diff options
author | Bryan Newbold <bnewbold@robocracy.org> | 2021-11-09 18:13:15 -0800 |
---|---|---|
committer | Bryan Newbold <bnewbold@robocracy.org> | 2021-11-09 18:49:46 -0800 |
commit | 1024e688bb12d64648ceb638daf049d508f87561 (patch) | |
tree | 8995645d370ac89d105a1cd7b1574fcba3bf5a3a /python/fatcat_tools/importers/crossref.py | |
parent | 2fd90ad2cc561fa743a617315824b2744f737575 (diff) | |
download | fatcat-1024e688bb12d64648ceb638daf049d508f87561.tar.gz fatcat-1024e688bb12d64648ceb638daf049d508f87561.zip |
importers: use clean_doi() in many more (all?) importers
Diffstat (limited to 'python/fatcat_tools/importers/crossref.py')
-rw-r--r-- | python/fatcat_tools/importers/crossref.py | 9 |
1 files changed, 8 insertions, 1 deletions
diff --git a/python/fatcat_tools/importers/crossref.py b/python/fatcat_tools/importers/crossref.py index 9c69fee3..c9f251fc 100644 --- a/python/fatcat_tools/importers/crossref.py +++ b/python/fatcat_tools/importers/crossref.py @@ -4,6 +4,8 @@ from typing import Any, Dict, List, Optional, Sequence import fatcat_openapi_client from fatcat_openapi_client import ApiClient, ReleaseContrib, ReleaseEntity +from fatcat_tools.normal import clean_doi + from .common import EntityImporter, clean # The docs/guide should be the canonical home for these mappings; update there @@ -467,6 +469,11 @@ class CrossrefImporter(EntityImporter): self.counts["skip-blank-title"] += 1 return None + doi = clean_doi(obj["DOI"].lower()) + if not doi: + self.counts["skip-bad-doi"] += 1 + return None + subtitle = None if obj.get("subtitle"): subtitle = clean(obj["subtitle"][0], force_xml=True) @@ -489,7 +496,7 @@ class CrossrefImporter(EntityImporter): release_year=release_year, publisher=publisher, ext_ids=fatcat_openapi_client.ReleaseExtIds( - doi=obj["DOI"].lower(), + doi=doi, isbn13=isbn13, ), volume=clean(obj.get("volume")), |