summaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools
diff options
context:
space:
mode:
authorbnewbold <bnewbold@archive.org>2020-02-25 05:09:09 +0000
committerbnewbold <bnewbold@archive.org>2020-02-25 05:09:09 +0000
commit1556cdd7f0f5f4bc4fe5ccc9764c1598c852bb9b (patch)
tree7f31e907be31ecb9c1ef1f6fe94f8f6f17ae86ac /python/fatcat_tools
parentbd6fe91108f7f99f7a88a6b33d9323ffc4d8a6e6 (diff)
parentdc4116f8ebd225eb4af1cecfc75f5c1291589694 (diff)
downloadfatcat-1556cdd7f0f5f4bc4fe5ccc9764c1598c852bb9b.tar.gz
fatcat-1556cdd7f0f5f4bc4fe5ccc9764c1598c852bb9b.zip
Merge branch 'bnewbold-more-ingest' into 'master'
entity worker: ingest more Datacite releases; filter some out See merge request webgroup/fatcat!29
Diffstat (limited to 'python/fatcat_tools')
-rw-r--r--python/fatcat_tools/workers/changelog.py38
1 files changed, 37 insertions, 1 deletions
diff --git a/python/fatcat_tools/workers/changelog.py b/python/fatcat_tools/workers/changelog.py
index 745ee85a..b84d5e70 100644
--- a/python/fatcat_tools/workers/changelog.py
+++ b/python/fatcat_tools/workers/changelog.py
@@ -123,6 +123,33 @@ class EntityUpdatesWorker(FatcatWorker):
ingest_type = ingest_request.get('ingest_type')
doi = ingest_request.get('ext_ids', {}).get('doi')
+ is_document = release.release_type in (
+ 'article-journal',
+ 'paper-conference',
+ 'article',
+ 'report',
+ 'chapter',
+ 'manuscript',
+ 'review',
+ 'thesis',
+ 'letter',
+ 'editorial',
+ 'abstract',
+ 'entry',
+ 'patent',
+ 'post',
+ 'review-book',
+ )
+ is_not_pdf = release.release_type in (
+ 'dataset',
+ 'stub',
+ 'software',
+ 'figure',
+ 'graphic',
+ )
+
+ # accept list sets a default "crawl it" despite OA metadata for
+ # known-OA DOI prefixes
in_acceptlist = False
if doi:
for prefix in self.live_pdf_ingest_doi_prefix_acceptlist:
@@ -131,9 +158,18 @@ class EntityUpdatesWorker(FatcatWorker):
if self.ingest_oa_only and link_source not in ('arxiv', 'pmc'):
es = release_to_elasticsearch(release)
- if not es['is_oa'] and not in_acceptlist:
+ # most datacite documents are in IRs and should be crawled
+ is_datacite_doc = False
+ if release.extra and ('datacite' in release.extra) and is_document:
+ is_datacite_doc = True
+ if not (es['is_oa'] or in_acceptlist or is_datacite_doc):
return False
+ # if ingest_type is pdf but release_type is almost certainly not a PDF,
+ # skip it. This is mostly a datacite thing.
+ if ingest_type == "pdf" and is_not_pdf:
+ return False
+
if ingest_type == "pdf" and doi:
for prefix in self.ingest_pdf_doi_prefix_blocklist:
if doi.startswith(prefix):