Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-08-22 04:39:36 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

fix merge error

Browse source
This commit is contained in:
Barbara Miller 2024-05-29 21:38:24 -07:00
parent f9ad480fae
commit 0f3d2b7172
1 changed files with 6 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

13
brozzler/ydl.py
View file

@ -51,7 +51,7 @@ def _timestamp4datetime(timestamp):
int(timestamp[-2:]) int(timestamp[-2:])
) )
def should_ytdlp(page, site): def should_ytdlp(worker, site, page):
# called only after we've passed needs_browsing() check # called only after we've passed needs_browsing() check
if page.status_code != 200: if page.status_code != 200:
logging.info("skipping ytdlp: non-200 page status") logging.info("skipping ytdlp: non-200 page status")
@ -60,6 +60,11 @@ def should_ytdlp(page, site):
logging.info("skipping ytdlp: site marked skip_ytdlp") logging.info("skipping ytdlp: site marked skip_ytdlp")
return False return False
ytdlp_url = page.redirect_url if page.redirect_url else page.url
if "chrome-error:" in ytdlp_url:
return False
ytdlp_seed = ( ytdlp_seed = (
site["metadata"]["ait_seed_id"] site["metadata"]["ait_seed_id"]
if "metadata" in site and "ait_seed_id" in site["metadata"] if "metadata" in site and "ait_seed_id" in site["metadata"]
@ -71,12 +76,6 @@ def should_ytdlp(page, site):
site.skip_ytdlp = True site.skip_ytdlp = True
return False return False
ytdlp_url = page.redirect_url if page.redirect_url else page.url
if "chrome-error:" in ytdlp_url:
return False
ytdlp_seed = site["metadata"]["ait_seed_id"]
logging.info("checking containing page %s for seed %s", ytdlp_url, ytdlp_seed) logging.info("checking containing page %s for seed %s", ytdlp_url, ytdlp_seed)
if ytdlp_seed and "youtube.com/watch?v" in ytdlp_url: if ytdlp_seed and "youtube.com/watch?v" in ytdlp_url: