Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-02-24 00:29:53 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity

skip .pdf

Browse Source
This commit is contained in:
Barbara Miller 2023-12-11 17:25:19 -08:00
parent 48681e2a93
commit 88223cee59
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
brozzler/worker.py
View File

@ -198,7 +198,7 @@ class BrozzlerWorker:
self.logger.info("brozzling {}".format(page)) self.logger.info("brozzling {}".format(page))
ydl_fetches = None ydl_fetches = None
outlinks = set() outlinks = set()
if enable_youtube_dl: if enable_youtube_dl and not page.url.lower().endswith(".pdf"):
try: try:
ydl_fetches, outlinks = ydl.do_youtube_dl(self, site, page) ydl_fetches, outlinks = ydl.do_youtube_dl(self, site, page)
except brozzler.ReachedLimit as e: except brozzler.ReachedLimit as e: