From 0b8047c7a0c451246ef768120bd8e10971730ba9 Mon Sep 17 00:00:00 2001
From: Gergo Moricz <mo.geryy@gmail.com>
Date: Thu, 18 Jul 2024 19:13:43 +0200
Subject: [PATCH] fix(WebScraper): infinite regex leading to fly.io instance
 hangs

---
 apps/api/src/scraper/WebScraper/utils/replacePaths.ts | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/apps/api/src/scraper/WebScraper/utils/replacePaths.ts b/apps/api/src/scraper/WebScraper/utils/replacePaths.ts
index 788916cd..25b43f0a 100644
--- a/apps/api/src/scraper/WebScraper/utils/replacePaths.ts
+++ b/apps/api/src/scraper/WebScraper/utils/replacePaths.ts
@@ -6,13 +6,13 @@ export const replacePathsWithAbsolutePaths = (documents: Document[]): Document[]
       const baseUrl = new URL(document.metadata.sourceURL).origin;
       const paths =
         document.content.match(
-          /(!?\[.*?\])\(((?:[^()]+|\((?:[^()]+|\([^()]*\))*\))*)\)|href="([^"]+)"/g
+          /!?\[.*?\]\(.*?\)|href=".+?"/g
         ) || [];
 
       paths.forEach((path: string) => {
         try {
           const isImage = path.startsWith("!");
-        let matchedUrl = path.match(/\(([^)]+)\)/) || path.match(/href="([^"]+)"/);
+        let matchedUrl = path.match(/\((.*?)\)/) || path.match(/href="([^"]+)"/);
         let url = matchedUrl[1];
 
         if (!url.startsWith("data:") && !url.startsWith("http")) {
@@ -50,11 +50,11 @@ export const replaceImgPathsWithAbsolutePaths = (documents: Document[]): Documen
       const baseUrl = new URL(document.metadata.sourceURL).origin;
       const images =
         document.content.match(
-          /!\[.*?\]\(((?:[^()]+|\((?:[^()]+|\([^()]*\))*\))*)\)/g
+          /!\[.*?\]\(.*?\)/g
         ) || [];
 
       images.forEach((image: string) => {
-        let imageUrl = image.match(/\(([^)]+)\)/)[1];
+        let imageUrl = image.match(/\((.*?)\)/)[1];
         let altText = image.match(/\[(.*?)\]/)[1];
 
         if (!imageUrl.startsWith("data:image")) {