From 2a492914082444690f853a55ab1394fc0cf50108 Mon Sep 17 00:00:00 2001
From: Bryan Newbold <bnewbold@robocracy.org>
Date: Tue, 11 Aug 2020 14:52:47 -0700
Subject: entity update: skip ingest of figshare+zenodo 'group' DOIs

---
 python/fatcat_tools/workers/changelog.py | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/python/fatcat_tools/workers/changelog.py b/python/fatcat_tools/workers/changelog.py
index 1ac7a865..dc5ef299 100644
--- a/python/fatcat_tools/workers/changelog.py
+++ b/python/fatcat_tools/workers/changelog.py
@@ -175,6 +175,7 @@ class EntityUpdatesWorker(FatcatWorker):
             'paper-conference',
             'patent',
             'peer_review',
+            'post',
             'report',
             'retraction',
             'review',
@@ -217,6 +218,20 @@ class EntityUpdatesWorker(FatcatWorker):
                 if doi.startswith(prefix):
                     return False
 
+        # figshare
+        if doi and doi.startswith('10.6084/') or doi.startswith('10.25384/'):
+            # don't crawl "most recent version" (aka "group") DOIs
+            if not release.version:
+                return False
+
+        # zenodo
+        if doi and doi.startswith('10.5281/'):
+            # if this is a "grouping" DOI of multiple "version" DOIs, do not crawl (will crawl the versioned DOIs)
+            if release.extra and release.extra.get('relations'):
+                for rel in release.extra['relations']:
+                    if (rel.get('relationType') == 'HasVersion' and rel.get('relatedIdentifier', '').startswith('10.5281/')):
+                        return False
+
         return True
 
     def run(self):
-- 
cgit v1.2.3