From cc91b10e048b022e05441bdb38e5bedc0d6a220f Mon Sep 17 00:00:00 2001 From: datechnoman Date: Tue, 2 Apr 2024 11:30:27 +0000 Subject: [PATCH] Updated to add pdf extraction --- archiveteam_project_url_extractor.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/archiveteam_project_url_extractor.py b/archiveteam_project_url_extractor.py index 72d0fee..44549e9 100644 --- a/archiveteam_project_url_extractor.py +++ b/archiveteam_project_url_extractor.py @@ -77,7 +77,7 @@ url_patterns = { r'(mediafire\.com|mfi\.re)\S+': 'filtered_mediafire_mfi.re.txt', r'\S*imgur\S*': 'filtered_imgur.txt', r'http(s)?://(www\.)?pastebin.com': 'filtered_pastebin.txt', - r'https://cdn.discordapp.com': 'filtered_cdn.discordapp.com.txt' + r'https?://\S+\.pdf\b'': 'filtered_cdn.discordapp.com.txt' } # Ask the user for the number of concurrent instances