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/jalc.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/jalc.py')
| -rw-r--r-- | python/fatcat_tools/importers/jalc.py | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/python/fatcat_tools/importers/jalc.py b/python/fatcat_tools/importers/jalc.py index 9bf2621c..12f5450f 100644 --- a/python/fatcat_tools/importers/jalc.py +++ b/python/fatcat_tools/importers/jalc.py @@ -5,6 +5,7 @@ import datetime from bs4 import BeautifulSoup import fatcat_openapi_client +from fatcat_tools.normal import clean_doi from .common import EntityImporter, clean, is_cjk, DATE_FMT @@ -171,7 +172,7 @@ class JalcImporter(EntityImporter): doi = None if record.doi: - doi = record.doi.string.lower().strip() + doi = clean_doi(record.doi.string.strip().lower()) if doi.startswith('http://dx.doi.org/'): doi = doi.replace('http://dx.doi.org/', '') elif doi.startswith('https://dx.doi.org/'): |