aboutsummaryrefslogtreecommitdiffstats
path: root/python/sandcrawler
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@archive.org>2020-10-27 15:52:54 -0700
committerBryan Newbold <bnewbold@archive.org>2020-10-27 15:52:54 -0700
commitc4cf72914560f92e914a5dbf7360637f6c24f323 (patch)
treeb3127216f69c5bc52d994a6b6ddf084e7971ad1c /python/sandcrawler
parent58f89d645063415bb9e1d36102cbf4dfc45cffda (diff)
downloadsandcrawler-c4cf72914560f92e914a5dbf7360637f6c24f323.tar.gz
sandcrawler-c4cf72914560f92e914a5dbf7360637f6c24f323.zip
HTML metadata: fix type warnings
Diffstat (limited to 'python/sandcrawler')
-rw-r--r--python/sandcrawler/html_metadata.py4
1 files changed, 3 insertions, 1 deletions
diff --git a/python/sandcrawler/html_metadata.py b/python/sandcrawler/html_metadata.py
index 71715c2..a9536a6 100644
--- a/python/sandcrawler/html_metadata.py
+++ b/python/sandcrawler/html_metadata.py
@@ -219,7 +219,9 @@ def html_extract_biblio(doc: HTMLParser) -> Optional[BiblioMetadata]:
raw_date = meta.pop('raw_date', None)
if raw_date:
- meta['release_date'] = dateparser.parse(raw_date).date()
+ parsed = dateparser.parse(raw_date)
+ if parsed:
+ meta['release_date'] = parsed.date()
raw_release_type = meta.pop('raw_release_type', None)
if raw_release_type: