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/grobid_metadata.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/grobid_metadata.py')
-rw-r--r-- | python/fatcat_tools/importers/grobid_metadata.py | 9 |
1 files changed, 6 insertions, 3 deletions
diff --git a/python/fatcat_tools/importers/grobid_metadata.py b/python/fatcat_tools/importers/grobid_metadata.py index e36e1b48..7c595787 100644 --- a/python/fatcat_tools/importers/grobid_metadata.py +++ b/python/fatcat_tools/importers/grobid_metadata.py @@ -7,6 +7,8 @@ from typing import Any, Dict, List, Optional import fatcat_openapi_client from fatcat_openapi_client import ApiClient, FileEntity, ReleaseEntity +from fatcat_tools.normal import clean_doi + from .common import EntityImporter, clean, make_rel_url MAX_ABSTRACT_BYTES = 4096 @@ -133,9 +135,10 @@ class GrobidMetadataImporter(EntityImporter): # only returns year, ever? release_year = int(obj["date"][:4]) - extra = dict() - if obj.get("doi"): - extra["doi"] = obj["doi"] + extra: Dict[str, Any] = dict() + doi = clean_doi(obj.get("doi")) + if doi: + extra["doi"] = doi if obj["journal"] and obj["journal"].get("name"): extra["container_name"] = clean(obj["journal"]["name"]) |