10 @@
 
 #include "book.h"
 #include "library.h"
-#include "reader.h"
 
 #include <string>
 #include <vector>
+#include <memory>
 
 namespace pugi {
 class xml_document;
@@ -34,26 +34,25 @@ class xml_document;
 namespace kiwix
 {
 
-class LibraryManipulator {
- public:
-  virtual ~LibraryManipulator() {}
-  virtual bool addBookToLibrary(Book book) = 0;
-  virtual void addBookmarkToLibrary(Bookmark bookmark) = 0;
-};
+class LibraryManipulator
+{
+ public: // functions
+  explicit LibraryManipulator(Library* library);
+  virtual ~LibraryManipulator();
 
-class DefaultLibraryManipulator : public LibraryManipulator {
- public:
-  DefaultLibraryManipulator(Library* library) :
-    library(library) {}
-  virtual ~DefaultLibraryManipulator() {}
-  bool addBookToLibrary(Book book) {
-    return library->addBook(book);
-  }
-  void addBookmarkToLibrary(Bookmark bookmark) {
-    library->addBookmark(bookmark);
-  }
- private:
-   kiwix::Library* library;
+  Library& getLibrary() const { return library; }
+
+  bool addBookToLibrary(const Book& book);
+  void addBookmarkToLibrary(const Bookmark& bookmark);
+  uint32_t removeBooksNotUpdatedSince(Library::Revision rev);
+
+ protected: // overrides
+  virtual void bookWasAddedToLibrary(const Book& book);
+  virtual void bookmarkWasAddedToLibrary(const Bookmark& bookmark);
+  virtual void booksWereRemovedFromLibrary();
+
+ private: // data
+  kiwix::Library& library;
 };
 
 /**
@@ -61,10 +60,12 @@ class DefaultLibraryManipulator : public LibraryManipulator {
  */
 class Manager
 {
- public:
-  Manager(LibraryManipulator* manipulator);
-  Manager(Library* library);
-  ~Manager();
+ public: // types
+  typedef std::vector<std::string> Paths;
+
+ public: // functions
+  explicit Manager(LibraryManipulator* manipulator);
+  explicit Manager(Library* library);
 
   /**
    * Read a `library.xml` and add book in the file to the library.
@@ -72,10 +73,22 @@ class Manager
    * @param path The (utf8) path to the `library.xml`.
    * @param readOnly Set if the libray path could be overwritten latter with
    *                 updated content.
+   * @param trustLibrary use book metadata coming from XML.
    * @return True if file has been properly parsed.
    */
   bool readFile(const std::string& path, bool readOnly = true, bool trustLibrary = true);
 
+  /**
+   * Sync the contents of the library with one or more `library.xml` files.
+   *
+   * The metadata of the library files is trusted unconditionally.
+   * Any books not present in the input library.xml files are removed
+   * from the library.
+   *
+   * @param paths The (utf8) paths to the `library.xml` files.
+   */
+  void reload(const Paths& paths);
+
   /**
    * Load a library content store in the string.
    *
@@ -150,8 +163,7 @@ class Manager
   uint64_t m_itemsPerPage = 0;
 
  protected:
-  kiwix::LibraryManipulator* manipulator;
-  bool mustDeleteManipulator;
+  std::shared_ptr<kiwix::LibraryManipulator> manipulator;
 
   bool readBookFromPath(const std::string& path, Book* book);
   bool parseXmlDom(const pugi::xml_document& doc,

include/meson.build

@@ -7,25 +7,12 @@ headers = [
   'libxml_dumper.h',
   'opds_dumper.h',
   'downloader.h',
-  'reader.h',
-  'entry.h',
-  'searcher.h',
   'search_renderer.h',
   'server.h',
   'kiwixserve.h',
-  'name_mapper.h'
+  'name_mapper.h',
+  'tools.h',
+  'version.h'
 ]
 
 install_headers(headers, subdir:'kiwix')
-
-install_headers(
-  'tools/base64.h',
-  'tools/networkTools.h',
-  'tools/otherTools.h',
-  'tools/pathTools.h',
-  'tools/regexTools.h',
-  'tools/stringTools.h',
-  'tools/lock.h',
-  subdir:'kiwix/tools'
-)
-

include/name_mapper.h

@@ -22,6 +22,8 @@
 
 #include <string>
 #include <map>
+#include <memory>
+#include <mutex>
 
 namespace kiwix
 {
@@ -31,15 +33,15 @@ class Library;
 class NameMapper {
   public:
     virtual ~NameMapper() = default;
-    virtual std::string getNameForId(const std::string& id) = 0;
-    virtual std::string getIdForName(const std::string& name) = 0;
+    virtual std::string getNameForId(const std::string& id) const = 0;
+    virtual std::string getIdForName(const std::string& name) const = 0;
 };
 
 
 class IdNameMapper : public NameMapper {
   public:
-    virtual std::string getNameForId(const std::string& id) { return id; };
-    virtual std::string getIdForName(const std::string& name) { return name; };
+    virtual std::string getNameForId(const std::string& id) const { return id; };
+    virtual std::string getIdForName(const std::string& name) const { return name; };
 };
 
 class HumanReadableNameMapper : public NameMapper {
@@ -50,11 +52,29 @@ class HumanReadableNameMapper : public NameMapper {
   public:
     HumanReadableNameMapper(kiwix::Library& library, bool withAlias);
     virtual ~HumanReadableNameMapper() = default;
-    virtual std::string getNameForId(const std::string& id);
-    virtual std::string getIdForName(const std::string& name);
+    virtual std::string getNameForId(const std::string& id) const;
+    virtual std::string getIdForName(const std::string& name) const;
 };
 
+class UpdatableNameMapper : public NameMapper {
+    typedef std::shared_ptr<NameMapper> NameMapperHandle;
+  public:
+    UpdatableNameMapper(Library& library, bool withAlias);
 
+    virtual std::string getNameForId(const std::string& id) const;
+    virtual std::string getIdForName(const std::string& name) const;
+
+    void update();
+
+  private:
+    NameMapperHandle currentNameMapper() const;
+
+  private:
+    mutable std::mutex mutex;
+    Library& library;
+    NameMapperHandle nameMapper;
+    const bool withAlias;
+};
 
 }
 

include/opds_catalog.h

@@ -1,6 +1,5 @@
 /*
- * Copyright (C) 2013 Emmanuel Engelhart <kelson@kiwix.org>
- * Copyright (C) 2017 Matthieu Gautier <mgautier@kymeria.fr>
+ * Copyright 2021 Veloman Yunkan <veloman.yunkan@gmail.com>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU  General Public License as published by
@@ -18,9 +17,17 @@
  * MA 02110-1301, USA.
  */
 
-package org.kiwix.kiwixlib;
+#ifndef KIWIX_OPDS_CATALOG_H
+#define KIWIX_OPDS_CATALOG_H
 
-public class JNIICU
+
+#include "library.h"
+
+namespace kiwix
 {
-  static public native void setDataDirectory(String icuDataDir);
-}
+
+std::string getSearchUrl(const Filter& f);
+
+} // namespace kiwix
+
+#endif // KIWIX_OPDS_CATALOG_H

include/opds_dumper.h

@@ -26,11 +26,7 @@
 
 #include <pugixml.hpp>
 
-#include "tools/base64.h"
-#include "tools/pathTools.h"
-#include "tools/regexTools.h"
 #include "library.h"
-#include "reader.h"
 
 using namespace std;
 
@@ -51,24 +47,50 @@ class OPDSDumper
   /**
    * Dump the OPDS feed.
    *
-   * @param id The id of the library.
+   * @param bookIds the ids of the books to include in the feed
+   * @param query the query used to obtain the list of book ids
    * @return The OPDS feed.
    */
-  std::string dumpOPDSFeed(const std::vector<std::string>& bookIds);
+  std::string dumpOPDSFeed(const std::vector<std::string>& bookIds, const std::string& query) const;
 
   /**
-   * Set the id of the opds stream.
+   * Dump the OPDS feed.
+   *
+   * @param bookIds the ids of the books to include in the feed
+   * @param query the query used to obtain the list of book ids
+   * @param partial whether the feed should include partial or complete entries
+   * @return The OPDS feed.
+   */
+  std::string dumpOPDSFeedV2(const std::vector<std::string>& bookIds, const std::string& query, bool partial) const;
+
+  /**
+   * Dump the OPDS complete entry document.
+   *
+   * @param bookId the id of the book
+   * @return The OPDS complete entry document.
+   */
+  std::string dumpOPDSCompleteEntry(const std::string& bookId) const;
+
+  /**
+   * Dump the categories OPDS feed.
+   *
+   * @return The OPDS feed.
+   */
+  std::string categoriesOPDSFeed() const;
+
+  /**
+   * Dump the languages OPDS feed.
+   *
+   * @return The OPDS feed.
+   */
+  std::string languagesOPDSFeed() const;
+
+  /**
+   * Set the id of the library.
    *
    * @param id the id to use.
    */
-  void setId(const std::string& id) { this->id = id;}
-
-  /**
-   * Set the title oft the opds stream.
-   *
-   * @param title the title to use.
-   */
-  void setTitle(const std::string& title) { this->title = title; }
+  void setLibraryId(const std::string& id) { this->libraryId = id;}
 
   /**
    * Set the root location used when generating url.
@@ -77,13 +99,6 @@ class OPDSDumper
    */
   void setRootLocation(const std::string& rootLocation) { this->rootLocation = rootLocation; }
 
-  /**
-   * Set the search url.
-   *
-   * @param searchUrl the search url to use.
-   */
-  void setSearchDescriptionUrl(const std::string& searchDescriptionUrl) { this->searchDescriptionUrl = searchDescriptionUrl; }
-
   /**
    * Set some informations about the search results.
    *
@@ -93,27 +108,13 @@ class OPDSDumper
    */
   void setOpenSearchInfo(int totalResult, int startIndex, int count);
 
-  /**
-   * Set the library to dump.
-   *
-   * @param library The library to dump.
-   */
-  void setLibrary(Library* library) { this->library = library; }
-
  protected:
   kiwix::Library* library;
-  std::string id;
-  std::string title;
-  std::string date;
+  std::string libraryId;
   std::string rootLocation;
-  std::string searchDescriptionUrl;
   int m_totalResults;
   int m_startIndex;
   int m_count;
-  bool m_isSearchResult = false;
-
- private:
-  pugi::xml_node handleBook(Book book, pugi::xml_node root_node);
 };
 }
 

include/reader.h

@@ -1,570 +0,0 @@
-/*
- * Copyright 2011 Emmanuel Engelhart <kelson@kiwix.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#ifndef KIWIX_READER_H
-#define KIWIX_READER_H
-
-#include <stdio.h>
-#include <zim/article.h>
-#include <zim/file.h>
-#include <zim/fileiterator.h>
-#include <zim/zim.h>
-#include <exception>
-#include <map>
-#include <sstream>
-#include <string>
-#include "common.h"
-#include "entry.h"
-#include "tools/pathTools.h"
-#include "tools/stringTools.h"
-
-using namespace std;
-
-namespace kiwix
-{
-
-/**
- * The Reader class is the class who allow to get an entry content from a zim
- * file.
- */
-class Reader
-{
- public:
-  /**
-   * Create a Reader to read a zim file specified by zimFilePath.
-   *
-   * @param zimFilePath The path to the zim file to read.
-   *                    The zim file can be splitted (.zimaa, .zimab, ...).
-   *                    In this case, the file path must still point to the
-   *                    unsplitted path as if the file were not splitted
-   *                    (.zim extesion).
-   */
-  Reader(const string zimFilePath);
-  ~Reader();
-
-  /**
-   * Get the number of "displayable" entries in the zim file.
-   *
-   * @return If the zim file has a /M/Counter metadata, return the number of
-   *         entries with the 'text/html' MIMEtype specified in the metadata.
-   *         Else return the number of entries in the 'A' namespace.
-   */
-  unsigned int getArticleCount() const;
-
-  /**
-   * Get the number of media in the zim file.
-   *
-   * @return If the zim file has a /M/Counter metadata, return the number of
-   *         entries with the 'image/jpeg', 'image/gif' and 'image/png' in
-   *         the metadata.
-   *         Else return the number of entries in the 'I' namespace.
-   */
-  unsigned int getMediaCount() const;
-
-  /**
-   * Get the number of all entries in the zim file.
-   *
-   * @return Return the number of all the entries, whatever their MIMEtype or
-   *         their namespace.
-   */
-  unsigned int getGlobalCount() const;
-
-  /**
-   * Get the path of the zim file.
-   *
-   * @return the path of the zim file as given in the constructor.
-   */
-  string getZimFilePath() const;
-
-  /**
-   * Get the Id of the zim file.
-   *
-   * @return The uuid stored in the zim file.
-   */
-  string getId() const;
-
-  /**
-   * Get the url of a random page.
-   *
-   * Deprecated : Use `getRandomPage` instead.
-   *
-   * @return Url of a random page. The page is picked from all entries in
-   *         the 'A' namespace.
-   *         The main page is excluded from the potential results.
-   */
-  DEPRECATED string getRandomPageUrl() const;
-
-  /**
-   * Get a random page.
-   *
-   * @return A random Entry. The entry is picked from all entries in
-   *         the 'A' namespace.
-   *         The main entry is excluded from the potential results.
-   */
-  Entry getRandomPage() const;
-
-  /**
-   * Get the url of the first page.
-   *
-   * Deprecated : Use `getFirstPage` instead.
-   *
-   * @return Url of the first entry in the 'A' namespace.
-   */
-  DEPRECATED string getFirstPageUrl() const;
-
-  /**
-   * Get the entry of the first page.
-   *
-   * @return The first entry in the 'A' namespace.
-   */
-  Entry getFirstPage() const;
-
-  /**
-   * Get the url of the main page.
-   *
-   * Deprecated : Use `getMainPage` instead.
-   *
-   * @return Url of the main page as specified in the zim file.
-   */
-  DEPRECATED string getMainPageUrl() const;
-
-  /**
-   * Get the entry of the main page.
-   *
-   * @return Entry of the main page as specified in the zim file.
-   */
-  Entry getMainPage() const;
-
-  /**
-   * Get the content of a metadata.
-   *
-   * @param[in]  name The name of the metadata.
-   * @param[out] value The value will be set to the content of the metadata.
-   * @return True if it was possible to get the content of the metadata.
-   */
-  bool getMetadata(const string& name, string& value) const;
-
-  /**
-   * Get the name of the zim file.
-   *
-   * @return The name of the zim file as specified in the zim metadata.
-   */
-  string getName() const;
-
-  /**
-   * Get the title of the zim file.
-   *
-   * @return The title of zim file as specified in the zim metadata.
-   *         If no title has been set, return a title computed from the
-   *         file path.
-   */
-  string getTitle() const;
-
-  /**
-   * Get the creator of the zim file.
-   *
-   * @return The creator of the zim file as specified in the zim metadata.
-   */
-  string getCreator() const;
-
-  /**
-   * Get the publisher of the zim file.
-   *
-   * @return The publisher of the zim file as specified in the zim metadata.
-   */
-  string getPublisher() const;
-
-  /**
-   * Get the date of the zim file.
-   *
-   * @return The date of the zim file as specified in the zim metadata.
-   */
-  string getDate() const;
-
-  /**
-   * Get the description of the zim file.
-   *
-   * @return The description of the zim file as specified in the zim metadata.
-   *         If no description has been set, return the subtitle.
-   */
-  string getDescription() const;
-
-  /**
-   * Get the long description of the zim file.
-   *
-   * @return The long description of the zim file as specifed in the zim metadata.
-   */
-  string getLongDescription() const;
-
-  /**
-   * Get the language of the zim file.
-   *
-   * @return The language of the zim file as specified in the zim metadata.
-   */
-  string getLanguage() const;
-
-  /**
-   * Get the license of the zim file.
-   *
-   * @return The license of the zim file as specified in the zim metadata.
-   */
-  string getLicense() const;
-
-  /**
-   * Get the tags of the zim file.
-   *
-   * @param original If true, return the original tags as specified in the zim metadata.
-   *                 Else, try to convert it to the new 'normalized' format.
-   * @return The tags of the zim file.
-   */
-  string getTags(bool original=false) const;
-
-  /**
-   * Get the value (as a string) of a specific tag.
-   *
-   * According to https://wiki.openzim.org/wiki/Tags
-   *
-   * @return The value of the specified tag.
-   * @throw  std::out_of_range if the specified tag is not found.
-   */
-  string getTagStr(const std::string& tagName) const;
-
-  /**
-   * Get the boolean value of a specific tag.
-   *
-   * According to https://wiki.openzim.org/wiki/Tags
-   *
-   * @return The boolean value of the specified tag.
-   * @throw  std::out_of_range if the specified tag is not found.
-   *         std::domain_error if the value of the tag cannot be convert to bool.
-   */
-  bool getTagBool(const std::string& tagName) const;
-
-  /**
-   * Get the relations of the zim file.
-   *
-   * @return The relation of the zim file as specified in the zim metadata.
-   */
-  string getRelation() const;
-
-  /**
-   * Get the flavour of the zim file.
-   *
-   * @return The flavour of the zim file as specified in the zim metadata.
-   */
-  string getFlavour() const;
-
-  /**
-   * Get the source of the zim file.
-   *
-   * @return The source of the zim file as specified in the zim metadata.
-   */
-  string getSource() const;
-
-  /**
-   * Get the scraper of the zim file.
-   *
-   * @return The scraper of the zim file as specified in the zim metadata.
-   */
-  string getScraper() const;
-
-  /**
-   * Get the origId of the zim file.
-   *
-   * The origId is only used in the case of patch zim file and is the Id
-   * of the original zim file.
-   *
-   * @return The origId of the zim file as specified in the zim metadata.
-   */
-  string getOrigId() const;
-
-  /**
-   * Get the favicon of the zim file.
-   *
-   * @param[out] content The content of the favicon.
-   * @param[out] mimeType The mimeType of the favicon.
-   * @return True if a favicon has been found.
-   */
-  bool getFavicon(string& content, string& mimeType) const;
-
-  /**
-   * Get an entry associated to an path.
-   *
-   * @param path The path of the entry.
-   * @return The entry.
-   * @throw NoEntry If no entry correspond to the path.
-   */
-  Entry getEntryFromPath(const std::string& path) const;
-
-  /**
-   * Get an entry associated to an url encoded path.
-   *
-   * Equivalent to `getEntryFromPath(urlDecode(path));`
-   *
-   * @param path The url encoded path.
-   * @return The entry.
-   * @throw NoEntry If no entry correspond to the path.
-   */
-  Entry getEntryFromEncodedPath(const std::string& path) const;
-
-  /**
-   * Get un entry associated to a title.
-   *
-   * @param title The title.
-   * @return The entry
-   * throw NoEntry If no entry correspond to the url.
-   */
-  Entry getEntryFromTitle(const std::string& title) const;
-
-  /**
-   * Get the url of a page specified by a title.
-   *
-   * @param[in] title the title of the page.
-   * @param[out] url the url of the page.
-   * @return True if the page can be found.
-   */
-  DEPRECATED bool getPageUrlFromTitle(const string& title, string& url) const;
-
-  /**
-   * Get the mimetype of a entry specified by a url.
-   *
-   * @param[in] url the url of the entry.
-   * @param[out] mimeType the mimeType of the entry.
-   * @return True if the mimeType has been found.
-   */
-  DEPRECATED bool getMimeTypeByUrl(const string& url, string& mimeType) const;
-
-  /**
-   * Get the content of an entry specifed by a url.
-   *
-   * Alias to `getContentByEncodedUrl`
-   */
-  DEPRECATED bool getContentByUrl(const string& url,
-                       string& content,
-                       string& title,
-                       unsigned int& contentLength,
-                       string& contentType) const;
-
-  /**
-   * Get the content of an entry specified by a url encoded url.
-   *
-   * Equivalent to getContentByDecodedUrl(urlDecode(url), ...).
-   */
-  DEPRECATED bool getContentByEncodedUrl(const string& url,
-                              string& content,
-                              string& title,
-                              unsigned int& contentLength,
-                              string& contentType,
-                              string& baseUrl) const;
-
-  /**
-   * Get the content of an entry specified by an url encoded url.
-   *
-   * Equivalent to getContentByEncodedUrl but without baseUrl.
-   */
-  DEPRECATED bool getContentByEncodedUrl(const string& url,
-                              string& content,
-                              string& title,
-                              unsigned int& contentLength,
-                              string& contentType) const;
-
-  /**
-   * Get the content of an entry specified by a url.
-   *
-   * @param[in] url The url of the entry.
-   * @param[out] content The content of the entry.
-   * @param[out] title the title of the entry.
-   * @param[out] contentLength The size of the entry (size of content).
-   * @param[out] contentType The mimeType of the entry.
-   * @param[out] baseUrl Return the true url of the entry.
-   *                     If the specified entry is a redirection, contains
-   *                     the url of the targeted entry.
-   * @return True if the entry has been found.
-   */
-  DEPRECATED bool getContentByDecodedUrl(const string& url,
-                              string& content,
-                              string& title,
-                              unsigned int& contentLength,
-                              string& contentType,
-                              string& baseUrl) const;
-  /**
-   * Get the content of an entry specified by a url.
-   *
-   * Equivalent to getContentByDecodedUrl but withou the baseUrl.
-   */
-  DEPRECATED bool getContentByDecodedUrl(const string& url,
-                              string& content,
-                              string& title,
-                              unsigned int& contentLength,
-                              string& contentType) const;
-
-  /**
-   * Search for entries with title starting with prefix (case sensitive).
-   *
-   * Suggestions are stored in an internal vector and can be retrieved using
-   * `getNextSuggestion` method.
-   *
-   * @param prefix The prefix to search.
-   * @param suggestionsCount How many suggestions to search for.
-   * @param reset If true, remove previous suggestions in the internal vector.
-   *              If false, add suggestions to the internal vector
-   *              (until internal vector size is suggestionCount (or no more
-   *               suggestion))
-   * @return True if some suggestions where added to the internal vector.
-   */
-  bool searchSuggestions(const string& prefix,
-                         unsigned int suggestionsCount,
-                         const bool reset = true);
-
-  /**
-   * Search for entries for the given prefix.
-   *
-   * If the zim file has a internal fulltext index, the suggestions will be
-   * searched using it.
-   * Else the suggestions will be search using `searchSuggestions` while trying
-   * to be smart about case sensitivity (using `getTitleVariants`).
-   *
-   * In any case, suggestions are stored in an internal vector and can be
-   * retrieved using `getNextSuggestion` method.
-   * The internal vector will be reset.
-   *
-   * @param prefix The prefix to search for.
-   * @param suggestionsCount How many suggestions to search for.
-   */
-  bool searchSuggestionsSmart(const string& prefix,
-                              unsigned int suggestionsCount);
-
-  /**
-   * Check if the url exists in the zim file.
-   *
-   * Deprecated : Use `pathExists` instead.
-   *
-   * @param url the url to check.
-   * @return True if the url exits in the zim file.
-   */
-  DEPRECATED bool urlExists(const string& url) const;
-
-  /**
-   * Check if the path exists in the zim file.
-   *
-   * @param path the path to check.
-   * @return True if the path exists in the zim file.
-   */
-  bool pathExists(const string& path) const;
-
-  /**
-   * Check if the zim file has a embedded fulltext index.
-   *
-   * @return True if the zim file has a embedded fulltext index
-   *         and is not split (else the fulltext is not accessible).
-   */
-  bool hasFulltextIndex() const;
-
-  /**
-   * Get potential case title variations for a title.
-   *
-   * @param title a title.
-   * @return the list of variantions.
-   */
-  std::vector<std::string> getTitleVariants(const std::string& title) const;
-
-  /**
-   * Get the next suggestion title.
-   *
-   * @param[out] title the title of the suggestion.
-   * @return True if title has been set.
-   */
-  bool getNextSuggestion(string& title);
-
-  /**
-   * Get the next suggestion title and url.
-   *
-   * @param[out] title the title of the suggestion.
-   * @param[out] url the url of the suggestion.
-   * @return True if title and url have been set.
-   */
-  bool getNextSuggestion(string& title, string& url);
-
-  /**
-   * Get if we can check zim file integrity (has a checksum).
-   *
-   * @return True if zim file have a checksum.
-   */
-  bool canCheckIntegrity() const;
-
-  /**
-   * Check is zim file is corrupted.
-   *
-   * @return True if zim file is corrupted.
-   */
-  bool isCorrupted() const;
-
-  /**
-   * Parse a full url into a namespace and url.
-   *
-   * @param[in] url The full url ("/N/url").
-   * @param[out] ns The namespace (N).
-   * @param[out] title The url (url).
-   * @return True
-   */
-  DEPRECATED bool parseUrl(const string& url, char* ns, string& title) const;
-
-  /**
-   * Return the total size of the zim file.
-   *
-   * If zim file is split, return the sum of all parts' size.
-   *
-   * @return Size of the size file is KiB.
-   */
-  unsigned int getFileSize() const;
-
-  /**
-   * Get the zim file handler.
-   *
-   * @return The libzim file handler.
-   */
-  zim::File* getZimFileHandler() const;
-
-  /**
-   * Get the zim article object associated to a url.
-   *
-   * @param[in] url The url of the article.
-   * @param[out] article The libzim article object.
-   * @return True if the url is good (article.good()).
-   */
-  DEPRECATED bool getArticleObjectByDecodedUrl(const string& url,
-                                    zim::Article& article) const;
-
- protected:
-  zim::File* zimFileHandler;
-  zim::size_type firstArticleOffset;
-  zim::size_type lastArticleOffset;
-  zim::size_type nsACount;
-  zim::size_type nsICount;
-  std::string zimFilePath;
-
-  std::vector<std::vector<std::string>> suggestions;
-  std::vector<std::vector<std::string>>::iterator suggestionsOffset;
-
- private:
-  std::map<const std::string, unsigned int> parseCounterMetadata() const;
-};
-}
-
-#endif

include/search_renderer.h

@@ -21,11 +21,12 @@
 #define KIWIX_SEARCH_RENDERER_H
 
 #include <string>
+#include <zim/search.h>
+#include "library.h"
 
 namespace kiwix
 {
 
-class Searcher;
 class NameMapper;
 /**
  * The SearcherRenderer class is used to render a search result to a html page.
@@ -34,21 +35,43 @@ class SearchRenderer
 {
  public:
   /**
-   * The default constructor.
+   * Construct a SearchRenderer from a SearchResultSet.
    *
-   * @param humanReadableName The global zim's humanReadableName.
-   *                          Used to generate pagination links.
+   * The constructed version of the SearchRenderer will not introduce
+   * the book name for each result. It is better to use the other constructor
+   * with a Library pointer to have a better html page.
+   *
+   * @param srs The `SearchResultSet` to render.
+   * @param mapper The `NameMapper` to use to do the rendering.
+   * @param start The start offset used for the srs.
+   * @param estimatedResultCount The estimatedResultCount of the whole search
    */
-  SearchRenderer(Searcher* searcher, NameMapper* mapper);
+  SearchRenderer(zim::SearchResultSet srs, NameMapper* mapper,
+                 unsigned int start, unsigned int estimatedResultCount);
+
+  /**
+   * Construct a SearchRenderer from a SearchResultSet.
+   *
+   * @param srs The `SearchResultSet` to render.
+   * @param mapper The `NameMapper` to use to do the rendering.
+   * @param library The `Library` to use to look up book details for search results.
+   * @param start The start offset used for the srs.
+   * @param estimatedResultCount The estimatedResultCount of the whole search
+   */
+  SearchRenderer(zim::SearchResultSet srs, NameMapper* mapper, Library* library,
+                 unsigned int start, unsigned int estimatedResultCount);
 
   ~SearchRenderer();
 
+  /**
+   * Set the search pattern used to do the search
+   */
   void setSearchPattern(const std::string& pattern);
 
   /**
-   * Set the search content id.
+   * Set the querystring used to select books
    */
-  void setSearchContent(const std::string& name);
+  void setSearchBookQuery(const std::string& bookQuery);
 
   /**
    * Set protocol prefix.
@@ -60,23 +83,38 @@ class SearchRenderer
    */
   void setSearchProtocolPrefix(const std::string& prefix);
 
+  /**
+   * set result count per page
+   */
+  void setPageLength(unsigned int pageLength){
+    this->pageLength  = pageLength;
+  }
+
+  std::string renderTemplate(const std::string& tmpl_str);
+
   /**
    * Generate the html page with the resutls of the search.
    */
   std::string getHtml();
 
+    /**
+   * Generate the xml page with the resutls of the search.
+   */
+  std::string getXml();
+
+
  protected:
   std::string beautifyInteger(const unsigned int number);
-  Searcher* mp_searcher;
+  zim::SearchResultSet m_srs;
   NameMapper* mp_nameMapper;
-  std::string searchContent;
+  Library* mp_library;
+  std::string searchBookQuery;
   std::string searchPattern;
   std::string protocolPrefix;
   std::string searchProtocolPrefix;
-  unsigned int resultCountPerPage;
+  unsigned int pageLength;
   unsigned int estimatedResultCount;
   unsigned int resultStart;
-  unsigned int resultEnd;
 };
 
 

include/searcher.h

@@ -1,172 +0,0 @@
-/*
- * Copyright 2011 Emmanuel Engelhart <kelson@kiwix.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#ifndef KIWIX_SEARCHER_H
-#define KIWIX_SEARCHER_H
-
-#include <stdio.h>
-#include <stdlib.h>
-#include <unicode/putil.h>
-#include <algorithm>
-#include <cctype>
-#include <locale>
-#include <string>
-#include <vector>
-#include <vector>
-#include "tools/pathTools.h"
-#include "tools/stringTools.h"
-#include "kiwix_config.h"
-
-using namespace std;
-
-namespace kiwix
-{
-class Reader;
-class Result
-{
- public:
-  virtual ~Result(){};
-  virtual std::string get_url() = 0;
-  virtual std::string get_title() = 0;
-  virtual int get_score() = 0;
-  virtual std::string get_snippet() = 0;
-  virtual std::string get_content() = 0;
-  virtual int get_wordCount() = 0;
-  virtual int get_size() = 0;
-  virtual int get_readerIndex() = 0;
-};
-
-struct SearcherInternal;
-/**
- * The Searcher class is reponsible to do different kind of search using the
- * fulltext index.
- */
-class Searcher
-{
- public:
-  /**
-   * The default constructor.
-   */
-  Searcher();
-
-  ~Searcher();
-
-  /**
-   * Add a reader (containing embedded fulltext index) to the search.
-   *
-   * @param reader The Reader for the zim containing the fulltext index.
-   * @return true if the reader has been added.
-   *         false if the reader cannot be added (no embedded fulltext index present)
-   */
-  bool add_reader(Reader* reader);
-
-
-  Reader* get_reader(int index);
-
-  /**
-   * Start a search on the zim associated to the Searcher.
-   *
-   * Search results should be retrived using the getNextResult method.
-   *
-   * @param search The search query.
-   * @param resultStart the start offset of the search results (used for pagination).
-   * @param resultEnd the end offset of the search results (used for pagination).
-   * @param verbose print some info on stdout if true.
-   */
-  void search(const std::string& search,
-              unsigned int resultStart,
-              unsigned int resultEnd,
-              const bool verbose = false);
-
-  /**
-   * Start a geographique search.
-   * The search return result for entry in a disc of center latitude/longitude
-   * and radius distance.
-   *
-   * Search results should be retrived using the getNextResult method.
-   *
-   * @param latitude The latitude of the center point.
-   * @param longitude The longitude of the center point.
-   * @param distance The radius of the disc.
-   * @param resultStart the start offset of the search results (used for pagination).
-   * @param resultEnd the end offset of the search results (used for pagination).
-   * @param verbose print some info on stdout if true.
-   */
-  void geo_search(float latitude, float longitude, float distance,
-                  unsigned int resultStart,
-                  unsigned int resultEnd,
-                  const bool verbose = false);
-
-  /**
-   * Start a suggestion search.
-   * The search made depend of the "version" of the embedded index.
-   *  - If the index is newer enough and have a title namespace, the search is
-   *    made in the titles only.
-   *  - Else the search is made on the whole article content.
-   * In any case, the search is made "partial" (as adding '*' at the end of the query)
-   *
-   * @param search The search query.
-   * @param verbose print some info on stdout if true.
-   */
-  void suggestions(std::string& search, const bool verbose = false);
-
-  /**
-   * Get the next result of a started search.
-   * This is the method to use to loop hover the search results.
-   */
-  Result* getNextResult();
-
-  /**
-   * Restart the previous search.
-   * Next call to getNextResult will return the first result.
-   */
-  void restart_search();
-
-  /**
-   * Get a estimation of the result count.
-   */
-  unsigned int getEstimatedResultCount();
-
-  unsigned int getResultStart() { return resultStart; }
-  unsigned int getResultEnd() { return resultEnd; }
-
- protected:
-  std::string beautifyInteger(const unsigned int number);
-  void closeIndex();
-  void searchInIndex(string& search,
-                     const unsigned int resultStart,
-                     const unsigned int resultEnd,
-                     const bool verbose = false);
-
-  std::vector<Reader*> readers;
-  SearcherInternal* internal;
-  std::string searchPattern;
-  unsigned int estimatedResultCount;
-  unsigned int resultStart;
-  unsigned int resultEnd;
-
- private:
-  void reset();
-
-};
-
-
-}
-
-#endif

include/server.h

@@ -54,23 +54,31 @@ namespace kiwix
        void setAddress(const std::string& addr) { m_addr = addr; }
        void setPort(int port) { m_port = port; }
        void setNbThreads(int threads) { m_nbThreads = threads; }
+       void setMultiZimSearchLimit(unsigned int limit) { m_multizimSearchLimit = limit; }
+       void setIpConnectionLimit(int limit) { m_ipConnectionLimit = limit; }
        void setVerbose(bool verbose) { m_verbose = verbose; }
+       void setIndexTemplateString(const std::string& indexTemplateString) { m_indexTemplateString = indexTemplateString; }
        void setTaskbar(bool withTaskbar, bool withLibraryButton)
         { m_withTaskbar = withTaskbar; m_withLibraryButton = withLibraryButton; }
        void setBlockExternalLinks(bool blockExternalLinks)
         { m_blockExternalLinks = blockExternalLinks; }
+       int getPort();
+       std::string getAddress();
 
      protected:
        Library* mp_library;
        NameMapper* mp_nameMapper;
        std::string m_root = "";
        std::string m_addr = "";
+       std::string m_indexTemplateString = "";
        int m_port = 80;
        int m_nbThreads = 1;
+       unsigned int m_multizimSearchLimit = 0;
        bool m_verbose = false;
        bool m_withTaskbar = true;
        bool m_withLibraryButton = true;
        bool m_blockExternalLinks = false;
+       int m_ipConnectionLimit = 0;
        std::unique_ptr<InternalServer> mp_server;
   };
 }

include/tools.h

@@ -0,0 +1,220 @@
+/*
+ * Copyright 2021 Matthieu Gautier <mgautier@kymeria.fr>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#ifndef KIWIX_TOOLS_H
+#define KIWIX_TOOLS_H
+
+#include <string>
+#include <vector>
+#include <map>
+
+namespace kiwix {
+
+/**
+ * Return the current directory.
+ *
+ * @return the current directory (utf8 encoded)
+ */
+std::string getCurrentDirectory();
+
+/**
+ * Return the data directory.
+ *
+ * The data directory is a directory where to put data (zim files, ...)
+ * It depends of the platform and it may be changed by user using environment variable.
+ *
+ * The resolution order is :
+ * - `KIWIX_DATA_DIR` env variable (if set).
+ * - On Windows :
+ *    . `$APPDATA/kiwix` if $APPDATA is set
+ *    . `$USERPROFILE/kiwix` if $USERPROFILE is set
+ * - Else :
+ *    . `$XDG_DATA_HOME/kiwix`if $XDG_DATA_HOME is set
+ *    . `$HOME/.local/share/kiwx` if $HOWE is set
+ * - current directory
+ *
+ * @return the path of the data directory (utf8 encoded)
+ */
+std::string getDataDirectory();
+
+/** Return the path of the executable
+ *
+ * Some application may be packaged in auto extractible archive (Appimage) and the
+ * real executable is different of the path of the archive.
+ * If `realPathOnly` is true, return the path of the real executable instead of the
+ * archive launched by the user.
+ *
+ * @param realPathOnly If we must return the real path of the executable.
+ * @return the path of the executable (utf8 encoded)
+ */
+std::string getExecutablePath(bool realPathOnly = false);
+
+/** Tell if the path is a relative path.
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate paths.
+ *
+ * @param path A utf8 encoded path.
+ * @return true if the path is relative.
+ */
+bool isRelativePath(const std::string& path);
+
+/** Append a path to another one.
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate paths.
+ *
+ * @param basePath the base path.
+ * @param relativePath a path to add to the base path, must be a relative path.
+ * @return The concatenation of the paths, using the right separator.
+ */
+std::string appendToDirectory(const std::string& basePath, const std::string& relativePath);
+
+/** Remove the last element of a path.
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate paths.
+ *
+ * @param path a path.
+ * @return The parent directory (or empty string if none).
+ */
+std::string removeLastPathElement(const std::string& path);
+
+/** Get the last element of a path.
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate paths.
+ *
+ * @param path a path.
+ * @return The base name of the path or empty string if none (ending with a separator).
+ */
+std::string getLastPathElement(const std::string& path);
+
+/** Compute the absolute path of a relative path based on another one
+ *
+ * Equivalent to appendToDirectory followed by a normalization of the path.
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate paths.
+ *
+ * @param path the base path (if empty, current directory is taken).
+ * @param relativePath the relative path.
+ * @return a absolute path.
+ */
+std::string computeAbsolutePath(const std::string& path, const std::string& relativePath);
+
+/** Compute the relative path of a path relative to another one
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate paths.
+ *
+ * @param path the base path.
+ * @param absolutePath the absolute path to find the relative path for.
+ * @return a relative path (pointing to absolutePath, relative to path).
+ */
+std::string computeRelativePath(const std::string& path, const std::string& absolutePath);
+
+/** Sleep the current thread.
+ *
+ * This function is provided as a small helper. It is probably better to use native tools.
+ *
+ * @param milliseconds The number of milliseconds to wait for.
+ */
+void sleep(unsigned int milliseconds);
+
+/** Split a string
+ *
+ * This function is provided as a small helper. It is probably better to use native tools.
+ *
+ * Assuming text = "foo:;bar;baz,oups;"
+ *
+ * split(text, ":;", true, true) => ["foo", ":", ";", "bar", ";", "baz,oups", ";"]
+ * split(text, ":;", true, false) => ["foo", "bar", "baz,oups"] (default)
+ * split(text, ":;", false, true) => ["foo", ":", "", ";", "bar", ";", "baz,oups", ";", ""]
+ * split(text, ":;", false, false) => ["foo", "", "bar", "baz,oups", ""]
+ *
+ * @param str The string to split.
+ * @param delims A string of potential delimiters.
+ *               Each charater in the string can be a individual delimiters.
+ * @param dropEmpty true if empty part must be dropped from the result.
+ * @param keepDelim true if delimiter must be included from the result.
+ * @return a list of part (potentially containing delimiters)
+ */
+std::vector<std::string> split(const std::string& str, const std::string& delims, bool dropEmpty=true, bool keepDelim = false);
+
+/** Convert language code from iso2 code to iso3
+ *
+ * This function is provided as a small helper. It is probably better to use native tools
+ * to manipulate locales.
+ *
+ * @param a2code a iso2 code string.
+ * @return the corresponding iso3 code.
+ * @throw std::out_of_range if iso2 code is not known.
+ */
+std::string converta2toa3(const std::string& a2code);
+
+/** Extracts content from given file.
+ *
+ * This function provides content of a file provided it's path.
+ *
+ * @param path The absolute path provided in string format.
+ * @return Content of corresponding file in string format.
+ */
+std::string getFileContent(const std::string& path);
+
+/** Checks if file exists.
+ *
+ * This function returns boolean stating if file exists.
+ *
+ * @param path The absolute path provided in string format.
+ * @return Boolean representing if file exists or not.
+ */
+bool fileExists(const std::string& path);
+
+/** Checks if file is readable.
+ *
+ * This function returns boolean stating if file is readable.
+ *
+ * @param path The absolute path provided in string format.
+ * @return Boolean representing if file is readale or not.
+ */
+bool fileReadable(const std::string& path);
+
+/** Provides mimetype from filename.
+ *
+ * This function provides mimetype from file-name.
+ *
+ * @param filename string containing filename.
+ * @return mimetype from filename in string format.
+ */
+std::string getMimeTypeForFile(const std::string& filename);
+
+/** Provides all available network interfaces
+ *
+ * This function provides the available IPv4 network interfaces
+ */
+std::map<std::string, std::string> getNetworkInterfaces();
+
+/** Provides the best IP address
+ * This function provides the best IP address from the list given by getNetworkInterfaces
+ */
+std::string getBestPublicIp();
+
+}
+#endif // KIWIX_TOOLS_H

include/tools/lock.h

@@ -1,46 +0,0 @@
-
-
-#ifndef KIWIXLIB_TOOL_LOCK_H
-#define KIWIXLIB_TOOL_LOCK_H
-
-#include <pthread.h>
-
-namespace kiwix {
-
-class Lock
-{
-  public:
-    explicit Lock(pthread_mutex_t* mutex) :
-      mp_mutex(mutex)
-    {
-      pthread_mutex_lock(mp_mutex);
-    }
-    ~Lock() {
-      if (mp_mutex != nullptr) {
-        pthread_mutex_unlock(mp_mutex);
-      }
-    }
-    Lock(Lock && other) :
-      mp_mutex(other.mp_mutex)
-    {
-      other.mp_mutex = nullptr;
-    }
-    Lock & operator=(Lock && other)
-    {
-      mp_mutex = other.mp_mutex;
-      other.mp_mutex = nullptr;
-      return *this;
-    }
-
-
-  private:
-    pthread_mutex_t* mp_mutex;
-
-    Lock(Lock const &) = delete;
-    Lock & operator=(Lock const &) = delete;
-};
-
-
-}
-
-#endif //KIWIXLIB_TOOL_LOCK_H

include/version.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2013 Emmanuel Engelhart <kelson@kiwix.org>
+ * Copyright 2021 Emmanuel Engelhart <kelson@kiwix.org>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU  General Public License as published by
@@ -17,21 +17,18 @@
  * MA 02110-1301, USA.
  */
 
-package org.kiwix.kiwixlib;
+#ifndef KIWIX_VERSION_H
+#define KIWIX_VERSION_H
 
-public class JNIKiwixString
+#include <string>
+#include <vector>
+#include <iostream>
+
+namespace kiwix
 {
-  public String value;
-
-  public JNIKiwixString(String value) {
-    this.value = value;
-  }
-
-  public JNIKiwixString() {
-    this("");
-  }
-
-  public String getValue() {
-    return value;
-  }
+  typedef std::vector<std::pair<std::string, std::string>> LibVersions;
+  LibVersions getVersions();
+  void printVersions(std::ostream& out = std::cout);
 }
+
+#endif // KIWIX_VERSION_H

meson.build

@@ -1,46 +1,62 @@
-project('kiwix-lib', 'cpp',
-  version : '9.2', # Also change this in android-kiwix-lib-publisher/kiwixLibAndroid/build.gradle
-  license : 'GPL',
+project('libkiwix', 'cpp',
+  version : '11.0.0',
+  license : 'GPLv3+',
   default_options : ['c_std=c11', 'cpp_std=c++11', 'werror=true'])
 
 compiler = meson.get_compiler('cpp')
 
-wrapper = get_option('wrapper')
+static_deps = get_option('static-linkage') or get_option('default_library') == 'static'
 
-static_deps = 'android' in wrapper or 'java' in wrapper or get_option('default_library') == 'static'
-if 'android' in wrapper
-  extra_libs = ['-llog']
+# See https://github.com/kiwix/libkiwix/issues/371
+if ['arm', 'mips', 'm68k', 'ppc', 'sh4'].contains(host_machine.cpu_family())
+  extra_libs = ['-latomic']
 else
   extra_libs = []
 endif
 
-if 'java' in wrapper
-  add_languages('java')
+if (compiler.get_id() == 'gcc' and build_machine.system() == 'linux') or host_machine.system() == 'freebsd'
+  # C++ std::thread is implemented using pthread on linux by gcc
+  thread_dep = dependency('threads')
+else
+  thread_dep = dependency('', required:false)
 endif
-
-thread_dep = dependency('threads')
 libicu_dep = dependency('icu-i18n', static:static_deps)
-libzim_dep = dependency('libzim', version : '>=6.1.1', static:static_deps)
 pugixml_dep = dependency('pugixml', static:static_deps)
 libcurl_dep = dependency('libcurl', static:static_deps)
 microhttpd_dep = dependency('libmicrohttpd', static:static_deps)
+zlib_dep = dependency('zlib', static:static_deps)
+xapian_dep = dependency('xapian-core', static:static_deps)
 
-if not compiler.has_header('mustache.hpp')
+if compiler.has_header('mustache.hpp')
+  extra_include = []
+elif compiler.has_header('mustache.hpp', args: '-I/usr/include/kainjow')
+  extra_include = ['/usr/include/kainjow']
+else
   error('Cannot found header mustache.hpp')
 endif
 
+libzim_dep = dependency('libzim', version : '>=7.2.0', static:static_deps)
+if not compiler.has_header_symbol('zim/zim.h', 'LIBZIM_WITH_XAPIAN')
+  error('Libzim seems to be compiled without xapian. Xapian support is mandatory.')
+endif
+
+
 extra_cflags = ''
-if target_machine.system() == 'windows' and static_deps
+if host_machine.system() == 'windows' and static_deps
   add_project_arguments('-DCURL_STATICLIB', language : 'cpp')
   extra_cflags += '-DCURL_STATICLIB'
 endif
 
-all_deps = [thread_dep, libicu_dep, libzim_dep, pugixml_dep, libcurl_dep, microhttpd_dep]
+if host_machine.system() == 'windows'
+  add_project_arguments('-DNOMINMAX', language: 'cpp')
+endif
 
-inc = include_directories('include')
+all_deps = [thread_dep, libicu_dep, libzim_dep, pugixml_dep, libcurl_dep, microhttpd_dep, zlib_dep, xapian_dep]
+
+inc = include_directories('include', extra_include)
 
 conf = configuration_data()
-conf.set('VERSION', '"@0@"'.format(meson.project_version()))
+conf.set('LIBKIWIX_VERSION', '"@0@"'.format(meson.project_version()))
 
 if build_machine.system() == 'windows'
      extra_link_args = ['-lshlwapi', '-lwinmm']
@@ -53,8 +69,11 @@ subdir('scripts')
 subdir('static')
 subdir('src')
 subdir('test')
+if get_option('doc')
+  subdir('docs')
+endif
 
-pkg_requires = ['libzim', 'icu-i18n', 'pugixml', 'libcurl']
+pkg_requires = ['libzim', 'icu-i18n', 'pugixml', 'libcurl', 'libmicrohttpd', 'xapian-core']
 
 pkg_conf = configuration_data()
 pkg_conf.set('prefix', get_option('prefix'))

meson_options.txt

@@ -1,2 +1,4 @@
-option('wrapper', type:'array', choices:['java', 'android'], value:[],
-  description: 'The wrapper to generate.')
+option('static-linkage', type : 'boolean', value : false,
+  description : 'Link statically with the dependencies.')
+option('doc', type : 'boolean', value : false,
+  description : 'Build the documentations.')

scripts/kiwix-compile-i18n

@@ -0,0 +1,148 @@
+#!/usr/bin/env python3
+
+'''
+Copyright 2022 Veloman Yunkan <veloman.yunkan@gmail.com>
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 3 of the License, or any
+later version.
+
+This program is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+02110-1301, USA.
+'''
+
+import argparse
+import os.path
+import re
+import json
+
+def to_identifier(name):
+    ident = re.sub(r'[^0-9a-zA-Z]', '_', name)
+    if ident[0].isnumeric():
+        return "_"+ident
+    return ident
+
+def lang_code(filename):
+    filename = os.path.basename(filename)
+    lang = to_identifier(os.path.splitext(filename)[0])
+    print(filename, '->', lang)
+    return lang
+
+from string import Template
+
+def expand_cxx_template(t, **kwargs):
+    return Template(t).substitute(**kwargs)
+
+def cxx_string_literal(s):
+    # Taking advantage of the fact the JSON string escape rules match
+    # those of C++
+    return 'u8' + json.dumps(s)
+
+string_table_cxx_template = '''
+const I18nString $TABLE_NAME[] = {
+  $TABLE_ENTRIES
+};
+'''
+
+lang_table_entry_cxx_template = '''
+    {
+        $LANG_STRING_LITERAL,
+        ARRAY_ELEMENT_COUNT($STRING_TABLE_NAME),
+        $STRING_TABLE_NAME
+    }'''
+
+cxxfile_template = '''// This file is automatically generated. Do not modify it.
+
+#include "server/i18n.h"
+
+namespace kiwix {
+namespace i18n {
+
+namespace
+{
+
+$STRING_DATA
+
+} // unnamed namespace
+
+#define ARRAY_ELEMENT_COUNT(a) (sizeof(a)/sizeof(a[0]))
+
+extern const I18nStringTable stringTables[] = {
+  $LANG_TABLE
+};
+
+extern const size_t langCount = $LANG_COUNT;
+
+} // namespace i18n
+} // namespace kiwix
+'''
+
+class Resource:
+    def __init__(self, filename):
+        filename = filename.strip()
+        self.filename = filename
+        self.lang_code = lang_code(filename)
+        with open(filename, 'r', encoding='utf-8') as f:
+            self.data = f.read()
+
+    def get_string_table_name(self):
+        return "string_table_for_" + self.lang_code
+
+    def get_string_table(self):
+        table_entries = ",\n  ".join(self.get_string_table_entries())
+        return expand_cxx_template(string_table_cxx_template,
+                    TABLE_NAME=self.get_string_table_name(),
+                    TABLE_ENTRIES=table_entries)
+
+    def get_string_table_entries(self):
+        d = json.loads(self.data)
+        for k in sorted(d.keys()):
+            if k != "@metadata":
+                key_string = cxx_string_literal(k)
+                value_string = cxx_string_literal(d[k])
+                yield '{ ' + key_string + ', ' + value_string + ' }'
+
+    def get_lang_table_entry(self):
+        return expand_cxx_template(lang_table_entry_cxx_template,
+                LANG_STRING_LITERAL=cxx_string_literal(self.lang_code),
+                STRING_TABLE_NAME=self.get_string_table_name())
+
+
+
+def gen_c_file(resources):
+    string_data = []
+    lang_table = []
+    for r in resources:
+        string_data.append(r.get_string_table())
+        lang_table.append(r.get_lang_table_entry())
+
+    return expand_cxx_template(cxxfile_template,
+                STRING_DATA="\n".join(string_data),
+                LANG_TABLE=",\n  ".join(lang_table),
+                LANG_COUNT=len(resources)
+           )
+
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--cxxfile',
+                        required=True,
+                        help='The Cpp file name to generate')
+    parser.add_argument('i18n_resource_files', nargs='+',
+                        help='The list of resources to compile.')
+    args = parser.parse_args()
+
+    resources = [Resource(filename) for filename in args.i18n_resource_files]
+
+    with open(args.cxxfile, 'w') as f:
+        f.write(gen_c_file(resources))
+

scripts/kiwix-compile-i18n.1

@@ -0,0 +1,18 @@
+.TH KIWIX-COMPILE-I18N "1" "January 2022" "Kiwix" "User Commands"
+.SH NAME
+kiwix-compile-i18n \- helper to compile Kiwix i18n (internationalization) data
+.SH SYNOPSIS
+\fBkiwix\-compile\-i18n\fR [\-h] \-\-cxxfile CXXFILE i18n_resource_files ...\fR
+.SH DESCRIPTION
+.TP
+i18n_resource_files ...
+The list of i18n resources to compile.
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+show a help message and exit
+.TP
+\fB\-\-cxxfile\fR CXXFILE
+The Cpp file name to generate
+.TP
+.SH AUTHOR
+Veloman Yunkan <veloman.yunkan@gmail.com>

scripts/kiwix-compile-resources

@@ -102,7 +102,7 @@ class Resource:
 
 
 
-master_c_template = """//This file is automaically generated. Do not modify it.
+master_c_template = """//This file is automatically generated. Do not modify it.
 
 #include <stdlib.h>
 #include <fstream>

scripts/kiwix-compile-resources.1

@@ -0,0 +1,20 @@
+.TH KIWIX-COMPILE-RESOURCES "1" "August 2017" "Kiwix" "User Commands"
+.SH NAME
+kiwix-compile-resources \- helper to compile and generate some Kiwix resources
+.SH SYNOPSIS
+\fBkiwix\-compile\-resources\fR [\-h] [\-\-cxxfile CXXFILE] [\-\-hfile HFILE] resource_file\fR
+.SH DESCRIPTION
+.TP
+resource_file
+The list of resources to compile.
+.TP
+\fB\-h\fR, \fB\-\-help\fR
+show a help message and exit
+.TP
+\fB\-\-cxxfile\fR CXXFILE
+The Cpp file name to generate
+.TP
+\fB\-\-hfile\fR HFILE
+The h file name to generate
+.SH AUTHOR
+Matthieu Gautier <mgautier@kymeria.fr>

scripts/kiwix-resources

@@ -0,0 +1,127 @@
+#!/usr/bin/env python3
+
+'''
+Copyright 2022 Veloman Yunkan <veloman.yunkan@gmail.com>
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 3 of the License, or any
+later version.
+
+This program is distributed in the hope that it will be useful, but
+WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
+General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
+02110-1301, USA.
+'''
+
+import argparse
+import hashlib
+import os.path
+import re
+
+def read_resource_file(resource_file_path):
+    with open(resource_file_path, 'r') as f:
+        return [line.strip() for line in f]
+
+def list_resources(resource_file_path):
+    for resource_path in read_resource_file(resource_file_path):
+        print(resource_path)
+
+def compute_resource_revision(resource_path):
+    with open(os.path.join(OUT_DIR, resource_path), 'rb') as f:
+        return hashlib.sha1(f.read()).hexdigest()[:8]
+
+resource_revisions = {}
+
+def get_resource_revision(res):
+    if not res in resource_revisions:
+        preprocess_resource(res)
+        resource_revisions[res] = compute_resource_revision(res)
+    return resource_revisions[res]
+
+RESOURCE_WITH_CACHEID_URL_PATTERN=r'(?P<pre>.*/(?P<resource>skin/[^"?]+)\?)KIWIXCACHEID(?P<post>[^"]*)'
+
+def set_cacheid(resource_matchobj):
+    pre = resource_matchobj.group('pre')
+    resource = resource_matchobj.group('resource')
+    post = resource_matchobj.group('post')
+    cacheid = 'cacheid=' + get_resource_revision(resource)
+    return pre + cacheid + post
+
+def preprocess_text(s):
+    if 'KIWIXCACHEID' in s:
+        s = re.sub(RESOURCE_WITH_CACHEID_URL_PATTERN, set_cacheid, s)
+        assert not 'KIWIXCACHEID' in s
+    return s
+
+def get_preprocessed_resource(srcpath):
+    """Get the transformed content of a resource
+
+    If the resource at srcpath is modified by preprocessing then this function
+    returns the transformed content of the resource. Otherwise it returns None.
+    """
+    try:
+        with open(srcpath, 'r') as resource_file:
+            content = resource_file.read()
+            preprocessed_content = preprocess_text(content)
+        return preprocessed_content if preprocessed_content != content else None
+    except UnicodeDecodeError:
+        # It was a binary resource
+        return None
+
+
+def symlink_resource(src, resource_path):
+    if os.path.exists(resource_path):
+        if os.path.islink(resource_path) and os.readlink(resource_path) == src:
+            return
+        os.remove(resource_path)
+    os.symlink(src, resource_path)
+
+def preprocess_resource(resource_path):
+    print('Preprocessing', resource_path, '...')
+    resource_dir = os.path.dirname(resource_path)
+    if resource_dir != '':
+        os.makedirs(os.path.join(OUT_DIR, resource_dir), exist_ok=True)
+    srcpath = os.path.join(BASE_DIR, resource_path)
+    outpath = os.path.join(OUT_DIR, resource_path)
+    if os.path.exists(outpath):
+        os.remove(outpath)
+    preprocessed_content = get_preprocessed_resource(srcpath)
+    if preprocessed_content is None:
+        symlink_resource(srcpath, outpath)
+    else:
+        with open(outpath, 'w') as target:
+            print(preprocessed_content, end='', file=target)
+
+
+def copy_file(src_path, dst_path):
+    with open(src_path, 'rb') as src:
+        with open(dst_path, 'wb') as dst:
+            dst.write(src.read())
+
+def preprocess_resources(resource_file_path):
+    resource_filename = os.path.basename(resource_file_path)
+    for resource in read_resource_file(resource_file_path):
+        preprocess_resource(resource)
+    copy_file(resource_file_path, os.path.join(OUT_DIR, resource_filename))
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    commands = parser.add_mutually_exclusive_group()
+    commands.add_argument('--list-all', action='store_true')
+    commands.add_argument('--preprocess', action='store_true')
+    parser.add_argument('--outdir')
+    parser.add_argument('resource_file')
+    args = parser.parse_args()
+    BASE_DIR = os.path.dirname(os.path.realpath(args.resource_file))
+    OUT_DIR = args.outdir
+
+    if args.list_all:
+        list_resources(args.resource_file)
+    elif args.preprocess:
+        preprocess_resources(args.resource_file)

scripts/meson.build

@@ -1,4 +1,13 @@
 
+res_manager = find_program('kiwix-resources')
 res_compiler = find_program('kiwix-compile-resources')
 
 install_data(res_compiler.path(), install_dir:get_option('bindir'))
+
+install_man('kiwix-compile-resources.1')
+
+i18n_compiler = find_program('kiwix-compile-i18n')
+
+install_data(i18n_compiler.path(), install_dir:get_option('bindir'))
+
+install_man('kiwix-compile-i18n.1')

src/aria2.cpp

@@ -3,13 +3,15 @@
 #include "aria2.h"
 #include "xmlrpc.h"
 #include <iostream>
+#include <algorithm>
 #include <sstream>
 #include <thread>
 #include <chrono>
-#include <tools/otherTools.h>
-#include <tools/pathTools.h>
-#include <tools/stringTools.h>
-#include <downloader.h> // For AriaError
+#include "tools.h"
+#include "tools/pathTools.h"
+#include "tools/stringTools.h"
+#include "tools/otherTools.h"
+#include "downloader.h" // For AriaError
 
 #ifdef _WIN32
 # define ARIA2_CMD "aria2c.exe"
@@ -19,15 +21,20 @@
 #endif
 
 
+#define LOG_ARIA_ERROR() \
+  { \
+    std::cerr << "ERROR: aria2 RPC request failed. (" << res << ")." << std::endl; \
+    std::cerr << (m_curlErrorBuffer[0] ? m_curlErrorBuffer.get() : curl_easy_strerror(res)) << std::endl; \
+  }
 
 namespace kiwix {
 
 Aria2::Aria2():
   mp_aria(nullptr),
   m_port(42042),
-  m_secret("kiwixariarpc"),
-  mp_curl(nullptr),
-  m_lock(PTHREAD_MUTEX_INITIALIZER)
+  m_secret(getNewRpcSecret()),
+  m_curlErrorBuffer(new char[CURL_ERROR_SIZE]),
+  mp_curl(nullptr)
 {
   m_downloadDir = getDataDirectory();
   makeDirectory(m_downloadDir);
@@ -58,11 +65,12 @@ Aria2::Aria2():
     // Try to use a potential installed aria2c.
     callCmd.push_back(ARIA2_CMD);
   }
+  callCmd.push_back("--follow-metalink=mem");
   callCmd.push_back("--enable-rpc");
   callCmd.push_back(rpc_secret.c_str());
   callCmd.push_back(rpc_port.c_str());
   callCmd.push_back(download_dir.c_str());
-  if (fileExists(session_file)) {
+  if (fileReadable(session_file)) {
     callCmd.push_back(inputFile.c_str());
   }
   callCmd.push_back(session.c_str());
@@ -84,27 +92,21 @@ Aria2::Aria2():
   }
   mp_aria = Subprocess::run(callCmd);
   mp_curl = curl_easy_init();
-  char errbuf[CURL_ERROR_SIZE];
+
   curl_easy_setopt(mp_curl, CURLOPT_URL, "http://localhost/rpc");
   curl_easy_setopt(mp_curl, CURLOPT_PORT, m_port);
   curl_easy_setopt(mp_curl, CURLOPT_POST, 1L);
-  curl_easy_setopt(mp_curl, CURLOPT_ERRORBUFFER, errbuf);
+  curl_easy_setopt(mp_curl, CURLOPT_ERRORBUFFER, m_curlErrorBuffer.get());
 
   int watchdog = 50;
   while(--watchdog) {
     sleep(10);
-    errbuf[0] = 0;
+    m_curlErrorBuffer[0] = 0;
     auto res = curl_easy_perform(mp_curl);
     if (res == CURLE_OK) {
       break;
     } else if (watchdog == 1) {
-      std::cerr <<" curl_easy_perform() failed." << std::endl;
-      fprintf(stderr, "\nlibcurl: (%d) ", res);
-      if (errbuf[0] != 0) {
-        std::cerr << errbuf << std::endl;
-      } else {
-        std::cerr << curl_easy_strerror(res) << std::endl;
-      }
+      LOG_ARIA_ERROR();
     }
   }
   if (!watchdog) {
@@ -115,6 +117,7 @@ Aria2::Aria2():
 
 Aria2::~Aria2()
 {
+  std::unique_lock<std::mutex> lock(m_lock);
   curl_easy_cleanup(mp_curl);
 }
 
@@ -126,38 +129,44 @@ void Aria2::close()
 
 size_t write_callback_to_iss(char* ptr, size_t size, size_t nmemb, void* userdata)
 {
-  auto str = static_cast<std::stringstream*>(userdata);
-  str->write(ptr, nmemb);
+  auto outStream = static_cast<std::stringstream*>(userdata);
+  outStream->write(ptr, nmemb);
   return nmemb;
 }
 
 std::string Aria2::doRequest(const MethodCall& methodCall)
 {
-  pthread_mutex_lock(&m_lock);
   auto requestContent = methodCall.toString();
-  std::stringstream stringstream;
+  std::stringstream outStream;
   CURLcode res;
-  curl_easy_setopt(mp_curl, CURLOPT_POSTFIELDSIZE, requestContent.size());
-  curl_easy_setopt(mp_curl, CURLOPT_POSTFIELDS, requestContent.c_str());
-  curl_easy_setopt(mp_curl, CURLOPT_WRITEFUNCTION, &write_callback_to_iss);
-  curl_easy_setopt(mp_curl, CURLOPT_WRITEDATA, &stringstream);
-  res = curl_easy_perform(mp_curl);
-  if (res == CURLE_OK) {
-    long response_code;
+  long response_code;
+  {
+    std::unique_lock<std::mutex> lock(m_lock);
+    curl_easy_setopt(mp_curl, CURLOPT_POSTFIELDSIZE, requestContent.size());
+    curl_easy_setopt(mp_curl, CURLOPT_POSTFIELDS, requestContent.c_str());
+    curl_easy_setopt(mp_curl, CURLOPT_WRITEFUNCTION, &write_callback_to_iss);
+    curl_easy_setopt(mp_curl, CURLOPT_WRITEDATA, &outStream);
+    m_curlErrorBuffer[0] = 0;
+    res = curl_easy_perform(mp_curl);
+    if (res != CURLE_OK) {
+      LOG_ARIA_ERROR();
+      throw std::runtime_error("Cannot perform request");
+    }
     curl_easy_getinfo(mp_curl, CURLINFO_RESPONSE_CODE, &response_code);
-    pthread_mutex_unlock(&m_lock);
-    if (response_code != 200) {
-      throw std::runtime_error("Invalid return code from aria");
-    }
-    auto responseContent = stringstream.str();
-    MethodResponse response(responseContent);
-    if (response.isFault()) {
-      throw AriaError(response.getFault().getFaultString());
-    }
-    return responseContent;
   }
-  pthread_mutex_unlock(&m_lock);
-  throw std::runtime_error("Cannot perform request");
+
+  auto responseContent = outStream.str();
+  if (response_code != 200) {
+    std::cerr << "ERROR: Invalid return code (" << response_code << ") from aria :" << std::endl;
+    std::cerr << responseContent << std::endl;
+    throw std::runtime_error("Invalid return code from aria");
+  }
+
+  MethodResponse response(responseContent);
+  if (response.isFault()) {
+    throw AriaError(response.getFault().getFaultString());
+  }
+  return responseContent;
 }
 
 std::string Aria2::addUri(const std::vector<std::string>& uris, const std::vector<std::pair<std::string, std::string>>& options)
@@ -188,6 +197,13 @@ std::string Aria2::tellStatus(const std::string& gid, const std::vector<std::str
   return doRequest(methodCall);
 }
 
+std::string Aria2::getNewRpcSecret()
+{
+  std::string uuid = gen_uuid("");
+  uuid.erase(std::remove(uuid.begin(), uuid.end(), '-'));
+  return uuid.substr(0, 9);
+}
+
 std::vector<std::string> Aria2::tellActive()
 {
   MethodCall methodCall("aria2.tellActive", m_secret);

src/aria2.h

@@ -12,8 +12,8 @@
 #include "xmlrpc.h"
 
 #include <memory>
+#include <mutex>
 #include <curl/curl.h>
-#include <pthread.h>
 
 namespace kiwix {
 
@@ -24,8 +24,9 @@ class Aria2
     int m_port;
     std::string m_secret;
     std::string m_downloadDir;
+    std::unique_ptr<char[]> m_curlErrorBuffer;
     CURL* mp_curl;
-    pthread_mutex_t m_lock;
+    std::mutex m_lock;
 
     std::string doRequest(const MethodCall& methodCall);
 
@@ -36,6 +37,7 @@ class Aria2
 
     std::string addUri(const std::vector<std::string>& uri, const std::vector<std::pair<std::string, std::string>>& options = {});
     std::string tellStatus(const std::string& gid, const std::vector<std::string>& statusKey);
+    static std::string getNewRpcSecret();
     std::vector<std::string> tellActive();
     std::vector<std::string> tellWaiting();
     void saveSession();

src/book.cpp

@@ -18,13 +18,18 @@
  */
 
 #include "book.h"
-#include "reader.h"
 
+#include "tools.h"
 #include "tools/base64.h"
 #include "tools/regexTools.h"
 #include "tools/networkTools.h"
 #include "tools/otherTools.h"
+#include "tools/stringTools.h"
+#include "tools/pathTools.h"
+#include "tools/archiveTools.h"
 
+#include <zim/archive.h>
+#include <zim/item.h>
 #include <pugixml.hpp>
 
 namespace kiwix
@@ -35,11 +40,17 @@ Book::Book() :
   m_readOnly(false)
 {
 }
+
 /* Destructor */
 Book::~Book()
 {
 }
 
+Book::Illustrations Book::getIllustrations() const
+{
+  return m_illustrations;
+}
+
 bool Book::update(const kiwix::Book& other)
 {
   if (m_readOnly)
@@ -48,53 +59,38 @@ bool Book::update(const kiwix::Book& other)
   if (m_id != other.m_id)
     return false;
 
-  m_readOnly = other.m_readOnly;
-  m_path = other.m_path;
-  m_pathValid = other.m_pathValid;
-  m_title = other.m_title;
-  m_description = other.m_description;
-  m_language = other.m_language;
-  m_creator = other.m_creator;
-  m_publisher = other.m_publisher;
-  m_date = other.m_date;
-  m_url = other.m_url;
-  m_name = other.m_name;
-  m_flavour = other.m_flavour;
-  m_tags = other.m_tags;
-  m_origId = other.m_origId;
-  m_articleCount = other.m_articleCount;
-  m_mediaCount = other.m_mediaCount;
-  m_size = other.m_size;
-  m_favicon = other.m_favicon;
-  m_faviconMimeType = other.m_faviconMimeType;
-  m_faviconUrl = other.m_faviconUrl;
-
-  m_downloadId = other.m_downloadId;
-
+  *this = other;
   return true;
 }
 
-void Book::update(const kiwix::Reader& reader)
-{
-  m_path = reader.getZimFilePath();
-  m_pathValid = true;
-  m_id = reader.getId();
-  m_title = reader.getTitle();
-  m_description = reader.getDescription();
-  m_language = reader.getLanguage();
-  m_creator = reader.getCreator();
-  m_publisher = reader.getPublisher();
-  m_date = reader.getDate();
-  m_name = reader.getName();
-  m_flavour = reader.getFlavour();
-  m_tags = reader.getTags();
-  m_origId = reader.getOrigId();
-  m_articleCount = reader.getArticleCount();
-  m_mediaCount = reader.getMediaCount();
-  m_size = static_cast<uint64_t>(reader.getFileSize()) << 10;
+void Book::update(const zim::Archive& archive) {
+  m_path = archive.getFilename();
   m_pathValid = true;
+  m_id = getArchiveId(archive);
+  m_title = getArchiveTitle(archive);
+  m_description = getMetaDescription(archive);
+  m_language = getMetaLanguage(archive);
+  m_creator = getMetaCreator(archive);
+  m_publisher = getMetaPublisher(archive);
+  m_date = getMetaDate(archive);
+  m_name = getMetaName(archive);
+  m_flavour = getMetaFlavour(archive);
+  m_tags = getMetaTags(archive);
+  m_category = getCategoryFromTags();
+  m_articleCount = getArchiveArticleCount(archive);
+  m_mediaCount = getArchiveMediaCount(archive);
+  m_size = static_cast<uint64_t>(getArchiveFileSize(archive)) << 10;
 
-  reader.getFavicon(m_favicon, m_faviconMimeType);
+  m_illustrations.clear();
+  for ( const auto illustrationSize : archive.getIllustrationSizes() ) {
+    const auto illustration = std::make_shared<Illustration>();
+    const zim::Item illustrationItem = archive.getIllustrationItem(illustrationSize);
+    illustration->width = illustration->height = illustrationSize;
+    illustration->mimeType = illustrationItem.getMimetype();
+    illustration->data = illustrationItem.getData();
+    // NOTE: illustration->url is left uninitialized
+    m_illustrations.push_back(illustration);
+  }
 }
 
 #define ATTR(name) node.attribute(name).value()
@@ -106,7 +102,7 @@ void Book::updateFromXml(const pugi::xml_node& node, const std::string& baseDir)
     path = computeAbsolutePath(baseDir, path);
   }
   m_path = path;
-  m_pathValid = fileExists(path);
+  m_pathValid = fileReadable(path);
   m_title = ATTR("title");
   m_description = ATTR("description");
   m_language = ATTR("language");
@@ -121,12 +117,19 @@ void Book::updateFromXml(const pugi::xml_node& node, const std::string& baseDir)
   m_articleCount = strtoull(ATTR("articleCount"), 0, 0);
   m_mediaCount = strtoull(ATTR("mediaCount"), 0, 0);
   m_size = strtoull(ATTR("size"), 0, 0) << 10;
-  m_favicon = base64_decode(ATTR("favicon"));
-  m_faviconMimeType = ATTR("faviconMimeType");
-  m_faviconUrl = ATTR("faviconUrl");
+  std::string favicon_mimetype = ATTR("faviconMimeType");
+  if (! favicon_mimetype.empty()) {
+    const auto favicon = std::make_shared<Illustration>();
+    favicon->data = base64_decode(ATTR("favicon"));
+    favicon->mimeType = favicon_mimetype;
+    favicon->url = ATTR("faviconUrl");
+    m_illustrations.assign(1, favicon);
+  }
   try {
     m_downloadId = ATTR("downloadId");
   } catch(...) {}
+  const auto catattr = node.attribute("category");
+  m_category = catattr.empty() ? getCategoryFromTags() : catattr.value();
 }
 #undef ATTR
 
@@ -152,10 +155,14 @@ void Book::updateFromOpds(const pugi::xml_node& node, const std::string& urlHost
   m_language = VALUE("language");
   m_creator = node.child("author").child("name").child_value();
   m_publisher = node.child("publisher").child("name").child_value();
-  m_date = fromOpdsDate(VALUE("updated"));
+  const std::string dcIssuedDate = VALUE("dc:issued");
+  m_date = dcIssuedDate.empty() ? VALUE("updated") : dcIssuedDate;
+  m_date = fromOpdsDate(m_date);
   m_name = VALUE("name");
   m_flavour = VALUE("flavour");
   m_tags = VALUE("tags");
+  const auto catnode = node.child("category");
+  m_category = catnode.empty() ? getCategoryFromTags() : catnode.child_value();
   m_articleCount = strtoull(VALUE("articleCount"), 0, 0);
   m_mediaCount = strtoull(VALUE("mediaCount"), 0, 0);
   for(auto linkNode = node.child("link"); linkNode;
@@ -167,8 +174,11 @@ void Book::updateFromOpds(const pugi::xml_node& node, const std::string& urlHost
       m_size = strtoull(linkNode.attribute("length").value(), 0, 0);
     }
     if (rel == "http://opds-spec.org/image/thumbnail") {
-      m_faviconUrl = urlHost + linkNode.attribute("href").value();
-      m_faviconMimeType = linkNode.attribute("type").value();
+      const auto favicon = std::make_shared<Illustration>();
+      favicon->data.clear();
+      favicon->url = urlHost + linkNode.attribute("href").value();
+      favicon->mimeType = linkNode.attribute("type").value();
+      m_illustrations.assign(1, favicon);
     }
  }
 
@@ -179,7 +189,7 @@ std::string Book::getHumanReadableIdFromPath() const
 {
   std::string id = m_path;
   if (!id.empty()) {
-    kiwix::removeAccents(id);
+    id = kiwix::removeAccents(id);
 
 #ifdef _WIN32
     id = replaceRegex(id, "", "^.*\\\\");
@@ -201,15 +211,54 @@ void Book::setPath(const std::string& path)
    : path;
 }
 
-const std::string& Book::getFavicon() const {
-  if (m_favicon.empty() && !m_faviconUrl.empty()) {
-    try {
-      m_favicon = download(m_faviconUrl);
-    } catch(...) {
-      std::cerr << "Cannot download favicon from " << m_faviconUrl;
+const Book::Illustration Book::missingDefaultIllustration;
+
+std::shared_ptr<const Book::Illustration> Book::getIllustration(unsigned int size) const
+{
+  for ( const auto& ilPtr : m_illustrations ) {
+    if (ilPtr->width == size && ilPtr->height == size) {
+      return ilPtr;
     }
   }
-  return m_favicon;
+  throw std::runtime_error("Cannot find illustration");
+}
+
+const Book::Illustration& Book::getDefaultIllustration() const
+{
+  try {
+    return *getIllustration(48);
+  } catch (...) {
+    return missingDefaultIllustration;
+  }
+}
+
+const std::string& Book::Illustration::getData() const
+{
+  if (data.empty() && !url.empty()) {
+    const std::lock_guard<std::mutex> l(mutex);
+    if ( data.empty() ) {
+      try {
+        data = download(url);
+      } catch(...) {
+        std::cerr << "Cannot download favicon from " << url;
+      }
+    }
+  }
+  return data;
+}
+
+const std::string& Book::getFavicon() const {
+  return getDefaultIllustration().getData();
+}
+
+const std::string& Book::getFaviconUrl() const
+{
+  return getDefaultIllustration().url;
+}
+
+const std::string& Book::getFaviconMimeType() const
+{
+  return getDefaultIllustration().mimeType;
 }
 
 std::string Book::getTagStr(const std::string& tagName) const {
@@ -220,4 +269,21 @@ bool Book::getTagBool(const std::string& tagName) const {
   return convertStrToBool(getTagStr(tagName));
 }
 
+std::string Book::getCategory() const
+{
+  return m_category;
+}
+
+std::string Book::getCategoryFromTags() const
+{
+  try
+  {
+    return getTagStr("category");
+  }
+  catch ( const std::out_of_range& )
+  {
+    return "";
+  }
+}
+
 }

src/config.h.in

@@ -1,3 +1,3 @@
 
-#mesondefine VERSION
+#mesondefine LIBKIWIX_VERSION
 

src/downloader.cpp

@@ -133,7 +133,7 @@ Downloader::Downloader() :
       m_knownDownloads[gid]->updateStatus();
     }
   } catch (std::exception& e) {
-    std::cerr << "aria2 tellActive failed : " << e.what();
+    std::cerr << "aria2 tellActive failed : " << e.what() << std::endl;
   }
   try {
     for (auto gid : mp_aria->tellWaiting()) {
@@ -141,7 +141,7 @@ Downloader::Downloader() :
       m_knownDownloads[gid]->updateStatus();
     }
   } catch (std::exception& e) {
-    std::cerr << "aria2 tellWaiting failed : " << e.what();
+    std::cerr << "aria2 tellWaiting failed : " << e.what() << std::endl;
   }
 }
 

src/entry.cpp

@@ -1,140 +0,0 @@
-/*
- * Copyright 2011 Emmanuel Engelhart <kelson@kiwix.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#include "reader.h"
-#include <time.h>
-
-#include <zim/search.h>
-
-namespace kiwix
-{
-
-Entry::Entry(zim::Article article)
-  : article(article)
-{
-}
-
-#define RETURN_IF_INVALID(WHAT) if(!good()) { return (WHAT); }
-
-std::string Entry::getPath() const
-{
-  RETURN_IF_INVALID("");
-  return article.getLongUrl();
-}
-
-std::string Entry::getTitle() const
-{
-  RETURN_IF_INVALID("");
-  return article.getTitle();
-}
-
-std::string Entry::getContent() const
-{
-  RETURN_IF_INVALID("");
-  return article.getData();
-}
-
-zim::Blob Entry::getBlob(offset_type offset) const
-{
-  RETURN_IF_INVALID(zim::Blob());
-  return article.getData(offset);
-}
-
-zim::Blob Entry::getBlob(offset_type offset, size_type size) const
-{
-  RETURN_IF_INVALID(zim::Blob());
-  return article.getData(offset, size);
-}
-
-std::pair<std::string, offset_type> Entry::getDirectAccessInfo() const
-{
-  RETURN_IF_INVALID(std::make_pair("", 0));
-  return article.getDirectAccessInformation();
-}
-
-size_type Entry::getSize() const
-{
-  RETURN_IF_INVALID(0);
-  return article.getArticleSize();
-}
-
-std::string Entry::getMimetype() const
-{
-  RETURN_IF_INVALID("");
-  try {
-    return article.getMimeType();
-  } catch (exception& e) {
-    return "application/octet-stream";
-  }
-}
-
-bool Entry::isRedirect() const
-{
-  RETURN_IF_INVALID(false);
-  return article.isRedirect();
-}
-
-bool Entry::isLinkTarget() const
-{
-  RETURN_IF_INVALID(false);
-  return article.isLinktarget();
-}
-
-bool Entry::isDeleted() const
-{
-  RETURN_IF_INVALID(false);
-  return article.isDeleted();
-}
-
-Entry Entry::getRedirectEntry() const
-{
-  RETURN_IF_INVALID(Entry());
-  if ( !article.isRedirect() ) {
-    throw NoEntry();
-  }
-
-  auto targeted_article = article.getRedirectArticle();
-  if ( !targeted_article.good()) {
-    throw NoEntry();
-  }
-  return targeted_article;
-}
-
-Entry Entry::getFinalEntry() const
-{
-  RETURN_IF_INVALID(Entry());
-  if (final_article.good()) {
-    return final_article;
-  }
-  int loopCounter = 42;
-  final_article = article;
-  while (final_article.isRedirect() && loopCounter--) {
-    final_article = final_article.getRedirectArticle();
-    if ( !final_article.good()) {
-      throw NoEntry();
-    }
-  }
-  // Prevent infinite loops.
-  if (final_article.isRedirect()) {
-    throw NoEntry();
-  }
-  return final_article;
-}
-
-}

src/kiwixserve.cpp

@@ -9,6 +9,7 @@
     # include <unistd.h>
 #endif
 
+#include "tools.h"
 #include "tools/pathTools.h"
 #include "tools/stringTools.h"
 

src/library.cpp

@@ -19,54 +19,184 @@
 
 #include "library.h"
 #include "book.h"
-#include "reader.h"
 #include "libxml_dumper.h"
 
+#include "tools.h"
 #include "tools/base64.h"
 #include "tools/regexTools.h"
 #include "tools/pathTools.h"
 #include "tools/stringTools.h"
+#include "tools/otherTools.h"
+#include "tools/concurrent_cache.h"
 
 #include <pugixml.hpp>
 #include <algorithm>
 #include <set>
+#include <cmath>
+#include <unicode/locid.h>
+#include <xapian.h>
 
 namespace kiwix
 {
 
-/* Constructor */
-Library::Library()
+namespace
 {
+
+std::string iso639_3ToXapian(const std::string& lang) {
+  return icu::Locale(lang.c_str()).getLanguage();
+};
+
+std::string normalizeText(const std::string& text)
+{
+  return removeAccents(text);
 }
-/* Destructor */
-Library::~Library()
+
+bool booksReferToTheSameArchive(const Book& book1, const Book& book2)
+{
+  return book1.isPathValid()
+      && book2.isPathValid()
+      && book1.getPath() == book2.getPath();
+}
+
+template<typename Key, typename Value>
+class MultiKeyCache: public ConcurrentCache<std::set<Key>, Value>
+{
+  public:
+    explicit MultiKeyCache(size_t maxEntries)
+      : ConcurrentCache<std::set<Key>, Value>(maxEntries)
+    {}
+
+    bool drop(const Key& key)
+    {
+      std::unique_lock<std::mutex> l(this->lock_);
+      bool removed = false;
+      for(auto& cache_key: this->impl_.keys()) {
+        if(cache_key.find(key)!=cache_key.end()) {
+          removed |= this->impl_.drop(cache_key);
+        }
+      }
+      return removed;
+    }
+};
+
+} // unnamed namespace
+
+struct Library::Impl
+{
+  struct Entry : Book
+  {
+    Library::Revision lastUpdatedRevision = 0;
+  };
+
+  Library::Revision m_revision;
+  std::map<std::string, Entry> m_books;
+  using ArchiveCache = ConcurrentCache<std::string, std::shared_ptr<zim::Archive>>;
+  std::unique_ptr<ArchiveCache> mp_archiveCache;
+  using SearcherCache = MultiKeyCache<std::string, std::shared_ptr<ZimSearcher>>;
+  std::unique_ptr<SearcherCache> mp_searcherCache;
+  std::vector<kiwix::Bookmark> m_bookmarks;
+  Xapian::WritableDatabase m_bookDB;
+
+  unsigned int getBookCount(const bool localBooks, const bool remoteBooks) const;
+
+  Impl();
+  ~Impl();
+
+  Impl(Impl&& );
+  Impl& operator=(Impl&& );
+};
+
+Library::Impl::Impl()
+  : mp_archiveCache(new ArchiveCache(std::max(getEnvVar<int>("KIWIX_ARCHIVE_CACHE_SIZE", 1), 1))),
+    mp_searcherCache(new SearcherCache(std::max(getEnvVar<int>("KIWIX_SEARCHER_CACHE_SIZE", 1), 1))),
+    m_bookDB("", Xapian::DB_BACKEND_INMEMORY)
 {
 }
 
+Library::Impl::~Impl()
+{
+}
+
+Library::Impl::Impl(Library::Impl&& ) = default;
+Library::Impl& Library::Impl::operator=(Library::Impl&& ) = default;
+
+unsigned int
+Library::Impl::getBookCount(const bool localBooks, const bool remoteBooks) const
+{
+  unsigned int result = 0;
+  for (auto& pair: m_books) {
+    auto& book = pair.second;
+    if ((!book.getPath().empty() && localBooks)
+        || (!book.getUrl().empty() && remoteBooks)) {
+      result++;
+    }
+  }
+  return result;
+}
+
+/* Constructor */
+Library::Library()
+  : mp_impl(new Library::Impl)
+{
+}
+
+Library::Library(Library&& other)
+  : mp_impl(std::move(other.mp_impl))
+{
+}
+
+Library& Library::operator=(Library&& other)
+{
+  mp_impl = std::move(other.mp_impl);
+  return *this;
+}
+
+/* Destructor */
+Library::~Library() = default;
 
 bool Library::addBook(const Book& book)
 {
+  std::lock_guard<std::mutex> lock(m_mutex);
+  ++mp_impl->m_revision;
   /* Try to find it */
+  updateBookDB(book);
   try {
-    auto& oldbook = m_books.at(book.getId());
-    oldbook.update(book);
+    auto& oldbook = mp_impl->m_books.at(book.getId());
+    if ( ! booksReferToTheSameArchive(oldbook, book) ) {
+      dropCache(book.getId());
+    }
+    oldbook.update(book); // XXX: This may have no effect if oldbook is readonly
+                          // XXX: Then m_bookDB will become out-of-sync with
+                          // XXX: the real contents of the library.
+    oldbook.lastUpdatedRevision = mp_impl->m_revision;
     return false;
   } catch (std::out_of_range&) {
-    m_books[book.getId()] = book;
+    auto& newEntry = mp_impl->m_books[book.getId()];
+    static_cast<Book&>(newEntry) = book;
+    newEntry.lastUpdatedRevision = mp_impl->m_revision;
+    size_t new_cache_size = static_cast<size_t>(std::ceil(mp_impl->getBookCount(true, true)*0.1));
+    if (getEnvVar<int>("KIWIX_ARCHIVE_CACHE_SIZE", -1) <= 0) {
+      mp_impl->mp_archiveCache->setMaxSize(new_cache_size);
+    }
+    if (getEnvVar<int>("KIWIX_SEARCHER_CACHE_SIZE", -1) <= 0) {
+      mp_impl->mp_searcherCache->setMaxSize(new_cache_size);
+    }
     return true;
   }
 }
 
 void Library::addBookmark(const Bookmark& bookmark)
 {
-  m_bookmarks.push_back(bookmark);
+  std::lock_guard<std::mutex> lock(m_mutex);
+  mp_impl->m_bookmarks.push_back(bookmark);
 }
 
 bool Library::removeBookmark(const std::string& zimId, const std::string& url)
 {
-  for(auto it=m_bookmarks.begin(); it!=m_bookmarks.end(); it++) {
+  std::lock_guard<std::mutex> lock(m_mutex);
+  for(auto it=mp_impl->m_bookmarks.begin(); it!=mp_impl->m_bookmarks.end(); it++) {
     if (it->getBookId() == zimId && it->getUrl() == url) {
-      m_bookmarks.erase(it);
+      mp_impl->m_bookmarks.erase(it);
       return true;
     }
   }
@@ -74,20 +204,70 @@ bool Library::removeBookmark(const std::string& zimId, const std::string& url)
 }
 
 
+void Library::dropCache(const std::string& id)
+{
+  mp_impl->mp_archiveCache->drop(id);
+  mp_impl->mp_searcherCache->drop(id);
+}
 
 bool Library::removeBookById(const std::string& id)
 {
-  return m_books.erase(id) == 1;
+  std::lock_guard<std::mutex> lock(m_mutex);
+  mp_impl->m_bookDB.delete_document("Q" + id);
+  dropCache(id);
+  // We do not change the cache size here
+  // Most of the time, the book is remove in case of library refresh, it is
+  // often associated with addBook calls (which will properly set the cache size)
+  // Having a too big cache is not a problem here (or it would have been before)
+  // (And setMaxSize doesn't actually reduce the cache size, extra cached items
+  //  will be removed in put or getOrPut).
+  return mp_impl->m_books.erase(id) == 1;
 }
 
-Book& Library::getBookById(const std::string& id)
+Library::Revision Library::getRevision() const
 {
-  return m_books.at(id);
+  std::lock_guard<std::mutex> lock(m_mutex);
+  return mp_impl->m_revision;
 }
 
-Book& Library::getBookByPath(const std::string& path)
+uint32_t Library::removeBooksNotUpdatedSince(Revision libraryRevision)
 {
-  for(auto& it: m_books) {
+  BookIdCollection booksToRemove;
+  {
+    std::lock_guard<std::mutex> lock(m_mutex);
+    for ( const auto& entry : mp_impl->m_books) {
+      if ( entry.second.lastUpdatedRevision <= libraryRevision ) {
+        booksToRemove.push_back(entry.first);
+      }
+    }
+  }
+
+  uint32_t countOfRemovedBooks = 0;
+  for ( const auto& id : booksToRemove ) {
+    if ( removeBookById(id) )
+      ++countOfRemovedBooks;
+  }
+  return countOfRemovedBooks;
+}
+
+const Book& Library::getBookById(const std::string& id) const
+{
+  // XXX: Doesn't make sense to lock this operation since it cannot
+  // XXX: guarantee thread-safety because of its return type
+  return mp_impl->m_books.at(id);
+}
+
+Book Library::getBookByIdThreadSafe(const std::string& id) const
+{
+  std::lock_guard<std::mutex> lock(m_mutex);
+  return getBookById(id);
+}
+
+const Book& Library::getBookByPath(const std::string& path) const
+{
+  // XXX: Doesn't make sense to lock this operation since it cannot
+  // XXX: guarantee thread-safety because of its return type
+  for(auto& it: mp_impl->m_books) {
     auto& book = it.second;
     if (book.getPath() == path)
       return book;
@@ -97,113 +277,141 @@ Book& Library::getBookByPath(const std::string& path)
   throw std::out_of_range(ss.str());
 }
 
-std::shared_ptr<Reader> Library::getReaderById(const std::string& id)
+std::shared_ptr<zim::Archive> Library::getArchiveById(const std::string& id)
 {
   try {
-    return m_readers.at(id);
-  } catch (std::out_of_range& e) {}
-
-  auto book = getBookById(id);
-  if (!book.isPathValid())
+    return mp_impl->mp_archiveCache->getOrPut(id,
+    [&](){
+      auto book = getBookById(id);
+      if (!book.isPathValid()) {
+        throw std::invalid_argument("");
+      }
+      return std::make_shared<zim::Archive>(book.getPath());
+    });
+  } catch (std::invalid_argument&) {
     return nullptr;
-  auto sptr = make_shared<Reader>(book.getPath());
-  m_readers[id] = sptr;
-  return sptr;
+  }
+}
+
+std::shared_ptr<ZimSearcher> Library::getSearcherByIds(const BookIdSet& ids)
+{
+  assert(!ids.empty());
+  try {
+    return mp_impl->mp_searcherCache->getOrPut(ids,
+    [&](){
+      std::vector<zim::Archive> archives;
+      for(auto& id:ids) {
+        auto archive = getArchiveById(id);
+        if(!archive) {
+          throw std::invalid_argument("");
+        }
+        archives.push_back(*archive);
+      }
+      return std::make_shared<ZimSearcher>(zim::Searcher(archives));
+    });
+  } catch (std::invalid_argument&) {
+    return nullptr;
+  }
 }
 
 unsigned int Library::getBookCount(const bool localBooks,
-                                   const bool remoteBooks)
+                                   const bool remoteBooks) const
 {
-  unsigned int result = 0;
-  for (auto& pair: m_books) {
-    auto& book = pair.second;
-    if ((!book.getPath().empty() && localBooks)
-        || (book.getPath().empty() && remoteBooks)) {
-      result++;
+  std::lock_guard<std::mutex> lock(m_mutex);
+  return mp_impl->getBookCount(localBooks, remoteBooks);
+}
+
+bool Library::writeToFile(const std::string& path) const
+{
+  const auto allBookIds = getBooksIds();
+
+  auto baseDir = removeLastPathElement(path);
+  LibXMLDumper dumper(this);
+  dumper.setBaseDir(baseDir);
+  std::string xml;
+  {
+    std::lock_guard<std::mutex> lock(m_mutex);
+    xml = dumper.dumpLibXMLContent(allBookIds);
+  };
+  return writeTextFile(path, xml);
+}
+
+bool Library::writeBookmarksToFile(const std::string& path) const
+{
+  LibXMLDumper dumper(this);
+  // NOTE: LibXMLDumper::dumpLibXMLBookmark uses Library in a thread-safe way
+  const std::string xml = dumper.dumpLibXMLBookmark();
+  return writeTextFile(path, xml);
+}
+
+Library::AttributeCounts Library::getBookAttributeCounts(BookStrPropMemFn p) const
+{
+  std::lock_guard<std::mutex> lock(m_mutex);
+  AttributeCounts propValueCounts;
+
+  for (const auto& pair: mp_impl->m_books) {
+    const auto& book = pair.second;
+    if (book.getOrigId().empty()) {
+      propValueCounts[(book.*p)()] += 1;
     }
   }
+  return propValueCounts;
+}
+
+std::vector<std::string> Library::getBookPropValueSet(BookStrPropMemFn p) const
+{
+  std::vector<std::string> result;
+  for ( const auto& kv : getBookAttributeCounts(p) ) {
+    result.push_back(kv.first);
+  }
   return result;
 }
 
-bool Library::writeToFile(const std::string& path)
+std::vector<std::string> Library::getBooksLanguages() const
 {
-  auto baseDir = removeLastPathElement(path);
-  LibXMLDumper dumper(this);
-  dumper.setBaseDir(baseDir);
-  return writeTextFile(path, dumper.dumpLibXMLContent(getBooksIds()));
+  return getBookPropValueSet(&Book::getLanguage);
 }
 
-bool Library::writeBookmarksToFile(const std::string& path)
+Library::AttributeCounts Library::getBooksLanguagesWithCounts() const
 {
-  LibXMLDumper dumper(this);
-  return writeTextFile(path, dumper.dumpLibXMLBookmark());
+  return getBookAttributeCounts(&Book::getLanguage);
 }
 
-std::vector<std::string> Library::getBooksLanguages()
+std::vector<std::string> Library::getBooksCategories() const
 {
-  std::vector<std::string> booksLanguages;
-  std::map<std::string, bool> booksLanguagesMap;
+  std::lock_guard<std::mutex> lock(m_mutex);
+  std::set<std::string> categories;
 
-  for (auto& pair: m_books) {
-    auto& book = pair.second;
-    auto& language = book.getLanguage();
-    if (booksLanguagesMap.find(language) == booksLanguagesMap.end()) {
-      if (book.getOrigId().empty()) {
-        booksLanguagesMap[language] = true;
-        booksLanguages.push_back(language);
-      }
+  for (const auto& pair: mp_impl->m_books) {
+    const auto& book = pair.second;
+    const auto& c = book.getCategory();
+    if ( !c.empty() ) {
+      categories.insert(c);
     }
   }
 
-  return booksLanguages;
+  return std::vector<std::string>(categories.begin(), categories.end());
 }
 
-std::vector<std::string> Library::getBooksCreators()
+std::vector<std::string> Library::getBooksCreators() const
 {
-  std::vector<std::string> booksCreators;
-  std::map<std::string, bool> booksCreatorsMap;
-
-  for (auto& pair: m_books) {
-    auto& book = pair.second;
-    auto& creator = book.getCreator();
-    if (booksCreatorsMap.find(creator) == booksCreatorsMap.end()) {
-      if (book.getOrigId().empty()) {
-        booksCreatorsMap[creator] = true;
-        booksCreators.push_back(creator);
-      }
-    }
-  }
-
-  return booksCreators;
+  return getBookPropValueSet(&Book::getCreator);
 }
 
-std::vector<std::string> Library::getBooksPublishers()
+std::vector<std::string> Library::getBooksPublishers() const
 {
-  std::vector<std::string> booksPublishers;
-  std::map<std::string, bool> booksPublishersMap;
-
-  for (auto& pair:m_books) {
-    auto& book = pair.second;
-    auto& publisher = book.getPublisher();
-    if (booksPublishersMap.find(publisher) == booksPublishersMap.end()) {
-      if (book.getOrigId().empty()) {
-        booksPublishersMap[publisher] = true;
-        booksPublishers.push_back(publisher);
-      }
-    }
-  }
-
-  return booksPublishers;
+  return getBookPropValueSet(&Book::getPublisher);
 }
 
-const std::vector<kiwix::Bookmark> Library::getBookmarks(bool onlyValidBookmarks)
+const std::vector<kiwix::Bookmark> Library::getBookmarks(bool onlyValidBookmarks) const
 {
   if (!onlyValidBookmarks) {
-    return m_bookmarks;
+    return mp_impl->m_bookmarks;
   }
   std::vector<kiwix::Bookmark> validBookmarks;
   auto booksId = getBooksIds();
-  for(auto& bookmark:m_bookmarks) {
+  std::lock_guard<std::mutex> lock(m_mutex);
+  for(auto& bookmark:mp_impl->m_bookmarks) {
     if (std::find(booksId.begin(), booksId.end(), bookmark.getBookId()) != booksId.end()) {
       validBookmarks.push_back(bookmark);
     }
@@ -211,37 +419,216 @@ const std::vector<kiwix::Bookmark> Library::getBookmarks(bool onlyValidBookmarks
   return validBookmarks;
 }
 
-std::vector<std::string> Library::getBooksIds()
+Library::BookIdCollection Library::getBooksIds() const
 {
-  std::vector<std::string> bookIds;
+  std::lock_guard<std::mutex> lock(m_mutex);
+  BookIdCollection bookIds;
 
-  for (auto& pair: m_books) {
+  for (auto& pair: mp_impl->m_books) {
     bookIds.push_back(pair.first);
   }
 
   return bookIds;
 }
 
-std::vector<std::string> Library::filter(const std::string& search)
+
+void Library::updateBookDB(const Book& book)
 {
-  if (search.empty()) {
-    return getBooksIds();
+  Xapian::Stem stemmer;
+  Xapian::TermGenerator indexer;
+  const std::string lang = book.getLanguage();
+  try {
+    stemmer = Xapian::Stem(iso639_3ToXapian(lang));
+    indexer.set_stemmer(stemmer);
+    indexer.set_stemming_strategy(Xapian::TermGenerator::STEM_SOME);
+  } catch (...) {}
+  Xapian::Document doc;
+  indexer.set_document(doc);
+
+  const std::string title = normalizeText(book.getTitle());
+  const std::string desc = normalizeText(book.getDescription());
+
+  // Index title and description without prefixes for general search
+  indexer.index_text(title);
+  indexer.increase_termpos();
+  indexer.index_text(desc);
+
+  // Index all fields for field-based search
+  indexer.index_text(title, 1, "S");
+  indexer.index_text(desc,  1, "XD");
+  indexer.index_text(lang,  1, "L");
+  indexer.index_text(normalizeText(book.getCreator()),   1, "A");
+  indexer.index_text(normalizeText(book.getPublisher()), 1, "XP");
+  indexer.index_text(normalizeText(book.getName()),      1, "XN");
+  indexer.index_text(normalizeText(book.getCategory()),  1, "XC");
+
+  for ( const auto& tag : split(normalizeText(book.getTags()), ";") ) {
+    doc.add_boolean_term("XT" + tag);
+    if ( tag[0] != '_' ) {
+      indexer.increase_termpos();
+      indexer.index_text(tag);
+    }
   }
 
-  return filter(Filter().query(search));
+  const std::string idterm = "Q" + book.getId();
+  doc.add_boolean_term(idterm);
+
+  doc.set_data(book.getId());
+
+  mp_impl->m_bookDB.replace_document(idterm, doc);
+}
+
+namespace
+{
+
+bool willSelectEverything(const Xapian::Query& query)
+{
+  return query.get_type() == Xapian::Query::LEAF_MATCH_ALL;
 }
 
 
-std::vector<std::string> Library::filter(const Filter& filter)
+Xapian::Query buildXapianQueryFromFilterQuery(const Filter& filter)
 {
-  std::vector<std::string> bookIds;
-  for(auto& pair:m_books) {
-    auto book = pair.second;
-    if(filter.accept(book)) {
-      bookIds.push_back(pair.first);
+  if ( !filter.hasQuery() || filter.getQuery().empty() ) {
+    // This is a thread-safe way to construct an equivalent of
+    // a Xapian::Query::MatchAll query
+    return Xapian::Query(std::string());
+  }
+
+  Xapian::QueryParser queryParser;
+  queryParser.set_default_op(Xapian::Query::OP_AND);
+  queryParser.add_prefix("title", "S");
+  queryParser.add_prefix("description", "XD");
+  queryParser.add_prefix("name", "XN");
+  queryParser.add_prefix("category", "XC");
+  queryParser.add_prefix("lang", "L");
+  queryParser.add_prefix("publisher", "XP");
+  queryParser.add_prefix("creator", "A");
+  queryParser.add_prefix("tag", "XT");
+  const auto partialQueryFlag = filter.queryIsPartial()
+                              ? Xapian::QueryParser::FLAG_PARTIAL
+                              : 0;
+  // Language assumed for the query is not known for sure so stemming
+  // is not applied
+  //queryParser.set_stemmer(Xapian::Stem(iso639_3ToXapian(???)));
+  //queryParser.set_stemming_strategy(Xapian::QueryParser::STEM_SOME);
+  const auto flags = Xapian::QueryParser::FLAG_PHRASE
+                   | Xapian::QueryParser::FLAG_BOOLEAN
+                   | Xapian::QueryParser::FLAG_BOOLEAN_ANY_CASE
+                   | Xapian::QueryParser::FLAG_LOVEHATE
+                   | Xapian::QueryParser::FLAG_WILDCARD
+                   | partialQueryFlag;
+  return queryParser.parse_query(normalizeText(filter.getQuery()), flags);
+}
+
+Xapian::Query nameQuery(const std::string& name)
+{
+  return Xapian::Query("XN" + normalizeText(name));
+}
+
+Xapian::Query categoryQuery(const std::string& category)
+{
+  return Xapian::Query("XC" + normalizeText(category));
+}
+
+Xapian::Query langQuery(const std::string& lang)
+{
+  return Xapian::Query("L" + normalizeText(lang));
+}
+
+Xapian::Query publisherQuery(const std::string& publisher)
+{
+  Xapian::QueryParser queryParser;
+  queryParser.set_default_op(Xapian::Query::OP_OR);
+  queryParser.set_stemming_strategy(Xapian::QueryParser::STEM_NONE);
+  const auto flags = 0;
+  const auto q = queryParser.parse_query(normalizeText(publisher), flags, "XP");
+  return Xapian::Query(Xapian::Query::OP_PHRASE, q.get_terms_begin(), q.get_terms_end(), q.get_length());
+}
+
+Xapian::Query creatorQuery(const std::string& creator)
+{
+  Xapian::QueryParser queryParser;
+  queryParser.set_default_op(Xapian::Query::OP_OR);
+  queryParser.set_stemming_strategy(Xapian::QueryParser::STEM_NONE);
+  const auto flags = 0;
+  const auto q = queryParser.parse_query(normalizeText(creator), flags, "A");
+  return Xapian::Query(Xapian::Query::OP_PHRASE, q.get_terms_begin(), q.get_terms_end(), q.get_length());
+}
+
+Xapian::Query tagsQuery(const Filter::Tags& acceptTags, const Filter::Tags& rejectTags)
+{
+  Xapian::Query q = Xapian::Query(std::string());
+  if (!acceptTags.empty()) {
+    for ( const auto& tag : acceptTags )
+      q &= Xapian::Query("XT" + normalizeText(tag));
+  }
+
+  if (!rejectTags.empty()) {
+    for ( const auto& tag : rejectTags )
+      q = Xapian::Query(Xapian::Query::OP_AND_NOT, q, "XT" + normalizeText(tag));
+  }
+  return q;
+}
+
+Xapian::Query buildXapianQuery(const Filter& filter)
+{
+  auto q = buildXapianQueryFromFilterQuery(filter);
+  if ( filter.hasName() ) {
+    q = Xapian::Query(Xapian::Query::OP_AND, q, nameQuery(filter.getName()));
+  }
+  if ( filter.hasCategory() ) {
+    q = Xapian::Query(Xapian::Query::OP_AND, q, categoryQuery(filter.getCategory()));
+  }
+  if ( filter.hasLang() ) {
+    q = Xapian::Query(Xapian::Query::OP_AND, q, langQuery(filter.getLang()));
+  }
+  if ( filter.hasPublisher() ) {
+    q = Xapian::Query(Xapian::Query::OP_AND, q, publisherQuery(filter.getPublisher()));
+  }
+  if ( filter.hasCreator() ) {
+    q = Xapian::Query(Xapian::Query::OP_AND, q, creatorQuery(filter.getCreator()));
+  }
+  if ( !filter.getAcceptTags().empty() || !filter.getRejectTags().empty() ) {
+    const auto tq = tagsQuery(filter.getAcceptTags(), filter.getRejectTags());
+    q = Xapian::Query(Xapian::Query::OP_AND, q, tq);;
+  }
+  return q;
+}
+
+} // unnamed namespace
+
+Library::BookIdCollection Library::filterViaBookDB(const Filter& filter) const
+{
+  const auto query = buildXapianQuery(filter);
+
+  if ( willSelectEverything(query) )
+    return getBooksIds();
+
+  BookIdCollection bookIds;
+
+  std::lock_guard<std::mutex> lock(m_mutex);
+  Xapian::Enquire enquire(mp_impl->m_bookDB);
+  enquire.set_query(query);
+  const auto results = enquire.get_mset(0, mp_impl->m_books.size());
+  for ( auto it = results.begin(); it != results.end(); ++it  ) {
+    bookIds.push_back(it.get_document().get_data());
+  }
+
+  return bookIds;
+}
+
+Library::BookIdCollection Library::filter(const Filter& filter) const
+{
+  BookIdCollection result;
+  const auto preliminaryResult = filterViaBookDB(filter);
+  std::lock_guard<std::mutex> lock(m_mutex);
+  for(auto id : preliminaryResult) {
+    if(filter.accept(mp_impl->m_books.at(id))) {
+      result.push_back(id);
     }
   }
-  return bookIds;
+  return result;
 }
 
 template<supportedListSortBy SORT>
@@ -257,13 +644,13 @@ struct KEY_TYPE<SIZE> {
 template<supportedListSortBy sort>
 class Comparator {
   private:
-    Library* lib;
-    bool     ascending;
+    const Library* const lib;
+    const bool     ascending;
 
     inline typename KEY_TYPE<sort>::TYPE get_key(const std::string& id);
 
   public:
-    Comparator(Library* lib, bool ascending) : lib(lib), ascending(ascending) {}
+    Comparator(const Library* lib, bool ascending) : lib(lib), ascending(ascending) {}
     inline bool operator() (const std::string& id1, const std::string& id2) {
       if (ascending) {
         return get_key(id1) < get_key(id2);
@@ -303,8 +690,13 @@ std::string Comparator<PUBLISHER>::get_key(const std::string& id)
   return lib->getBookById(id).getPublisher();
 }
 
-void Library::sort(std::vector<std::string>& bookIds, supportedListSortBy sort, bool ascending)
+void Library::sort(BookIdCollection& bookIds, supportedListSortBy sort, bool ascending) const
 {
+  // NOTE: Can reimplement this method in a way that doesn't require locking
+  // NOTE: for the entire duration of the sort. Will need to obtain (under a
+  // NOTE: lock) the required atributes from the books once, and then the
+  // NOTE: sorting will run on a copy of data without locking.
+  std::lock_guard<std::mutex> lock(m_mutex);
   switch(sort) {
     case TITLE:
       std::sort(bookIds.begin(), bookIds.end(), Comparator<TITLE>(this, ascending));
@@ -327,48 +719,6 @@ void Library::sort(std::vector<std::string>& bookIds, supportedListSortBy sort,
 }
 
 
-std::vector<std::string> Library::listBooksIds(
-    int mode,
-    supportedListSortBy sortBy,
-    const std::string& search,
-    const std::string& language,
-    const std::string& creator,
-    const std::string& publisher,
-    const std::vector<std::string>& tags,
-    size_t maxSize) {
-
-  Filter _filter;
-  if (mode & LOCAL)
-    _filter.local(true);
-  if (mode & NOLOCAL)
-    _filter.local(false);
-  if (mode & VALID)
-    _filter.valid(true);
-  if (mode & NOVALID)
-    _filter.valid(false);
-  if (mode & REMOTE)
-    _filter.remote(true);
-  if (mode & NOREMOTE)
-    _filter.remote(false);
-  if (!tags.empty())
-    _filter.acceptTags(tags);
-  if (maxSize != 0)
-    _filter.maxSize(maxSize);
-  if (!language.empty())
-    _filter.lang(language);
-  if (!publisher.empty())
-    _filter.publisher(publisher);
-  if (!creator.empty())
-    _filter.creator(creator);
-  if (!search.empty())
-    _filter.query(search);
-
-  auto bookIds = filter(_filter);
-
-  sort(bookIds, sortBy, true);
-  return bookIds;
-}
-
 Filter::Filter()
   : activeFilters(0),
     _maxSize(0)
@@ -391,6 +741,7 @@ enum filterTypes {
   MAXSIZE = FLAG(11),
   QUERY = FLAG(12),
   NAME = FLAG(13),
+  CATEGORY = FLAG(14),
 };
 
 Filter& Filter::local(bool accept)
@@ -429,20 +780,27 @@ Filter& Filter::valid(bool accept)
   return *this;
 }
 
-Filter& Filter::acceptTags(std::vector<std::string> tags)
+Filter& Filter::acceptTags(const Tags& tags)
 {
   _acceptTags = tags;
   activeFilters |= ACCEPTTAGS;
   return *this;
 }
 
-Filter& Filter::rejectTags(std::vector<std::string> tags)
+Filter& Filter::rejectTags(const Tags& tags)
 {
   _rejectTags = tags;
   activeFilters |= REJECTTAGS;
   return *this;
 }
 
+Filter& Filter::category(std::string category)
+{
+  _category = category;
+  activeFilters |= CATEGORY;
+  return *this;
+}
+
 Filter& Filter::lang(std::string lang)
 {
   _lang = lang;
@@ -471,9 +829,10 @@ Filter& Filter::maxSize(size_t maxSize)
   return *this;
 }
 
-Filter& Filter::query(std::string query)
+Filter& Filter::query(std::string query, bool partial)
 {
   _query = query;
+  _queryIsPartial = partial;
   activeFilters |= QUERY;
   return *this;
 }
@@ -487,6 +846,36 @@ Filter& Filter::name(std::string name)
 
 #define ACTIVE(X) (activeFilters & (X))
 #define FILTER(TAG, TEST) if (ACTIVE(TAG) && !(TEST)) { return false; }
+bool Filter::hasQuery() const
+{
+  return ACTIVE(QUERY);
+}
+
+bool Filter::hasName() const
+{
+  return ACTIVE(NAME);
+}
+
+bool Filter::hasCategory() const
+{
+  return ACTIVE(CATEGORY);
+}
+
+bool Filter::hasLang() const
+{
+  return ACTIVE(LANG);
+}
+
+bool Filter::hasPublisher() const
+{
+  return ACTIVE(_PUBLISHER);
+}
+
+bool Filter::hasCreator() const
+{
+  return ACTIVE(_CREATOR);
+}
+
 bool Filter::accept(const Book& book) const
 {
   auto local = !book.getPath().empty();
@@ -502,40 +891,8 @@ bool Filter::accept(const Book& book) const
   FILTER(_NOREMOTE, !remote)
 
   FILTER(MAXSIZE, book.getSize() <= _maxSize)
-  FILTER(LANG, book.getLanguage() == _lang)
-  FILTER(_PUBLISHER, book.getPublisher() == _publisher)
-  FILTER(_CREATOR, book.getCreator() == _creator)
-  FILTER(NAME, book.getName() == _name)
-
-  if (ACTIVE(ACCEPTTAGS)) {
-    if (!_acceptTags.empty()) {
-      auto vBookTags = split(book.getTags(), ";");
-      std::set<std::string> sBookTags(vBookTags.begin(), vBookTags.end());
-      for (auto& t: _acceptTags) {
-        if (sBookTags.find(t) == sBookTags.end()) {
-          return false;
-        }
-      }
-    }
-  }
-  if (ACTIVE(REJECTTAGS)) {
-    if (!_rejectTags.empty()) {
-      auto vBookTags = split(book.getTags(), ";");
-      std::set<std::string> sBookTags(vBookTags.begin(), vBookTags.end());
-      for (auto& t: _rejectTags) {
-        if (sBookTags.find(t) != sBookTags.end()) {
-          return false;
-        }
-      }
-    }
-  }
-  if ( ACTIVE(QUERY)
-    && !(matchRegex(book.getTitle(), "\\Q" + _query + "\\E")
-        || matchRegex(book.getDescription(), "\\Q" + _query + "\\E")))
-    return false;
 
   return true;
-
 }
 
 }

src/libxml_dumper.cpp

@@ -20,15 +20,15 @@
 #include "libxml_dumper.h"
 #include "book.h"
 
+#include "tools.h"
 #include "tools/base64.h"
 #include "tools/stringTools.h"
 #include "tools/otherTools.h"
-#include "tools/pathTools.h"
 
 namespace kiwix
 {
 /* Constructor */
-LibXMLDumper::LibXMLDumper(Library* library)
+LibXMLDumper::LibXMLDumper(const Library* library)
   : library(library)
 {
 }
@@ -60,10 +60,13 @@ void LibXMLDumper::handleBook(Book book, pugi::xml_node root_node) {
     ADD_ATTR_NOT_EMPTY(entry_node, "name", book.getName());
     ADD_ATTR_NOT_EMPTY(entry_node, "flavour", book.getFlavour());
     ADD_ATTR_NOT_EMPTY(entry_node, "tags", book.getTags());
-    ADD_ATTR_NOT_EMPTY(entry_node, "faviconMimeType", book.getFaviconMimeType());
-    ADD_ATTR_NOT_EMPTY(entry_node, "faviconUrl", book.getFaviconUrl());
-    if (!book.getFavicon().empty())
-      ADD_ATTRIBUTE(entry_node, "favicon", base64_encode(book.getFavicon()));
+    try {
+      auto defaultIllustration = book.getIllustration(48);
+      ADD_ATTR_NOT_EMPTY(entry_node, "faviconMimeType", defaultIllustration->mimeType);
+      ADD_ATTR_NOT_EMPTY(entry_node, "faviconUrl", defaultIllustration->url);
+      if (!defaultIllustration->getData().empty())
+        ADD_ATTRIBUTE(entry_node, "favicon", base64_encode(defaultIllustration->getData()));
+    } catch(...) {}
   } else {
     ADD_ATTRIBUTE(entry_node, "origId", book.getOrigId());
   }
@@ -91,7 +94,7 @@ void LibXMLDumper::handleBookmark(Bookmark bookmark, pugi::xml_node root_node) {
   auto book_node = entry_node.append_child("book");
 
   try {
-    auto book = library->getBookById(bookmark.getBookId());
+    auto book = library->getBookByIdThreadSafe(bookmark.getBookId());
     ADD_TEXT_ENTRY(book_node, "id", book.getId());
     ADD_TEXT_ENTRY(book_node, "title", book.getTitle());
     ADD_TEXT_ENTRY(book_node, "language", book.getLanguage());

src/manager.cpp

@@ -19,34 +19,88 @@
 
 #include "manager.h"
 
+#include "tools.h"
 #include "tools/pathTools.h"
 
 #include <pugixml.hpp>
 
 namespace kiwix
 {
+
+namespace
+{
+
+struct NoDelete
+{
+  template<class T> void operator()(T*) {}
+};
+
+} // unnamed namespace
+
+////////////////////////////////////////////////////////////////////////////////
+// LibraryManipulator
+////////////////////////////////////////////////////////////////////////////////
+
+LibraryManipulator::LibraryManipulator(Library* library)
+  : library(*library)
+{}
+
+LibraryManipulator::~LibraryManipulator()
+{}
+
+bool LibraryManipulator::addBookToLibrary(const Book& book)
+{
+  const auto ret = library.addBook(book);
+  if ( ret ) {
+    bookWasAddedToLibrary(book);
+  }
+  return ret;
+}
+
+void LibraryManipulator::addBookmarkToLibrary(const Bookmark& bookmark)
+{
+  library.addBookmark(bookmark);
+  bookmarkWasAddedToLibrary(bookmark);
+}
+
+uint32_t LibraryManipulator::removeBooksNotUpdatedSince(Library::Revision rev)
+{
+  const auto n = library.removeBooksNotUpdatedSince(rev);
+  if ( n != 0 ) {
+    booksWereRemovedFromLibrary();
+  }
+  return n;
+}
+
+void LibraryManipulator::bookWasAddedToLibrary(const Book& book)
+{
+}
+
+void LibraryManipulator::bookmarkWasAddedToLibrary(const Bookmark& bookmark)
+{
+}
+
+void LibraryManipulator::booksWereRemovedFromLibrary()
+{
+}
+
+////////////////////////////////////////////////////////////////////////////////
+// Manager
+////////////////////////////////////////////////////////////////////////////////
+
 /* Constructor */
 Manager::Manager(LibraryManipulator* manipulator):
   writableLibraryPath(""),
-  manipulator(manipulator),
-  mustDeleteManipulator(false)
+  manipulator(manipulator, NoDelete())
 {
 }
 
 Manager::Manager(Library* library) :
   writableLibraryPath(""),
-  manipulator(new DefaultLibraryManipulator(library)),
-  mustDeleteManipulator(true)
+  manipulator(new LibraryManipulator(library))
 {
 }
 
-/* Destructor */
-Manager::~Manager()
-{
-  if (mustDeleteManipulator) {
-    delete manipulator;
-  }
-}
 bool Manager::parseXmlDom(const pugi::xml_document& doc,
                           bool readOnly,
                           const std::string& libraryPath,
@@ -80,7 +134,7 @@ bool Manager::readXml(const std::string& xml,
 {
   pugi::xml_document doc;
   pugi::xml_parse_result result
-      = doc.load_buffer_inplace((void*)xml.data(), xml.size());
+      = doc.load_buffer((void*)xml.data(), xml.size());
 
   if (result) {
     this->parseXmlDom(doc, readOnly, libraryPath, trustLibrary);
@@ -124,7 +178,7 @@ bool Manager::readOpds(const std::string& content, const std::string& urlHost)
 {
   pugi::xml_document doc;
   pugi::xml_parse_result result
-      = doc.load_buffer_inplace((void*)content.data(), content.size());
+      = doc.load_buffer((void*)content.data(), content.size());
 
   if (result) {
     this->parseOpdsDom(doc, urlHost);
@@ -214,8 +268,8 @@ bool Manager::readBookFromPath(const std::string& path, kiwix::Book* book)
     tmp_path = computeAbsolutePath(getCurrentDirectory(), path);
   }
   try {
-    kiwix::Reader reader(tmp_path);
-    book->update(reader);
+    zim::Archive archive(tmp_path);
+    book->update(archive);
     book->setPathValid(true);
   } catch (const std::exception& e) {
     book->setPathValid(false);
@@ -248,4 +302,21 @@ bool Manager::readBookmarkFile(const std::string& path)
   return true;
 }
 
+void Manager::reload(const Paths& paths)
+{
+  const auto libRevision = manipulator->getLibrary().getRevision();
+  for (std::string path : paths) {
+    if (!path.empty()) {
+      if ( kiwix::isRelativePath(path) )
+        path = kiwix::computeAbsolutePath(kiwix::getCurrentDirectory(), path);
+
+      if (!readFile(path, false, true)) {
+        throw std::runtime_error("Failed to load the XML library file '" + path + "'.");
+      }
+    }
+  }
+
+  manipulator->removeBooksNotUpdatedSince(libRevision);
+}
+
 }

src/meson.build

@@ -6,10 +6,7 @@ kiwix_sources = [
   'libxml_dumper.cpp',
   'opds_dumper.cpp',
   'downloader.cpp',
-  'reader.cpp',
-  'entry.cpp',
   'server.cpp',
-  'searcher.cpp',
   'search_renderer.cpp',
   'subprocess.cpp',
   'aria2.cpp',
@@ -19,13 +16,21 @@ kiwix_sources = [
   'tools/stringTools.cpp',
   'tools/networkTools.cpp',
   'tools/otherTools.cpp',
+  'tools/archiveTools.cpp',
   'kiwixserve.cpp',
   'name_mapper.cpp',
+  'server/byte_range.cpp',
   'server/etag.cpp',
   'server/request_context.cpp',
-  'server/response.cpp'
+  'server/response.cpp',
+  'server/internalServer.cpp',
+  'server/internalServer_catalog_v2.cpp',
+  'server/i18n.cpp',
+  'opds_catalog.cpp',
+  'version.cpp'
 ]
 kiwix_sources += lib_resources
+kiwix_sources += i18n_resources
 
 if host_machine.system() == 'windows'
   kiwix_sources += 'subprocess_windows.cpp'
@@ -33,15 +38,7 @@ else
   kiwix_sources += 'subprocess_unix.cpp'
 endif
 
-if 'android' in wrapper
-  install_dir = 'kiwix-lib/jniLibs/' + meson.get_cross_property('android_abi')
-else
-  install_dir = get_option('libdir')
-endif
-
-if 'android' in wrapper or 'java' in wrapper
-  subdir('wrapper/java')
-endif
+install_dir = get_option('libdir')
 
 config_h = configure_file(output : 'kiwix_config.h',
                configuration : conf,
@@ -52,6 +49,7 @@ kiwixlib = library('kiwix',
                    kiwix_sources,
                    include_directories : inc,
                    dependencies : all_deps,
+                   link_args: extra_libs,
                    version: meson.project_version(),
                    install: true,
                    install_dir: install_dir)

src/microhttpd_wrapper.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2013 Emmanuel Engelhart <kelson@kiwix.org>
+ * Copyright 2020 Emmanuel Engelhart <kelson@kiwix.org>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU  General Public License as published by
@@ -17,9 +17,8 @@
  * MA 02110-1301, USA.
  */
 
-package org.kiwix.kiwixlib;
+#include <microhttpd.h>
 
-public class JNIKiwixInt
-{
-  public int value;
-}
+#if MHD_VERSION < 0x00097002
+typedef int MHD_Result;
+#endif

src/name_mapper.cpp

@@ -51,12 +51,54 @@ HumanReadableNameMapper::HumanReadableNameMapper(kiwix::Library& library, bool w
   }
 }
 
-std::string HumanReadableNameMapper::getNameForId(const std::string& id) {
+std::string HumanReadableNameMapper::getNameForId(const std::string& id) const {
   return m_idToName.at(id);
 }
 
-std::string HumanReadableNameMapper::getIdForName(const std::string& name) {
+std::string HumanReadableNameMapper::getIdForName(const std::string& name) const {
   return m_nameToId.at(name);
 }
 
+////////////////////////////////////////////////////////////////////////////////
+// UpdatableNameMapper
+////////////////////////////////////////////////////////////////////////////////
+
+UpdatableNameMapper::UpdatableNameMapper(Library& lib, bool withAlias)
+  : library(lib)
+  , withAlias(withAlias)
+{
+  update();
+}
+
+void UpdatableNameMapper::update()
+{
+  const auto newNameMapper = new HumanReadableNameMapper(library, withAlias);
+  std::lock_guard<std::mutex> lock(mutex);
+  nameMapper.reset(newNameMapper);
+}
+
+UpdatableNameMapper::NameMapperHandle
+UpdatableNameMapper::currentNameMapper() const
+{
+  // Return a copy of the handle to the current NameMapper object. It will
+  // ensure that the object survives any call to UpdatableNameMapper::update()
+  // made before the completion of any pending operation on that object.
+  std::lock_guard<std::mutex> lock(mutex);
+  return nameMapper;
+}
+
+std::string UpdatableNameMapper::getNameForId(const std::string& id) const
+{
+  // Ensure that the current nameMapper object survives a concurrent call
+  // to UpdatableNameMapper::update()
+  return currentNameMapper()->getNameForId(id);
+}
+
+std::string UpdatableNameMapper::getIdForName(const std::string& name) const
+{
+  // Ensure that the current nameMapper object survives a concurrent call
+  // to UpdatableNameMapper::update()
+  return currentNameMapper()->getIdForName(name);
+}
+
 }

src/opds_catalog.cpp

@@ -0,0 +1,74 @@
+/*
+ * Copyright 2021 Veloman Yunkan <veloman.yunkan@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#include "opds_catalog.h"
+#include "tools/stringTools.h"
+
+#include <sstream>
+
+namespace kiwix
+{
+
+namespace
+{
+const char opdsSearchEndpoint[] = "/catalog/v2/entries";
+
+enum Separator { AMP };
+
+std::ostringstream& operator<<(std::ostringstream& oss, Separator sep)
+{
+  if ( oss.tellp() > 0 )
+    oss << "&";
+  return oss;
+}
+
+std::string buildSearchString(const Filter& f)
+{
+  std::ostringstream oss;
+  if ( f.hasQuery() )
+    oss << AMP << "q=" << urlEncode(f.getQuery());
+
+  if ( f.hasCategory() )
+    oss << AMP << "category=" << urlEncode(f.getCategory());
+
+  if ( f.hasLang() )
+    oss << AMP << "lang=" << urlEncode(f.getLang());
+
+  if ( f.hasName() )
+    oss << AMP << "name=" << urlEncode(f.getName());
+
+  if ( !f.getAcceptTags().empty() )
+    oss << AMP << "tag=" << urlEncode(join(f.getAcceptTags(), ";"));
+
+  return oss.str();
+}
+
+} // unnamed namespace
+
+std::string getSearchUrl(const Filter& f)
+{
+  const std::string searchString = buildSearchString(f);
+
+  if ( searchString.empty() )
+    return opdsSearchEndpoint;
+  else
+    return opdsSearchEndpoint + ("?" + searchString);
+}
+
+} // namespace kiwix

src/opds_dumper.cpp

@@ -20,11 +20,15 @@
 #include "opds_dumper.h"
 #include "book.h"
 
+#include "kiwixlib-resources.h"
+#include <mustache.hpp>
+
+#include "tools/stringTools.h"
 #include "tools/otherTools.h"
-#include <iomanip>
 
 namespace kiwix
 {
+
 /* Constructor */
 OPDSDumper::OPDSDumper(Library* library)
   : library(library)
@@ -35,120 +39,240 @@ OPDSDumper::~OPDSDumper()
 {
 }
 
-std::string gen_date_str()
-{
-  auto now = time(0);
-  auto tm = localtime(&now);
-
-  std::stringstream is;
-  is << std::setw(2) << std::setfill('0')
-     << 1900+tm->tm_year << "-"
-     << std::setw(2) << std::setfill('0') << tm->tm_mon << "-"
-     << std::setw(2) << std::setfill('0') << tm->tm_mday << "T"
-     << std::setw(2) << std::setfill('0') << tm->tm_hour << ":"
-     << std::setw(2) << std::setfill('0') << tm->tm_min << ":"
-     << std::setw(2) << std::setfill('0') << tm->tm_sec << "Z";
-  return is.str();
-}
-
-static std::string gen_date_from_yyyy_mm_dd(const std::string& date)
-{
-  std::stringstream is;
-  is << date << "T00:00::00:Z";
-  return is.str();
-}
-
 void OPDSDumper::setOpenSearchInfo(int totalResults, int startIndex, int count)
 {
   m_totalResults = totalResults;
   m_startIndex = startIndex,
   m_count = count;
-  m_isSearchResult = true;
 }
 
-#define ADD_TEXT_ENTRY(node, child, value) (node).append_child((child)).append_child(pugi::node_pcdata).set_value((value).c_str())
-
-pugi::xml_node OPDSDumper::handleBook(Book book, pugi::xml_node root_node) {
-  auto entry_node = root_node.append_child("entry");
-  ADD_TEXT_ENTRY(entry_node, "id", "urn:uuid:"+book.getId());
-  ADD_TEXT_ENTRY(entry_node, "title", book.getTitle());
-  ADD_TEXT_ENTRY(entry_node, "summary", book.getDescription());
-  ADD_TEXT_ENTRY(entry_node, "language", book.getLanguage());
-  ADD_TEXT_ENTRY(entry_node, "updated", gen_date_from_yyyy_mm_dd(book.getDate()));
-  ADD_TEXT_ENTRY(entry_node, "name", book.getName());
-  ADD_TEXT_ENTRY(entry_node, "flavour", book.getFlavour());
-  ADD_TEXT_ENTRY(entry_node, "tags", book.getTags());
-  ADD_TEXT_ENTRY(entry_node, "articleCount", to_string(book.getArticleCount()));
-  ADD_TEXT_ENTRY(entry_node, "mediaCount", to_string(book.getMediaCount()));
-  ADD_TEXT_ENTRY(entry_node, "icon", rootLocation + "/meta?name=favicon&content=" + book.getHumanReadableIdFromPath());
-
-  auto content_node = entry_node.append_child("link");
-  content_node.append_attribute("type") = "text/html";
-  content_node.append_attribute("href") = (rootLocation + "/" + book.getHumanReadableIdFromPath()).c_str();
-
-  auto author_node = entry_node.append_child("author");
-  ADD_TEXT_ENTRY(author_node, "name", book.getCreator());
-
-  auto publisher_node = entry_node.append_child("publisher");
-  ADD_TEXT_ENTRY(publisher_node, "name", book.getPublisher());
-
-  if (! book.getUrl().empty()) {
-    auto acquisition_link = entry_node.append_child("link");
-    acquisition_link.append_attribute("rel") = "http://opds-spec.org/acquisition/open-access";
-    acquisition_link.append_attribute("type") = "application/x-zim";
-    acquisition_link.append_attribute("href") = book.getUrl().c_str();
-    acquisition_link.append_attribute("length") = to_string(book.getSize()).c_str();
-  }
-
-  if (! book.getFaviconMimeType().empty() ) {
-    auto image_link = entry_node.append_child("link");
-    image_link.append_attribute("rel") = "http://opds-spec.org/image/thumbnail";
-    image_link.append_attribute("type") = book.getFaviconMimeType().c_str();
-    image_link.append_attribute("href") = (rootLocation + "/meta?name=favicon&content=" + book.getHumanReadableIdFromPath()).c_str();
-  }
-  return entry_node;
-}
-
-string OPDSDumper::dumpOPDSFeed(const std::vector<std::string>& bookIds)
+namespace
 {
-  date = gen_date_str();
-  pugi::xml_document doc;
 
-  auto root_node = doc.append_child("feed");
-  root_node.append_attribute("xmlns") = "http://www.w3.org/2005/Atom";
-  root_node.append_attribute("xmlns:opds") = "http://opds-spec.org/2010/catalog";
+typedef kainjow::mustache::data MustacheData;
+typedef kainjow::mustache::list BooksData;
+typedef kainjow::mustache::list IllustrationInfo;
 
-  ADD_TEXT_ENTRY(root_node, "id", id);
+IllustrationInfo getBookIllustrationInfo(const Book& book)
+{
+    kainjow::mustache::list illustrations;
+    if ( book.isPathValid() ) {
+      for ( const auto& illustration : book.getIllustrations() ) {
+        // For now, we are handling only sizexsize@1 illustration.
+        // So we can simply pass one size to mustache.
+        illustrations.push_back(kainjow::mustache::object{
+          {"icon_size", to_string(illustration->width)},
+          {"icon_mimetype", illustration->mimeType}
+        });
+      }
+    }
+    return illustrations;
+}
 
-  ADD_TEXT_ENTRY(root_node, "title", title);
-  ADD_TEXT_ENTRY(root_node, "updated", date);
+kainjow::mustache::object getSingleBookData(const Book& book)
+{
+    const auto bookDate = book.getDate() + "T00:00:00Z";
+    return kainjow::mustache::object{
+      {"id", book.getId()},
+      {"name", book.getName()},
+      {"title", book.getTitle()},
+      {"description", book.getDescription()},
+      {"language", book.getLanguage()},
+      {"content_id",  urlEncode(book.getHumanReadableIdFromPath(), true)},
+      {"updated", bookDate}, // XXX: this should be the entry update datetime
+      {"book_date", bookDate},
+      {"category", book.getCategory()},
+      {"flavour", book.getFlavour()},
+      {"tags", book.getTags()},
+      {"article_count", to_string(book.getArticleCount())},
+      {"media_count", to_string(book.getMediaCount())},
+      {"author_name", book.getCreator()},
+      {"publisher_name", book.getPublisher()},
+      {"url", onlyAsNonEmptyMustacheValue(book.getUrl())},
+      {"size", to_string(book.getSize())},
+      {"icons", getBookIllustrationInfo(book)},
+    };
+}
 
-  if (m_isSearchResult) {
-    ADD_TEXT_ENTRY(root_node, "totalResults", to_string(m_totalResults));
-    ADD_TEXT_ENTRY(root_node, "startIndex", to_string(m_startIndex));
-    ADD_TEXT_ENTRY(root_node, "itemsPerPage", to_string(m_count));
-  }
+std::string getSingleBookEntryXML(const Book& book, bool withXMLHeader, const std::string& rootLocation, const std::string& endpointRoot, bool partial)
+{
+  auto data = getSingleBookData(book);
+  data["with_xml_header"] = MustacheData(withXMLHeader);
+  data["dump_partial_entries"] = MustacheData(partial);
+  data["endpoint_root"] = endpointRoot;
+  data["root"] = rootLocation;
+  return render_template(RESOURCE::templates::catalog_v2_entry_xml, data);
+}
 
-  auto self_link_node = root_node.append_child("link");
-  self_link_node.append_attribute("rel") = "self";
-  self_link_node.append_attribute("href") = "";
-  self_link_node.append_attribute("type") = "application/atom+xml";
-
-
-  if (!searchDescriptionUrl.empty() ) {
-    auto search_link = root_node.append_child("link");
-    search_link.append_attribute("rel") = "search";
-    search_link.append_attribute("type") = "application/opensearchdescription+xml";
-    search_link.append_attribute("href") = searchDescriptionUrl.c_str();
-  }
-
-  if (library) {
-    for (auto& bookId: bookIds) {
-      handleBook(library->getBookById(bookId), root_node);
+BooksData getBooksData(const Library* library, const std::vector<std::string>& bookIds, const std::string& rootLocation, const std::string& endpointRoot, bool partial)
+{
+  BooksData booksData;
+  for ( const auto& bookId : bookIds ) {
+    try {
+      const Book book = library->getBookByIdThreadSafe(bookId);
+      booksData.push_back(kainjow::mustache::object{
+          {"entry", getSingleBookEntryXML(book, false, rootLocation, endpointRoot, partial)}
+      });
+    } catch ( const std::out_of_range& ) {
+      // the book was removed from the library since its id was obtained
+      // ignore it
     }
   }
 
-  return nodeToString(root_node);
+  return booksData;
+}
+
+std::map<std::string, std::string> iso639_3 = {
+  {"atj", "atikamekw"},
+  {"azb", "آذربایجان دیلی"},
+  {"bcl", "central bikol"},
+  {"bgs", "tagabawa"},
+  {"bxr", "буряад хэлэн"},
+  {"cbk", "chavacano"},
+  {"cdo", "閩東語"},
+  {"dag", "Dagbani"},
+  {"diq", "dimli"},
+  {"dty", "डोटेली"},
+  {"eml", "emiliân-rumagnōl"},
+  {"fbs", "српскохрватски"},
+  {"ido", "ido"},
+  {"kbp", "kabɩyɛ"},
+  {"kld", "Gamilaraay"},
+  {"lbe", "лакку маз"},
+  {"lbj", "ལ་དྭགས་སྐད་"},
+  {"map", "Austronesian"},
+  {"mhr", "марий йылме"},
+  {"mnw", "ဘာသာမန်"},
+  {"myn", "mayan"},
+  {"nah", "nahuatl"},
+  {"nai", "north American Indian"},
+  {"nds", "plattdütsch"},
+  {"nrm", "bhasa narom"},
+  {"olo", "livvi"},
+  {"pih", "Pitcairn-Norfolk"},
+  {"pnb", "Western Panjabi"},
+  {"rmr", "Caló"},
+  {"rmy", "romani shib"},
+  {"roa", "romance languages"},
+  {"twi", "twi"}
+};
+
+std::once_flag fillLanguagesFlag;
+
+void fillLanguagesMap()
+{
+  for (auto icuLangPtr = icu::Locale::getISOLanguages(); *icuLangPtr != NULL; ++icuLangPtr) {
+    const ICULanguageInfo lang(*icuLangPtr);
+    iso639_3.insert({lang.iso3Code(), lang.selfName()});
+  }
+}
+
+std::string getLanguageSelfName(const std::string& lang) {
+  const auto itr = iso639_3.find(lang);
+  if (itr != iso639_3.end()) {
+    return itr->second;
+  }
+  return lang;
+};
+
+} // unnamed namespace
+
+string OPDSDumper::dumpOPDSFeed(const std::vector<std::string>& bookIds, const std::string& query) const
+{
+  const auto booksData = getBooksData(library, bookIds, rootLocation, "", false);
+  const kainjow::mustache::object template_data{
+     {"date", gen_date_str()},
+     {"root", rootLocation},
+     {"feed_id", gen_uuid(libraryId + "/catalog/search?"+query)},
+     {"filter", onlyAsNonEmptyMustacheValue(query)},
+     {"totalResults", to_string(m_totalResults)},
+     {"startIndex", to_string(m_startIndex)},
+     {"itemsPerPage", to_string(m_count)},
+     {"books", booksData }
+  };
+
+  return render_template(RESOURCE::templates::catalog_entries_xml, template_data);
+}
+
+string OPDSDumper::dumpOPDSFeedV2(const std::vector<std::string>& bookIds, const std::string& query, bool partial) const
+{
+  const auto endpointRoot = rootLocation + "/catalog/v2";
+  const auto booksData = getBooksData(library, bookIds, rootLocation, endpointRoot, partial);
+
+  const char* const endpoint = partial ? "/partial_entries" : "/entries";
+  const kainjow::mustache::object template_data{
+     {"date", gen_date_str()},
+     {"endpoint_root", endpointRoot},
+     {"feed_id", gen_uuid(libraryId + endpoint + "?" + query)},
+     {"filter", onlyAsNonEmptyMustacheValue(query)},
+     {"query", query.empty() ? "" : "?" + urlEncode(query)},
+     {"totalResults", to_string(m_totalResults)},
+     {"startIndex", to_string(m_startIndex)},
+     {"itemsPerPage", to_string(m_count)},
+     {"books", booksData },
+     {"dump_partial_entries", MustacheData(partial)}
+  };
+
+  return render_template(RESOURCE::templates::catalog_v2_entries_xml, template_data);
+}
+
+std::string OPDSDumper::dumpOPDSCompleteEntry(const std::string& bookId) const
+{
+  return getSingleBookEntryXML(library->getBookById(bookId), true, rootLocation, "", false);
+}
+
+std::string OPDSDumper::categoriesOPDSFeed() const
+{
+  const auto now = gen_date_str();
+  kainjow::mustache::list categoryData;
+  for ( const auto& category : library->getBooksCategories() ) {
+    const auto urlencodedCategoryName = urlEncode(category);
+    categoryData.push_back(kainjow::mustache::object{
+      {"name", category},
+      {"urlencoded_name",  urlencodedCategoryName},
+      {"updated", now},
+      {"id", gen_uuid(libraryId + "/categories/" + urlencodedCategoryName)}
+    });
+  }
+
+  return render_template(
+             RESOURCE::templates::catalog_v2_categories_xml,
+             kainjow::mustache::object{
+               {"date", now},
+               {"endpoint_root", rootLocation + "/catalog/v2"},
+               {"feed_id", gen_uuid(libraryId + "/categories")},
+               {"categories", categoryData }
+             }
+  );
+}
+
+std::string OPDSDumper::languagesOPDSFeed() const
+{
+  const auto now = gen_date_str();
+  kainjow::mustache::list languageData;
+  std::call_once(fillLanguagesFlag, fillLanguagesMap);
+  for ( const auto& langAndBookCount : library->getBooksLanguagesWithCounts() ) {
+    const std::string languageCode = langAndBookCount.first;
+    const int bookCount = langAndBookCount.second;
+    const auto languageSelfName = getLanguageSelfName(languageCode);
+    languageData.push_back(kainjow::mustache::object{
+      {"lang_code",  languageCode},
+      {"lang_self_name", languageSelfName},
+      {"book_count", to_string(bookCount)},
+      {"updated", now},
+      {"id", gen_uuid(libraryId + "/languages/" + languageCode)}
+    });
+  }
+
+  return render_template(
+             RESOURCE::templates::catalog_v2_languages_xml,
+             kainjow::mustache::object{
+               {"date", now},
+               {"endpoint_root", rootLocation + "/catalog/v2"},
+               {"feed_id", gen_uuid(libraryId + "/languages")},
+               {"languages", languageData }
+             }
+  );
 }
 
 }

src/reader.cpp

@@ -1,904 +0,0 @@
-/*
- * Copyright 2011 Emmanuel Engelhart <kelson@kiwix.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-#include "reader.h"
-#include <time.h>
-
-#include <zim/search.h>
-
-#include "tools/otherTools.h"
-
-inline char hi(char v)
-{
-  char hex[] = "0123456789abcdef";
-  return hex[(v >> 4) & 0xf];
-}
-
-inline char lo(char v)
-{
-  char hex[] = "0123456789abcdef";
-  return hex[v & 0xf];
-}
-
-std::string hexUUID(std::string in)
-{
-  std::ostringstream out;
-  for (unsigned n = 0; n < 4; ++n) {
-    out << hi(in[n]) << lo(in[n]);
-  }
-  out << '-';
-  for (unsigned n = 4; n < 6; ++n) {
-    out << hi(in[n]) << lo(in[n]);
-  }
-  out << '-';
-  for (unsigned n = 6; n < 8; ++n) {
-    out << hi(in[n]) << lo(in[n]);
-  }
-  out << '-';
-  for (unsigned n = 8; n < 10; ++n) {
-    out << hi(in[n]) << lo(in[n]);
-  }
-  out << '-';
-  for (unsigned n = 10; n < 16; ++n) {
-    out << hi(in[n]) << lo(in[n]);
-  }
-  std::string op = out.str();
-  return op;
-}
-
-namespace kiwix
-{
-/* Constructor */
-Reader::Reader(const string zimFilePath) : zimFileHandler(NULL)
-{
-  string tmpZimFilePath = zimFilePath;
-
-  /* Remove potential trailing zimaa */
-  size_t found = tmpZimFilePath.rfind("zimaa");
-  if (found != string::npos && tmpZimFilePath.size() > 5
-      && found == tmpZimFilePath.size() - 5) {
-    tmpZimFilePath.resize(tmpZimFilePath.size() - 2);
-  }
-
-  this->zimFileHandler = new zim::File(tmpZimFilePath);
-
-  if (this->zimFileHandler != NULL) {
-    this->firstArticleOffset
-        = this->zimFileHandler->getNamespaceBeginOffset('A');
-    this->lastArticleOffset = this->zimFileHandler->getNamespaceEndOffset('A');
-    this->nsACount = this->zimFileHandler->getNamespaceCount('A');
-    this->nsICount = this->zimFileHandler->getNamespaceCount('I');
-    this->zimFilePath = zimFilePath;
-  }
-
-  /* initialize random seed: */
-  srand(time(NULL));
-}
-
-/* Destructor */
-Reader::~Reader()
-{
-  if (this->zimFileHandler != NULL) {
-    delete this->zimFileHandler;
-  }
-}
-
-zim::File* Reader::getZimFileHandler() const
-{
-  return this->zimFileHandler;
-}
-std::map<const std::string, unsigned int> Reader::parseCounterMetadata() const
-{
-  std::map<const std::string, unsigned int> counters;
-  string mimeType, item, counterString;
-  unsigned int counter;
-
-  zim::Article article = this->zimFileHandler->getArticle('M', "Counter");
-
-  if (article.good()) {
-    stringstream ssContent(article.getData());
-
-    while (getline(ssContent, item, ';')) {
-      stringstream ssItem(item);
-      getline(ssItem, mimeType, '=');
-      getline(ssItem, counterString, '=');
-      if (!counterString.empty() && !mimeType.empty()) {
-        sscanf(counterString.c_str(), "%u", &counter);
-        counters.insert(pair<string, int>(mimeType, counter));
-      }
-    }
-  }
-
-  return counters;
-}
-
-/* Get the count of articles which can be indexed/displayed */
-unsigned int Reader::getArticleCount() const
-{
-  std::map<const std::string, unsigned int> counterMap
-      = this->parseCounterMetadata();
-  unsigned int counter = 0;
-
-  if (counterMap.empty()) {
-    counter = this->nsACount;
-  } else {
-    auto it = counterMap.find("text/html");
-    if (it != counterMap.end()) {
-      counter = it->second;
-    }
-  }
-
-  return counter;
-}
-
-/* Get the count of medias content in the ZIM file */
-unsigned int Reader::getMediaCount() const
-{
-  std::map<const std::string, unsigned int> counterMap
-      = this->parseCounterMetadata();
-  unsigned int counter = 0;
-
-  if (counterMap.empty()) {
-    counter = this->nsICount;
-  } else {
-    auto it = counterMap.find("image/jpeg");
-    if (it != counterMap.end()) {
-      counter += it->second;
-    }
-
-    it = counterMap.find("image/gif");
-    if (it != counterMap.end()) {
-      counter += it->second;
-    }
-
-    it = counterMap.find("image/png");
-    if (it != counterMap.end()) {
-      counter += it->second;
-    }
-  }
-  return counter;
-}
-
-/* Get the total of all items of a ZIM file, redirects included */
-unsigned int Reader::getGlobalCount() const
-{
-  return this->zimFileHandler->getCountArticles();
-}
-
-/* Return the UID of the ZIM file */
-string Reader::getId() const
-{
-  std::ostringstream s;
-  s << this->zimFileHandler->getFileheader().getUuid();
-  return s.str();
-}
-
-/* Return a page url from a title */
-bool Reader::getPageUrlFromTitle(const string& title, string& url) const
-{
-  try {
-    auto entry = getEntryFromTitle(title);
-    entry = entry.getFinalEntry();
-    url = entry.getPath();
-    return true;
-  } catch (NoEntry& e) {
-    return false;
-  }
-}
-
-/* Return an URL from a title */
-string Reader::getRandomPageUrl() const
-{
-  return getRandomPage().getPath();
-}
-
-Entry Reader::getRandomPage() const
-{
-  if (!this->zimFileHandler) {
-    throw NoEntry();
-  }
-
-  zim::Article article;
-  std::string mainPagePath = this->getMainPage().getPath();
-  int watchdog = 42;
-
-  do {
-    auto idx = this->firstArticleOffset
-          + (zim::size_type)((double)rand() / ((double)RAND_MAX + 1)
-                             * this->nsACount);
-    article = zimFileHandler->getArticle(idx);
-    if (!watchdog--) {
-      throw NoEntry();
-    }
-  } while (!article.good() && article.getLongUrl() == mainPagePath);
-
-  return article;
-}
-
-/* Return the welcome page URL */
-string Reader::getMainPageUrl() const
-{
-  return getMainPage().getPath();
-}
-
-Entry Reader::getMainPage() const
-{
-  if (!this->zimFileHandler) {
-    throw NoEntry();
-  }
-
-  zim::Article article;
-  if (this->zimFileHandler->getFileheader().hasMainPage())
-  {
-    article = zimFileHandler->getArticle(
-        this->zimFileHandler->getFileheader().getMainPage());
-  }
-
-  if (!article.good())
-  {
-    return getFirstPage();
-  }
-
-  return article;
-}
-
-bool Reader::getFavicon(string& content, string& mimeType) const
-{
-  static const char* const paths[] = {"-/favicon", "-/favicon.png", "I/favicon.png", "I/favicon"};
-
-  for (auto &path: paths) {
-    try {
-      auto entry = getEntryFromPath(path);
-      entry = entry.getFinalEntry();
-      content = entry.getContent();
-      mimeType = entry.getMimetype();
-      return true;
-    } catch(NoEntry& e) {};
-  }
-
-  return false;
-}
-
-string Reader::getZimFilePath() const
-{
-  return this->zimFilePath;
-}
-/* Return a metatag value */
-bool Reader::getMetadata(const string& name, string& value) const
-{
-  try {
-    auto entry = getEntryFromPath("M/"+name);
-    value = entry.getContent();
-    return true;
-  } catch(NoEntry& e) {
-    return false;
-  }
-}
-
-#define METADATA(NAME) std::string v; getMetadata(NAME, v); return v;
-
-string Reader::getName() const
-{
-  METADATA("Name")
-}
-
-string Reader::getTitle() const
-{
-  string value;
-  this->getMetadata("Title", value);
-  if (value.empty()) {
-    value = getLastPathElement(zimFileHandler->getFilename());
-    std::replace(value.begin(), value.end(), '_', ' ');
-    size_t pos = value.find(".zim");
-    value = value.substr(0, pos);
-  }
-  return value;
-}
-
-string Reader::getCreator() const
-{
-  METADATA("Creator")
-}
-
-string Reader::getPublisher() const
-{
-  METADATA("Publisher")
-}
-
-string Reader::getDate() const
-{
-  METADATA("Date")
-}
-
-string Reader::getDescription() const
-{
-  string value;
-  this->getMetadata("Description", value);
-
-  /* Mediawiki Collection tends to use the "Subtitle" name */
-  if (value.empty()) {
-    this->getMetadata("Subtitle", value);
-  }
-
-  return value;
-}
-
-string Reader::getLongDescription() const
-{
-  METADATA("LongDescription")
-}
-
-string Reader::getLanguage() const
-{
-  METADATA("Language")
-}
-
-string Reader::getLicense() const
-{
-  METADATA("License")
-}
-
-string Reader::getTags(bool original) const
-{
-  string tags_str;
-  getMetadata("Tags", tags_str);
-  if (original) {
-    return tags_str;
-  }
-  auto tags = convertTags(tags_str);
-  return join(tags, ";");
-}
-
-
-string Reader::getTagStr(const std::string& tagName) const
-{
-  string tags_str;
-  getMetadata("Tags", tags_str);
-  return getTagValueFromTagList(convertTags(tags_str), tagName);
-}
-
-bool Reader::getTagBool(const std::string& tagName) const
-{
-  return convertStrToBool(getTagStr(tagName));
-}
-
-string Reader::getRelation() const
-{
-  METADATA("Relation")
-}
-
-string Reader::getFlavour() const
-{
-  METADATA("Flavour")
-}
-
-string Reader::getSource() const
-{
-  METADATA("Source")
-}
-
-string Reader::getScraper() const
-{
-  METADATA("Scraper")
-}
-#undef METADATA
-
-string Reader::getOrigId() const
-{
-  string value;
-  this->getMetadata("startfileuid", value);
-  if (value.empty()) {
-    return "";
-  }
-  std::string id = value;
-  std::string origID;
-  std::string temp = "";
-  unsigned int k = 0;
-  char tempArray[16] = "";
-  for (unsigned int i = 0; i < id.size(); i++) {
-    if (id[i] == '\n') {
-      tempArray[k] = atoi(temp.c_str());
-      temp = "";
-      k++;
-    } else {
-      temp += id[i];
-    }
-  }
-  origID = hexUUID(tempArray);
-  return origID;
-}
-
-/* Return the first page URL */
-string Reader::getFirstPageUrl() const
-{
-  return getFirstPage().getPath();
-}
-
-Entry Reader::getFirstPage() const
-{
-  if (!this->zimFileHandler) {
-    throw NoEntry();
-  }
-
-  auto firstPageOffset = zimFileHandler->getNamespaceBeginOffset('A');
-  auto article = zimFileHandler->getArticle(firstPageOffset);
-
-  if (! article.good()) {
-    throw NoEntry();
-  }
-
-  return article;
-}
-
-bool _parseUrl(const string& url, char* ns, string& title)
-{
-  /* Offset to visit the url */
-  unsigned int urlLength = url.size();
-  unsigned int offset = 0;
-
-  /* Ignore the first '/' */
-  if (url[offset] == '/')
-    offset++;
-
-  if (url[offset] == '/' || offset >= urlLength)
-    return false;
-
-  /* Get namespace */
-  *ns = url[offset++];
-
-  if (url[offset] != '/' || offset >= urlLength)
-    return false;
-
-  offset++;
-
-  if ( offset >= urlLength)
-    return false;
-
-  /* Get content title */
-  title = url.substr(offset, urlLength - offset);
-
-  return true;
-}
-
-bool Reader::parseUrl(const string& url, char* ns, string& title) const
-{
-  return _parseUrl(url, ns, title);
-}
-
-Entry Reader::getEntryFromPath(const std::string& path) const
-{
-  char ns = 0;
-  std::string short_url;
-
-  if (!this->zimFileHandler) {
-    throw NoEntry();
-  }
-  _parseUrl(path, &ns, short_url);
-
-  if (short_url.empty() && ns == 0) {
-    return getMainPage();
-  }
-
-  auto article = zimFileHandler->getArticle(ns, short_url);
-  if (!article.good()) {
-    throw NoEntry();
-  }
-
-  return article;
-}
-
-Entry Reader::getEntryFromEncodedPath(const std::string& path) const
-{
-  return getEntryFromPath(urlDecode(path, true));
-}
-
-Entry Reader::getEntryFromTitle(const std::string& title) const
-{
-  if (!this->zimFileHandler) {
-    throw NoEntry();
-  }
-
-  auto article = this->zimFileHandler->getArticleByTitle('A', title);
-  if (!article.good()) {
-    throw NoEntry();
-  }
-
-  return article;
-}
-
-/* Return article by url */
-bool Reader::getArticleObjectByDecodedUrl(const string& url,
-                                          zim::Article& article) const
-{
-  if (this->zimFileHandler == NULL) {
-    return false;
-  }
-
-  /* Parse the url */
-  char ns = 0;
-  string urlStr;
-  _parseUrl(url, &ns, urlStr);
-
-  /* Main page */
-  if (urlStr.empty() && ns == 0) {
-    _parseUrl(this->getMainPage().getPath(), &ns, urlStr);
-  }
-
-  /* Extract the content from the zim file */
-  article = zimFileHandler->getArticle(ns, urlStr);
-  return article.good();
-}
-
-/* Return the mimeType without the content */
-bool Reader::getMimeTypeByUrl(const string& url, string& mimeType) const
-{
-  try {
-    auto entry = getEntryFromPath(url);
-    mimeType = entry.getMimetype();
-    return true;
-  } catch (NoEntry& e) {
-    mimeType = "";
-    return false;
-  }
-}
-
-bool get_content_by_decoded_url(const Reader& reader,
-                                const string& url,
-                                string& content,
-                                string& title,
-                                unsigned int& contentLength,
-                                string& contentType,
-                                string& baseUrl)
-{
-  content = "";
-  contentType = "";
-  contentLength = 0;
-
-  try {
-    auto entry = reader.getEntryFromPath(url);
-    entry = entry.getFinalEntry();
-    baseUrl = entry.getPath();
-    contentType = entry.getMimetype();
-    content = entry.getContent();
-    contentLength = entry.getSize();
-    title = entry.getTitle();
-
-    /* Try to set a stub HTML header/footer if necesssary */
-    if (contentType.find("text/html") != string::npos
-      && content.find("<body") == std::string::npos
-      && content.find("<BODY") == std::string::npos) {
-      content = "<html><head><title>" + title +
-              "</title><meta http-equiv=\"Content-Type\" content=\"text/html; "
-              "charset=utf-8\" /></head><body>" +
-              content + "</body></html>";
-    }
-    return true;
-  } catch (NoEntry& e) {
-    return false;
-  }
-}
-
-
-/* Get a content from a zim file */
-bool Reader::getContentByUrl(const string& url,
-                             string& content,
-                             string& title,
-                             unsigned int& contentLength,
-                             string& contentType) const
-{
-  std::string stubRedirectUrl;
-  return get_content_by_decoded_url(*this,
-                                kiwix::urlDecode(url),
-                                content,
-                                title,
-                                contentLength,
-                                contentType,
-                                stubRedirectUrl);
-}
-
-bool Reader::getContentByEncodedUrl(const string& url,
-                                    string& content,
-                                    string& title,
-                                    unsigned int& contentLength,
-                                    string& contentType,
-                                    string& baseUrl) const
-{
-  return get_content_by_decoded_url(*this,
-                                kiwix::urlDecode(url),
-                                content,
-                                title,
-                                contentLength,
-                                contentType,
-                                baseUrl);
-}
-
-bool Reader::getContentByEncodedUrl(const string& url,
-                                    string& content,
-                                    string& title,
-                                    unsigned int& contentLength,
-                                    string& contentType) const
-{
-  std::string stubRedirectUrl;
-  return get_content_by_decoded_url(*this,
-                                kiwix::urlDecode(url),
-                                content,
-                                title,
-                                contentLength,
-                                contentType,
-                                stubRedirectUrl);
-}
-
-bool Reader::getContentByDecodedUrl(const string& url,
-                                    string& content,
-                                    string& title,
-                                    unsigned int& contentLength,
-                                    string& contentType) const
-{
-  std::string stubRedirectUrl;
-  return get_content_by_decoded_url(*this,
-                                url,
-                                content,
-                                title,
-                                contentLength,
-                                contentType,
-                                stubRedirectUrl);
-}
-
-bool Reader::getContentByDecodedUrl(const string& url,
-                                    string& content,
-                                    string& title,
-                                    unsigned int& contentLength,
-                                    string& contentType,
-                                    string& baseUrl) const
-{
-  return get_content_by_decoded_url(*this,
-                                url,
-                                content,
-                                title,
-                                contentLength,
-                                contentType,
-                                baseUrl);
-}
-
-/* Check if an article exists */
-bool Reader::urlExists(const string& url) const
-{
-  return pathExists(url);
-}
-
-bool Reader::pathExists(const string& path) const
-{
-  if (!zimFileHandler)
-  {
-    return false;
-  }
-
-  char ns = 0;
-  string titleStr;
-  _parseUrl(path, &ns, titleStr);
-  zim::File::const_iterator findItr = zimFileHandler->find(ns, titleStr);
-  return findItr != zimFileHandler->end() && findItr->getUrl() == titleStr;
-}
-
-/* Does the ZIM file has a fulltext index */
-bool Reader::hasFulltextIndex() const
-{
-  if (!zimFileHandler || zimFileHandler->is_multiPart() )
-  {
-    return false;
-  }
-
-  return ( pathExists("Z//fulltextIndex/xapian")
-        || pathExists("X/fulltext/xapian"));
-}
-
-/* Search titles by prefix */
-bool Reader::searchSuggestions(const string& prefix,
-                               unsigned int suggestionsCount,
-                               const bool reset)
-{
-  bool retVal = false;
-
-  /* Reset the suggestions otherwise check if the suggestions number is less
-   * than the suggestionsCount */
-  if (reset) {
-    this->suggestions.clear();
-    this->suggestionsOffset = this->suggestions.begin();
-  } else {
-    if (this->suggestions.size() > suggestionsCount) {
-      return false;
-    }
-  }
-
-  /* Return if no prefix */
-  if (prefix.size() == 0) {
-    return false;
-  }
-
-  for (auto articleItr = zimFileHandler->findByTitle('A', prefix);
-       articleItr != zimFileHandler->end()
-       && articleItr->getTitle().compare(0, prefix.size(), prefix) == 0
-       && this->suggestions.size() < suggestionsCount;
-       ++articleItr) {
-    /* Extract the interesting part of article title & url */
-    std::string normalizedArticleTitle
-        = kiwix::normalize(articleItr->getTitle());
-    std::string articleFinalUrl = "/A/" + articleItr->getUrl();
-    if (articleItr->isRedirect()) {
-      zim::Article article = *articleItr;
-      unsigned int loopCounter = 0;
-      while (article.isRedirect() && loopCounter++ < 42) {
-        article = article.getRedirectArticle();
-      }
-      articleFinalUrl = "/A/" + article.getUrl();
-    }
-
-    /* Go through all already found suggestions and skip if this
-       article is already in the suggestions list (with an other
-       title) */
-    bool insert = true;
-    std::vector<std::vector<std::string>>::iterator suggestionItr;
-    for (suggestionItr = this->suggestions.begin();
-         suggestionItr != this->suggestions.end();
-         suggestionItr++) {
-      int result = normalizedArticleTitle.compare((*suggestionItr)[2]);
-      if (result == 0 && articleFinalUrl.compare((*suggestionItr)[1]) == 0) {
-        insert = false;
-        break;
-      } else if (result < 0) {
-        break;
-      }
-    }
-
-    /* Insert if possible */
-    if (insert) {
-      std::vector<std::string> suggestion;
-      suggestion.push_back(articleItr->getTitle());
-      suggestion.push_back(articleFinalUrl);
-      suggestion.push_back(normalizedArticleTitle);
-      this->suggestions.insert(suggestionItr, suggestion);
-    }
-
-    /* Suggestions where found */
-    retVal = true;
-  }
-
-  /* Set the cursor to the begining */
-  this->suggestionsOffset = this->suggestions.begin();
-
-  return retVal;
-}
-
-std::vector<std::string> Reader::getTitleVariants(
-    const std::string& title) const
-{
-  std::vector<std::string> variants;
-  variants.push_back(title);
-  variants.push_back(kiwix::ucFirst(title));
-  variants.push_back(kiwix::lcFirst(title));
-  variants.push_back(kiwix::toTitle(title));
-  return variants;
-}
-
-/* Try also a few variations of the prefix to have better results */
-bool Reader::searchSuggestionsSmart(const string& prefix,
-                                    unsigned int suggestionsCount)
-{
-  std::vector<std::string> variants = this->getTitleVariants(prefix);
-  bool retVal = false;
-
-  this->suggestions.clear();
-  this->suggestionsOffset = this->suggestions.begin();
-  /* Try to search in the title using fulltext search database */
-  const auto suggestionSearch
-      = this->getZimFileHandler()->suggestions(prefix, 0, suggestionsCount);
-
-  if (suggestionSearch->get_matches_estimated()) {
-    for (auto current = suggestionSearch->begin();
-         current != suggestionSearch->end();
-         current++) {
-      if (!current->good()) {
-          continue;
-      }
-      std::vector<std::string> suggestion;
-      suggestion.push_back(current->getTitle());
-      suggestion.push_back("/A/" + current->getUrl());
-      suggestion.push_back(kiwix::normalize(current->getTitle()));
-      this->suggestions.push_back(suggestion);
-    }
-    this->suggestionsOffset = this->suggestions.begin();
-    retVal = true;
-  } else {
-    for (std::vector<std::string>::iterator variantsItr = variants.begin();
-         variantsItr != variants.end();
-         variantsItr++) {
-      retVal = this->searchSuggestions(*variantsItr, suggestionsCount, false)
-               || retVal;
-    }
-  }
-
-  return retVal;
-}
-
-/* Get next suggestion */
-bool Reader::getNextSuggestion(string& title)
-{
-  if (this->suggestionsOffset != this->suggestions.end()) {
-    /* title */
-    title = (*(this->suggestionsOffset))[0];
-
-    /* increment the cursor for the next call */
-    this->suggestionsOffset++;
-
-    return true;
-  }
-
-  return false;
-}
-
-bool Reader::getNextSuggestion(string& title, string& url)
-{
-  if (this->suggestionsOffset != this->suggestions.end()) {
-    /* title */
-    title = (*(this->suggestionsOffset))[0];
-    url = (*(this->suggestionsOffset))[1];
-
-    /* increment the cursor for the next call */
-    this->suggestionsOffset++;
-
-    return true;
-  }
-
-  return false;
-}
-
-/* Check if the file has as checksum */
-bool Reader::canCheckIntegrity() const
-{
-  return this->zimFileHandler->getChecksum() != "";
-}
-
-/* Return true if corrupted, false otherwise */
-bool Reader::isCorrupted() const
-{
-  try {
-    if (this->zimFileHandler->verify() == true) {
-      return false;
-    }
-  } catch (exception& e) {
-    cerr << e.what() << endl;
-    return true;
-  }
-
-  return true;
-}
-
-/* Return the file size, works also for splitted files */
-unsigned int Reader::getFileSize() const
-{
-  zim::File* file = this->getZimFileHandler();
-  zim::size_type size = 0;
-
-  if (file != NULL) {
-    size = file->getFilesize();
-  }
-
-  return (size / 1024);
-}
-}

src/search_renderer.cpp

@@ -21,26 +21,35 @@
 #include <cmath>
 
 #include "search_renderer.h"
-#include "searcher.h"
-#include "reader.h"
 #include "library.h"
 #include "name_mapper.h"
 
+#include "tools/archiveTools.h"
+
 #include <zim/search.h>
 
 #include <mustache.hpp>
 #include "kiwixlib-resources.h"
-
+#include "tools/stringTools.h"
 
 namespace kiwix
 {
 
 /* Constructor */
-SearchRenderer::SearchRenderer(Searcher* searcher, NameMapper* mapper)
-    : mp_searcher(searcher),
+SearchRenderer::SearchRenderer(zim::SearchResultSet srs, NameMapper* mapper,
+                      unsigned int start, unsigned int estimatedResultCount)
+    : SearchRenderer(srs, mapper, nullptr, start, estimatedResultCount)
+{}
+
+SearchRenderer::SearchRenderer(zim::SearchResultSet srs, NameMapper* mapper, Library* library,
+                      unsigned int start, unsigned int estimatedResultCount)
+    : m_srs(srs),
       mp_nameMapper(mapper),
+      mp_library(library),
       protocolPrefix("zim://"),
-      searchProtocolPrefix("search://?")
+      searchProtocolPrefix("search://"),
+      estimatedResultCount(estimatedResultCount),
+      resultStart(start)
 {}
 
 /* Destructor */
@@ -48,12 +57,12 @@ SearchRenderer::~SearchRenderer() = default;
 
 void SearchRenderer::setSearchPattern(const std::string& pattern)
 {
-  this->searchPattern = pattern;
+  searchPattern = pattern;
 }
 
-void SearchRenderer::setSearchContent(const std::string& name)
+void SearchRenderer::setSearchBookQuery(const std::string& bookQuery)
 {
-  this->searchContent = name;
+  searchBookQuery = bookQuery;
 }
 
 void SearchRenderer::setProtocolPrefix(const std::string& prefix)
@@ -66,87 +75,161 @@ void SearchRenderer::setSearchProtocolPrefix(const std::string& prefix)
   this->searchProtocolPrefix = prefix;
 }
 
-std::string SearchRenderer::getHtml()
-{
-  kainjow::mustache::data results{kainjow::mustache::data::type::list};
-
-  mp_searcher->restart_search();
-  Result* p_result = NULL;
-  while ((p_result = mp_searcher->getNextResult())) {
-    kainjow::mustache::data result;
-    result.set("title", p_result->get_title());
-    result.set("url", p_result->get_url());
-    result.set("snippet", p_result->get_snippet());
-    auto readerIndex = p_result->get_readerIndex();
-    auto reader = mp_searcher->get_reader(readerIndex);
-    result.set("resultContentId", mp_nameMapper->getNameForId(reader->getId()));
-
-    if (p_result->get_wordCount() >= 0) {
-      result.set("wordCount", kiwix::beautifyInteger(p_result->get_wordCount()));
-    }
-
-    results.push_back(result);
-    delete p_result;
+std::string extractValueFromQuery(const std::string& query, const std::string& key) {
+  const std::string p = key + "=";
+  const size_t i = query.find(p);
+  if (i == std::string::npos) {
+    return "";
   }
+  std::string r = query.substr(i + p.size());
+  return r.substr(0, r.find("&"));
+}
 
-  // pages
+kainjow::mustache::data buildQueryData
+(
+  const std::string& searchProtocolPrefix,
+  const std::string& pattern,
+  const std::string& bookQuery
+) {
+  kainjow::mustache::data query;
+  query.set("pattern", kiwix::encodeDiples(pattern));
+  std::ostringstream ss;
+  ss << searchProtocolPrefix << "?pattern=" << urlEncode(pattern, true);
+  ss << "&" << bookQuery;
+  query.set("unpaginatedQuery", ss.str());
+  auto lang = extractValueFromQuery(bookQuery, "books.filter.lang");
+  if(!lang.empty()) {
+    query.set("lang", lang);
+  }
+  return query;
+}
+
+kainjow::mustache::data buildPagination(
+  unsigned int pageLength,
+  unsigned int resultsCount,
+  unsigned int resultsStart
+)
+{
+  assert(pageLength!=0);
+  kainjow::mustache::data pagination;
   kainjow::mustache::data pages{kainjow::mustache::data::type::list};
 
-  auto resultStart = mp_searcher->getResultStart();
-  auto resultEnd = mp_searcher->getResultEnd();
-  auto resultCountPerPage = resultEnd - resultStart;
-  auto estimatedResultCount = mp_searcher->getEstimatedResultCount();
-  auto currentPage = 0U;
-  auto pageStart = 0U;
-  auto pageEnd = 0U;
-  auto lastPageStart = 0U;
-  if (resultCountPerPage) {
-    currentPage = resultStart/resultCountPerPage;
-    pageStart = currentPage > 4 ? currentPage-4 : 0;
-    pageEnd = currentPage + 5;
-    if (pageEnd > estimatedResultCount / resultCountPerPage) {
-      pageEnd = estimatedResultCount / resultCountPerPage;
+  if (resultsCount == 0) {
+    // Easy case
+    pagination.set("itemsPerPage", to_string(pageLength));
+    pagination.set("hasPages", false);
+    pagination.set("pages", pages);
+    return pagination;
+  }
+
+  // First we want to display pages starting at a multiple of `pageLength`
+  // so, let's calculate the start index of the current page.
+  auto currentPage = resultsStart/pageLength;
+  auto lastPage = ((resultsCount-1)/pageLength);
+  auto lastPageStart = lastPage*pageLength;
+  auto nbPages = lastPage + 1;
+
+  auto firstPageGenerated = currentPage > 4 ? currentPage-4 : 0;
+  auto lastPageGenerated = std::min(currentPage+4, lastPage);
+
+  if (nbPages != 1) {
+    if (firstPageGenerated!=0) {
+      kainjow::mustache::data page;
+      page.set("label", "◀");
+      page.set("start", to_string(0));
+      page.set("current", false);
+      pages.push_back(page);
     }
-    if (estimatedResultCount > resultCountPerPage) {
-      lastPageStart = static_cast<int>(round(estimatedResultCount/resultCountPerPage)) * resultCountPerPage;
+
+    for (auto i=firstPageGenerated; i<=lastPageGenerated; i++) {
+      kainjow::mustache::data page;
+      page.set("label", to_string(i+1));
+      page.set("start", to_string(i*pageLength));
+      page.set("current", bool(i == currentPage));
+      pages.push_back(page);
+    }
+
+    if (lastPageGenerated!=lastPage) {
+      kainjow::mustache::data page;
+      page.set("label", "▶");
+      page.set("start", to_string(lastPageStart));
+      page.set("current", false);
+      pages.push_back(page);
     }
   }
 
-  for (unsigned int i = pageStart; i < pageEnd; i++) {
-    kainjow::mustache::data page;
-    page.set("label", to_string(i + 1));
-    page.set("start", to_string(i * resultCountPerPage));
-    page.set("end", to_string((i + 1) * resultCountPerPage));
+  pagination.set("itemsPerPage", to_string(pageLength));
+  pagination.set("hasPages", firstPageGenerated < lastPageGenerated);
+  pagination.set("pages", pages);
+  return pagination;
+}
 
-    if (i == currentPage) {
-      page.set("selected", true);
+std::string SearchRenderer::renderTemplate(const std::string& tmpl_str)
+{
+  const std::string absPathPrefix = protocolPrefix + "content/";
+  // Build the results list
+  kainjow::mustache::data items{kainjow::mustache::data::type::list};
+  for (auto it = m_srs.begin(); it != m_srs.end(); it++) {
+    kainjow::mustache::data result;
+    std::string zim_id(it.getZimId());
+    result.set("title", it.getTitle());
+    result.set("absolutePath", absPathPrefix + urlEncode(mp_nameMapper->getNameForId(zim_id), true) + "/" + urlEncode(it.getPath()));
+    result.set("snippet", it.getSnippet());
+    if (mp_library) {
+      result.set("bookTitle", mp_library->getBookById(zim_id).getTitle());
+    }
+    if (it.getWordCount() >= 0) {
+      result.set("wordCount", kiwix::beautifyInteger(it.getWordCount()));
     }
-    pages.push_back(page);
-  }
 
-  std::string template_str = RESOURCE::templates::search_result_html;
-  kainjow::mustache::mustache tmpl(template_str);
+    items.push_back(result);
+  }
+  kainjow::mustache::data results;
+  results.set("items", items);
+  results.set("count", kiwix::beautifyInteger(estimatedResultCount));
+  results.set("hasResults", estimatedResultCount != 0);
+  results.set("start", kiwix::beautifyInteger(resultStart));
+  results.set("end", kiwix::beautifyInteger(std::min(resultStart+pageLength-1, estimatedResultCount)));
+
+  // pagination
+  auto pagination = buildPagination(
+    pageLength,
+    estimatedResultCount,
+    resultStart
+  );
+
+  kainjow::mustache::data query = buildQueryData(
+    searchProtocolPrefix,
+    searchPattern,
+    searchBookQuery
+  );
+
 
   kainjow::mustache::data allData;
+  allData.set("protocolPrefix", protocolPrefix);
   allData.set("results", results);
-  allData.set("pages", pages);
-  allData.set("hasResults", estimatedResultCount != 0);
-  allData.set("hasPages", pageStart != pageEnd);
-  allData.set("count", kiwix::beautifyInteger(estimatedResultCount));
-  allData.set("searchPattern", kiwix::encodeDiples(this->searchPattern));
-  allData.set("searchPatternEncoded", urlEncode(this->searchPattern));
-  allData.set("resultStart", to_string(resultStart + 1));
-  allData.set("resultEnd", to_string(min(resultEnd, estimatedResultCount)));
-  allData.set("resultRange", to_string(resultCountPerPage));
-  allData.set("resultLastPageStart", to_string(lastPageStart));
-  allData.set("lastResult", to_string(estimatedResultCount));
-  allData.set("protocolPrefix", this->protocolPrefix);
-  allData.set("searchProtocolPrefix", this->searchProtocolPrefix);
-  allData.set("contentId", this->searchContent);
+  allData.set("pagination", pagination);
+  allData.set("query", query);
+
+  kainjow::mustache::mustache tmpl(tmpl_str);
 
   std::stringstream ss;
   tmpl.render(allData, [&ss](const std::string& str) { ss << str; });
+  if (!tmpl.is_valid()) {
+    throw std::runtime_error("Error while rendering search results: " + tmpl.error_message());
+  }
   return ss.str();
 }
 
+std::string SearchRenderer::getHtml()
+{
+  return renderTemplate(RESOURCE::templates::search_result_html);
+}
+
+std::string SearchRenderer::getXml()
+{
+  return renderTemplate(RESOURCE::templates::search_result_xml);
+}
+
+
 }

src/searcher.cpp

@@ -1,279 +0,0 @@
-/*
- * Copyright 2011 Emmanuel Engelhart <kelson@kiwix.org>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-
-#include <cmath>
-
-#include "searcher.h"
-#include "reader.h"
-
-#include <zim/search.h>
-
-#include <mustache.hpp>
-#include "kiwixlib-resources.h"
-
-#define MAX_SEARCH_LEN 140
-
-namespace kiwix
-{
-class _Result : public Result
-{
- public:
-  _Result(zim::Search::iterator& iterator);
-  virtual ~_Result(){};
-
-  virtual std::string get_url();
-  virtual std::string get_title();
-  virtual int get_score();
-  virtual std::string get_snippet();
-  virtual std::string get_content();
-  virtual int get_wordCount();
-  virtual int get_size();
-  virtual int get_readerIndex();
-
- private:
-  zim::Search::iterator iterator;
-};
-
-struct SearcherInternal {
-  const zim::Search* _search;
-  zim::Search::iterator current_iterator;
-
-  SearcherInternal() : _search(NULL) {}
-  ~SearcherInternal()
-  {
-    if (_search != NULL) {
-      delete _search;
-    }
-  }
-};
-
-/* Constructor */
-Searcher::Searcher()
-    : internal(new SearcherInternal()),
-      searchPattern(""),
-      estimatedResultCount(0),
-      resultStart(0),
-      resultEnd(0)
-{
-  loadICUExternalTables();
-}
-
-/* Destructor */
-Searcher::~Searcher()
-{
-  delete internal;
-}
-
-bool Searcher::add_reader(Reader* reader)
-{
-  if (!reader->hasFulltextIndex()) {
-      return false;
-  }
-  this->readers.push_back(reader);
-  return true;
-}
-
-
-Reader* Searcher::get_reader(int readerIndex)
-{
-  return readers.at(readerIndex);
-}
-
-/* Search strings in the database */
-void Searcher::search(const std::string& search,
-                      unsigned int resultStart,
-                      unsigned int resultEnd,
-                      const bool verbose)
-{
-  this->reset();
-
-  if (verbose == true) {
-    cout << "Performing query `" << search << "'" << endl;
-  }
-
-  this->searchPattern = search;
-  this->resultStart = resultStart;
-  this->resultEnd = resultEnd;
-  /* Try to find results */
-  if (resultStart != resultEnd) {
-    /* Perform the search */
-    string unaccentedSearch = removeAccents(search);
-    std::vector<const zim::File*> zims;
-    for (auto current = this->readers.begin(); current != this->readers.end();
-         current++) {
-      if ( (*current)->hasFulltextIndex() ) {
-          zims.push_back((*current)->getZimFileHandler());
-      }
-    }
-    zim::Search* search = new zim::Search(zims);
-    search->set_verbose(verbose);
-    search->set_query(unaccentedSearch);
-    search->set_range(resultStart, resultEnd);
-    internal->_search = search;
-    internal->current_iterator = internal->_search->begin();
-    this->estimatedResultCount = internal->_search->get_matches_estimated();
-  }
-
-  return;
-}
-
-
-void Searcher::geo_search(float latitude, float longitude, float distance,
-                          unsigned int resultStart,
-                          unsigned int resultEnd,
-                          const bool verbose)
-{
-  this->reset();
-
-  if (verbose == true) {
-    cout << "Performing geo query `" << distance << "&(" << latitude << ";" << longitude << ")'" << endl;
-  }
-
-  /* Perform the search */
-  std::ostringstream oss;
-  oss << "Articles located less than " << distance << " meters of " << latitude << ";" << longitude;
-  this->searchPattern = oss.str();
-  this->resultStart = resultStart;
-  this->resultEnd = resultEnd;
-
-  /* Try to find results */
-  if (resultStart == resultEnd) {
-    return;
-  }
-
-  std::vector<const zim::File*> zims;
-  for (auto current = this->readers.begin(); current != this->readers.end();
-       current++) {
-    zims.push_back((*current)->getZimFileHandler());
-  }
-  zim::Search* search = new zim::Search(zims);
-  search->set_verbose(verbose);
-  search->set_query("");
-  search->set_georange(latitude, longitude, distance);
-  search->set_range(resultStart, resultEnd);
-  internal->_search = search;
-  internal->current_iterator = internal->_search->begin();
-  this->estimatedResultCount = internal->_search->get_matches_estimated();
-}
-
-
-void Searcher::restart_search()
-{
-  if (internal->_search) {
-    internal->current_iterator = internal->_search->begin();
-  }
-}
-
-Result* Searcher::getNextResult()
-{
-  if (internal->_search &&
-             internal->current_iterator != internal->_search->end()) {
-    Result* result = new _Result(internal->current_iterator);
-    internal->current_iterator++;
-    return result;
-  }
-  return NULL;
-}
-
-/* Reset the results */
-void Searcher::reset()
-{
-  this->estimatedResultCount = 0;
-  this->searchPattern = "";
-  return;
-}
-
-void Searcher::suggestions(std::string& searchPattern, const bool verbose)
-{
-  this->reset();
-
-  if (verbose == true) {
-    cout << "Performing suggestion query `" << searchPattern << "`" << endl;
-  }
-
-  this->searchPattern = searchPattern;
-  this->resultStart = 0;
-  this->resultEnd = 10;
-  string unaccentedSearch = removeAccents(searchPattern);
-
-  std::vector<const zim::File*> zims;
-  for (auto current = this->readers.begin(); current != this->readers.end();
-       current++) {
-    zims.push_back((*current)->getZimFileHandler());
-  }
-  zim::Search* search = new zim::Search(zims);
-  search->set_verbose(verbose);
-  search->set_query(unaccentedSearch);
-  search->set_range(resultStart, resultEnd);
-  search->set_suggestion_mode(true);
-  internal->_search = search;
-  internal->current_iterator = internal->_search->begin();
-  this->estimatedResultCount = internal->_search->get_matches_estimated();
-}
-
-/* Return the result count estimation */
-unsigned int Searcher::getEstimatedResultCount()
-{
-  return this->estimatedResultCount;
-}
-
-_Result::_Result(zim::Search::iterator& iterator)
-    : iterator(iterator)
-{
-}
-
-std::string _Result::get_url()
-{
-  return iterator.get_url();
-}
-std::string _Result::get_title()
-{
-  return iterator.get_title();
-}
-int _Result::get_score()
-{
-  return iterator.get_score();
-}
-std::string _Result::get_snippet()
-{
-  return iterator.get_snippet();
-}
-std::string _Result::get_content()
-{
-  if (iterator->good()) {
-    return iterator->getData();
-  }
-  return "";
-}
-int _Result::get_size()
-{
-  return iterator.get_size();
-}
-int _Result::get_wordCount()
-{
-  return iterator.get_wordCount();
-}
-int _Result::get_readerIndex()
-{
-  return iterator.get_fileIndex();
-}
-
-
-}

src/server.cpp

@@ -19,140 +19,16 @@
 
 #include "server.h"
 
-#ifdef _WIN32
-# if !defined(__MINGW32__) && (_MSC_VER < 1600)
-#   include "stdint4win.h"
-# endif
-# include <winsock2.h>
-# include <ws2tcpip.h>
-# ifdef __GNUC__
-  // inet_pton is not declared in mingw, even if the function exists.
-  extern "C" {
-    WINSOCK_API_LINKAGE  INT WSAAPI inet_pton( INT Family, PCSTR pszAddrString, PVOID pAddrBuf);
-  }
-# endif
-  typedef UINT64 uint64_t;
-  typedef UINT16 uint16_t;
-#endif
-
-extern "C" {
-#include <microhttpd.h>
-}
-
-#include "tools/otherTools.h"
-#include "tools/pathTools.h"
-#include "tools/regexTools.h"
-#include "tools/stringTools.h"
 #include "library.h"
 #include "name_mapper.h"
-#include "entry.h"
-#include "searcher.h"
-#include "search_renderer.h"
-#include "opds_dumper.h"
 
-#include <zim/uuid.h>
-
-#include <mustache.hpp>
-
-#include <pthread.h>
-#include <atomic>
 #include <string>
-#include <vector>
-#include <chrono>
-#include "kiwixlib-resources.h"
 
-#ifndef _WIN32
-# include <arpa/inet.h>
-#endif
-
-#include "server/request_context.h"
-#include "server/response.h"
-
-#define MAX_SEARCH_LEN 140
-#define KIWIX_MIN_CONTENT_SIZE_TO_DEFLATE 100
+#include <zim/item.h>
+#include "server/internalServer.h"
 
 namespace kiwix {
 
-static IdNameMapper defaultNameMapper;
-
-typedef kainjow::mustache::data MustacheData;
-
-static int staticHandlerCallback(void* cls,
-                                 struct MHD_Connection* connection,
-                                 const char* url,
-                                 const char* method,
-                                 const char* version,
-                                 const char* upload_data,
-                                 size_t* upload_data_size,
-                                 void** cont_cls);
-
-
-class InternalServer {
-  public:
-    InternalServer(Library* library,
-                   NameMapper* nameMapper,
-                   std::string addr,
-                   int port,
-                   std::string root,
-                   int nbThreads,
-                   bool verbose,
-                   bool withTaskbar,
-                   bool withLibraryButton,
-                   bool blockExternalLinks);
-    virtual ~InternalServer() = default;
-
-    int handlerCallback(struct MHD_Connection* connection,
-                        const char* url,
-                        const char* method,
-                        const char* version,
-                        const char* upload_data,
-                        size_t* upload_data_size,
-                        void** cont_cls);
-    bool start();
-    void stop();
-
-  private: // functions
-    Response handle_request(const RequestContext& request);
-    Response build_500(const std::string& msg);
-    Response build_404(const RequestContext& request, const std::string& zimName);
-    Response build_304(const RequestContext& request, const ETag& etag) const;
-    Response build_redirect(const std::string& bookName, const kiwix::Entry& entry) const;
-    Response build_homepage(const RequestContext& request);
-    Response handle_skin(const RequestContext& request);
-    Response handle_catalog(const RequestContext& request);
-    Response handle_meta(const RequestContext& request);
-    Response handle_search(const RequestContext& request);
-    Response handle_suggest(const RequestContext& request);
-    Response handle_random(const RequestContext& request);
-    Response handle_captured_external(const RequestContext& request);
-    Response handle_content(const RequestContext& request);
-
-    MustacheData get_default_data() const;
-    MustacheData homepage_data() const;
-    Response get_default_response() const;
-
-    std::shared_ptr<Reader> get_reader(const std::string& bookName) const;
-    bool etag_not_needed(const RequestContext& r) const;
-    ETag get_matching_if_none_match_etag(const RequestContext& request) const;
-
-  private: // data
-    std::string m_addr;
-    int m_port;
-    std::string m_root;
-    int m_nbThreads;
-    std::atomic_bool m_verbose;
-    bool m_withTaskbar;
-    bool m_withLibraryButton;
-    bool m_blockExternalLinks;
-    struct MHD_Daemon* mp_daemon;
-
-    Library* mp_library;
-    NameMapper* mp_nameMapper;
-
-    std::string m_server_id;
-};
-
-
 Server::Server(Library* library, NameMapper* nameMapper) :
   mp_library(library),
   mp_nameMapper(nameMapper),
@@ -170,16 +46,21 @@ bool Server::start() {
     m_port,
     m_root,
     m_nbThreads,
+    m_multizimSearchLimit,
     m_verbose,
     m_withTaskbar,
     m_withLibraryButton,
-    m_blockExternalLinks));
+    m_blockExternalLinks,
+    m_indexTemplateString,
+    m_ipConnectionLimit));
   return mp_server->start();
 }
 
 void Server::stop() {
-  mp_server->stop();
-  mp_server.reset(nullptr);
+  if (mp_server) {
+    mp_server->stop();
+    mp_server.reset(nullptr);
+  }
 }
 
 void Server::setRoot(const std::string& root)
@@ -193,773 +74,14 @@ void Server::setRoot(const std::string& root)
   }
 }
 
-
-InternalServer::InternalServer(Library* library,
-                               NameMapper* nameMapper,
-                               std::string addr,
-                               int port,
-                               std::string root,
-                               int nbThreads,
-                               bool verbose,
-                               bool withTaskbar,
-                               bool withLibraryButton,
-                               bool blockExternalLinks) :
-  m_addr(addr),
-  m_port(port),
-  m_root(root),
-  m_nbThreads(nbThreads),
-  m_verbose(verbose),
-  m_withTaskbar(withTaskbar),
-  m_withLibraryButton(withLibraryButton),
-  m_blockExternalLinks(blockExternalLinks),
-  mp_daemon(nullptr),
-  mp_library(library),
-  mp_nameMapper(nameMapper ? nameMapper : &defaultNameMapper)
-{}
-
-bool InternalServer::start() {
-#ifdef _WIN32
-  int flags = MHD_USE_SELECT_INTERNALLY;
-#else
-  int flags = MHD_USE_POLL_INTERNALLY;
-#endif
-  if (m_verbose.load())
-    flags |= MHD_USE_DEBUG;
-
-
-  struct sockaddr_in sockAddr;
-  memset(&sockAddr, 0, sizeof(sockAddr));
-  sockAddr.sin_family = AF_INET;
-  sockAddr.sin_port = htons(m_port);
-  if (m_addr.empty()) {
-    if (0 != INADDR_ANY)
-      sockAddr.sin_addr.s_addr = htonl(INADDR_ANY);
-  } else {
-    if (inet_pton(AF_INET, m_addr.c_str(), &(sockAddr.sin_addr.s_addr)) == 0) {
-      std::cerr << "Ip address " << m_addr << "  is not a valid ip address" << std::endl;
-      return false;
-    }
-  }
-
-  mp_daemon = MHD_start_daemon(flags,
-                            m_port,
-                            NULL,
-                            NULL,
-                            &staticHandlerCallback,
-                            this,
-                            MHD_OPTION_SOCK_ADDR, &sockAddr,
-                            MHD_OPTION_THREAD_POOL_SIZE, m_nbThreads,
-                            MHD_OPTION_END);
-  if (mp_daemon == nullptr) {
-    std::cerr << "Unable to instantiate the HTTP daemon. The port " << m_port
-              << " is maybe already occupied or need more permissions to be open. "
-                 "Please try as root or with a port number higher or equal to 1024."
-              << std::endl;
-    return false;
-  }
-  auto server_start_time = std::chrono::system_clock::now().time_since_epoch();
-  m_server_id = kiwix::to_string(server_start_time.count());
-  return true;
-}
-
-void InternalServer::stop()
+int Server::getPort()
 {
-  MHD_stop_daemon(mp_daemon);
+  return mp_server->getPort();
 }
 
-static int staticHandlerCallback(void* cls,
-                                 struct MHD_Connection* connection,
-                                 const char* url,
-                                 const char* method,
-                                 const char* version,
-                                 const char* upload_data,
-                                 size_t* upload_data_size,
-                                 void** cont_cls)
+std::string Server::getAddress()
 {
-  InternalServer* _this = static_cast<InternalServer*>(cls);
-
-  return _this->handlerCallback(connection,
-                                url,
-                                method,
-                                version,
-                                upload_data,
-                                upload_data_size,
-                                cont_cls);
-}
-
-int InternalServer::handlerCallback(struct MHD_Connection* connection,
-                                    const char* url,
-                                    const char* method,
-                                    const char* version,
-                                    const char* upload_data,
-                                    size_t* upload_data_size,
-                                    void** cont_cls)
-{
-  auto start_time = std::chrono::steady_clock::now();
-  if (m_verbose.load() ) {
-    printf("======================\n");
-    printf("Requesting : \n");
-    printf("full_url  : %s\n", url);
-  }
-  RequestContext request(connection, m_root, url, method, version);
-
-  if (m_verbose.load() ) {
-    request.print_debug_info();
-  }
-  /* Unexpected method */
-  if (request.get_method() != RequestMethod::GET
-   && request.get_method() != RequestMethod::POST
-   && request.get_method() != RequestMethod::HEAD) {
-    printf("Reject request because of unhandled request method.\n");
-    printf("----------------------\n");
-    return MHD_NO;
-  }
-
-  auto response = handle_request(request);
-
-  if (response.getReturnCode() == MHD_HTTP_INTERNAL_SERVER_ERROR) {
-    printf("========== INTERNAL ERROR !! ============\n");
-    if (!m_verbose.load()) {
-      printf("Requesting : \n");
-      printf("full_url : %s\n", url);
-      request.print_debug_info();
-    }
-  }
-
-  if (response.getReturnCode() == MHD_HTTP_OK && !etag_not_needed(request))
-    response.set_server_id(m_server_id);
-
-  auto ret = response.send(request, connection);
-  auto end_time = std::chrono::steady_clock::now();
-  auto time_span = std::chrono::duration_cast<std::chrono::duration<double>>(end_time - start_time);
-  if (m_verbose.load()) {
-    printf("Request time : %fs\n", time_span.count());
-    printf("----------------------\n");
-  }
-  return ret;
-}
-
-Response InternalServer::build_304(const RequestContext& request, const ETag& etag) const
-{
-  auto response = get_default_response();
-  response.set_code(MHD_HTTP_NOT_MODIFIED);
-  response.set_etag(etag);
-  response.set_content("");
-  return response;
-}
-
-Response InternalServer::handle_request(const RequestContext& request)
-{
-  try {
-    if (! request.is_valid_url())
-      return build_404(request, "");
-
-    const ETag etag = get_matching_if_none_match_etag(request);
-    if ( etag )
-      return build_304(request, etag);
-
-    if (kiwix::startsWith(request.get_url(), "/skin/"))
-      return handle_skin(request);
-
-    if (startsWith(request.get_url(), "/catalog"))
-      return handle_catalog(request);
-
-    if (request.get_url() == "/meta")
-      return handle_meta(request);
-
-    if (request.get_url() == "/search")
-      return handle_search(request);
-
-    if (request.get_url() == "/suggest")
-     return handle_suggest(request);
-
-    if (request.get_url() == "/random")
-      return handle_random(request);
-
-    if (request.get_url() == "/catch/external")
-      return handle_captured_external(request);
-
-    return handle_content(request);
-  } catch (std::exception& e) {
-    fprintf(stderr, "===== Unhandled error : %s\n", e.what());
-    return build_500(e.what());
-  } catch (...) {
-    fprintf(stderr, "===== Unhandled unknown error\n");
-    return build_500("Unknown error");
-  }
-}
-
-MustacheData InternalServer::get_default_data() const
-{
-  MustacheData data;
-  data.set("root", m_root);
-  return data;
-}
-
-Response InternalServer::get_default_response() const
-{
-  return Response(m_root, m_verbose.load(), m_withTaskbar, m_withLibraryButton, m_blockExternalLinks);
-}
-
-
-Response InternalServer::build_404(const RequestContext& request,
-                                   const std::string& bookName)
-{
-  MustacheData results;
-  results.set("url", request.get_full_url());
-
-  auto response = get_default_response();
-  response.set_template(RESOURCE::templates::_404_html, results);
-  response.set_mimeType("text/html");
-  response.set_code(MHD_HTTP_NOT_FOUND);
-  response.set_compress(true);
-  response.set_taskbar(bookName, "");
-
-  return response;
-}
-
-Response InternalServer::build_500(const std::string& msg)
-{
-  MustacheData data;
-  data.set("error", msg);
-  Response response(m_root, true, false, false, false);
-  response.set_template(RESOURCE::templates::_500_html, data);
-  response.set_mimeType("text/html");
-  response.set_code(MHD_HTTP_INTERNAL_SERVER_ERROR);
-  return response;
-}
-
-MustacheData InternalServer::homepage_data() const
-{
-  auto data = get_default_data();
-
-  MustacheData books{MustacheData::type::list};
-  for (auto& bookId: mp_library->filter(kiwix::Filter().local(true).valid(true))) {
-    auto& currentBook = mp_library->getBookById(bookId);
-
-    MustacheData book;
-    book.set("name", mp_nameMapper->getNameForId(bookId));
-    book.set("title", currentBook.getTitle());
-    book.set("description", currentBook.getDescription());
-    book.set("articleCount", beautifyInteger(currentBook.getArticleCount()));
-    book.set("mediaCount", beautifyInteger(currentBook.getMediaCount()));
-    books.push_back(book);
-  }
-
-  data.set("books", books);
-  return data;
-}
-
-bool InternalServer::etag_not_needed(const RequestContext& request) const
-{
-  const std::string url = request.get_url();
-  return kiwix::startsWith(url, "/catalog")
-      || url == "/search"
-      || url == "/suggest"
-      || url == "/random"
-      || url == "/catch/external";
-}
-
-ETag
-InternalServer::get_matching_if_none_match_etag(const RequestContext& r) const
-{
-  try {
-    const std::string etag_list = r.get_header(MHD_HTTP_HEADER_IF_NONE_MATCH);
-    return ETag::match(etag_list, m_server_id);
-  } catch (const std::out_of_range&) {
-    return ETag();
-  }
-}
-
-Response InternalServer::build_homepage(const RequestContext& request)
-{
-  auto response = get_default_response();
-  response.set_template(RESOURCE::templates::index_html, homepage_data());
-  response.set_mimeType("text/html; charset=utf-8");
-  response.set_compress(true);
-  response.set_taskbar("", "");
-  return response;
-}
-
-Response InternalServer::handle_meta(const RequestContext& request)
-{
-  std::string bookName;
-  std::string bookId;
-  std::string meta_name;
-  std::shared_ptr<Reader> reader;
-  try {
-    bookName = request.get_argument("content");
-    bookId = mp_nameMapper->getIdForName(bookName);
-    meta_name = request.get_argument("name");
-    reader = mp_library->getReaderById(bookId);
-  } catch (const std::out_of_range& e) {
-    return build_404(request, bookName);
-  }
-
-  if (reader == nullptr) {
-    return build_404(request, bookName);
-  }
-
-  std::string content;
-  std::string mimeType = "text";
-
-  if (meta_name == "title") {
-    content = reader->getTitle();
-  } else if (meta_name == "description") {
-    content = reader->getDescription();
-  } else if (meta_name == "language") {
-    content = reader->getLanguage();
-  } else if (meta_name == "name") {
-    content = reader->getName();
-  } else if (meta_name == "tags") {
-    content = reader->getTags();
-  } else if (meta_name == "date") {
-    content = reader->getDate();
-  } else if (meta_name == "creator") {
-    content = reader->getCreator();
-  } else if (meta_name == "publisher") {
-    content = reader->getPublisher();
-  } else if (meta_name == "favicon") {
-    reader->getFavicon(content, mimeType);
-  } else {
-    return build_404(request, bookName);
-  }
-
-  auto response = get_default_response();
-  response.set_content(content);
-  response.set_mimeType(mimeType);
-  response.set_compress(false);
-  response.set_cacheable();
-  return response;
-}
-
-Response InternalServer::handle_suggest(const RequestContext& request)
-{
-  if (m_verbose.load()) {
-    printf("** running handle_suggest\n");
-  }
-
-  std::string content;
-  std::string mimeType;
-  unsigned int maxSuggestionCount = 10;
-  unsigned int suggestionCount = 0;
-  std::string suggestion;
-
-  std::string bookName;
-  std::string bookId;
-  std::string term;
-  std::shared_ptr<Reader> reader;
-  try {
-    bookName = request.get_argument("content");
-    bookId = mp_nameMapper->getIdForName(bookName);
-    term = request.get_argument("term");
-    reader = mp_library->getReaderById(bookId);
-  } catch (const std::out_of_range&) {
-    return build_404(request, bookName);
-  }
-
-  if (m_verbose.load()) {
-    printf("Searching suggestions for: \"%s\"\n", term.c_str());
-  }
-
-  MustacheData results{MustacheData::type::list};
-
-  bool first = true;
-  if (reader != nullptr) {
-    /* Get the suggestions */
-    reader->searchSuggestionsSmart(term, maxSuggestionCount);
-    while (reader->getNextSuggestion(suggestion)) {
-      MustacheData result;
-      result.set("label", suggestion);
-      result.set("value", suggestion);
-      result.set("first", first);
-      first = false;
-      results.push_back(result);
-      suggestionCount++;
-    }
-  }
-
-  /* Propose the fulltext search if possible */
-  if (reader->hasFulltextIndex()) {
-    MustacheData result;
-    result.set("label", "containing '" + term + "'...");
-    result.set("value", term + " ");
-    result.set("first", first);
-    results.push_back(result);
-  }
-
-  auto data = get_default_data();
-  data.set("suggestions", results);
-
-  auto response = get_default_response();
-  response.set_template(RESOURCE::templates::suggestion_json, data);
-  response.set_mimeType("application/json; charset=utf-8");
-  response.set_compress(true);
-  return response;
-}
-
-Response InternalServer::handle_skin(const RequestContext& request)
-{
-  if (m_verbose.load()) {
-    printf("** running handle_skin\n");
-  }
-
-  auto response = get_default_response();
-  auto resourceName = request.get_url().substr(1);
-  try {
-    response.set_content(getResource(resourceName));
-  } catch (const ResourceNotFound& e) {
-    return build_404(request, "");
-  }
-  response.set_mimeType(getMimeTypeForFile(resourceName));
-  response.set_compress(true);
-  response.set_cacheable();
-  return response;
-}
-
-Response InternalServer::handle_search(const RequestContext& request)
-{
-  if (m_verbose.load()) {
-    printf("** running handle_search\n");
-  }
-
-  std::string bookName;
-  std::string bookId;
-  try {
-    bookName = request.get_argument("content");
-    bookId = mp_nameMapper->getIdForName(bookName);
-  } catch (const std::out_of_range&) {}
-
-  std::string patternString;
-  try {
-    patternString = request.get_argument("pattern");
-  } catch (const std::out_of_range&) {}
-
-  /* Retrive geo search */
-  bool has_geo_query = false;
-  float latitude = 0;
-  float longitude = 0;
-  float distance = 0;
-  try {
-    latitude = request.get_argument<float>("latitude");
-    longitude = request.get_argument<float>("longitude");
-    distance = request.get_argument<float>("distance");
-    has_geo_query = true;
-  } catch(const std::out_of_range&) {}
-    catch(const std::invalid_argument&) {}
-
-  std::shared_ptr<Reader> reader(nullptr);
-  try {
-    reader = mp_library->getReaderById(bookId);
-  } catch (const std::out_of_range&) {}
-
-  /* Try first to load directly the article */
-  if (reader != nullptr && !patternString.empty()) {
-    std::string patternCorrespondingUrl;
-    auto variants = reader->getTitleVariants(patternString);
-    auto variantsItr = variants.begin();
-
-    while (patternCorrespondingUrl.empty() && variantsItr != variants.end()) {
-      try {
-        auto entry = reader->getEntryFromTitle(*variantsItr);
-        entry = entry.getFinalEntry();
-        patternCorrespondingUrl = entry.getPath();
-        break;
-      } catch(kiwix::NoEntry& e) {
-        variantsItr++;
-      }
-    }
-
-    /* If article found then redirect directly to it */
-    if (!patternCorrespondingUrl.empty()) {
-      auto response = get_default_response();
-      response.set_redirection(m_root + "/" + bookName + "/" + patternCorrespondingUrl);
-      return response;
-    }
-  }
-
-  /* Make the search */
-  auto response = get_default_response();
-  response.set_mimeType("text/html; charset=utf-8");
-  response.set_taskbar(bookName, reader ? reader->getTitle() : "");
-  response.set_compress(true);
-
-  if ( (!reader && !bookName.empty())
-    || (patternString.empty() && ! has_geo_query) ) {
-    auto data = get_default_data();
-    data.set("pattern", encodeDiples(patternString));
-    response.set_template(RESOURCE::templates::no_search_result_html, data);
-    response.set_code(MHD_HTTP_NOT_FOUND);
-    return response;
-  }
-
-  Searcher searcher;
-  if (reader) {
-    searcher.add_reader(reader.get());
-  } else {
-    for (auto& bookId: mp_library->filter(kiwix::Filter().local(true).valid(true))) {
-      auto currentReader = mp_library->getReaderById(bookId);
-      if (currentReader) {
-        searcher.add_reader(currentReader.get());
-      }
-    }
-  }
-
-  auto start = 0;
-  try {
-    start = request.get_argument<unsigned int>("start");
-  } catch (const std::exception&) {}
-  auto end = 25;
-  try {
-    end = request.get_argument<unsigned int>("end");
-  } catch (const std::exception&) {}
-  if (start>end) {
-    auto tmp = start;
-    start = end;
-    end = tmp;
-  }
-  if (end > start + MAX_SEARCH_LEN) {
-    end = start + MAX_SEARCH_LEN;
-  }
-
-  /* Get the results */
-  try {
-    if (patternString.empty()) {
-      searcher.geo_search(latitude, longitude, distance,
-                           start, end, m_verbose.load());
-    } else {
-      searcher.search(patternString,
-                       start, end, m_verbose.load());
-    }
-    SearchRenderer renderer(&searcher, mp_nameMapper);
-    renderer.setSearchPattern(patternString);
-    renderer.setSearchContent(bookName);
-    renderer.setProtocolPrefix(m_root + "/");
-    renderer.setSearchProtocolPrefix(m_root + "/search?");
-    response.set_content(renderer.getHtml());
-  } catch (const std::exception& e) {
-    std::cerr << e.what() << std::endl;
-  }
-  return response;
-}
-
-Response InternalServer::handle_random(const RequestContext& request)
-{
-  if (m_verbose.load()) {
-    printf("** running handle_random\n");
-  }
-
-  std::string bookName;
-  std::string bookId;
-  std::shared_ptr<Reader> reader;
-  try {
-    bookName = request.get_argument("content");
-    bookId = mp_nameMapper->getIdForName(bookName);
-    reader = mp_library->getReaderById(bookId);
-  } catch (const std::out_of_range&) {
-    return build_404(request, bookName);
-  }
-
-  if (reader == nullptr) {
-    return build_404(request, bookName);
-  }
-
-  try {
-    auto entry = reader->getRandomPage();
-    return build_redirect(bookName, entry.getFinalEntry());
-  } catch(kiwix::NoEntry& e) {
-    return build_404(request, bookName);
-  }
-}
-
-Response InternalServer::handle_captured_external(const RequestContext& request)
-{
-  std::string source = "";
-  try {
-    source = kiwix::urlDecode(request.get_argument("source"));
-  } catch (const std::out_of_range& e) {}
-
-  if (source.empty())
-    return build_404(request, "");
-
-  auto data = get_default_data();
-  data.set("source", source);
-  auto response = get_default_response();
-  response.set_template(RESOURCE::templates::captured_external_html, data);
-  response.set_mimeType("text/html; charset=utf-8");
-  response.set_compress(true);
-  response.set_taskbar("", "");
-  return response;
-}
-
-Response InternalServer::handle_catalog(const RequestContext& request)
-{
-  if (m_verbose.load()) {
-    printf("** running handle_catalog");
-  }
-
-  std::string host;
-  std::string url;
-  try {
-    host = request.get_header("Host");
-    url  = request.get_url_part(1);
-  } catch (const std::out_of_range&) {
-    return build_404(request, "");
-  }
-
-  if (url != "searchdescription.xml" && url != "root.xml" && url != "search") {
-    return build_404(request, "");
-  }
-
-  auto response = get_default_response();
-  response.set_compress(true);
-  if (url == "searchdescription.xml") {
-    response.set_template(RESOURCE::opensearchdescription_xml, get_default_data());
-    response.set_mimeType("application/opensearchdescription+xml");
-    return response;
-  }
-
-  zim::Uuid uuid;
-  kiwix::OPDSDumper opdsDumper;
-  opdsDumper.setRootLocation(m_root);
-  opdsDumper.setSearchDescriptionUrl("catalog/searchdescription.xml");
-  opdsDumper.setLibrary(mp_library);
-  response.set_mimeType("application/atom+xml; profile=opds-catalog; kind=acquisition; charset=utf-8");
-  std::vector<std::string> bookIdsToDump;
-  if (url == "root.xml") {
-    opdsDumper.setTitle("All zims");
-    uuid = zim::Uuid::generate(host);
-    bookIdsToDump = mp_library->filter(kiwix::Filter().valid(true).local(true).remote(true));
-  } else if (url == "search") {
-    auto filter = kiwix::Filter().valid(true).local(true).remote(true);
-    string query("<Empty query>");
-    size_t count(10);
-    size_t startIndex(0);
-    try {
-      query = request.get_argument("q");
-      filter.query(query);
-    } catch (const std::out_of_range&) {}
-    try {
-      filter.maxSize(extractFromString<unsigned long>(request.get_argument("maxsize")));
-    } catch (...) {}
-    try {
-      filter.name(request.get_argument("name"));
-    } catch (const std::out_of_range&) {}
-    try {
-      filter.lang(request.get_argument("lang"));
-    } catch (const std::out_of_range&) {}
-    try {
-      count = extractFromString<unsigned long>(request.get_argument("count"));
-    } catch (...) {}
-    try {
-      startIndex = extractFromString<unsigned long>(request.get_argument("start"));
-    } catch (...) {}
-    try {
-      filter.acceptTags(kiwix::split(request.get_argument("tag"), ";"));
-    } catch (...) {}
-    try {
-      filter.rejectTags(kiwix::split(request.get_argument("notag"), ";"));
-    } catch (...) {}
-    opdsDumper.setTitle("Search result for " + query);
-    uuid = zim::Uuid::generate();
-    bookIdsToDump = mp_library->filter(filter);
-    auto totalResults = bookIdsToDump.size();
-    bookIdsToDump.erase(bookIdsToDump.begin(), bookIdsToDump.begin()+startIndex);
-    if (count>0 && bookIdsToDump.size() > count) {
-      bookIdsToDump.resize(count);
-    }
-    opdsDumper.setOpenSearchInfo(totalResults, startIndex, bookIdsToDump.size());
-  }
-
-  opdsDumper.setId(kiwix::to_string(uuid));
-  response.set_content(opdsDumper.dumpOPDSFeed(bookIdsToDump));
-  return response;
-}
-
-namespace
-{
-
-std::string get_book_name(const RequestContext& request)
-{
-  try {
-    return request.get_url_part(0);
-  } catch (const std::out_of_range& e) {
-    return std::string();
-  }
-}
-
-} // unnamed namespace
-
-std::shared_ptr<Reader>
-InternalServer::get_reader(const std::string& bookName) const
-{
-  std::shared_ptr<Reader> reader;
-  try {
-    const std::string bookId = mp_nameMapper->getIdForName(bookName);
-    reader = mp_library->getReaderById(bookId);
-  } catch (const std::out_of_range& e) {
-  }
-  return reader;
-}
-
-Response
-InternalServer::build_redirect(const std::string& bookName, const kiwix::Entry& entry) const
-{
-  auto response = get_default_response();
-  response.set_redirection(m_root + "/" + bookName + "/" +
-                           kiwix::urlEncode(entry.getPath()));
-  return response;
-}
-
-Response InternalServer::handle_content(const RequestContext& request)
-{
-  if (m_verbose.load()) {
-    printf("** running handle_content\n");
-  }
-
-  const std::string bookName = get_book_name(request);
-  if (bookName.empty())
-    return build_homepage(request);
-
-  const std::shared_ptr<Reader> reader = get_reader(bookName);
-  if (reader == nullptr) {
-    return build_404(request, bookName);
-  }
-
-  auto urlStr = request.get_url().substr(bookName.size()+1);
-  if (urlStr[0] == '/') {
-    urlStr = urlStr.substr(1);
-  }
-
-  kiwix::Entry entry;
-
-  try {
-    entry = reader->getEntryFromPath(urlStr);
-    if (entry.isRedirect() || urlStr.empty()) {
-      // If urlStr is empty, we want to mainPage.
-      // We must do a redirection to the real page.
-      return build_redirect(bookName, entry.getFinalEntry());
-    }
-  } catch(kiwix::NoEntry& e) {
-    if (m_verbose.load())
-      printf("Failed to find %s\n", urlStr.c_str());
-
-    return build_404(request, bookName);
-  }
-
-  auto response = get_default_response();
-
-  response.set_entry(entry, request);
-
-  if (m_verbose.load()) {
-    printf("Found %s\n", entry.getPath().c_str());
-    printf("mimeType: %s\n", response.get_mimeType().c_str());
-  }
-
-  if (response.get_mimeType().find("text/html") != string::npos)
-    response.set_taskbar(bookName, reader->getTitle());
-
-  return response;
+  return mp_server->getAddress();
 }
 
 }

src/server/byte_range.cpp

@@ -0,0 +1,127 @@
+/*
+ * Copyright 2020 Veloman Yunkan <veloman.yunkan@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+
+#include "byte_range.h"
+
+#include "tools/stringTools.h"
+
+#include <cassert>
+#include <algorithm>
+
+namespace kiwix {
+
+namespace {
+
+ByteRange parseByteRange(const std::string& rangeStr)
+{
+  std::istringstream iss(rangeStr);
+
+  int64_t start, end = INT64_MAX;
+  if (iss >> start) {
+    if ( start < 0 ) {
+      if ( iss.eof() )
+        return ByteRange(-start);
+    } else {
+      char c;
+      if (iss >> c && c=='-') {
+        iss >> end; // if this fails, end is not modified, which is OK
+        if (iss.eof() && start <= end)
+          return ByteRange(ByteRange::PARSED, start, end);
+      }
+    }
+  }
+
+  return ByteRange(ByteRange::INVALID, 0, INT64_MAX);
+}
+
+} // unnamed namespace
+
+ByteRange::ByteRange()
+  : kind_(NONE)
+  , first_(0)
+  , last_(INT64_MAX)
+{}
+
+ByteRange::ByteRange(Kind kind, int64_t first, int64_t last)
+  : kind_(kind)
+  , first_(first)
+  , last_(last)
+{
+  assert(kind != NONE);
+  assert(first >= 0);
+  assert(last >= first || (first == 0 && last == -1));
+}
+
+ByteRange::ByteRange(int64_t suffix_length)
+  : kind_(PARSED)
+  , first_(-suffix_length)
+  , last_(INT64_MAX)
+{
+  assert(suffix_length > 0);
+}
+
+int64_t ByteRange::first() const
+{
+  assert(kind_ > PARSED);
+  return first_;
+}
+
+int64_t ByteRange::last() const
+{
+  assert(kind_ > PARSED);
+  return last_;
+}
+
+int64_t ByteRange::length() const
+{
+  assert(kind_ > PARSED);
+  return last_ + 1 - first_;
+}
+
+ByteRange ByteRange::parse(const std::string& rangeStr)
+{
+  const std::string byteUnitSpec("bytes=");
+  if ( ! kiwix::startsWith(rangeStr, byteUnitSpec) )
+    return ByteRange(INVALID, 0, INT64_MAX);
+
+  return parseByteRange(rangeStr.substr(byteUnitSpec.size()));
+}
+
+ByteRange ByteRange::resolve(int64_t contentSize) const
+{
+  if ( kind() == NONE )
+    return ByteRange(RESOLVED_FULL_CONTENT, 0, contentSize-1);
+
+  if ( kind() == INVALID )
+    return ByteRange(RESOLVED_UNSATISFIABLE, 0, contentSize-1);
+
+  const int64_t resolved_first = first_ < 0
+                               ? std::max(int64_t(0), contentSize + first_)
+                               : first_;
+
+  const int64_t resolved_last = std::min(contentSize-1, last_);
+
+  if ( resolved_first > resolved_last )
+    return ByteRange(RESOLVED_UNSATISFIABLE, 0, contentSize-1);
+
+  return ByteRange(RESOLVED_PARTIAL_CONTENT, resolved_first, resolved_last);
+}
+
+} // namespace kiwix

src/server/byte_range.h

@@ -0,0 +1,86 @@
+/*
+ * Copyright 2020 Veloman Yunkan <veloman.yunkan@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+
+#ifndef KIWIXLIB_SERVER_BYTE_RANGE_H
+#define KIWIXLIB_SERVER_BYTE_RANGE_H
+
+#include <cstdint>
+#include <string>
+
+namespace kiwix {
+
+class ByteRange
+{
+  public: // types
+    // ByteRange is parsed in a request, then it must be resolved (taking
+    // into account the actual size of the requested resource) before
+    // being applied in the response.
+    // The Kind enum represents possible states in such a lifecycle.
+    enum Kind {
+      // The request is not a range request (no Range header)
+      NONE,
+
+      // The value of the Range header is not a valid continuous
+      // range. Note that a valid (according to RFC7233) sequence of multiple
+      // byte ranges is considered invalid in the current implementation
+      // (i.e. only single-range partial requests are supported).
+      INVALID,
+
+      // This byte-range has been successfully parsed from the request
+      PARSED,
+
+      // This is a response to a regular (non-range) request
+      RESOLVED_FULL_CONTENT,
+
+      // The range request is invalid or unsatisfiable
+      RESOLVED_UNSATISFIABLE,
+
+      // This is a response to a (satisfiable) range request
+      RESOLVED_PARTIAL_CONTENT,
+    };
+
+  public: // functions
+    // Constructs a ByteRange object of NONE kind
+    ByteRange();
+
+    // Constructs a ByteRange object of the given kind (except NONE)
+    ByteRange(Kind kind, int64_t first, int64_t last);
+
+    // Constructs a ByteRange object of PARSED kind corresponding to a
+    // range request of the form "Range: bytes=-suffix_length"
+    explicit ByteRange(int64_t suffix_length);
+
+    Kind kind() const { return kind_; }
+    int64_t first() const;
+    int64_t last() const;
+    int64_t length() const;
+
+    static ByteRange parse(const std::string& rangeStr);
+    ByteRange resolve(int64_t contentSize) const;
+
+  private: // data
+    Kind kind_;
+    int64_t first_;
+    int64_t last_;
+};
+
+} // namespace kiwix
+
+#endif //KIWIXLIB_SERVER_BYTE_RANGE_H

src/server/i18n.cpp

@@ -0,0 +1,114 @@
+/*
+ * Copyright 2022 Veloman Yunkan <veloman.yunkan@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#include "i18n.h"
+
+#include "tools/otherTools.h"
+
+#include <algorithm>
+#include <map>
+
+namespace kiwix
+{
+
+const char* I18nStringTable::get(const std::string& key) const
+{
+  const I18nString* const begin = entries;
+  const I18nString* const end = begin + entryCount;
+  const I18nString* found = std::lower_bound(begin, end, key,
+      [](const I18nString& a, const std::string& k) {
+        return a.key < k;
+  });
+  return (found == end || found->key != key) ? nullptr : found->value;
+}
+
+namespace i18n
+{
+// this data is generated by the i18n resource compiler
+extern const I18nStringTable stringTables[];
+extern const size_t langCount;
+}
+
+namespace
+{
+
+class I18nStringDB
+{
+public: // functions
+  I18nStringDB() {
+    for ( size_t i = 0; i < kiwix::i18n::langCount; ++i ) {
+      const auto& t = kiwix::i18n::stringTables[i];
+      lang2TableMap[t.lang] = &t;
+    }
+    enStrings = lang2TableMap.at("en");
+  };
+
+  std::string get(const std::string& lang, const std::string& key) const {
+    const char* s = getStringsFor(lang)->get(key);
+    if ( s == nullptr ) {
+      s = enStrings->get(key);
+      if ( s == nullptr ) {
+        throw std::runtime_error("Invalid message id");
+      }
+    }
+    return s;
+  }
+
+private: // functions
+  const I18nStringTable* getStringsFor(const std::string& lang) const {
+    try {
+      return lang2TableMap.at(lang);
+    } catch(const std::out_of_range&) {
+      return enStrings;
+    }
+  }
+
+private: // data
+  std::map<std::string, const I18nStringTable*> lang2TableMap;
+  const I18nStringTable* enStrings;
+};
+
+} // unnamed namespace
+
+std::string getTranslatedString(const std::string& lang, const std::string& key)
+{
+  static const I18nStringDB stringDb;
+
+  return stringDb.get(lang, key);
+}
+
+namespace i18n
+{
+
+std::string expandParameterizedString(const std::string& lang,
+                                      const std::string& key,
+                                      const Parameters& params)
+{
+  const std::string tmpl = getTranslatedString(lang, key);
+  return render_template(tmpl, params);
+}
+
+} // namespace i18n
+
+std::string ParameterizedMessage::getText(const std::string& lang) const
+{
+  return i18n::expandParameterizedString(lang, msgId, params);
+}
+
+} // namespace kiwix

src/server/i18n.h

@@ -0,0 +1,94 @@
+/*
+ * Copyright 2022 Veloman Yunkan <veloman.yunkan@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#ifndef KIWIX_SERVER_I18N
+#define KIWIX_SERVER_I18N
+
+#include <string>
+#include <mustache.hpp>
+
+namespace kiwix
+{
+
+struct I18nString {
+  const char* const key;
+  const char* const value;
+};
+
+struct I18nStringTable {
+  const char* const lang;
+  const size_t entryCount;
+  const I18nString* const entries;
+
+  const char* get(const std::string& key) const;
+};
+
+std::string getTranslatedString(const std::string& lang, const std::string& key);
+
+namespace i18n
+{
+
+typedef kainjow::mustache::object Parameters;
+
+std::string expandParameterizedString(const std::string& lang,
+                                      const std::string& key,
+                                      const Parameters& params);
+
+class GetTranslatedString
+{
+public:
+  explicit GetTranslatedString(const std::string& lang) : m_lang(lang) {}
+
+  std::string operator()(const std::string& key) const
+  {
+    return getTranslatedString(m_lang, key);
+  }
+
+  std::string operator()(const std::string& key, const Parameters& params) const
+  {
+    return expandParameterizedString(m_lang, key, params);
+  }
+
+private:
+  const std::string m_lang;
+};
+
+} // namespace i18n
+
+struct ParameterizedMessage
+{
+public: // types
+  typedef kainjow::mustache::object Parameters;
+
+public: // functions
+  ParameterizedMessage(const std::string& msgId, const Parameters& params)
+    : msgId(msgId)
+    , params(params)
+  {}
+
+  std::string getText(const std::string& lang) const;
+
+private: // data
+  const std::string msgId;
+  const Parameters  params;
+};
+
+} // namespace kiwix
+
+#endif // KIWIX_SERVER_I18N

src/server/internalServer.h

@@ -0,0 +1,193 @@
+/*
+ * Copyright 2019 Matthieu Gautier <mgautier@kymeria.fr>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#ifndef KIWIXLIB_SERVER_INTERNALSERVER_H
+#define KIWIXLIB_SERVER_INTERNALSERVER_H
+
+
+extern "C" {
+#include "microhttpd_wrapper.h"
+}
+
+#include "library.h"
+#include "name_mapper.h"
+
+#include <zim/search.h>
+#include <zim/suggestion.h>
+
+#include <mustache.hpp>
+
+#include <atomic>
+#include <string>
+
+#include "server/request_context.h"
+#include "server/response.h"
+
+#include "tools/concurrent_cache.h"
+
+namespace kiwix {
+
+struct GeoQuery {
+  GeoQuery()
+    : GeoQuery(0, 0, -1)
+  {}
+
+  GeoQuery(float latitude, float longitude, float distance)
+    : latitude(latitude), longitude(longitude), distance(distance)
+  {}
+  float latitude;
+  float longitude;
+  float distance;
+
+  explicit operator bool() const {
+    return distance >= 0;
+  }
+
+  friend bool operator<(const GeoQuery& l, const GeoQuery& r)
+  {
+      return std::tie(l.latitude, l.longitude, l.distance)
+           < std::tie(r.latitude, r.longitude, r.distance); // keep the same order
+  }
+};
+
+class SearchInfo {
+  public:
+    SearchInfo(const std::string& pattern, GeoQuery geoQuery, const Library::BookIdSet& bookIds, const std::string& bookFilterString);
+
+    zim::Query getZimQuery(bool verbose) const;
+    const Library::BookIdSet& getBookIds() const { return bookIds; }
+
+    friend bool operator<(const SearchInfo& l, const SearchInfo& r)
+    {
+        return std::tie(l.bookIds, l.pattern, l.geoQuery)
+             < std::tie(r.bookIds, r.pattern, r.geoQuery); // keep the same order
+    }
+
+  public: //data
+    std::string pattern;
+    GeoQuery geoQuery;
+    Library::BookIdSet bookIds;
+    std::string bookFilterQuery;
+};
+
+
+typedef kainjow::mustache::data MustacheData;
+typedef ConcurrentCache<SearchInfo, std::shared_ptr<zim::Search>> SearchCache;
+typedef ConcurrentCache<std::string, std::shared_ptr<zim::SuggestionSearcher>> SuggestionSearcherCache;
+
+class OPDSDumper;
+
+class InternalServer {
+  public:
+    InternalServer(Library* library,
+                   NameMapper* nameMapper,
+                   std::string addr,
+                   int port,
+                   std::string root,
+                   int nbThreads,
+                   unsigned int multizimSearchLimit,
+                   bool verbose,
+                   bool withTaskbar,
+                   bool withLibraryButton,
+                   bool blockExternalLinks,
+                   std::string indexTemplateString,
+                   int ipConnectionLimit);
+    virtual ~InternalServer();
+
+    MHD_Result handlerCallback(struct MHD_Connection* connection,
+                               const char* url,
+                               const char* method,
+                               const char* version,
+                               const char* upload_data,
+                               size_t* upload_data_size,
+                               void** cont_cls);
+    bool start();
+    void stop();
+    std::string getAddress() { return m_addr; }
+    int getPort() { return m_port; }
+
+  private: // functions
+    std::unique_ptr<Response> handle_request(const RequestContext& request);
+    std::unique_ptr<Response> build_redirect(const std::string& bookName, const zim::Item& item) const;
+    std::unique_ptr<Response> build_homepage(const RequestContext& request);
+    std::unique_ptr<Response> handle_skin(const RequestContext& request);
+    std::unique_ptr<Response> handle_catalog(const RequestContext& request);
+    std::unique_ptr<Response> handle_catalog_v2(const RequestContext& request);
+    std::unique_ptr<Response> handle_catalog_v2_root(const RequestContext& request);
+    std::unique_ptr<Response> handle_catalog_v2_entries(const RequestContext& request, bool partial);
+    std::unique_ptr<Response> handle_catalog_v2_complete_entry(const RequestContext& request, const std::string& entryId);
+    std::unique_ptr<Response> handle_catalog_v2_categories(const RequestContext& request);
+    std::unique_ptr<Response> handle_catalog_v2_languages(const RequestContext& request);
+    std::unique_ptr<Response> handle_catalog_v2_illustration(const RequestContext& request);
+    std::unique_ptr<Response> handle_search(const RequestContext& request);
+    std::unique_ptr<Response> handle_suggest(const RequestContext& request);
+    std::unique_ptr<Response> handle_random(const RequestContext& request);
+    std::unique_ptr<Response> handle_catch(const RequestContext& request);
+    std::unique_ptr<Response> handle_captured_external(const RequestContext& request);
+    std::unique_ptr<Response> handle_content(const RequestContext& request);
+    std::unique_ptr<Response> handle_raw(const RequestContext& request);
+    std::unique_ptr<Response> handle_locally_customized_resource(const RequestContext& request);
+    std::unique_ptr<Response> handle_widget(const RequestContext& request);
+
+    std::vector<std::string> search_catalog(const RequestContext& request,
+                                            kiwix::OPDSDumper& opdsDumper);
+
+    MustacheData get_default_data() const;
+
+    bool etag_not_needed(const RequestContext& r) const;
+    ETag get_matching_if_none_match_etag(const RequestContext& request) const;
+    std::pair<std::string, Library::BookIdSet> selectBooks(const RequestContext& r) const;
+    SearchInfo getSearchInfo(const RequestContext& r) const;
+
+    bool isLocallyCustomizedResource(const std::string& url) const;
+
+  private: // data
+    std::string m_addr;
+    int m_port;
+    std::string m_root;
+    int m_nbThreads;
+    unsigned int m_multizimSearchLimit;
+    std::atomic_bool m_verbose;
+    bool m_withTaskbar;
+    bool m_withLibraryButton;
+    bool m_blockExternalLinks;
+    std::string m_indexTemplateString;
+    int m_ipConnectionLimit;
+    struct MHD_Daemon* mp_daemon;
+
+    Library* mp_library;
+    NameMapper* mp_nameMapper;
+
+    SearchCache searchCache;
+    SuggestionSearcherCache suggestionSearcherCache;
+
+    std::string m_server_id;
+    std::string m_library_id;
+
+    class CustomizedResources;
+    std::unique_ptr<CustomizedResources> m_customizedResources;
+
+    friend std::unique_ptr<Response> Response::build(const InternalServer& server);
+    friend std::unique_ptr<ContentResponse> ContentResponse::build(const InternalServer& server, const std::string& content, const std::string& mimetype, bool isHomePage, bool raw);
+    friend std::unique_ptr<Response> ItemResponse::build(const InternalServer& server, const RequestContext& request, const zim::Item& item, bool raw);
+};
+
+}
+
+#endif //KIWIXLIB_SERVER_INTERNALSERVER_H

src/server/internalServer_catalog_v2.cpp

@@ -0,0 +1,168 @@
+/*
+ * Copyright 2021 Veloman Yunkan <veloman.yunkan@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#include "internalServer.h"
+
+#include "library.h"
+#include "opds_dumper.h"
+#include "request_context.h"
+#include "response.h"
+#include "tools/otherTools.h"
+#include "kiwixlib-resources.h"
+
+#include <mustache.hpp>
+
+#include <string>
+#include <vector>
+
+namespace kiwix {
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2(const RequestContext& request)
+{
+  if (m_verbose.load()) {
+    printf("** running handle_catalog_v2");
+  }
+
+  std::string url;
+  try {
+    url  = request.get_url_part(2);
+  } catch (const std::out_of_range&) {
+    return HTTP404Response(*this, request)
+           + urlNotFoundMsg;
+  }
+
+  if (url == "root.xml") {
+    return handle_catalog_v2_root(request);
+  } else if (url == "searchdescription.xml") {
+    const std::string endpoint_root = m_root + "/catalog/v2";
+    return ContentResponse::build(*this,
+        RESOURCE::catalog_v2_searchdescription_xml,
+        kainjow::mustache::object({{"endpoint_root", endpoint_root}}),
+        "application/opensearchdescription+xml"
+    );
+  } else if (url == "entry") {
+    const std::string entryId  = request.get_url_part(3);
+    return handle_catalog_v2_complete_entry(request, entryId);
+  } else if (url == "entries") {
+    return handle_catalog_v2_entries(request, /*partial=*/false);
+  } else if (url == "partial_entries") {
+    return handle_catalog_v2_entries(request, /*partial=*/true);
+  } else if (url == "categories") {
+    return handle_catalog_v2_categories(request);
+  } else if (url == "languages") {
+    return handle_catalog_v2_languages(request);
+  } else if (url == "illustration") {
+    return handle_catalog_v2_illustration(request);
+  } else {
+    return HTTP404Response(*this, request)
+           + urlNotFoundMsg;
+  }
+}
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2_root(const RequestContext& request)
+{
+  return ContentResponse::build(
+             *this,
+             RESOURCE::templates::catalog_v2_root_xml,
+             kainjow::mustache::object{
+               {"date", gen_date_str()},
+               {"endpoint_root", m_root + "/catalog/v2"},
+               {"feed_id", gen_uuid(m_library_id)},
+               {"all_entries_feed_id", gen_uuid(m_library_id + "/entries")},
+               {"partial_entries_feed_id", gen_uuid(m_library_id + "/partial_entries")},
+               {"category_list_feed_id", gen_uuid(m_library_id + "/categories")},
+               {"language_list_feed_id", gen_uuid(m_library_id + "/languages")}
+             },
+             "application/atom+xml;profile=opds-catalog;kind=navigation"
+  );
+}
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2_entries(const RequestContext& request, bool partial)
+{
+  OPDSDumper opdsDumper(mp_library);
+  opdsDumper.setRootLocation(m_root);
+  opdsDumper.setLibraryId(m_library_id);
+  const auto bookIds = search_catalog(request, opdsDumper);
+  const auto opdsFeed = opdsDumper.dumpOPDSFeedV2(bookIds, request.get_query(), partial);
+  return ContentResponse::build(
+             *this,
+             opdsFeed,
+             "application/atom+xml;profile=opds-catalog;kind=acquisition"
+  );
+}
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2_complete_entry(const RequestContext& request, const std::string& entryId)
+{
+  try {
+    mp_library->getBookById(entryId);
+  } catch (const std::out_of_range&) {
+    return HTTP404Response(*this, request)
+           + urlNotFoundMsg;
+  }
+
+  OPDSDumper opdsDumper(mp_library);
+  opdsDumper.setRootLocation(m_root);
+  opdsDumper.setLibraryId(m_library_id);
+  const auto opdsFeed = opdsDumper.dumpOPDSCompleteEntry(entryId);
+  return ContentResponse::build(
+             *this,
+             opdsFeed,
+             "application/atom+xml;type=entry;profile=opds-catalog"
+  );
+}
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2_categories(const RequestContext& request)
+{
+  OPDSDumper opdsDumper(mp_library);
+  opdsDumper.setRootLocation(m_root);
+  opdsDumper.setLibraryId(m_library_id);
+  return ContentResponse::build(
+             *this,
+             opdsDumper.categoriesOPDSFeed(),
+             "application/atom+xml;profile=opds-catalog;kind=navigation"
+  );
+}
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2_languages(const RequestContext& request)
+{
+  OPDSDumper opdsDumper(mp_library);
+  opdsDumper.setRootLocation(m_root);
+  opdsDumper.setLibraryId(m_library_id);
+  return ContentResponse::build(
+             *this,
+             opdsDumper.languagesOPDSFeed(),
+             "application/atom+xml;profile=opds-catalog;kind=navigation"
+  );
+}
+
+std::unique_ptr<Response> InternalServer::handle_catalog_v2_illustration(const RequestContext& request)
+{
+  try {
+    const auto bookId  = request.get_url_part(3);
+    auto book = mp_library->getBookByIdThreadSafe(bookId);
+    auto size = request.get_argument<unsigned int>("size");
+    auto illustration = book.getIllustration(size);
+    return ContentResponse::build(*this, illustration->getData(), illustration->mimeType);
+  } catch(...) {
+    return HTTP404Response(*this, request)
+           + urlNotFoundMsg;
+  }
+}
+
+} // namespace kiwix

src/server/request_context.cpp

@@ -26,6 +26,8 @@
 #include <cstdio>
 #include <atomic>
 
+#include "tools/stringTools.h"
+
 namespace kiwix {
 
 static std::atomic_ullong s_requestIndex(0);
@@ -73,58 +75,38 @@ RequestContext::RequestContext(struct MHD_Connection* connection,
   method(str2RequestMethod(_method)),
   version(version),
   requestIndex(s_requestIndex++),
-  acceptEncodingDeflate(false),
-  accept_range(false),
-  range_pair(0, -1)
+  acceptEncodingGzip(false),
+  byteRange_()
 {
   MHD_get_connection_values(connection, MHD_HEADER_KIND, &RequestContext::fill_header, this);
   MHD_get_connection_values(connection, MHD_GET_ARGUMENT_KIND, &RequestContext::fill_argument, this);
 
   try {
-    acceptEncodingDeflate =
-        (get_header(MHD_HTTP_HEADER_ACCEPT_ENCODING).find("deflate") != std::string::npos);
+    acceptEncodingGzip =
+        (get_header(MHD_HTTP_HEADER_ACCEPT_ENCODING).find("gzip") != std::string::npos);
   } catch (const std::out_of_range&) {}
 
-  /*Check if range is requested. */
   try {
-    auto range = get_header(MHD_HTTP_HEADER_RANGE);
-    int start = 0;
-    int end = -1;
-    std::istringstream iss(range);
-    char c;
-
-    iss >> start >> c;
-    if (iss.good() && c=='-') {
-      iss >> end;
-      if (iss.fail()) {
-        // Something went wrong will extracting.
-        end = -1;
-      }
-      if (iss.eof()) {
-        accept_range = true;
-        range_pair = std::pair<int, int>(start, end);
-      }
-    }
+    byteRange_ = ByteRange::parse(get_header(MHD_HTTP_HEADER_RANGE));
   } catch (const std::out_of_range&) {}
 }
 
 RequestContext::~RequestContext()
 {}
 
-
-int RequestContext::fill_header(void *__this, enum MHD_ValueKind kind,
-                                 const char *key, const char *value)
+MHD_Result RequestContext::fill_header(void *__this, enum MHD_ValueKind kind,
+                                       const char *key, const char *value)
 {
   RequestContext *_this = static_cast<RequestContext*>(__this);
-  _this->headers[key] = value;
+  _this->headers[lcAll(key)] = value;
   return MHD_YES;
 }
 
-int RequestContext::fill_argument(void *__this, enum MHD_ValueKind kind,
-                                   const char *key, const char* value)
+MHD_Result RequestContext::fill_argument(void *__this, enum MHD_ValueKind kind,
+                                         const char *key, const char* value)
 {
   RequestContext *_this = static_cast<RequestContext*>(__this);
-  _this->arguments[key] = value == nullptr ? "" : value;
+  _this->arguments[key].push_back(value == nullptr ? "" : value);
   return MHD_YES;
 }
 
@@ -139,14 +121,20 @@ void RequestContext::print_debug_info() const {
     printf(" - %s : '%s'\n", it->first.c_str(), it->second.c_str());
   }
   printf("arguments :\n");
-  for (auto it=arguments.begin(); it!=arguments.end(); it++) {
-    printf(" - %s : '%s'\n", it->first.c_str(), it->second.c_str());
+  for (auto& pair:arguments) {
+    printf(" - %s :", pair.first.c_str());
+    bool first = true;
+    for (auto& v: pair.second) {
+      printf("%s %s", first?"":",", v.c_str());
+      first = false;
+    }
+    printf("\n");
   }
   printf("Parsed : \n");
   printf("full_url: %s\n", full_url.c_str());
   printf("url   : %s\n", url.c_str());
-  printf("acceptEncodingDeflate : %d\n", acceptEncodingDeflate);
-  printf("has_range : %d\n", accept_range);
+  printf("acceptEncodingGzip : %d\n", acceptEncodingGzip);
+  printf("has_range : %d\n", byteRange_.kind() != ByteRange::NONE);
   printf("is_valid_url : %d\n", is_valid_url());
   printf(".............\n");
 }
@@ -188,21 +176,35 @@ bool RequestContext::is_valid_url() const {
   return !url.empty();
 }
 
-bool RequestContext::has_range() const {
-  return accept_range;
-}
-
-std::pair<int, int> RequestContext::get_range() const {
-  return range_pair;
+ByteRange RequestContext::get_range() const {
+  return byteRange_;
 }
 
 template<>
 std::string RequestContext::get_argument(const std::string& name) const {
-  return arguments.at(name);
+  return arguments.at(name)[0];
 }
 
 std::string RequestContext::get_header(const std::string& name) const {
-  return headers.at(name);
+  return headers.at(lcAll(name));
+}
+
+std::string RequestContext::get_user_language() const
+{
+  try {
+    return get_argument("userlang");
+  } catch(const std::out_of_range&) {}
+
+  try {
+    return get_header("Accept-Language");
+  } catch(const std::out_of_range&) {}
+
+  return "en";
+}
+
+std::string RequestContext::get_requested_format() const
+{
+  return get_optional_param<std::string>("format", "html");
 }
 
 }

src/server/request_context.h

@@ -25,10 +25,14 @@
 #include <string>
 #include <sstream>
 #include <map>
+#include <vector>
 #include <stdexcept>
 
+#include "byte_range.h"
+#include "tools/stringTools.h"
+
 extern "C" {
-#include <microhttpd.h>
+#include "microhttpd_wrapper.h"
 }
 
 namespace kiwix {
@@ -51,9 +55,6 @@ class IndexError: public std::runtime_error {};
 
 
 class RequestContext {
-  public: // types
-  typedef std::pair<int, int> ByteRange;
-
   public: // functions
     RequestContext(struct MHD_Connection* connection,
                    std::string rootLocation,
@@ -69,10 +70,20 @@ class RequestContext {
     std::string get_header(const std::string& name) const;
     template<typename T=std::string>
     T get_argument(const std::string& name) const {
-        std::istringstream stream(arguments.at(name));
-        T v;
-        stream >> v;
-        return v;
+        return extractFromString<T>(get_argument(name));
+    }
+
+    std::vector<std::string> get_arguments(const std::string& name) const {
+      return arguments.at(name);
+    }
+
+    template<class T>
+    T get_optional_param(const std::string& name, T default_value) const
+    {
+      try {
+        return get_argument<T>(name);
+      } catch (...) {}
+      return default_value;
     }
 
 
@@ -81,10 +92,33 @@ class RequestContext {
     std::string get_url_part(int part) const;
     std::string get_full_url() const;
 
-    bool has_range() const;
+    std::string get_query(bool mustEncode = false) const {
+      return get_query([](const std::string& key) {return true;}, mustEncode);
+    }
+
+    template<class F>
+    std::string get_query(F filter, bool mustEncode) const {
+      std::string q;
+      const char* sep = "";
+      auto encode = [=](const std::string& value) { return mustEncode?urlEncode(value, true):value; };
+      for ( const auto& a : arguments ) {
+        if (!filter(a.first)) {
+          continue;
+        }
+        for (const auto& v: a.second) {
+          q += sep + encode(a.first) + '=' + encode(v);
+          sep = "&";
+        }
+      }
+      return q;
+    }
+
     ByteRange get_range() const;
 
-    bool can_compress() const { return acceptEncodingDeflate; }
+    bool can_compress() const { return acceptEncodingGzip; }
+
+    std::string get_user_language() const;
+    std::string get_requested_format() const;
 
   private: // data
     std::string full_url;
@@ -93,16 +127,15 @@ class RequestContext {
     std::string version;
     unsigned long long requestIndex;
 
-    bool acceptEncodingDeflate;
+    bool acceptEncodingGzip;
 
-    bool accept_range;
-    ByteRange range_pair;
+    ByteRange byteRange_;
     std::map<std::string, std::string> headers;
-    std::map<std::string, std::string> arguments;
+    std::map<std::string, std::vector<std::string>> arguments;
 
   private: // functions
-    static int fill_header(void *, enum MHD_ValueKind, const char*, const char*);
-    static int fill_argument(void *, enum MHD_ValueKind, const char*, const char*);
+    static MHD_Result fill_header(void *, enum MHD_ValueKind, const char*, const char*);
+    static MHD_Result fill_argument(void *, enum MHD_ValueKind, const char*, const char*);
 };
 
 template<> std::string RequestContext::get_argument(const std::string& name) const;

src/server/response.cpp

@@ -1,19 +1,47 @@
-
-
+/*
+ * Copyright 2019 Matthieu Gautier <mgautier@kymeria.fr>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
 
 #include "response.h"
 #include "request_context.h"
+#include "internalServer.h"
 #include "kiwixlib-resources.h"
 
 #include "tools/regexTools.h"
 #include "tools/stringTools.h"
+#include "tools/otherTools.h"
+#include "tools/archiveTools.h"
 
 #include "string.h"
 #include <mustache.hpp>
 #include <zlib.h>
 
+#include <array>
 
-#define KIWIX_MIN_CONTENT_SIZE_TO_DEFLATE 100
+// This is somehow a magic value.
+// If this value is too small, we will compress (and lost cpu time) too much
+// content.
+// If this value is too big, we will not compress enough content and send too
+// much data.
+// If we assume that MTU is 1500 Bytes it is useless to compress
+// content smaller as the content will be sent in one packet anyway.
+// 1400 Bytes seems to be a common accepted limit.
+#define KIWIX_MIN_CONTENT_SIZE_TO_COMPRESS 1400
 
 namespace kiwix {
 
@@ -21,52 +49,240 @@ namespace
 {
 // some utilities
 
-std::string get_mime_type(const kiwix::Entry& entry)
+std::string get_mime_type(const zim::Item& item)
 {
   try {
-    return entry.getMimetype();
-  } catch (exception& e) {
+    return item.getMimetype();
+  } catch (std::exception& e) {
     return "application/octet-stream";
   }
 }
 
 bool is_compressible_mime_type(const std::string& mimeType)
 {
-  return mimeType.find("text/") != string::npos
-      || mimeType.find("application/javascript") != string::npos
-      || mimeType.find("application/atom") != string::npos
-      || mimeType.find("application/opensearchdescription") != string::npos
-      || mimeType.find("application/json") != string::npos;
+  return mimeType.find("text/") != std::string::npos
+      || mimeType.find("application/javascript") != std::string::npos
+      || mimeType.find("application/atom") != std::string::npos
+      || mimeType.find("application/opensearchdescription") != std::string::npos
+      || mimeType.find("application/json") != std::string::npos;
 }
 
-int get_range_len(const kiwix::Entry& entry, RequestContext::ByteRange range)
-{
-  return range.second == -1
-       ? entry.getSize() - range.first
-       : range.second - range.first;
+bool compress(std::string &content) {
+  z_stream strm;
+  strm.zalloc = Z_NULL;
+  strm.zfree = Z_NULL;
+  strm.opaque = Z_NULL;
+
+  auto ret = deflateInit2(&strm, Z_DEFAULT_COMPRESSION, Z_DEFLATED, 31, 8,
+                          Z_DEFAULT_STRATEGY);
+  if (ret != Z_OK) { return false; }
+
+  strm.avail_in = static_cast<decltype(strm.avail_in)>(content.size());
+  strm.next_in =
+      const_cast<Bytef *>(reinterpret_cast<const Bytef *>(content.data()));
+
+  std::string compressed;
+
+  std::array<char, 16384> buff{};
+  do {
+    strm.avail_out = buff.size();
+    strm.next_out = reinterpret_cast<Bytef *>(buff.data());
+    ret = deflate(&strm, Z_FINISH);
+    assert(ret != Z_STREAM_ERROR);
+    compressed.append(buff.data(), buff.size() - strm.avail_out);
+  } while (strm.avail_out == 0);
+
+  assert(ret == Z_STREAM_END);
+  assert(strm.avail_in == 0);
+
+  content.swap(compressed);
+
+  deflateEnd(&strm);
+  return true;
 }
 
+
+
 } // unnamed namespace
 
-Response::Response(const std::string& root, bool verbose, bool withTaskbar, bool withLibraryButton, bool blockExternalLinks)
+Response::Response(bool verbose)
   : m_verbose(verbose),
-    m_root(root),
-    m_content(""),
-    m_mimeType(""),
-    m_returnCode(MHD_HTTP_OK),
-    m_withTaskbar(withTaskbar),
-    m_withLibraryButton(withLibraryButton),
-    m_blockExternalLinks(blockExternalLinks),
-    m_addTaskbar(false),
-    m_bookName(""),
-    m_startRange(0),
-    m_lenRange(0)
+    m_returnCode(MHD_HTTP_OK)
+{
+  add_header(MHD_HTTP_HEADER_ACCESS_CONTROL_ALLOW_ORIGIN, "*");
+}
+
+std::unique_ptr<Response> Response::build(const InternalServer& server)
+{
+  return std::unique_ptr<Response>(new Response(server.m_verbose.load()));
+}
+
+std::unique_ptr<Response> Response::build_304(const InternalServer& server, const ETag& etag)
+{
+  auto response = Response::build(server);
+  response->set_code(MHD_HTTP_NOT_MODIFIED);
+  response->m_etag = etag;
+  if ( etag.get_option(ETag::COMPRESSED_CONTENT) ) {
+    response->add_header(MHD_HTTP_HEADER_VARY, "Accept-Encoding");
+  }
+  return response;
+}
+
+const UrlNotFoundMsg urlNotFoundMsg;
+const InvalidUrlMsg invalidUrlMsg;
+
+std::string ContentResponseBlueprint::getMessage(const std::string& msgId) const
+{
+  return getTranslatedString(m_request.get_user_language(), msgId);
+}
+
+std::unique_ptr<ContentResponse> ContentResponseBlueprint::generateResponseObject() const
+{
+  auto r = ContentResponse::build(m_server, m_template, m_data, m_mimeType);
+  r->set_code(m_httpStatusCode);
+  if ( m_taskbarInfo ) {
+    r->set_taskbar(m_taskbarInfo->bookName, m_taskbarInfo->archive);
+  }
+  return r;
+}
+
+HTTPErrorResponse::HTTPErrorResponse(const InternalServer& server,
+                                     const RequestContext& request,
+                                     int httpStatusCode,
+                                     const std::string& pageTitleMsgId,
+                                     const std::string& headingMsgId,
+                                     const std::string& cssUrl)
+  : ContentResponseBlueprint(&server,
+                             &request,
+                             httpStatusCode,
+                             request.get_requested_format() == "html" ? "text/html; charset=utf-8" : "application/xml; charset=utf-8",
+                             request.get_requested_format() == "html" ? RESOURCE::templates::error_html : RESOURCE::templates::error_xml)
+{
+  kainjow::mustache::list emptyList;
+  this->m_data = kainjow::mustache::object{
+                    {"CSS_URL", onlyAsNonEmptyMustacheValue(cssUrl) },
+                    {"PAGE_TITLE",   getMessage(pageTitleMsgId)},
+                    {"PAGE_HEADING", getMessage(headingMsgId)},
+                    {"details", emptyList}
+  };
+}
+
+HTTP404Response::HTTP404Response(const InternalServer& server,
+                                 const RequestContext& request)
+  : HTTPErrorResponse(server,
+                      request,
+                      MHD_HTTP_NOT_FOUND,
+                      "404-page-title",
+                      "404-page-heading")
 {
 }
 
+HTTPErrorResponse& HTTP404Response::operator+(UrlNotFoundMsg /*unused*/)
+{
+  const std::string requestUrl = m_request.get_full_url();
+  return *this + ParameterizedMessage("url-not-found", {{"url", requestUrl}});
+}
 
-static int print_key_value (void *cls, enum MHD_ValueKind kind,
-                            const char *key, const char *value)
+HTTPErrorResponse& HTTPErrorResponse::operator+(const std::string& msg)
+{
+  m_data["details"].push_back({"p", msg});
+  return *this;
+}
+
+HTTPErrorResponse& HTTPErrorResponse::operator+(const ParameterizedMessage& details)
+{
+  return *this + details.getText(m_request.get_user_language());
+}
+
+HTTPErrorResponse& HTTPErrorResponse::operator+=(const ParameterizedMessage& details)
+{
+  // operator+() is already a state-modifying operator (akin to operator+=)
+  return *this + details;
+}
+
+
+HTTP400Response::HTTP400Response(const InternalServer& server,
+                                 const RequestContext& request)
+  : HTTPErrorResponse(server,
+                      request,
+                      MHD_HTTP_BAD_REQUEST,
+                      "400-page-title",
+                      "400-page-heading")
+{
+}
+
+HTTPErrorResponse& HTTP400Response::operator+(InvalidUrlMsg /*unused*/)
+{
+  std::string requestUrl = m_request.get_full_url();
+  const auto query = m_request.get_query();
+  if (!query.empty()) {
+    requestUrl += "?" + encodeDiples(query);
+  }
+  kainjow::mustache::mustache msgTmpl(R"(The requested URL "{{{url}}}" is not a valid request.)");
+  return *this + msgTmpl.render({"url", requestUrl});
+}
+
+HTTP500Response::HTTP500Response(const InternalServer& server,
+                                 const RequestContext& request)
+  : HTTPErrorResponse(server,
+                      request,
+                      MHD_HTTP_INTERNAL_SERVER_ERROR,
+                      "500-page-title",
+                      "500-page-heading")
+{
+  // operator+() is a state-modifying operator (akin to operator+=)
+  *this + "An internal server error occured. We are sorry about that :/";
+}
+
+std::unique_ptr<ContentResponse> HTTP500Response::generateResponseObject() const
+{
+  // We want a 500 response to be a minimalistic one (so that the server doesn't
+  // have to provide additional resources required for its proper rendering)
+  // ";raw=true" in the MIME-type below disables response decoration
+  // (see ContentResponse::contentDecorationAllowed())
+  const std::string mimeType = "text/html;charset=utf-8;raw=true";
+  auto r = ContentResponse::build(m_server, m_template, m_data, mimeType);
+  r->set_code(m_httpStatusCode);
+  return r;
+}
+
+ContentResponseBlueprint& ContentResponseBlueprint::operator+(const TaskbarInfo& taskbarInfo)
+{
+  this->m_taskbarInfo.reset(new TaskbarInfo(taskbarInfo));
+  return *this;
+}
+
+ContentResponseBlueprint& ContentResponseBlueprint::operator+=(const TaskbarInfo& taskbarInfo)
+{
+  // operator+() is already a state-modifying operator (akin to operator+=)
+  return *this + taskbarInfo;
+}
+
+
+std::unique_ptr<Response> Response::build_416(const InternalServer& server, size_t resourceLength)
+{
+  auto response = Response::build(server);
+// [FIXME] (compile with recent enough version of libmicrohttpd)
+//  response->set_code(MHD_HTTP_RANGE_NOT_SATISFIABLE);
+  response->set_code(416);
+  std::ostringstream oss;
+  oss << "bytes */" << resourceLength;
+  response->add_header(MHD_HTTP_HEADER_CONTENT_RANGE, oss.str());
+
+  return response;
+}
+
+
+std::unique_ptr<Response> Response::build_redirect(const InternalServer& server, const std::string& redirectUrl)
+{
+  auto response = Response::build(server);
+  response->m_returnCode = MHD_HTTP_FOUND;
+  response->add_header(MHD_HTTP_HEADER_LOCATION, redirectUrl);
+  return response;
+}
+
+static MHD_Result print_key_value (void *cls, enum MHD_ValueKind kind,
+                                   const char *key, const char *value)
 {
   printf (" - %s: '%s'\n", key, value);
   return MHD_YES;
@@ -74,32 +290,32 @@ static int print_key_value (void *cls, enum MHD_ValueKind kind,
 
 
 struct RunningResponse {
-   kiwix::Entry entry;
+   zim::Item item;
    int range_start;
 
-   RunningResponse(kiwix::Entry entry,
+   RunningResponse(zim::Item item,
                    int range_start) :
-     entry(entry),
+     item(item),
      range_start(range_start)
    {}
 };
 
-static ssize_t callback_reader_from_entry(void* cls,
+static ssize_t callback_reader_from_item(void* cls,
                                   uint64_t pos,
                                   char* buf,
                                   size_t max)
 {
   RunningResponse* response = static_cast<RunningResponse*>(cls);
 
-  size_t max_size_to_set = min<size_t>(
+  size_t max_size_to_set = std::min<size_t>(
     max,
-    response->entry.getSize() - pos - response->range_start);
+    response->item.getSize() - pos - response->range_start);
 
   if (max_size_to_set <= 0) {
     return MHD_CONTENT_READER_END_WITH_ERROR;
   }
 
-  zim::Blob blob = response->entry.getBlob(response->range_start+pos, max_size_to_set);
+  zim::Blob blob = response->item.getData(response->range_start+pos, max_size_to_set);
   memcpy(buf, blob.data(), max_size_to_set);
   return max_size_to_set;
 }
@@ -121,33 +337,24 @@ void print_response_info(int retCode, MHD_Response* response)
 }
 
 
-std::string render_template(const std::string& template_str, kainjow::mustache::data data)
+void ContentResponse::introduce_taskbar(const std::string& lang)
 {
-  kainjow::mustache::mustache tmpl(template_str);
-  kainjow::mustache::data urlencode{kainjow::mustache::lambda2{
-                               [](const std::string& str,const kainjow::mustache::renderer& r) { return urlEncode(r(str), true); }}};
-  data.set("urlencoded", urlencode);
-  std::stringstream ss;
-  tmpl.render(data, [&ss](const std::string& str) { ss << str; });
-  return ss.str();
-}
-
-
-void Response::introduce_taskbar()
-{
-  if (! m_withTaskbar)
-    // Taskbar is globally disabled.
-    return;
-  kainjow::mustache::data data;
-  data.set("root", m_root);
-  data.set("content", m_bookName);
-  data.set("hascontent", !m_bookName.empty());
-  data.set("title", m_bookTitle);
-  data.set("withlibrarybutton", m_withLibraryButton);
-  auto head_content = render_template(RESOURCE::templates::head_part_html, data);
-  m_content = appendToFirstOccurence(
+  i18n::GetTranslatedString t(lang);
+  kainjow::mustache::object data{
+    {"root", m_root},
+    {"content", m_bookName},
+    {"hascontent", (!m_bookName.empty() && !m_bookTitle.empty())},
+    {"title", m_bookTitle},
+    {"withlibrarybutton", m_withLibraryButton},
+    {"LIBRARY_BUTTON_TEXT", t("library-button-text")},
+    {"HOME_BUTTON_TEXT", t("home-button-text", {{"BOOK_TITLE", m_bookTitle}}) },
+    {"RANDOM_PAGE_BUTTON_TEXT", t("random-page-button-text") },
+    {"SEARCHBOX_TOOLTIP", t("searchbox-tooltip", {{"BOOK_TITLE", m_bookTitle}}) },
+  };
+  auto head_content = render_template(RESOURCE::templates::head_taskbar_html, data);
+  m_content = prependToFirstOccurence(
     m_content,
-    "<head>",
+    "</head[ \\t]*>",
     head_content);
 
   auto taskbar_part = render_template(RESOURCE::templates::taskbar_part_html, data);
@@ -158,136 +365,92 @@ void Response::introduce_taskbar()
 }
 
 
-void Response::inject_externallinks_blocker()
+void ContentResponse::inject_externallinks_blocker()
 {
   kainjow::mustache::data data;
   data.set("root", m_root);
   auto script_tag = render_template(RESOURCE::templates::external_blocker_part_html, data);
-  m_content = appendToFirstOccurence(
+  m_content = prependToFirstOccurence(
     m_content,
-    "<head>",
+    "</head[ \\t]*>",
     script_tag);
 }
 
+void ContentResponse::inject_root_link(){
+  m_content = prependToFirstOccurence(
+    m_content,
+    "</head[ \\t]*>",
+    "<link type=\"root\" href=\"" + m_root + "\">");
+}
+
 bool
-Response::can_compress(const RequestContext& request) const
+ContentResponse::can_compress(const RequestContext& request) const
 {
   return request.can_compress()
       && is_compressible_mime_type(m_mimeType)
-      && (m_content.size() > KIWIX_MIN_CONTENT_SIZE_TO_DEFLATE);
+      && (m_content.size() > KIWIX_MIN_CONTENT_SIZE_TO_COMPRESS);
 }
 
-MHD_Response*
-Response::create_raw_content_mhd_response(const RequestContext& request)
+bool
+ContentResponse::contentDecorationAllowed() const
 {
-  if (m_addTaskbar) {
-    introduce_taskbar();
+  if (m_raw) {
+    return false;
   }
-  if ( m_blockExternalLinks ) {
-    inject_externallinks_blocker();
-  }
-
-  bool shouldCompress = m_compress && can_compress(request);
-  if (shouldCompress) {
-    std::vector<Bytef> compr_buffer(compressBound(m_content.size()));
-    uLongf comprLen = compr_buffer.capacity();
-    int err = compress(&compr_buffer[0],
-                       &comprLen,
-                       (const Bytef*)(m_content.data()),
-                       m_content.size());
-    if (err == Z_OK && comprLen > 2 && comprLen < (m_content.size() + 2)) {
-      /* /!\ Internet Explorer has a bug with deflate compression.
-         It can not handle the first two bytes (compression headers)
-         We need to chunk them off (move the content 2bytes)
-         It has no incidence on other browsers
-         See http://www.subbu.org/blog/2008/03/ie7-deflate-or-not and comments */
-      m_content = string((char*)&compr_buffer[2], comprLen - 2);
-      m_etag.set_option(ETag::COMPRESSED_CONTENT);
-    } else {
-      shouldCompress = false;
-    }
-  }
-
-  MHD_Response* response = MHD_create_response_from_buffer(
-    m_content.size(), const_cast<char*>(m_content.data()), MHD_RESPMEM_MUST_COPY);
-
-  // At shis point m_etag.get_option(ETag::COMPRESSED_CONTENT) and
-  // shouldCompress can have different values. This can happen for a 304 (Not
-  // Modified) response generated while handling a conditional If-None-Match
-  // request. In that case the m_etag (together with its COMPRESSED_CONTENT
-  // option) is obtained from the ETag list of the If-None-Match header and the
-  // response has no body (which shouldn't be compressed).
-  if ( m_etag.get_option(ETag::COMPRESSED_CONTENT) ) {
-    MHD_add_response_header(
-        response, MHD_HTTP_HEADER_VARY, "Accept-Encoding");
-  }
-  if (shouldCompress) {
-    MHD_add_response_header(
-        response, MHD_HTTP_HEADER_CONTENT_ENCODING, "deflate");
-  }
-  MHD_add_response_header(response, MHD_HTTP_HEADER_CONTENT_TYPE, m_mimeType.c_str());
-  return response;
-}
-
-MHD_Response*
-Response::create_redirection_mhd_response() const
-{
-  MHD_Response* response = MHD_create_response_from_buffer(0, nullptr, MHD_RESPMEM_MUST_COPY);
-  MHD_add_response_header(response, MHD_HTTP_HEADER_LOCATION, m_content.c_str());
-  return response;
-}
-
-MHD_Response*
-Response::create_entry_mhd_response() const
-{
-  MHD_Response* response = MHD_create_response_from_callback(m_entry.getSize(),
-                                               16384,
-                                               callback_reader_from_entry,
-                                               new RunningResponse(m_entry, m_startRange),
-                                               callback_free_response);
-  MHD_add_response_header(response,
-    MHD_HTTP_HEADER_CONTENT_TYPE, m_mimeType.c_str());
-  MHD_add_response_header(response, MHD_HTTP_HEADER_ACCEPT_RANGES, "bytes");
-  std::ostringstream oss;
-  oss << "bytes " << m_startRange << "-" << m_startRange + m_lenRange - 1
-      << "/" << m_entry.getSize();
-
-  MHD_add_response_header(response,
-    MHD_HTTP_HEADER_CONTENT_RANGE, oss.str().c_str());
-
-  MHD_add_response_header(response,
-    MHD_HTTP_HEADER_CONTENT_LENGTH, kiwix::to_string(m_lenRange).c_str());
-  return response;
+  return (startsWith(m_mimeType, "text/html")
+    && m_mimeType.find(";raw=true") == std::string::npos);
 }
 
 MHD_Response*
 Response::create_mhd_response(const RequestContext& request)
 {
-  switch (m_mode) {
-    case ResponseMode::RAW_CONTENT :
-      return create_raw_content_mhd_response(request);
-
-    case ResponseMode::REDIRECTION :
-      return create_redirection_mhd_response();
-
-    case ResponseMode::ENTRY :
-      return create_entry_mhd_response();
-  }
-  return nullptr;
+  MHD_Response* response = MHD_create_response_from_buffer(0, nullptr, MHD_RESPMEM_PERSISTENT);
+  return response;
 }
 
-int Response::send(const RequestContext& request, MHD_Connection* connection)
+MHD_Response*
+ContentResponse::create_mhd_response(const RequestContext& request)
+{
+  if (contentDecorationAllowed()) {
+    inject_root_link();
+
+    if (m_withTaskbar) {
+      introduce_taskbar(request.get_user_language());
+    }
+    if (m_blockExternalLinks) {
+      inject_externallinks_blocker();
+    }
+  }
+
+  const bool isCompressed = can_compress(request) && compress(m_content);
+
+  MHD_Response* response = MHD_create_response_from_buffer(
+    m_content.size(), const_cast<char*>(m_content.data()), MHD_RESPMEM_MUST_COPY);
+
+  if (isCompressed) {
+    m_etag.set_option(ETag::COMPRESSED_CONTENT);
+    MHD_add_response_header(
+        response, MHD_HTTP_HEADER_VARY, "Accept-Encoding");
+    MHD_add_response_header(
+        response, MHD_HTTP_HEADER_CONTENT_ENCODING, "gzip");
+  }
+  return response;
+}
+
+MHD_Result Response::send(const RequestContext& request, MHD_Connection* connection)
 {
   MHD_Response* response = create_mhd_response(request);
 
-  MHD_add_response_header(response, "Access-Control-Allow-Origin", "*");
   MHD_add_response_header(response, MHD_HTTP_HEADER_CACHE_CONTROL,
     m_etag.get_option(ETag::CACHEABLE_ENTITY) ? "max-age=2723040, public" : "no-cache, no-store, must-revalidate");
   const std::string etag = m_etag.get_etag();
   if ( ! etag.empty() )
-      MHD_add_response_header(response, MHD_HTTP_HEADER_ETAG, etag.c_str());
+    MHD_add_response_header(response, MHD_HTTP_HEADER_ETAG, etag.c_str());
+  for(auto& p: m_customHeaders) {
+    MHD_add_response_header(response, p.first.c_str(), p.second.c_str());
+  }
 
-  if (m_returnCode == MHD_HTTP_OK && request.has_range())
+  if (m_returnCode == MHD_HTTP_OK && m_byteRange.kind() == ByteRange::RESOLVED_PARTIAL_CONTENT)
     m_returnCode = MHD_HTTP_PARTIAL_CONTENT;
 
   if (m_verbose)
@@ -298,47 +461,115 @@ int Response::send(const RequestContext& request, MHD_Connection* connection)
   return ret;
 }
 
-void Response::set_template(const std::string& template_str, kainjow::mustache::data data) {
-  set_content(render_template(template_str, data));
-}
-
-void Response::set_content(const std::string& content) {
-  m_content = content;
-  m_mode = ResponseMode::RAW_CONTENT;
-}
-
-void Response::set_redirection(const std::string& url) {
-  m_content = url;
-  m_mode = ResponseMode::REDIRECTION;
-  m_returnCode = MHD_HTTP_FOUND;
-}
-
-void Response::set_entry(const Entry& entry, const RequestContext& request) {
-  m_entry = entry;
-  m_mode = ResponseMode::ENTRY;
-
-  const std::string mimeType = get_mime_type(entry);
-  set_mimeType(mimeType);
-  set_cacheable();
-
-  if ( is_compressible_mime_type(mimeType) ) {
-    zim::Blob raw_content = entry.getBlob();
-    const std::string content = string(raw_content.data(), raw_content.size());
-
-    set_content(content);
-    set_compress(true);
-  } else {
-    const int range_len = get_range_len(entry, request.get_range());
-    set_range_first(request.get_range().first);
-    set_range_len(range_len);
-  }
-}
-
-void Response::set_taskbar(const std::string& bookName, const std::string& bookTitle)
+void ContentResponse::set_taskbar(const std::string& bookName, const zim::Archive* archive)
 {
-  m_addTaskbar = true;
   m_bookName = bookName;
-  m_bookTitle = bookTitle;
+  m_bookTitle = archive ? getArchiveTitle(*archive) : "";
+}
+
+
+ContentResponse::ContentResponse(const std::string& root, bool verbose, bool raw, bool withTaskbar, bool withLibraryButton, bool blockExternalLinks, const std::string& content, const std::string& mimetype) :
+  Response(verbose),
+  m_root(root),
+  m_content(content),
+  m_mimeType(mimetype),
+  m_raw(raw),
+  m_withTaskbar(withTaskbar),
+  m_withLibraryButton(withLibraryButton),
+  m_blockExternalLinks(blockExternalLinks),
+  m_bookName(""),
+  m_bookTitle("")
+{
+  add_header(MHD_HTTP_HEADER_CONTENT_TYPE, m_mimeType);
+}
+
+std::unique_ptr<ContentResponse> ContentResponse::build(
+  const InternalServer& server,
+  const std::string& content,
+  const std::string& mimetype,
+  bool isHomePage,
+  bool raw)
+{
+   return std::unique_ptr<ContentResponse>(new ContentResponse(
+        server.m_root,
+        server.m_verbose.load(),
+        raw,
+        server.m_withTaskbar && !isHomePage,
+        server.m_withLibraryButton,
+        server.m_blockExternalLinks,
+        content,
+        mimetype));
+}
+
+std::unique_ptr<ContentResponse> ContentResponse::build(
+  const InternalServer& server,
+  const std::string& template_str,
+  kainjow::mustache::data data,
+  const std::string& mimetype,
+  bool isHomePage)
+{
+  auto content = render_template(template_str, data);
+  return ContentResponse::build(server, content, mimetype, isHomePage);
+}
+
+ItemResponse::ItemResponse(bool verbose, const zim::Item& item, const std::string& mimetype, const ByteRange& byterange) :
+  Response(verbose),
+  m_item(item),
+  m_mimeType(mimetype)
+{
+  m_byteRange = byterange;
+  set_cacheable();
+  add_header(MHD_HTTP_HEADER_CONTENT_TYPE, m_mimeType);
+}
+
+std::unique_ptr<Response> ItemResponse::build(const InternalServer& server, const RequestContext& request, const zim::Item& item, bool raw)
+{
+  const std::string mimetype = get_mime_type(item);
+  auto byteRange = request.get_range().resolve(item.getSize());
+  const bool noRange = byteRange.kind() == ByteRange::RESOLVED_FULL_CONTENT;
+  if (noRange && is_compressible_mime_type(mimetype)) {
+    // Return a contentResponse
+    auto response = ContentResponse::build(server, item.getData(), mimetype, /*isHomePage=*/false, raw);
+    response->set_cacheable();
+    response->m_byteRange = byteRange;
+    return std::move(response);
+  }
+
+  if (byteRange.kind() == ByteRange::RESOLVED_UNSATISFIABLE) {
+    auto response = Response::build_416(server, item.getSize());
+    response->set_cacheable();
+    return response;
+  }
+
+  return std::unique_ptr<Response>(new ItemResponse(
+        server.m_verbose.load(),
+        item,
+        mimetype,
+        byteRange));
+}
+
+MHD_Response*
+ItemResponse::create_mhd_response(const RequestContext& request)
+{
+  const auto content_length = m_byteRange.length();
+  MHD_Response* response = MHD_create_response_from_callback(content_length,
+                                               16384,
+                                               callback_reader_from_item,
+                                               new RunningResponse(m_item, m_byteRange.first()),
+                                               callback_free_response);
+  MHD_add_response_header(response, MHD_HTTP_HEADER_ACCEPT_RANGES, "bytes");
+  if ( m_byteRange.kind() == ByteRange::RESOLVED_PARTIAL_CONTENT ) {
+    std::ostringstream oss;
+    oss << "bytes " << m_byteRange.first() << "-" << m_byteRange.last()
+        << "/" << m_item.getSize();
+
+    MHD_add_response_header(response,
+      MHD_HTTP_HEADER_CONTENT_RANGE, oss.str().c_str());
+  }
+
+  MHD_add_response_header(response,
+    MHD_HTTP_HEADER_CONTENT_LENGTH, kiwix::to_string(content_length).c_str());
+  return response;
 }
 
 

src/server/response.h

@@ -22,80 +22,229 @@
 #define KIWIXLIB_SERVER_RESPONSE_H
 
 #include <string>
+#include <map>
 
 #include <mustache.hpp>
-#include "entry.h"
+#include "byte_range.h"
 #include "etag.h"
+#include "i18n.h"
+
+#include <zim/item.h>
 
 extern "C" {
-#include <microhttpd.h>
+#include "microhttpd_wrapper.h"
 }
 
+namespace zim {
+class Archive;
+} // namespace zim
+
 namespace kiwix {
 
-enum class ResponseMode {
-  RAW_CONTENT,
-  REDIRECTION,
-  ENTRY
-};
-
+class InternalServer;
 class RequestContext;
 
 class Response {
   public:
-    Response(const std::string& root, bool verbose, bool withTaskbar, bool withLibraryButton, bool blockExternalLinks);
-    ~Response() = default;
+    Response(bool verbose);
+    virtual ~Response() = default;
 
-    int send(const RequestContext& request, MHD_Connection* connection);
+    static std::unique_ptr<Response> build(const InternalServer& server);
+    static std::unique_ptr<Response> build_304(const InternalServer& server, const ETag& etag);
+    static std::unique_ptr<Response> build_416(const InternalServer& server, size_t resourceLength);
+    static std::unique_ptr<Response> build_redirect(const InternalServer& server, const std::string& redirectUrl);
 
-    void set_template(const std::string& template_str, kainjow::mustache::data data);
-    void set_content(const std::string& content);
-    void set_redirection(const std::string& url);
-    void set_entry(const Entry& entry, const RequestContext& request);
+    MHD_Result send(const RequestContext& request, MHD_Connection* connection);
 
-
-    void set_mimeType(const std::string& mimeType) { m_mimeType = mimeType; }
     void set_code(int code) { m_returnCode = code; }
     void set_cacheable() { m_etag.set_option(ETag::CACHEABLE_ENTITY); }
     void set_server_id(const std::string& id) { m_etag.set_server_id(id); }
-    void set_etag(const ETag& etag) { m_etag = etag; }
-    void set_compress(bool compress) { m_compress = compress; }
-    void set_taskbar(const std::string& bookName, const std::string& bookTitle);
-    void set_range_first(uint64_t start) { m_startRange = start; }
-    void set_range_len(uint64_t len) { m_lenRange = len; }
+    void add_header(const std::string& name, const std::string& value) { m_customHeaders[name] = value; }
 
     int getReturnCode() const { return m_returnCode; }
-    std::string get_mimeType() const { return m_mimeType; }
-
-    void introduce_taskbar();
-    void inject_externallinks_blocker();
-
-    bool can_compress(const RequestContext& request) const;
 
   private: // functions
-    MHD_Response* create_mhd_response(const RequestContext& request);
-    MHD_Response* create_raw_content_mhd_response(const RequestContext& request);
-    MHD_Response* create_redirection_mhd_response() const;
-    MHD_Response* create_entry_mhd_response() const;
+    virtual MHD_Response* create_mhd_response(const RequestContext& request);
+    MHD_Response* create_error_response(const RequestContext& request) const;
 
-  private: // data
+  protected: // data
     bool m_verbose;
-    ResponseMode m_mode;
+    int m_returnCode;
+    ByteRange m_byteRange;
+    ETag m_etag;
+    std::map<std::string, std::string> m_customHeaders;
+
+    friend class ItemResponse;
+};
+
+
+class ContentResponse : public Response {
+  public:
+    ContentResponse(
+      const std::string& root,
+      bool verbose,
+      bool raw,
+      bool withTaskbar,
+      bool withLibraryButton,
+      bool blockExternalLinks,
+      const std::string& content,
+      const std::string& mimetype);
+    static std::unique_ptr<ContentResponse> build(
+      const InternalServer& server,
+      const std::string& content,
+      const std::string& mimetype,
+      bool isHomePage = false,
+      bool raw = false);
+    static std::unique_ptr<ContentResponse> build(
+      const InternalServer& server,
+      const std::string& template_str,
+      kainjow::mustache::data data,
+      const std::string& mimetype,
+      bool isHomePage = false);
+
+    void set_taskbar(const std::string& bookName, const zim::Archive* archive);
+
+  private:
+    MHD_Response* create_mhd_response(const RequestContext& request);
+
+    void introduce_taskbar(const std::string& lang);
+    void inject_externallinks_blocker();
+    void inject_root_link();
+    bool can_compress(const RequestContext& request) const;
+    bool contentDecorationAllowed() const;
+
+
+  private:
     std::string m_root;
     std::string m_content;
-    Entry m_entry;
     std::string m_mimeType;
-    int m_returnCode;
+    bool m_raw;
     bool m_withTaskbar;
     bool m_withLibraryButton;
     bool m_blockExternalLinks;
-    bool m_compress;
-    bool m_addTaskbar;
     std::string m_bookName;
     std::string m_bookTitle;
-    uint64_t m_startRange;
-    uint64_t m_lenRange;
-    ETag m_etag;
+ };
+
+struct TaskbarInfo
+{
+  const std::string bookName;
+  const zim::Archive* const archive;
+
+  TaskbarInfo(const std::string& bookName, const zim::Archive* a = nullptr)
+    : bookName(bookName)
+    , archive(a)
+  {}
+};
+
+class ContentResponseBlueprint
+{
+public: // functions
+  ContentResponseBlueprint(const InternalServer* server,
+                           const RequestContext* request,
+                           int httpStatusCode,
+                           const std::string& mimeType,
+                           const std::string& templateStr)
+    : m_server(*server)
+    , m_request(*request)
+    , m_httpStatusCode(httpStatusCode)
+    , m_mimeType(mimeType)
+    , m_template(templateStr)
+  {}
+
+  virtual ~ContentResponseBlueprint() = default;
+
+  operator std::unique_ptr<ContentResponse>() const
+  {
+    return generateResponseObject();
+  }
+
+  operator std::unique_ptr<Response>() const
+  {
+    return operator std::unique_ptr<ContentResponse>();
+  }
+
+
+  ContentResponseBlueprint& operator+(const TaskbarInfo& taskbarInfo);
+  ContentResponseBlueprint& operator+=(const TaskbarInfo& taskbarInfo);
+
+protected: // functions
+  std::string getMessage(const std::string& msgId) const;
+  virtual std::unique_ptr<ContentResponse> generateResponseObject() const;
+
+public: //data
+  const InternalServer& m_server;
+  const RequestContext& m_request;
+  const int m_httpStatusCode;
+  const std::string m_mimeType;
+  const std::string m_template;
+  kainjow::mustache::data m_data;
+  std::unique_ptr<TaskbarInfo> m_taskbarInfo;
+};
+
+struct HTTPErrorResponse : ContentResponseBlueprint
+{
+  HTTPErrorResponse(const InternalServer& server,
+                    const RequestContext& request,
+                    int httpStatusCode,
+                    const std::string& pageTitleMsgId,
+                    const std::string& headingMsgId,
+                    const std::string& cssUrl = "");
+
+  using ContentResponseBlueprint::operator+;
+  using ContentResponseBlueprint::operator+=;
+  HTTPErrorResponse& operator+(const std::string& msg);
+  HTTPErrorResponse& operator+(const ParameterizedMessage& errorDetails);
+  HTTPErrorResponse& operator+=(const ParameterizedMessage& errorDetails);
+};
+
+class UrlNotFoundMsg {};
+
+extern const UrlNotFoundMsg urlNotFoundMsg;
+
+struct HTTP404Response : HTTPErrorResponse
+{
+  HTTP404Response(const InternalServer& server,
+                  const RequestContext& request);
+
+  using HTTPErrorResponse::operator+;
+  HTTPErrorResponse& operator+(UrlNotFoundMsg /*unused*/);
+};
+
+class InvalidUrlMsg {};
+
+extern const InvalidUrlMsg invalidUrlMsg;
+
+struct HTTP400Response : HTTPErrorResponse
+{
+  HTTP400Response(const InternalServer& server,
+                  const RequestContext& request);
+
+  using HTTPErrorResponse::operator+;
+  HTTPErrorResponse& operator+(InvalidUrlMsg /*unused*/);
+};
+
+struct HTTP500Response : HTTPErrorResponse
+{
+  HTTP500Response(const InternalServer& server,
+                  const RequestContext& request);
+
+private: // overrides
+  // generateResponseObject() is overriden in order to produce a minimal
+  // response without any need for additional resources from the server
+  std::unique_ptr<ContentResponse> generateResponseObject() const override;
+};
+
+class ItemResponse : public Response {
+  public:
+    ItemResponse(bool verbose, const zim::Item& item, const std::string& mimetype, const ByteRange& byterange);
+    static std::unique_ptr<Response> build(const InternalServer& server, const RequestContext& request, const zim::Item& item, bool raw = false);
+
+  private:
+    MHD_Response* create_mhd_response(const RequestContext& request);
+
+    zim::Item m_item;
+    std::string m_mimeType;
 };
 
 }

src/subprocess_unix.cpp

@@ -12,45 +12,31 @@
 UnixImpl::UnixImpl():
   m_pid(0),
   m_running(false),
-  m_mutex(PTHREAD_MUTEX_INITIALIZER),
-  m_waitingThread()
+  m_shouldQuit(false)
 {
 }
 
 UnixImpl::~UnixImpl()
 {
   kill();
-// Android has no pthread_cancel :(
-#ifdef __ANDROID__
-  pthread_kill(m_waitingThread, SIGUSR1);
-#else
-  pthread_cancel(m_waitingThread);
-#endif
+  m_shouldQuit = true;
+  m_waitingThread.join();
 }
 
-#ifdef __ANDROID__
-void thread_exit_handler(int sig) {
-  pthread_exit(0);
-}
-#endif
-
 void* UnixImpl::waitForPID(void* _self)
 {
-#ifdef __ANDROID__
-  struct sigaction actions;
-  memset(&actions, 0, sizeof(actions));
-  sigemptyset(&actions.sa_mask);
-  actions.sa_flags = 0;
-  actions.sa_handler = thread_exit_handler;
-  sigaction(SIGUSR1, &actions, NULL);
-#endif
-
   UnixImpl* self = static_cast<UnixImpl*>(_self);
-  waitpid(self->m_pid, NULL, 0);
+  while (true) {
+    if (!waitpid(self->m_pid, NULL, WNOHANG)) {
+      break;
+    }
+    if (self->m_shouldQuit) {
+      return nullptr;
+    }
+    std::this_thread::sleep_for(std::chrono::milliseconds(100));
+  }
 
-  pthread_mutex_lock(&self->m_mutex);
   self->m_running = false;
-  pthread_mutex_unlock(&self->m_mutex);
 
   return self;
 }
@@ -74,7 +60,7 @@ void UnixImpl::run(commandLine_t& commandLine)
     default:
       m_pid = pid;
       m_running = true;
-      pthread_create(&m_waitingThread, NULL, waitForPID, this);
+      m_waitingThread = std::thread(waitForPID, this);
       break;
   }
 }
@@ -86,8 +72,5 @@ bool UnixImpl::kill()
 
 bool UnixImpl::isRunning()
 {
-  pthread_mutex_lock(&m_mutex);
-  bool ret = m_running;
-  pthread_mutex_unlock(&m_mutex);
-  return ret;
+  return m_running;
 }

src/subprocess_unix.h

@@ -3,16 +3,16 @@
 
 #include "subprocess.h"
 
-#include <pthread.h>
-
+#include <atomic>
+#include <thread>
 
 class UnixImpl : public SubprocessImpl
 {
   private:
     int m_pid;
-    bool m_running;
-    pthread_mutex_t m_mutex;
-    pthread_t m_waitingThread;
+    std::atomic<bool> m_running;
+    std::atomic<bool> m_shouldQuit;
+    std::thread m_waitingThread;
 
   public:
     UnixImpl();

src/subprocess_windows.cpp

@@ -11,7 +11,8 @@
 WinImpl::WinImpl():
   m_pid(0),
   m_running(false),
-  m_handle(INVALID_HANDLE_VALUE)
+  m_subprocessHandle(INVALID_HANDLE_VALUE),
+  m_waitingThreadHandle(INVALID_HANDLE_VALUE)
 {
   InitializeCriticalSection(&m_criticalSection);
 }
@@ -19,14 +20,15 @@ WinImpl::WinImpl():
 WinImpl::~WinImpl()
 {
   kill();
-  CloseHandle(m_handle);
+  WaitForSingleObject(m_waitingThreadHandle, INFINITE);
+  CloseHandle(m_subprocessHandle);
   DeleteCriticalSection(&m_criticalSection);
 }
 
 DWORD WINAPI WinImpl::waitForPID(void* _self)
 {
   WinImpl* self = static_cast<WinImpl*>(_self);
-  WaitForSingleObject(self->m_handle, INFINITE);
+  WaitForSingleObject(self->m_subprocessHandle, INFINITE);
 
   EnterCriticalSection(&self->m_criticalSection);
   self->m_running = false;
@@ -79,16 +81,16 @@ void WinImpl::run(commandLine_t& commandLine)
     &procInfo))
   {
     m_pid = procInfo.dwProcessId;
-    m_handle = procInfo.hProcess;
+    m_subprocessHandle = procInfo.hProcess;
     CloseHandle(procInfo.hThread);
     m_running = true;
-    CreateThread(NULL, 0, &waitForPID, this, 0, NULL	);
+    m_waitingThreadHandle = CreateThread(NULL, 0, &waitForPID, this, 0, NULL);
   }
 }
 
 bool WinImpl::kill()
 {
-  return TerminateProcess(m_handle, 0);
+  return TerminateProcess(m_subprocessHandle, 0);
 }
 
 bool WinImpl::isRunning()

src/subprocess_windows.h

@@ -11,7 +11,8 @@ class WinImpl : public SubprocessImpl
   private:
     int m_pid;
     bool m_running;
-    HANDLE m_handle;
+    HANDLE m_subprocessHandle;
+    HANDLE  m_waitingThreadHandle;
     CRITICAL_SECTION m_criticalSection;
 
   public:

src/tools/archiveTools.cpp

@@ -0,0 +1,182 @@
+/*
+ * Copyright 2021 Maneesh P M <manu.pm55@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#include "archiveTools.h"
+
+#include "tools.h"
+#include "pathTools.h"
+#include "otherTools.h"
+#include "stringTools.h"
+
+#include <zim/error.h>
+#include <zim/item.h>
+
+namespace kiwix
+{
+std::string getMetadata(const zim::Archive& archive, const std::string& name) {
+    try {
+        return archive.getMetadata(name);
+    } catch (zim::EntryNotFound& e) {
+        return "";
+    }
+}
+
+std::string getArchiveTitle(const zim::Archive& archive) {
+  std::string value = getMetadata(archive, "Title");
+  if (value.empty()) {
+    value = getLastPathElement(archive.getFilename());
+    std::replace(value.begin(), value.end(), '_', ' ');
+    size_t pos = value.find(".zim");
+    value = value.substr(0, pos);
+  }
+  return value;
+}
+
+std::string getMetaDescription(const zim::Archive& archive) {
+  std::string value;
+  value = getMetadata(archive, "Description");
+
+  /* Mediawiki Collection tends to use the "Subtitle" name */
+  if (value.empty()) {
+    value = getMetadata(archive, "Subtitle");
+  }
+
+  return value;
+}
+
+std::string getMetaTags(const zim::Archive& archive, bool original) {
+  std::string tags_str = getMetadata(archive, "Tags");
+  if (original) {
+    return tags_str;
+  }
+  auto tags = convertTags(tags_str);
+  return join(tags, ";");
+}
+
+std::string getMetaLanguage(const zim::Archive& archive) {
+  return getMetadata(archive, "Language");
+}
+
+std::string getMetaName(const zim::Archive& archive) {
+  return getMetadata(archive, "Name");
+}
+
+std::string getMetaDate(const zim::Archive& archive) {
+  return getMetadata(archive, "Date");
+}
+
+std::string getMetaCreator(const zim::Archive& archive) {
+  return getMetadata(archive, "Creator");
+}
+
+std::string getMetaPublisher(const zim::Archive& archive) {
+  return getMetadata(archive, "Publisher");
+}
+
+std::string getMetaFlavour(const zim::Archive& archive) {
+  return getMetadata(archive, "Flavour");
+}
+
+std::string getArchiveId(const zim::Archive& archive) {
+  return (std::string) archive.getUuid();
+}
+
+bool getArchiveFavicon(const zim::Archive& archive, unsigned size,
+                           std::string& content, std::string& mimeType){
+  try {
+    auto item = archive.getIllustrationItem(size);
+    content = item.getData();
+    mimeType = item.getMimetype();
+    return true;
+  } catch(zim::EntryNotFound& e) {};
+
+  return false;
+}
+
+// should this be in libzim
+unsigned int getArchiveMediaCount(const zim::Archive& archive) {
+  std::map<const std::string, unsigned int> counterMap = parseArchiveCounter(archive);
+  unsigned int counter = 0;
+
+  for (auto &pair:counterMap) {
+    if (startsWith(pair.first, "image/") ||
+        startsWith(pair.first, "video/") ||
+        startsWith(pair.first, "audio/")) {
+      counter += pair.second;
+    }
+  }
+
+  return counter;
+}
+
+unsigned int getArchiveArticleCount(const zim::Archive& archive) {
+  // [HACK]
+  // getArticleCount() returns different things depending of the "version" of the zim.
+  // On old zim (<=6), it returns the number of entry in `A` namespace
+  // On recent zim (>=7), it returns:
+  //  - the number of entry in `C` namespace (==getEntryCount) if no frontArticleIndex is present
+  //  - the number of front article if a frontArticleIndex is present
+  // The use case >=7 without frontArticleIndex is pretty rare so we don't care
+  // We can detect if we are reading a zim <= 6 by checking if we have a newNamespaceScheme.
+  if (archive.hasNewNamespaceScheme()) {
+    //The articleCount is "good"
+    return archive.getArticleCount();
+  } else {
+    // We have to parse the `M/Counter` metadata
+    unsigned int counter = 0;
+    for(const auto& pair:parseArchiveCounter(archive)) {
+      if (startsWith(pair.first, "text/html")) {
+        counter += pair.second;
+      }
+    }
+    return counter;
+  }
+}
+
+unsigned int getArchiveFileSize(const zim::Archive& archive) {
+  return archive.getFilesize() / 1024;
+}
+
+zim::Item getFinalItem(const zim::Archive& archive, const zim::Entry& entry)
+{
+  return entry.getItem(true);
+}
+
+zim::Entry getEntryFromPath(const zim::Archive& archive, const std::string& path)
+{
+  try {
+    return archive.getEntryByPath(path);
+  } catch (zim::EntryNotFound& e) {
+    if (path.empty() || path == "/") {
+      return archive.getMainEntry();
+    }
+  }
+  throw zim::EntryNotFound("Cannot find entry for non empty path");
+}
+
+MimeCounterType parseArchiveCounter(const zim::Archive& archive) {
+  try {
+    auto counterContent = archive.getMetadata("Counter");
+    return parseMimetypeCounter(counterContent);
+  } catch (zim::EntryNotFound& e) {
+    return {};
+  }
+}
+
+} // kiwix

src/tools/archiveTools.h

@@ -0,0 +1,60 @@
+/*
+ * Copyright 2021 Maneesh P M <manu.pm55@gmail.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#ifndef KIWIX_ARCHIVETOOLS_H
+#define KIWIX_ARCHIVETOOLS_H
+
+#include <zim/archive.h>
+#include <tools/otherTools.h>
+
+/**
+ * This file contains all the functions that would make handling data related to
+ * an archive easier.
+ **/
+
+namespace kiwix
+{
+    std::string getMetadata(const zim::Archive& archive, const std::string& name);
+    std::string getArchiveTitle(const zim::Archive& archive);
+    std::string getMetaDescription(const zim::Archive& archive);
+    std::string getMetaTags(const zim::Archive& archive, bool original = false);
+    std::string getMetaLanguage(const zim::Archive& archive);
+    std::string getMetaName(const zim::Archive& archive);
+    std::string getMetaDate(const zim::Archive& archive);
+    std::string getMetaCreator(const zim::Archive& archive);
+    std::string getMetaPublisher(const zim::Archive& archive);
+    std::string getMetaFlavour(const zim::Archive& archive);
+    std::string getArchiveId(const zim::Archive& archive);
+
+    bool getArchiveFavicon(const zim::Archive& archive, unsigned size,
+                           std::string& content, std::string& mimeType);
+
+    unsigned int getArchiveMediaCount(const zim::Archive& archive);
+    unsigned int getArchiveArticleCount(const zim::Archive& archive);
+    unsigned int getArchiveFileSize(const zim::Archive& archive);
+
+    zim::Item getFinalItem(const zim::Archive& archive, const zim::Entry& entry);
+
+    zim::Entry getEntryFromPath(const zim::Archive& archive, const std::string& path);
+
+    MimeCounterType parseArchiveCounter(const zim::Archive& archive);
+
+}
+
+#endif

src/tools/concurrent_cache.h

@@ -0,0 +1,208 @@
+
+/*
+ * Copyright (C) 2021 Matthieu Gautier <mgautier@kymeria.fr>
+ * Copyright (C) 2020 Veloman Yunkan
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * is provided AS IS, WITHOUT ANY WARRANTY; without even the implied
+ * warranty of MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, and
+ * NON-INFRINGEMENT.  See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA
+ *
+ */
+
+#ifndef ZIM_CONCURRENT_CACHE_H
+#define ZIM_CONCURRENT_CACHE_H
+
+#include "lrucache.h"
+
+#include <future>
+#include <mutex>
+
+namespace kiwix
+{
+
+/**
+   ConcurrentCache implements a concurrent thread-safe cache
+
+   Compared to kiwix::lru_cache, each access operation is slightly more expensive.
+   However, different slots of the cache can be safely accessed concurrently
+   with minimal blocking. Concurrent access to the same element is also
+   safe, and, in case of a cache miss, will block until that element becomes
+   available.
+ */
+template <typename Key, typename Value>
+class ConcurrentCache
+{
+private: // types
+  typedef std::shared_future<Value> ValuePlaceholder;
+  typedef lru_cache<Key, ValuePlaceholder> Impl;
+
+public: // types
+  explicit ConcurrentCache(size_t maxEntries)
+    : impl_(maxEntries)
+  {}
+
+  // Gets the entry corresponding to the given key. If the entry is not in the
+  // cache, it is obtained by calling f() (without any arguments) and the
+  // result is put into the cache.
+  //
+  // The cache as a whole is locked only for the duration of accessing
+  // the respective slot. If, in the case of the a cache miss, the generation
+  // of the missing element takes a long time, only attempts to access that
+  // element will block - the rest of the cache remains open to concurrent
+  // access.
+  template<class F>
+  Value getOrPut(const Key& key, F f)
+  {
+    std::promise<Value> valuePromise;
+    std::unique_lock<std::mutex> l(lock_);
+    const auto x = impl_.getOrPut(key, valuePromise.get_future().share());
+    l.unlock();
+    if ( x.miss() ) {
+      try {
+        valuePromise.set_value(f());
+      } catch (std::exception& e) {
+        drop(key);
+        throw;
+      }
+    }
+
+    return x.value().get();
+  }
+
+  bool drop(const Key& key)
+  {
+    std::unique_lock<std::mutex> l(lock_);
+    return impl_.drop(key);
+  }
+
+  size_t setMaxSize(size_t new_size) {
+    std::unique_lock<std::mutex> l(lock_);
+    return  impl_.setMaxSize(new_size);
+  }
+
+protected: // data
+  Impl impl_;
+  std::mutex lock_;
+};
+
+
+/**
+  WeakStore represent a thread safe store (map) of weak ptr.
+  It allows to store weak_ptr from shared_ptr and retrieve shared_ptr from
+  potential non expired weak_ptr.
+  It is not limited in size.
+*/
+template<typename Key, typename Value>
+class WeakStore {
+  private: // types
+    typedef std::weak_ptr<Value> WeakValue;
+
+  public:
+    explicit WeakStore() = default;
+
+    std::shared_ptr<Value> get(const Key& key)
+    {
+      std::lock_guard<std::mutex> l(m_lock);
+      auto it = m_weakMap.find(key);
+      if (it != m_weakMap.end()) {
+        auto shared = it->second.lock();
+        if (shared) {
+          return shared;
+        } else {
+          m_weakMap.erase(it);
+        }
+      }
+      throw std::runtime_error("No weak ptr");
+    }
+
+    void add(const Key& key, std::shared_ptr<Value> shared)
+    {
+      std::lock_guard<std::mutex> l(m_lock);
+      m_weakMap[key] = WeakValue(shared);
+    }
+
+
+  private: //data
+    std::map<Key, WeakValue> m_weakMap;
+    std::mutex m_lock;
+};
+
+template <typename Key, typename RawValue>
+class ConcurrentCache<Key, std::shared_ptr<RawValue>>
+{
+private: // types
+  typedef std::shared_ptr<RawValue> Value;
+  typedef std::shared_future<Value> ValuePlaceholder;
+  typedef lru_cache<Key, ValuePlaceholder> Impl;
+
+public: // types
+  explicit ConcurrentCache(size_t maxEntries)
+    : impl_(maxEntries)
+  {}
+
+  // Gets the entry corresponding to the given key. If the entry is not in the
+  // cache, it is obtained by calling f() (without any arguments) and the
+  // result is put into the cache.
+  //
+  // The cache as a whole is locked only for the duration of accessing
+  // the respective slot. If, in the case of the a cache miss, the generation
+  // of the missing element takes a long time, only attempts to access that
+  // element will block - the rest of the cache remains open to concurrent
+  // access.
+  template<class F>
+  Value getOrPut(const Key& key, F f)
+  {
+    std::promise<Value> valuePromise;
+    std::unique_lock<std::mutex> l(lock_);
+    const auto x = impl_.getOrPut(key, valuePromise.get_future().share());
+    l.unlock();
+    if ( x.miss() ) {
+      // Try to get back the shared_ptr from the weak_ptr first.
+      try {
+        valuePromise.set_value(m_weakStore.get(key));
+      } catch(const std::runtime_error& e) {
+        try {
+          const auto value = f();
+          valuePromise.set_value(value);
+          m_weakStore.add(key, value);
+        } catch (std::exception& e) {
+          drop(key);
+          throw;
+        }
+      }
+    }
+
+    return x.value().get();
+  }
+
+  bool drop(const Key& key)
+  {
+    std::unique_lock<std::mutex> l(lock_);
+    return impl_.drop(key);
+  }
+
+  size_t setMaxSize(size_t new_size) {
+    std::unique_lock<std::mutex> l(lock_);
+    return  impl_.setMaxSize(new_size);
+  }
+
+protected: // data
+  std::mutex lock_;
+  Impl impl_;
+  WeakStore<Key, RawValue> m_weakStore;
+};
+
+} // namespace kiwix
+
+#endif // ZIM_CONCURRENT_CACHE_H
+

src/tools/lrucache.h

@@ -0,0 +1,175 @@
+/*
+ * Copyrigth (c) 2021, Matthieu Gautier <mgautier@kymeria.fr>
+ * Copyright (c) 2020, Veloman Yunkan
+ * Copyright (c) 2014, lamerman
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * * Redistributions of source code must retain the above copyright notice, this
+ *   list of conditions and the following disclaimer.
+ *
+ * * Redistributions in binary form must reproduce the above copyright notice,
+ *   this list of conditions and the following disclaimer in the documentation
+ *   and/or other materials provided with the distribution.
+ *
+ * * Neither the name of lamerman nor the names of its
+ *   contributors may be used to endorse or promote products derived from
+ *   this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+ * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+ * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+ * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+ * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ * File:   lrucache.hpp
+ * Author: Alexander Ponomarev
+ *
+ * Created on June 20, 2013, 5:09 PM
+ */
+
+#ifndef _LRUCACHE_HPP_INCLUDED_
+#define _LRUCACHE_HPP_INCLUDED_
+
+#include <map>
+#include <list>
+#include <set>
+#include <cstddef>
+#include <stdexcept>
+#include <cassert>
+
+namespace kiwix {
+
+template<typename key_t, typename value_t>
+class lru_cache {
+public: // types
+  typedef typename std::pair<key_t, value_t> key_value_pair_t;
+  typedef typename std::list<key_value_pair_t>::iterator list_iterator_t;
+
+  enum AccessStatus {
+    HIT, // key was found in the cache
+    PUT, // key was not in the cache but was created by the getOrPut() access
+    MISS // key was not in the cache; get() access failed
+  };
+
+  class AccessResult
+  {
+    const AccessStatus status_;
+    const value_t val_;
+  public:
+    AccessResult(const value_t& val, AccessStatus status)
+      : status_(status), val_(val)
+    {}
+    AccessResult() : status_(MISS), val_() {}
+
+    bool hit() const { return status_ == HIT; }
+    bool miss() const { return !hit(); }
+    const value_t& value() const
+    {
+      if ( status_ == MISS )
+        throw std::range_error("There is no such key in cache");
+      return val_;
+    }
+
+    operator const value_t& () const { return value(); }
+  };
+
+public: // functions
+  explicit lru_cache(size_t max_size) :
+    _max_size(max_size) {
+  }
+
+  // If 'key' is present in the cache, returns the associated value,
+  // otherwise puts the given value into the cache (and returns it with
+  // a status of a cache miss).
+  AccessResult getOrPut(const key_t& key, const value_t& value) {
+    auto it = _cache_items_map.find(key);
+    if (it != _cache_items_map.end()) {
+      _cache_items_list.splice(_cache_items_list.begin(), _cache_items_list, it->second);
+      return AccessResult(it->second->second, HIT);
+    } else {
+      putMissing(key, value);
+      return AccessResult(value, PUT);
+    }
+  }
+
+  void put(const key_t& key, const value_t& value) {
+    auto it = _cache_items_map.find(key);
+    if (it != _cache_items_map.end()) {
+      _cache_items_list.splice(_cache_items_list.begin(), _cache_items_list, it->second);
+      it->second->second = value;
+    } else {
+      putMissing(key, value);
+    }
+  }
+
+  AccessResult get(const key_t& key) {
+    auto it = _cache_items_map.find(key);
+    if (it == _cache_items_map.end()) {
+      return AccessResult();
+    } else {
+      _cache_items_list.splice(_cache_items_list.begin(), _cache_items_list, it->second);
+      return AccessResult(it->second->second, HIT);
+    }
+  }
+
+  bool drop(const key_t& key) {
+    try {
+      auto list_it = _cache_items_map.at(key);
+      _cache_items_list.erase(list_it);
+      _cache_items_map.erase(key);
+      return true;
+    } catch (std::out_of_range& e) {
+      return false;
+    }
+  }
+
+  bool exists(const key_t& key) const {
+    return _cache_items_map.find(key) != _cache_items_map.end();
+  }
+
+  size_t size() const {
+    return _cache_items_map.size();
+  }
+
+  size_t setMaxSize(size_t new_size) {
+    size_t previous = _max_size;
+    _max_size = new_size;
+    return previous;
+  }
+
+  std::set<key_t> keys() const  {
+    std::set<key_t> keys;
+    for(auto& item:_cache_items_map) {
+      keys.insert(item.first);
+    }
+    return keys;
+  }
+
+private: // functions
+  void putMissing(const key_t& key, const value_t& value) {
+    assert(_cache_items_map.find(key) == _cache_items_map.end());
+    _cache_items_list.push_front(key_value_pair_t(key, value));
+    _cache_items_map[key] = _cache_items_list.begin();
+    while (_cache_items_map.size() > _max_size) {
+      _cache_items_map.erase(_cache_items_list.back().first);
+      _cache_items_list.pop_back();
+    }
+  }
+
+private: // data
+  std::list<key_value_pair_t> _cache_items_list;
+  std::map<key_t, list_iterator_t> _cache_items_map;
+  size_t _max_size;
+};
+
+} // namespace kiwix
+
+#endif  /* _LRUCACHE_HPP_INCLUDED_ */

src/tools/networkTools.cpp

@@ -1,5 +1,6 @@
 /*
  * Copyright 2012 Emmanuel Engelhart <kelson@kiwix.org>
+ * Copyright 2021 Nikhil Tanwar <2002nikhiltanwar@gmail.com>
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU  General Public License as published by
@@ -17,6 +18,7 @@
  * MA 02110-1301, USA.
  */
 
+#include "tools.h"
 #include <tools/networkTools.h>
 
 #include <stdio.h>
@@ -29,6 +31,17 @@
 #include <iostream>
 #include <stdexcept>
 
+#ifdef _WIN32
+#include <winsock2.h>
+#include <ws2tcpip.h>
+#include <iostream>
+#else
+#include <unistd.h>
+#include <sys/ioctl.h>
+#include <sys/socket.h>
+#include <net/if.h>
+#include <netdb.h>
+#endif
 
 size_t write_callback_to_iss(char* ptr, size_t size, size_t nmemb, void* userdata)
 {
@@ -57,3 +70,104 @@ std::string kiwix::download(const std::string& url) {
   }
   return ss.str();
 }
+
+std::map<std::string, std::string> kiwix::getNetworkInterfaces() {
+  std::map<std::string, std::string> interfaces;
+
+#ifdef _WIN32
+  SOCKET sd = WSASocket(AF_INET, SOCK_DGRAM, 0, 0, 0, 0);
+  if (sd == INVALID_SOCKET) {
+    std::cerr << "Failed to get a socket. Error " << WSAGetLastError() << std::endl;
+    return interfaces;
+  }
+
+  INTERFACE_INFO InterfaceList[20];
+  unsigned long nBytesReturned;
+  if (WSAIoctl(sd, SIO_GET_INTERFACE_LIST, 0, 0, &InterfaceList,
+      sizeof(InterfaceList), &nBytesReturned, 0, 0) == SOCKET_ERROR) {
+    std::cerr << "Failed calling WSAIoctl: error " << WSAGetLastError() << std::endl;
+    return interfaces;
+  }
+
+  int nNumInterfaces = nBytesReturned / sizeof(INTERFACE_INFO);
+  for (int i = 0; i < nNumInterfaces; ++i) {
+    sockaddr_in *pAddress;
+    pAddress = (sockaddr_in *) & (InterfaceList[i].iiAddress.AddressIn);
+    if(pAddress->sin_family == AF_INET) {
+    /* Add to the map */
+    std::string interfaceName = std::string(inet_ntoa(pAddress->sin_addr));
+    interfaces[interfaceName] = interfaceName;
+    }
+  }
+#else
+  /* Get Network interfaces information */
+  char buf[16384];
+  struct ifconf ifconf;
+  int fd = socket(PF_INET, SOCK_DGRAM, 0); /* Only IPV4 */
+  ifconf.ifc_len = sizeof(buf);
+  ifconf.ifc_buf=buf;
+  if(ioctl(fd, SIOCGIFCONF, &ifconf)!=0) {
+    perror("ioctl(SIOCGIFCONF)");
+  }
+
+  /* Go through each interface */
+  struct ifreq *ifreq;
+  ifreq = ifconf.ifc_req;
+  for (int i = 0; i < ifconf.ifc_len; ) {
+    if (ifreq->ifr_addr.sa_family == AF_INET) {
+      /* Get the network interface ip */
+      char host[128] = { 0 };
+      const int error = getnameinfo(&(ifreq->ifr_addr), sizeof(ifreq->ifr_addr),
+                                    host, sizeof(host),
+                                    0, 0, NI_NUMERICHOST);
+      if (!error) {
+        std::string interfaceName = std::string(ifreq->ifr_name);
+        std::string interfaceIp = std::string(host);
+        /* Add to the map */
+        interfaces[interfaceName] = interfaceIp;
+      } else {
+        perror("getnameinfo()");
+      }
+    }
+
+    /* some systems have ifr_addr.sa_len and adjust the length that
+     * way, but not mine. weird */
+    size_t len;
+#ifndef __linux__
+    len = IFNAMSIZ + ifreq->ifr_addr.sa_len;
+#else
+    len = sizeof(*ifreq);
+#endif
+    ifreq = (struct ifreq*)((char*)ifreq+len);
+    i += len;
+  }
+#endif
+  return interfaces;
+}
+
+std::string kiwix::getBestPublicIp() {
+  auto interfaces = getNetworkInterfaces();
+
+#ifndef _WIN32
+  const char* const prioritizedNames[] =
+      { "eth0", "eth1", "wlan0", "wlan1", "en0", "en1" };
+  for(auto name: prioritizedNames) {
+    auto it = interfaces.find(name);
+    if(it != interfaces.end()) {
+      return it->second;
+    }
+  }
+#endif
+
+  const char* const prefixes[] = { "192.168", "172.16.", "10.0" };
+  for(auto prefix : prefixes){
+    for(auto& itr : interfaces) {
+      auto interfaceIp = itr.second;
+      if (interfaceIp.find(prefix) == 0) {
+        return interfaceIp;
+      }
+    }
+  }
+
+  return "127.0.0.1";
+}

src/tools/otherTools.cpp

@@ -17,8 +17,14 @@
  * MA 02110-1301, USA.
  */
 
+// Implement function declared in tools.h and tools/otherTools.h
+#include "tools.h"
 #include "tools/otherTools.h"
 
+
+#include <algorithm>
+#include <iomanip>
+
 #ifdef _WIN32
 #include <windows.h>
 #else
@@ -31,6 +37,8 @@
 #include <sstream>
 #include <pugixml.hpp>
 
+#include <zim/uuid.h>
+
 
 static std::map<std::string, std::string> codeisomapping {
 { "aa", "aar" },
@@ -280,3 +288,102 @@ bool kiwix::convertStrToBool(const std::string& value)
   throw std::domain_error(ss.str());
 }
 
+namespace
+{
+// The counter metadata format is a list of item separated by a `;` :
+// item0;item1;item2
+// Each item is a "tuple" mimetype=number.
+// However, the mimetype may contains parameters:
+// text/html;raw=true;foo=bar
+// So the final format may be complex to parse:
+// key0=value0;key1;foo=bar=value1;key2=value2
+
+typedef kiwix::MimeCounterType::value_type MimetypeAndCounter;
+
+std::string readFullMimetypeAndCounterString(std::istream& in)
+{
+  std::string mtcStr, params;
+  getline(in, mtcStr, ';');
+  if ( mtcStr.find('=') == std::string::npos )
+  {
+    do
+    {
+      if ( !getline(in, params, ';' ) )
+        return std::string();
+      mtcStr += ";" + params;
+    }
+    while ( std::count(params.begin(), params.end(), '=') != 2 );
+  }
+  return mtcStr;
+}
+
+MimetypeAndCounter parseASingleMimetypeCounter(const std::string& s)
+{
+  const std::string::size_type k = s.find_last_of("=");
+  if ( k != std::string::npos )
+  {
+    const std::string mimeType = s.substr(0, k);
+    std::istringstream counterSS(s.substr(k+1));
+    unsigned int counter;
+    if (counterSS >> counter && counterSS.eof())
+      return MimetypeAndCounter{mimeType, counter};
+  }
+  return MimetypeAndCounter{"", 0};
+}
+
+} // unnamed namespace
+
+kiwix::MimeCounterType kiwix::parseMimetypeCounter(const std::string& counterData)
+{
+  kiwix::MimeCounterType counters;
+  std::istringstream ss(counterData);
+
+  while (ss)
+  {
+    const std::string mtcStr = readFullMimetypeAndCounterString(ss);
+    const MimetypeAndCounter mtc = parseASingleMimetypeCounter(mtcStr);
+    if ( !mtc.first.empty() )
+      counters.insert(mtc);
+  }
+
+  return counters;
+}
+
+std::string kiwix::gen_date_str()
+{
+  auto now = std::time(0);
+  auto tm = std::localtime(&now);
+
+  std::stringstream is;
+  is << std::setw(2) << std::setfill('0')
+     << 1900+tm->tm_year << "-"
+     << std::setw(2) << std::setfill('0') << tm->tm_mon+1 << "-"
+     << std::setw(2) << std::setfill('0') << tm->tm_mday << "T"
+     << std::setw(2) << std::setfill('0') << tm->tm_hour << ":"
+     << std::setw(2) << std::setfill('0') << tm->tm_min << ":"
+     << std::setw(2) << std::setfill('0') << tm->tm_sec << "Z";
+  return is.str();
+}
+
+std::string kiwix::gen_uuid(const std::string& s)
+{
+  return kiwix::to_string(zim::Uuid::generate(s));
+}
+
+kainjow::mustache::data kiwix::onlyAsNonEmptyMustacheValue(const std::string& s)
+{
+  return s.empty()
+       ? kainjow::mustache::data(false)
+       : kainjow::mustache::data(s);
+}
+
+std::string kiwix::render_template(const std::string& template_str, kainjow::mustache::data data)
+{
+  kainjow::mustache::mustache tmpl(template_str);
+  kainjow::mustache::data urlencode{kainjow::mustache::lambda2{
+                               [](const std::string& str,const kainjow::mustache::renderer& r) { return urlEncode(r(str), true); }}};
+  data.set("urlencoded", urlencode);
+  std::stringstream ss;
+  tmpl.render(data, [&ss](const std::string& str) { ss << str; });
+  return ss.str();
+}

src/tools/otherTools.h

@@ -22,6 +22,12 @@
 
 #include <string>
 #include <vector>
+#include <map>
+#include <cstdlib>
+#include <zim/zim.h>
+#include <mustache.hpp>
+
+#include "stringTools.h"
 
 namespace pugi {
   class xml_node;
@@ -29,9 +35,7 @@ namespace pugi {
 
 namespace kiwix
 {
-  void sleep(unsigned int milliseconds);
   std::string nodeToString(const pugi::xml_node& node);
-  std::string converta2toa3(const std::string& a2code);
 
   /*
    * Convert all format tag string to new format
@@ -41,6 +45,31 @@ namespace kiwix
                                      const std::string& tagName);
   bool convertStrToBool(const std::string& value);
 
+  using MimeCounterType = std::map<const std::string, zim::entry_index_type>;
+  MimeCounterType parseMimetypeCounter(const std::string& counterData);
+
+  std::string gen_date_str();
+  std::string gen_uuid(const std::string& s);
+
+  // if s is empty then returns kainjow::mustache::data(false)
+  // otherwise kainjow::mustache::data(value)
+  kainjow::mustache::data onlyAsNonEmptyMustacheValue(const std::string& s);
+
+  std::string render_template(const std::string& template_str, kainjow::mustache::data data);
+
+  template<typename T>
+  T getEnvVar(const char* name, const T& defaultValue)
+  {
+    try {
+      const char* envString = std::getenv(name);
+      if (envString == nullptr) {
+        throw std::runtime_error("Environment variable not set");
+      }
+      return extractFromString<T>(envString);
+    } catch (...) {}
+
+    return defaultValue;
+  }
 }
 
 #endif

src/tools/pathTools.cpp

@@ -17,7 +17,10 @@
  * MA 02110-1301, USA.
  */
 
+// Implement method defined in <kiwix/tools.h> and "tools/pathTools.h"
+#include "tools.h"
 #include "tools/pathTools.h"
+
 #include <stdexcept>
 
 #ifdef __APPLE__
@@ -59,7 +62,6 @@
 #define PATH_MAX 1024
 #endif
 
-
 #ifdef _WIN32
 std::string WideToUtf8(const std::wstring& wstr)
 {
@@ -78,16 +80,22 @@ std::wstring Utf8ToWide(const std::string& str)
 }
 #endif
 
-bool isRelativePath(const std::string& path)
+bool kiwix::isRelativePath(const std::string& path)
 {
 #ifdef _WIN32
-  return path.empty() || path.substr(1, 2) == ":\\" ? false : true;
+  if (path.size() < 3 ) {
+    return true;
+  }
+  if (path.substr(1, 2) == ":\\" || path.substr(0, 2) == "\\\\") {
+    return false;
+  }
+  return true;
 #else
   return path.empty() || path.substr(0, 1) == "/" ? false : true;
 #endif
 }
 
-std::vector<std::string> normalizeParts(std::vector<std::string> parts, bool absolute)
+std::vector<std::string> normalizeParts(std::vector<std::string>& parts, bool absolute)
 {
   std::vector<std::string> ret;
 #ifdef _WIN32
@@ -95,10 +103,35 @@ std::vector<std::string> normalizeParts(std::vector<std::string> parts, bool abs
   //Starts from there.
   auto it = find_if(parts.rbegin(), parts.rend(),
                     [](const std::string& p) ->bool
-                      { return p.length() == 2 && p[1] == ':'; });
+                      { return ((p.length() == 2 && p[1] == ':')
+                             || (p.length() > 2 && p[0] == '\\' && p[1] == '\\')); });
   if (it != parts.rend()) {
     parts.erase(parts.begin(), it.base()-1);
   }
+  // Special case for samba mount point starting with two "\\" ("\\\\samba\\foo")
+  if (parts.size() > 2  && parts[0].empty() && parts[1].empty()) {
+    parts.erase(parts.begin(), parts.begin()+2);
+    parts[0] = "\\\\" + parts[0];
+  }
+  // Special case if we have a samba drive not at first.
+  // Path is "..\\\\sambdadrive\\..\\.." So we will have an empty part.
+  auto previous_empty = false;
+  for (it = parts.rbegin(); it!=parts.rend(); it++) {
+    if(it->empty()) {
+      if (previous_empty) {
+        it++;
+        break;
+      } else {
+        previous_empty = true;
+      }
+    } else {
+      previous_empty = false;
+    }
+  }
+  if (it != parts.rend()) {
+    parts.erase(parts.begin(), it.base()-1);
+    parts[0] = "\\\\" + parts[0];
+  }
 #endif
 
   size_t index = 0;
@@ -142,10 +175,12 @@ std::vector<std::string> normalizeParts(std::vector<std::string> parts, bool abs
   return ret;
 }
 
-std::string computeRelativePath(const std::string& path, const std::string& absolutePath)
+std::string kiwix::computeRelativePath(const std::string& path, const std::string& absolutePath)
 {
-  auto pathParts = normalizeParts(kiwix::split(path, SEPARATOR, false), false);
-  auto absolutePathParts = kiwix::split(absolutePath, SEPARATOR, false);
+  auto parts = kiwix::split(path, SEPARATOR, false);
+  auto pathParts = normalizeParts(parts, false);
+  parts = kiwix::split(absolutePath, SEPARATOR, false);
+  auto absolutePathParts = normalizeParts(parts, true);
 
   unsigned int commonCount = 0;
   while (commonCount < pathParts.size()
@@ -165,24 +200,27 @@ std::string computeRelativePath(const std::string& path, const std::string& abso
   return ret;
 }
 
-std::string computeAbsolutePath(const std::string& path, const std::string& relativePath)
+std::string kiwix::computeAbsolutePath(const std::string& path, const std::string& relativePath)
 {
   std::string absolutePath = path;
   if (path.empty()) {
-    absolutePath = getCurrentDirectory();
+    absolutePath = kiwix::getCurrentDirectory();
   }
 
-  auto absoluteParts = normalizeParts(kiwix::split(absolutePath, SEPARATOR, false), true);
-  auto relativeParts = kiwix::split(relativePath, SEPARATOR, false);
+  auto parts = kiwix::split(absolutePath, SEPARATOR, false);
+  auto absoluteParts = normalizeParts(parts, true);
+  parts = kiwix::split(relativePath, SEPARATOR, false);
+  auto relativeParts = normalizeParts(parts, false);
 
   absoluteParts.insert(absoluteParts.end(), relativeParts.begin(), relativeParts.end());
   auto ret = kiwix::join(normalizeParts(absoluteParts, true), SEPARATOR);
   return ret;
 }
 
-std::string removeLastPathElement(const std::string& path)
+std::string kiwix::removeLastPathElement(const std::string& path)
 {
-  auto parts = normalizeParts(kiwix::split(path, SEPARATOR, false), false);
+  auto parts_ = kiwix::split(path, SEPARATOR, false);
+  auto parts = normalizeParts(parts_, false);
   if (!parts.empty()) {
     parts.pop_back();
   }
@@ -190,7 +228,7 @@ std::string removeLastPathElement(const std::string& path)
   return ret;
 }
 
-std::string appendToDirectory(const std::string& directoryPath, const std::string& filename)
+std::string kiwix::appendToDirectory(const std::string& directoryPath, const std::string& filename)
 {
   std::string newPath = directoryPath;
   if (!directoryPath.empty() && directoryPath.back() != SEPARATOR[0]) {
@@ -200,9 +238,10 @@ std::string appendToDirectory(const std::string& directoryPath, const std::strin
   return newPath;
 }
 
-std::string getLastPathElement(const std::string& path)
+std::string kiwix::getLastPathElement(const std::string& path)
 {
-  auto parts = normalizeParts(kiwix::split(path, SEPARATOR), false);
+  auto parts_ = kiwix::split(path, SEPARATOR);
+  auto parts = normalizeParts(parts_, false);
   if (parts.empty()) {
     return "";
   }
@@ -230,7 +269,7 @@ std::string getFileSizeAsString(const std::string& path)
   return convert.str();
 }
 
-std::string getFileContent(const std::string& path)
+std::string kiwix::getFileContent(const std::string& path)
 {
 #ifdef _WIN32
   auto wpath = Utf8ToWide(path);
@@ -263,19 +302,21 @@ std::string getFileContent(const std::string& path)
   return content;
 }
 
-bool fileExists(const std::string& path)
+bool kiwix::fileExists(const std::string& path)
 {
 #ifdef _WIN32
-  return PathFileExistsW(Utf8ToWide(path).c_str());
+  return (_waccess_s(Utf8ToWide(path).c_str(), 0) == 0);
 #else
-  bool flag = false;
-  std::fstream fin;
-  fin.open(path.c_str(), std::ios::in);
-  if (fin.is_open()) {
-    flag = true;
-  }
-  fin.close();
-  return flag;
+  return (access(path.c_str(), F_OK) == 0);
+#endif
+}
+
+bool kiwix::fileReadable(const std::string& path)
+{
+#ifdef _WIN32
+  return (_waccess_s(Utf8ToWide(path).c_str(), 4) == 0);
+#else
+  return (access(path.c_str(), R_OK) == 0);
 #endif
 }
 
@@ -302,7 +343,7 @@ std::string makeTmpDirectory()
   _wmkdir(ctmp);
   return WideToUtf8(ctmp);
 #else
-  char _template_array[] = {"/tmp/kiwix-lib_XXXXXX"};
+  char _template_array[] = {"/tmp/libkiwix_XXXXXX"};
   std::string dir = mkdtemp(_template_array);
   return dir;
 #endif
@@ -329,7 +370,7 @@ bool copyFile(const std::string& sourcePath, const std::string& destPath)
 #endif
 }
 
-std::string getExecutablePath(bool realPathOnly)
+std::string kiwix::getExecutablePath(bool realPathOnly)
 {
   if (!realPathOnly) {
     char* cAppImage = ::getenv("APPIMAGE");
@@ -383,7 +424,7 @@ bool writeTextFile(const std::string& path, const std::string& content)
   return true;
 }
 
-std::string getCurrentDirectory()
+std::string kiwix::getCurrentDirectory()
 {
 #ifdef _WIN32
   wchar_t* a_cwd = _wgetcwd(NULL, 0);
@@ -397,7 +438,7 @@ std::string getCurrentDirectory()
   return ret;
 }
 
-std::string getDataDirectory()
+std::string kiwix::getDataDirectory()
 {
 // Try to get the dataDir from the `KIWIX_DATA_DIR` env var
 #ifdef _WIN32
@@ -466,7 +507,7 @@ static std::map<std::string, std::string> extMimeTypes = {
 };
 
 /* Try to get the mimeType from the file extension */
-std::string getMimeTypeForFile(const std::string& filename)
+std::string kiwix::getMimeTypeForFile(const std::string& filename)
 {
   std::string mimeType = "text/plain";
   auto pos = filename.find_last_of(".");
@@ -487,4 +528,3 @@ std::string getMimeTypeForFile(const std::string& filename)
 
   return mimeType;
 }
-

src/tools/pathTools.h

@@ -26,23 +26,13 @@
 std::string WideToUtf8(const std::wstring& wstr);
 std::wstring Utf8ToWide(const std::string& str);
 #endif
-bool isRelativePath(const std::string& path);
-std::string computeAbsolutePath(const std::string& path, const std::string& relativePath);
-std::string computeRelativePath(const std::string& path, const std::string& absolutePath);
-std::string removeLastPathElement(const std::string& path);
-std::string appendToDirectory(const std::string& directoryPath, const std::string& filename);
 
 unsigned int getFileSize(const std::string& path);
 std::string getFileSizeAsString(const std::string& path);
-std::string getFileContent(const std::string& path);
-bool fileExists(const std::string& path);
 bool makeDirectory(const std::string& path);
 std::string makeTmpDirectory();
 bool copyFile(const std::string& sourcePath, const std::string& destPath);
-std::string getLastPathElement(const std::string& path);
-std::string getExecutablePath(bool realPathOnly = false);
-std::string getCurrentDirectory();
-std::string getDataDirectory();
 bool writeTextFile(const std::string& path, const std::string& content);
-std::string getMimeTypeForFile(const std::string& filename);
+
 #endif
+

src/tools/regexTools.cpp

@@ -18,7 +18,6 @@
  */
 
 #include <tools/regexTools.h>
-#include <tools/lock.h>
 
 #include <unicode/regex.h>
 #include <unicode/ucnv.h>
@@ -26,10 +25,10 @@
 #include <memory>
 #include <map>
 #include <stdexcept>
-#include <pthread.h>
+#include <mutex>
 
 std::map<std::string, std::shared_ptr<icu::RegexPattern>> regexCache;
-static pthread_mutex_t regexLock = PTHREAD_MUTEX_INITIALIZER;
+static std::mutex regexLock;
 
 std::unique_ptr<icu::RegexMatcher> buildMatcher(const std::string& regex, icu::UnicodeString& content)
 {
@@ -39,7 +38,7 @@ std::unique_ptr<icu::RegexMatcher> buildMatcher(const std::string& regex, icu::U
     pattern = regexCache.at(regex);
   } catch (std::out_of_range&) {
     // Redo the search with a lock to avoid race condition.
-    kiwix::Lock l(&regexLock);
+    std::lock_guard<std::mutex> l(regexLock);
     try {
       pattern = regexCache.at(regex);
     } catch (std::out_of_range&) {
@@ -95,3 +94,22 @@ std::string appendToFirstOccurence(const std::string& content,
 
   return content;
 }
+
+std::string prependToFirstOccurence(const std::string& content,
+                                   const std::string& regex,
+                                   const std::string& replacement)
+{
+  ucnv_setDefaultName("UTF-8");
+  icu::UnicodeString ucontent(content.c_str());
+  icu::UnicodeString ureplacement(replacement.c_str());
+  auto matcher = buildMatcher(regex, ucontent);
+  if (matcher->find()) {
+    UErrorCode status = U_ZERO_ERROR;
+    ucontent.insert(matcher->start(status), ureplacement);
+    std::string tmp;
+    ucontent.toUTF8String(tmp);
+    return tmp;
+  }
+
+  return content;
+}

src/tools/regexTools.h

@@ -29,5 +29,8 @@ std::string replaceRegex(const std::string& content,
 std::string appendToFirstOccurence(const std::string& content,
                                    const std::string& regex,
                                    const std::string& replacement);
+std::string prependToFirstOccurence(const std::string& content,
+                                    const std::string& regex,
+                                    const std::string& replacement);
 
 #endif

src/tools/stringTools.cpp

@@ -17,9 +17,11 @@
  * MA 02110-1301, USA.
  */
 
-#include <tools/stringTools.h>
+// Implement function declared in tools.h and tools/stringTools.h
+#include "tools.h"
+#include "tools/stringTools.h"
 
-#include <tools/pathTools.h>
+#include "tools/pathTools.h"
 #include <unicode/normlzr.h>
 #include <unicode/rep.h>
 #include <unicode/translit.h>
@@ -47,6 +49,24 @@ void kiwix::loadICUExternalTables()
 #endif
 }
 
+kiwix::ICULanguageInfo::ICULanguageInfo(const std::string& langCode)
+  : locale(langCode.c_str())
+{}
+
+std::string kiwix::ICULanguageInfo::iso3Code() const
+{
+  return locale.getISO3Language();
+}
+
+std::string kiwix::ICULanguageInfo::selfName() const
+{
+  icu::UnicodeString langSelfNameICUString;
+  locale.getDisplayLanguage(locale, langSelfNameICUString);
+  std::string langSelfName;
+  langSelfNameICUString.toUTF8String(langSelfName);
+  return langSelfName;
+}
+
 std::string kiwix::removeAccents(const std::string& text)
 {
   loadICUExternalTables();
@@ -268,7 +288,8 @@ std::string kiwix::urlDecode(const std::string& value, bool component)
 /* Split string in a token array */
 std::vector<std::string> kiwix::split(const std::string& str,
                                       const std::string& delims,
-                                      bool trimEmpty)
+                                      bool dropEmpty,
+                                      bool keepDelim)
 {
   std::string::size_type lastPos = 0;
   std::string::size_type pos = 0;
@@ -276,14 +297,17 @@ std::vector<std::string> kiwix::split(const std::string& str,
   while( (pos = str.find_first_of(delims, lastPos)) < str.length() )
   {
     auto token = str.substr(lastPos, pos - lastPos);
-    if (!trimEmpty || !token.empty()) {
+    if (!dropEmpty || !token.empty()) {
       tokens.push_back(token);
     }
+    if (keepDelim) {
+      tokens.push_back(str.substr(pos, 1));
+    }
     lastPos = pos + 1;
   }
 
   auto token = str.substr(lastPos);
-  if (!trimEmpty || !token.empty()) {
+  if (!dropEmpty || !token.empty()) {
     tokens.push_back(token);
   }
   return tokens;
@@ -391,3 +415,16 @@ bool kiwix::startsWith(const std::string& base, const std::string& start)
         && std::equal(start.begin(), start.end(), base.begin());
 }
 
+std::vector<std::string> kiwix::getTitleVariants(const std::string& title) {
+  std::vector<std::string> variants;
+  variants.push_back(title);
+  variants.push_back(kiwix::ucFirst(title));
+  variants.push_back(kiwix::lcFirst(title));
+  variants.push_back(kiwix::toTitle(title));
+  return variants;
+}
+
+template<>
+std::string kiwix::extractFromString(const std::string& str) {
+  return str;
+}

src/tools/stringTools.h

@@ -21,10 +21,12 @@
 #define KIWIX_STRINGTOOLS_H
 
 #include <unicode/unistr.h>
+#include <unicode/locid.h>
 
 #include <string>
 #include <vector>
 #include <sstream>
+#include <stdexcept>
 
 namespace kiwix
 {
@@ -40,10 +42,22 @@ std::string encodeDiples(const std::string& str);
 std::string removeAccents(const std::string& text);
 void loadICUExternalTables();
 
+class ICULanguageInfo
+{
+public:
+  explicit ICULanguageInfo(const std::string& langCode);
+
+  std::string iso3Code() const;
+  std::string selfName() const;
+
+private:
+  const icu::Locale locale;
+};
+
+
 std::string urlEncode(const std::string& value, bool encodeReserved = false);
 std::string urlDecode(const std::string& value, bool component = false);
 
-std::vector<std::string> split(const std::string&, const std::string&, bool trimEmpty = true);
 std::string join(const std::vector<std::string>& list, const std::string& sep);
 
 std::string ucAll(const std::string& word);
@@ -66,9 +80,17 @@ T extractFromString(const std::string& str) {
     std::istringstream iss(str);
     T ret;
     iss >> ret;
+    if(iss.fail() || !iss.eof()) {
+        throw std::invalid_argument("no conversion");
+    }
     return ret;
 }
 
+template<>
+std::string extractFromString(const std::string& str);
+
 bool startsWith(const std::string& base, const std::string& start);
+
+std::vector<std::string> getTitleVariants(const std::string& title);
 } //namespace kiwix
 #endif

src/version.cpp

@@ -0,0 +1,77 @@
+/*
+ * Copyright 2021 Emmanuel Engelhart <kelson@kiwix.org>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU  General Public License as published by
+ * the Free Software Foundation; either version 3 of the License, or
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
+ * MA 02110-1301, USA.
+ */
+
+#include <iostream>
+#include <sstream>
+
+#include <version.h>
+#include <zim/zim.h>
+#include <kiwix_config.h>
+#include <unicode/uversion.h>
+#include <pugixml.hpp>
+#include <curl/curl.h>
+#include <microhttpd.h>
+#include <xapian.h>
+#include <mustache.hpp>
+#include <zlib.h>
+
+namespace kiwix
+{
+  LibVersions getVersions() {
+    LibVersions versions = {
+      { "libkiwix",      LIBKIWIX_VERSION    },
+      { "libzim",        LIBZIM_VERSION      },
+      { "libxapian",     XAPIAN_VERSION      },
+      { "libcurl",       LIBCURL_VERSION     },
+      { "libmicrohttpd", MHD_get_version()   },
+      { "libz",          ZLIB_VERSION        }
+    };
+
+    // U_ICU_VERSION does not include the patch level if 0
+    std::ostringstream libicu_version;
+    libicu_version << U_ICU_VERSION_MAJOR_NUM << "." << U_ICU_VERSION_MINOR_NUM << "." << U_ICU_VERSION_PATCHLEVEL_NUM;
+    versions.push_back({ "libicu", libicu_version.str() });
+
+    // No human readable version string for pugixml
+    const unsigned pugixml_major = (PUGIXML_VERSION - PUGIXML_VERSION % 1000) / 1000;
+    const unsigned pugixml_minor = (PUGIXML_VERSION - pugixml_major * 1000 - PUGIXML_VERSION % 10) / 10;
+    const unsigned pugixml_patch = PUGIXML_VERSION - pugixml_major * 1000 - pugixml_minor * 10;
+    std::ostringstream libpugixml_version;
+    libpugixml_version << pugixml_major << "." << pugixml_minor << "." << pugixml_patch;
+    versions.push_back({ "libpugixml", libpugixml_version.str() });
+
+    // Needs version 5.0 of Mustache
+#if defined(KAINJOW_MUSTACHE_VERSION_MAJOR)
+    std::ostringstream libmustache_version;
+    libmustache_version << KAINJOW_MUSTACHE_VERSION_MAJOR << "." <<
+      KAINJOW_MUSTACHE_VERSION_MINOR << "." << KAINJOW_MUSTACHE_VERSION_PATCH;
+    versions.push_back({ "libmustache", libmustache_version.str() });
+#endif
+
+    return versions;
+  }
+
+  void printVersions(std::ostream& out) {
+    LibVersions versions = getVersions();
+    for (const auto& iter : versions) {
+      out << (iter != versions.front() ? "+ " : "")
+                << iter.first << " " << iter.second << std::endl;
+    }
+  }
+} //namespace kiwix

src/wrapper/java/AndroidManifest.xml

@@ -1,13 +0,0 @@
-<manifest xmlns:android="http://schemas.android.com/apk/res/android"
-
-  package="kiwix.org.kiwixlib"
->
-
-  <application android:allowBackup="true"
-    android:label="@string/app_name"
-    android:supportsRtl="true"
-  >
-
-  </application>
-
-</manifest>

src/wrapper/java/book.cpp

@@ -1,87 +0,0 @@
-/*
- * Copyright (C) 2020 Matthieu Gautier <mgautier@kymeria.fr>
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU  General Public License as published by
- * the Free Software Foundation; either version 3 of the License, or
- * any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
- * MA 02110-1301, USA.
- */
-
-
-#include <jni.h>
-#include "org_kiwix_kiwixlib_Book.h"
-
-#include "utils.h"
-#include "book.h"
-
-JNIEXPORT void JNICALL
-Java_org_kiwix_kiwixlib_Book_allocate(
-  JNIEnv* env, jobject thisObj)
-{
-  allocate<kiwix::Book>(env, thisObj);
-}
-
-JNIEXPORT void JNICALL
-Java_org_kiwix_kiwixlib_Book_dispose(JNIEnv* env, jobject thisObj)
-{
-  dispose<kiwix::Book>(env, thisObj);
-}
-
-#define BOOK (getPtr<kiwix::Book>(env, thisObj))
-
-METHOD(void, Book, update__Lorg_kiwix_kiwixlib_Book_2, jobject otherBook)
-{
-  BOOK->update(*getPtr<kiwix::Book>(env, otherBook));
-}
-
-METHOD(void, Book, update__Lorg_kiwix_kiwixlib_JNIKiwixReader_2, jobject reader)
-{
-  BOOK->update(**Handle<kiwix::Reader>::getHandle(env, reader));
-}
-
-#define GETTER(retType, name) JNIEXPORT retType JNICALL \
-Java_org_kiwix_kiwixlib_Book_##name (JNIEnv* env, jobject thisObj) \
-{ \
-  auto cRet = BOOK->name(); \
-  retType ret = c2jni(cRet, env); \
-  return ret; \
-}
-
-GETTER(jstring, getId)
-GETTER(jstring, getPath)
-GETTER(jboolean, isPathValid)
-GETTER(jstring, getTitle)
-GETTER(jstring, getDescription)
-GETTER(jstring, getLanguage)
-GETTER(jstring, getCreator)
-GETTER(jstring, getPublisher)
-GETTER(jstring, getDate)
-GETTER(jstring, getUrl)
-GETTER(jstring, getName)
-GETTER(jstring, getFlavour)
-GETTER(jstring, getTags)
-GETTER(jlong, getArticleCount)
-GETTER(jlong, getMediaCount)
-GETTER(jlong, getSize)
-GETTER(jstring, getFavicon)
-GETTER(jstring, getFaviconUrl)
-GETTER(jstring, getFaviconMimeType)
-
-METHOD(jstring, Book, getTagStr, jstring tagName) try {
-  auto cRet = BOOK->getTagStr(jni2c(tagName, env));
-  return c2jni(cRet, env);
-} catch(...) {
-  return c2jni<std::string>("", env);
-}
-
-#undef GETTER