fix: add DOI to prefix_mapping in _get_not_found_ids

The `_get_not_found_ids` method was missing DOI from its
`prefix_mapping` dict. When a paper was looked up using a
DOI-prefixed ID (e.g., `DOI:10.1145/792550.792552`), the method
would add only the bare DOI value to `found_ids` (without the
`DOI:` prefix), causing the input ID to never match. This resulted
in a false "IDs not found" warning for every DOI-prefixed lookup,
even when the paper was successfully returned.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: lior-airis

semanticscholar/AsyncSemanticScholar.py

@@ -228,6 +228,7 @@ def _get_not_found_ids(self, paper_ids, papers):
 
         prefix_mapping = {
             'ARXIV': 'ArXiv',
+            'DOI': 'DOI',
             'MAG': 'MAG',
             'ACL': 'ACL',
             'PMID': 'PubMed',

tests/test_semanticscholar.py

@@ -284,6 +284,18 @@ def test_get_papers_return_not_found(self):
         self.assertEqual(len(not_found), 1)
         self.assertEqual(not_found[0], 'CorpusId:211530585')
 
+    def test_get_papers_doi_prefix_not_false_positive(self):
+        paper = Paper({
+            'paperId': 'abc123',
+            'externalIds': {
+                'DOI': '10.1145/792550.792552',
+                'CorpusId': 12345
+            }
+        })
+        not_found = self.sch._AsyncSemanticScholar._get_not_found_ids(
+            ['DOI:10.1145/792550.792552'], [paper])
+        self.assertEqual(not_found, [])
+
     @test_vcr.use_cassette
     def test_get_paper_authors(self):
         data = self.sch.get_paper_authors('10.2139/ssrn.2250500')
@@ -829,7 +841,19 @@ async def test_get_papers_return_not_found_async(self):
         not_found = data[1]
         self.assertEqual(len(not_found), 1)
         self.assertEqual(not_found[0], 'CorpusId:211530585')
-    
+
+    def test_get_papers_doi_prefix_not_false_positive_async(self):
+        paper = Paper({
+            'paperId': 'abc123',
+            'externalIds': {
+                'DOI': '10.1145/792550.792552',
+                'CorpusId': 12345
+            }
+        })
+        not_found = self.sch._get_not_found_ids(
+            ['DOI:10.1145/792550.792552'], [paper])
+        self.assertEqual(not_found, [])
+
     @test_vcr.use_cassette
     async def test_get_paper_authors_async(self):
         data = await self.sch.get_paper_authors('10.2139/ssrn.2250500')