backup/warcprox
1
0
Fork 0
mirror of https://github.com/internetarchive/warcprox.git synced 2025-01-18 13:22:09 +01:00
Code Issues Projects Releases Activity

python 3.5 version, mostly

Browse Source
This commit is contained in:
Barbara Miller 2021-12-02 11:46:30 -08:00
parent 1476bfec8c
commit e744075913
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
warcprox/dedup.py
View File

@ -418,8 +418,8 @@ class BatchTroughLoader(warcprox.BaseBatchPostfetchProcessor):
if (recorded_url.response_recorder
and recorded_url.payload_digest
and self.trough_dedup_db.should_dedup(recorded_url)
and f'{recorded_url.payload_digest}{recorded_url.url}' not in hash_plus_urls):
hash_plus_urls.add(f'{recorded_url.payload_digest}{recorded_url.url}')
and '{}{}'.format(recorded_url.payload_digest, recorded_url.url) not in hash_plus_urls):
hash_plus_urls.add('{}{}'.format(recorded_url.payload_digest, recorded_url.url))
if (recorded_url.warcprox_meta
and 'dedup-buckets' in recorded_url.warcprox_meta):
for bucket, bucket_mode in recorded_url.warcprox_meta["dedup-buckets"].items():
@ -427,9 +427,9 @@ class BatchTroughLoader(warcprox.BaseBatchPostfetchProcessor):
else:
buckets['__unspecified__'].append(recorded_url)
else:
if f'{recorded_url.payload_digest}{recorded_url.url}' in hash_plus_urls:
if recorded_url.payload_digest and '{}{}'.format(recorded_url.payload_digest, recorded_url.url) in hash_plus_urls:
self.logger.debug(
f'discarding duplicate {recorded_url.payload_digest} {recorded_url.url}')
'discarding duplicate {} {}'.format(recorded_url.payload_digest, recorded_url.url))
discards.append(
warcprox.digest_str(
recorded_url.payload_digest, self.options.base32)