From ffdccd5754bd47b99a235228b1364015fbcf74a4 Mon Sep 17 00:00:00 2001
From: jliddev <jliddev@gmail.com>
Date: Tue, 8 Dec 2020 16:49:28 -0600
Subject: [PATCH] Fix some issues with the CurseForge fingerprint scanner

---
 .../src/common/curse/curse-folder-scanner.ts  | 87 ++++++++++++++++---
 1 file changed, 77 insertions(+), 10 deletions(-)

diff --git a/wowup-electron/src/common/curse/curse-folder-scanner.ts b/wowup-electron/src/common/curse/curse-folder-scanner.ts
index 44e5b609..48cac8da 100644
--- a/wowup-electron/src/common/curse/curse-folder-scanner.ts
+++ b/wowup-electron/src/common/curse/curse-folder-scanner.ts
@@ -12,8 +12,43 @@ export class CurseFolderScanner {
   // This map is required for solving for case sensitive mismatches from addon authors on Linux
   private _fileMap: { [key: string]: string } = {};
 
+  private readonly _invalidPathChars = [
+    "|",
+    "\0",
+    "\u0001",
+    "\u0002",
+    "\u0003",
+    "\u0004",
+    "\u0005",
+    "\u0006",
+    "\b",
+    "\t",
+    "\n",
+    "\v",
+    "\f",
+    "\r",
+    "\u000e",
+    "\u000f",
+    "\u0010",
+    "\u0011",
+    "\u0012",
+    "\u0013",
+    "\u0014",
+    "\u0015",
+    "\u0016",
+    "\u0017",
+    "\u0018",
+    "\u0019",
+    "\u001a",
+    "\u001b",
+    "\u001c",
+    "\u001d",
+    "\u001e",
+    "\u001f",
+  ];
+
   private get tocFileCommentsRegex() {
-    return /\s*#.*$/gm;
+    return /\s*#.*$/gim;
   }
 
   private get tocFileIncludesRegex() {
@@ -21,31 +56,32 @@ export class CurseFolderScanner {
   }
 
   private get tocFileRegex() {
-    return /^([^\/]+)[\\\/]\1\.toc$/i;
+    return /^([^\/]+)[\\\/]\1\.toc$/gim;
   }
 
   private get bindingsXmlRegex() {
-    return /^[^\/\\]+[\/\\]Bindings\.xml$/i;
+    return /^[^\/\\]+[\/\\]Bindings\.xml$/gim;
   }
 
   private get bindingsXmlIncludesRegex() {
-    return /<(?:Include|Script)\s+file=[\""\""']((?:(?<!\.\.).)+)[\""\""']\s*\/>/gi;
+    return /<(?:Include|Script)\s+file=[\""\""']((?:(?<!\.\.).)+)[\""\""']\s*\/>/gis;
   }
 
   private get bindingsXmlCommentsRegex() {
-    return /<!--.*?-->/gs;
+    return /<!--.*?-->/gims;
   }
 
   async scanFolder(folderPath: string): Promise<CurseScanResult> {
     const fileList = await readDirRecursive(folderPath);
     fileList.forEach((fp) => (this._fileMap[fp.toLowerCase()] = fp));
-
     // log.debug("listAllFiles", folderPath, fileList.length);
 
     let matchingFiles = await this.getMatchingFiles(folderPath, fileList);
-    matchingFiles = _.sortBy(matchingFiles, (f) => f.toLowerCase());
+    matchingFiles = _.orderBy(matchingFiles, [(f) => f.toLowerCase()], ["asc"]);
     // log.debug("matchingFiles", matchingFiles.length);
 
+    const sq = matchingFiles.map((mf) => mf.toLowerCase()).join("\n");
+
     let individualFingerprints = await async.mapLimit<string, number>(matchingFiles, 4, async (path, callback) => {
       try {
         const fileHash = await this.getFileHash(path);
@@ -72,7 +108,7 @@ export class CurseFolderScanner {
   }
 
   private async getMatchingFiles(folderPath: string, filePaths: string[]): Promise<string[]> {
-    const parentDir = path.dirname(folderPath) + path.sep;
+    const parentDir = path.normalize(path.dirname(folderPath) + path.sep);
     const matchingFileList: string[] = [];
     const fileInfoList: string[] = [];
     for (let filePath of filePaths) {
@@ -119,18 +155,27 @@ export class CurseFolderScanner {
 
     const dirname = path.dirname(nativePath);
     for (let include of inclusions) {
+      if (this.hasInvalidPathChars(include)) {
+        log.debug(`Invalid include file ${include}`);
+        break;
+      }
+
       const fileName = path.join(dirname, include.replace(/\\/g, path.sep));
       await this.processIncludeFile(matchingFileList, fileName);
     }
   }
 
+  private hasInvalidPathChars(path: string) {
+    return this._invalidPathChars.some((c) => path.indexOf(c) !== -1);
+  }
+
   private getFileInclusionMatches(fileInfo: string, fileContent: string): string[] | null {
     const ext = path.extname(fileInfo);
     switch (ext) {
       case ".xml":
-        return this.matchAll(fileContent, this.bindingsXmlIncludesRegex);
+        return this.ripMatch(fileContent, () => this.bindingsXmlIncludesRegex);
       case ".toc":
-        return this.matchAll(fileContent, this.tocFileIncludesRegex);
+        return this.ripMatch(fileContent, () => this.tocFileIncludesRegex);
       default:
         return null;
     }
@@ -148,6 +193,28 @@ export class CurseFolderScanner {
     }
   }
 
+  /**
+   *  Recreate a strange behavior for .net regex regarding how it treats
+   *  lines that end in \r\t vs lines with \r\n\t
+   */
+  private ripMatch(str: string, regex: () => RegExp): string[] {
+    const splitStrings = str.split("\n");
+    const matches = [];
+    try {
+      for (const splitStr of splitStrings) {
+        const trimmedStr = splitStr.trim();
+        const match = regex().exec(trimmedStr);
+        if (match && match.length > 1) {
+          matches.push(match[1]);
+        }
+      }
+    } catch (e) {
+      log.error(e);
+    }
+
+    return matches.map((s) => s.trim());
+  }
+
   private matchAll(str: string, regex: RegExp): string[] {
     const matches: string[] = [];
     let currentMatch: RegExpExecArray;