From d5b2e24423c654ebdfbb0abd230874e2d4fa5f7f Mon Sep 17 00:00:00 2001
From: CaliBrain <calibrain@l4n.xyz>
Date: Sun, 16 Mar 2025 02:40:09 -0400
Subject: [PATCH] feat: Add CUSTOM_SCRIPT documentation and implementation
 details (#90)

---
 backend.py |  8 ++++++--
 config.py  | 11 +++++++++++
 readme.md  | 23 +++++++++++++++++++++++
 3 files changed, 40 insertions(+), 2 deletions(-)

diff --git a/backend.py b/backend.py
index 351682f..4ebb57d 100644
--- a/backend.py
+++ b/backend.py
@@ -6,7 +6,7 @@ from pathlib import Path
 from typing import Dict, List, Optional, Any, Tuple
 
 from logger import setup_logger
-from config import TMP_DIR, MAIN_LOOP_SLEEP_TIME, INGEST_DIR
+from config import TMP_DIR, MAIN_LOOP_SLEEP_TIME, INGEST_DIR, CUSTOM_SCRIPT
 from models import book_queue, BookInfo, QueueStatus, SearchFilters
 import book_manager
 
@@ -118,9 +118,13 @@ def _download_book(book_id: str) -> bool:
         if not success:
             raise Exception("Unkown error downloading book")
 
+        if CUSTOM_SCRIPT:
+            subprocess.run([CUSTOM_SCRIPT, book_path])
+
         final_path = INGEST_DIR /  f"{book_id}.{book_info.format}"
         
-        shutil.move(book_path, final_path)
+        if os.path.exists(book_path):
+            shutil.move(book_path, final_path)
         return True
     except Exception as e:
         logger.error(f"Error downloading book: {e}")
diff --git a/config.py b/config.py
index 3ceb124..d7a2203 100644
--- a/config.py
+++ b/config.py
@@ -66,6 +66,17 @@ BOOK_LANGUAGE = [l for l in BOOK_LANGUAGE if l in [lang['code'] for lang in _SUP
 if len(BOOK_LANGUAGE) == 0:
     BOOK_LANGUAGE = ['en']
 
+# Custom script settings
+CUSTOM_SCRIPT = os.getenv("CUSTOM_SCRIPT", None).strip()
+# check if the script is valid
+if CUSTOM_SCRIPT:
+    if not os.path.exists(CUSTOM_SCRIPT):
+        logger.error(f"Custom script {CUSTOM_SCRIPT} does not exist")
+        CUSTOM_SCRIPT = None
+    elif not os.access(CUSTOM_SCRIPT, os.X_OK):
+        logger.error(f"Custom script {CUSTOM_SCRIPT} is not executable")
+        CUSTOM_SCRIPT = None
+
 # API settings
 FLASK_HOST = os.getenv("FLASK_HOST",  "0.0.0.0")
 FLASK_PORT = int(os.getenv("FLASK_PORT", 5003))
diff --git a/readme.md b/readme.md
index 361821a..6d9c839 100644
--- a/readme.md
+++ b/readme.md
@@ -105,6 +105,29 @@ HTTP_PROXY=http://username:password@proxy.example.com:8080
 HTTPS_PROXY=http://username:password@proxy.example.com:8080
 ```
 
+
+#### Custom configuration
+
+| Variable               | Description                                                 | Default Value           |
+| ---------------------- | ----------------------------------------------------------- | ----------------------- |
+| `CUSTOM_SCRIPT`        | Path to an executable script that tuns after each download  | ``                      |
+
+If `CUSTOM_SCRIPT` is set, it will be executed after each successful download but before the file is moved to the ingest directory. This allows for custom processing like format conversion or validation.
+
+The script is called with the full path of the downloaded file as its argument. Important notes:
+- The script must preserve the original filename for proper processing
+- The file can be modified or even deleted if needed
+- The file will be moved to `/cwa-book-ingest` after the script execution (if not deleted)
+
+You can specify these configuration in this format :
+```
+environment:
+  - CUSTOM_SCRIPT=/scripts/process-book.sh
+
+volumes:
+  - local/scripts/custom_script.sh:/scripts/process-book.sh
+```
+
 ### Volume Configuration
 
 ```yaml