Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-04-20 23:56:34 -04:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

ytdlp_seed = site["metadata"]["ait_seed_id"] (?!)

Browse Source
This commit is contained in:
Barbara Miller 2024-04-05 14:46:19 -07:00
parent dfd4d7da84
commit 8c4f774504
2 changed files with 3 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
brozzler/model.py
View File

@ -111,7 +111,6 @@ def new_job(frontier, job_conf):
merged_conf["seed"] = merged_conf.pop("url")
site = brozzler.Site(frontier.rr, merged_conf)
site.id = str(uuid.uuid4())
site.seed_id = seed_conf['metadata']['ait_seed_id']
sites.append(site)
pages.append(new_seed_page(frontier, site))

6
brozzler/ydl.py
View File

@ -64,11 +64,11 @@ def _timestamp4datetime(timestamp):
def should_ytdlp(worker, page, site):
ytdlp_url = page.redirect_url if page.redirect_url else page.url
ytdlp_seed = site.seed_id if site.seed_id else None
# ytdlp_seed = site.get(site.id).pluck("metadata", "ait_seed_id").default(None) if site.rr else None ???
logging.info("ytdlp_seed: %s", ytdlp_seed)
ytdlp_seed = site["metadata"]["ait_seed_id"]
logging.info("checking containing page %r, site %r", ytdlp_url, ytdlp_seed)
if ytdlp_seed and "youtube.com/watch?v" in ytdlp_url:
logging.info("found youtube watch page %r", ytdlp_url)
# connect to bmiller-dev cluster, keyspace video; we can modify default timeout in cassandra.yaml
cluster = Cluster(["207.241.235.189"], protocol_version=5)
session = cluster.connect("video")