summaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools
diff options
context:
space:
mode:
authorBryan Newbold <bnewbold@robocracy.org>2019-01-24 15:21:43 -0800
committerBryan Newbold <bnewbold@robocracy.org>2019-01-24 15:21:48 -0800
commitf955f66789b0078dcb973ce587d2d3b3184e73a7 (patch)
treee048502b461fdfa973d4697b3043f4a516694b2f /python/fatcat_tools
parent206acf1b37a1a34d5338c744e17ef2035cd2db58 (diff)
downloadfatcat-f955f66789b0078dcb973ce587d2d3b3184e73a7.tar.gz
fatcat-f955f66789b0078dcb973ce587d2d3b3184e73a7.zip
allow importing contrib/refs lists
The motivation here isn't really to support these gigantic lists on principle, but to be able to ingest large corpuses without having to decide whether to filter out or crop such lists.
Diffstat (limited to 'python/fatcat_tools')
-rw-r--r--python/fatcat_tools/importers/crossref.py18
1 files changed, 13 insertions, 5 deletions
diff --git a/python/fatcat_tools/importers/crossref.py b/python/fatcat_tools/importers/crossref.py
index 00c719f1..4a0322e7 100644
--- a/python/fatcat_tools/importers/crossref.py
+++ b/python/fatcat_tools/importers/crossref.py
@@ -303,9 +303,12 @@ class CrossrefImporter(EntityImporter):
# external identifiers
extids = self.lookup_ext_ids(doi=obj['DOI'].lower())
- # TODO: filter out huge releases; we'll get them later (and fix bug in
- # fatcatd)
- if max(len(contribs), len(refs), len(abstracts)) > 750:
+ # filter out unreasonably huge releases
+ if len(abstracts) > 100:
+ return None
+ if len(refs) > 2000:
+ return None
+ if len(refs) > 5000:
return None
# release date parsing is amazingly complex
@@ -322,11 +325,16 @@ class CrossrefImporter(EntityImporter):
release_year = raw_date[0]
release_date = None
+ original_title = None
+ if obj.get('original-title'):
+ original_title = clean(obj.get('original-title')[0], force_xml=True)
+ if obj.get('title'):
+ title = clean(obj.get('title')[0], force_xml=True)
re = fatcat_client.ReleaseEntity(
work_id=None,
container_id=container_id,
- title=clean(obj.get('title', [None])[0], force_xml=True),
- original_title=clean(obj.get('original-title', [None])[0]),
+ title=title,
+ original_title=original_title,
release_type=release_type,
release_status=release_status,
release_date=release_date,