diff options
| author | Bryan Newbold <bnewbold@robocracy.org> | 2021-06-17 16:26:50 -0700 |
|---|---|---|
| committer | Bryan Newbold <bnewbold@robocracy.org> | 2021-06-23 17:51:15 -0700 |
| commit | fa11747574f086e99459914f93d24bad7a8eacce (patch) | |
| tree | 9e05825c26de60042142dd846cee31d85661a6b4 /python/fatcat_tools/importers/ingest.py | |
| parent | cf2c3432d3c28d2ba0a92c62fabda4ca434afe46 (diff) | |
| download | fatcat-fa11747574f086e99459914f93d24bad7a8eacce.tar.gz fatcat-fa11747574f086e99459914f93d24bad7a8eacce.zip | |
more consistent and defensive lower-casing of DOIs
After noticing more upper/lower ambiguity in production. In particular,
we have some old ingest requests in sandcrawler DB, which get
re-submitted/re-tried, which have capitalized DOIs in the link source id
field.
Diffstat (limited to 'python/fatcat_tools/importers/ingest.py')
| -rw-r--r-- | python/fatcat_tools/importers/ingest.py | 4 |
1 files changed, 4 insertions, 0 deletions
diff --git a/python/fatcat_tools/importers/ingest.py b/python/fatcat_tools/importers/ingest.py index 483932ad..ae3e147a 100644 --- a/python/fatcat_tools/importers/ingest.py +++ b/python/fatcat_tools/importers/ingest.py @@ -133,6 +133,8 @@ class IngestFileResultImporter(EntityImporter): extid = request['ext_ids'].get(extid_type) if not extid: continue + if extid_type == 'doi': + extid = extid.lower() try: release = self.api.lookup_release(**{extid_type: extid}) except fatcat_openapi_client.rest.ApiException as err: @@ -217,6 +219,8 @@ class IngestFileResultImporter(EntityImporter): if request.get('link_source') and request.get('link_source_id'): edit_extra['link_source'] = request['link_source'] edit_extra['link_source_id'] = request['link_source_id'] + if edit_extra['link_source'] == 'doi': + edit_extra['link_source_id'] = edit_extra['link_source_id'].lower() return edit_extra |