academic-refchecker 1.2.62__py3-none-any.whl → 1.2.64__py3-none-any.whl

{academic_refchecker-1.2.62.dist-info → academic_refchecker-1.2.64.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: academic-refchecker
-Version: 1.2.62
+Version: 1.2.64
 Summary: A comprehensive tool for validating reference accuracy in academic papers
 Author-email: Mark Russinovich <markrussinovich@hotmail.com>
 License-Expression: MIT

{academic_refchecker-1.2.62.dist-info → academic_refchecker-1.2.64.dist-info}/RECORD

@@ -1,7 +1,7 @@
-academic_refchecker-1.2.62.dist-info/licenses/LICENSE,sha256=Kwrx3fePVCeEFDCZvCW4OuoTNBiSoYbpGBI6qzGhWF0,1067
+academic_refchecker-1.2.64.dist-info/licenses/LICENSE,sha256=Kwrx3fePVCeEFDCZvCW4OuoTNBiSoYbpGBI6qzGhWF0,1067
 refchecker/__init__.py,sha256=Pg5MrtLxDBRcNYcI02N-bv3tzURVd1S3nQ8IyF7Zw7E,322
 refchecker/__main__.py,sha256=agBbT9iKN0g2xXtRNCoh29Nr7z2n5vU-r0MCVJKi4tI,232
-refchecker/__version__.py,sha256=82eUrwP5gqSW8BlwZxv0gZNfD9rQJb-uAsG91hYOXK0,89
+refchecker/__version__.py,sha256=5hnxQDsQrNttLvZ_a6svY53X2wfPUJeWaf_x62fqQt8,89
 refchecker/checkers/__init__.py,sha256=T0PAHTFt6UiGvn-WGoJU8CdhXNmf6zaHmcGVoWHhmJQ,533
 refchecker/checkers/crossref.py,sha256=5BeSCK8K_S_-iwgQaNAbxZGNsxaxOyBzUQ3AD0Rc6nU,20433
 refchecker/checkers/enhanced_hybrid_checker.py,sha256=c5I_h8w6xD7XkBNkbneffeAnrO8B-uXH99edWBJvDMo,27788
@@ -29,7 +29,7 @@ refchecker/scripts/start_vllm_server.py,sha256=ZepWp2y2cKFW0Kgsoima2RbmF02fTU29U
 refchecker/services/__init__.py,sha256=jGi9S74Msak3YR-C4Qb68VU7HB4oLaX9o1rlVAFpOFI,187
 refchecker/services/pdf_processor.py,sha256=7i5x043qfnyzE5EQmytfy_uPjbeCJp4Ka5OPyH-bwOE,10577
 refchecker/utils/__init__.py,sha256=1RrGoIIn1_gVzxd56b6a7HeAS-wu7uDP-nxLbR3fJ-8,1199
-refchecker/utils/arxiv_utils.py,sha256=idlCzkTApYwH-kdTiH9nrfo4GMmwdtUAv7cAGtoEG-0,19799
+refchecker/utils/arxiv_utils.py,sha256=PvMBJGubvDvo9_WxrXUH4OWMNrEAD7z7NCcRl0la2iI,20601
 refchecker/utils/author_utils.py,sha256=DLTo1xsxef2wxoe4s_MWrh36maj4fgnvFlsDLpDE-qQ,5507
 refchecker/utils/biblatex_parser.py,sha256=IKRUMtRsjdXIktyk9XGArt_ms0asmqP549uhFvvumuE,25581
 refchecker/utils/bibliography_utils.py,sha256=d6kqDOQou_PX6WQkOzrGyN5GpzaOjhu54w9wGfBRQZw,11760
@@ -42,8 +42,8 @@ refchecker/utils/mock_objects.py,sha256=QxU-UXyHSY27IZYN8Sb8ei0JtNkpGSdMXoErrRLH
 refchecker/utils/text_utils.py,sha256=d_X4r1nVvkL7i0DhxfLaVK3CzbMP2oZvqX3kxfDudQw,220978
 refchecker/utils/unicode_utils.py,sha256=-WBKarXO756p7fd7gCeNsMag4ztDNURwFX5IVniOtwY,10366
 refchecker/utils/url_utils.py,sha256=7b0rWCQJSajzqOvD7ghsBZPejiq6mUIz6SGhvU_WGDs,9441
-academic_refchecker-1.2.62.dist-info/METADATA,sha256=t8RL0LnlyGiAHVyhAgsH08J9oXyZt44TOtomQJj2n5s,24104
-academic_refchecker-1.2.62.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-academic_refchecker-1.2.62.dist-info/entry_points.txt,sha256=kG6k2JwFIRvmKe0oZTr2RYStyfl79BirJxyaO6kjIxA,72
-academic_refchecker-1.2.62.dist-info/top_level.txt,sha256=ZdIg_PFHiATpVT5Uvp4L17Q0d8mk8ZBsINXKf1tE0bo,11
-academic_refchecker-1.2.62.dist-info/RECORD,,
+academic_refchecker-1.2.64.dist-info/METADATA,sha256=qiFF2C0mnmv6PVve58sKwIT_obVcgTIU2LGTgXJR4mk,24104
+academic_refchecker-1.2.64.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+academic_refchecker-1.2.64.dist-info/entry_points.txt,sha256=kG6k2JwFIRvmKe0oZTr2RYStyfl79BirJxyaO6kjIxA,72
+academic_refchecker-1.2.64.dist-info/top_level.txt,sha256=ZdIg_PFHiATpVT5Uvp4L17Q0d8mk8ZBsINXKf1tE0bo,11
+academic_refchecker-1.2.64.dist-info/RECORD,,

refchecker/__version__.py

@@ -1,5 +1,5 @@
 """Version information for RefChecker."""
 
-__version__ = "1.2.62"
+__version__ = "1.2.64"
 
-__version__ = "1.2.62"
+__version__ = "1.2.64"

refchecker/utils/arxiv_utils.py

@@ -458,7 +458,16 @@ def get_bibtex_content(paper):
             
             logger.debug(f"Bibliography comparison: .bbl has {bbl_entry_count} entries, .bib has {bib_entry_count} entries")
             
-            if uses_bibtex and bib_entry_count > 0:
+            # IMPORTANT: Prefer .bbl when .bib is excessively large (e.g., includes full ACL Anthology)
+            # The .bbl file contains only the actually-cited references, while .bib may contain
+            # entire bibliography databases. Parsing 80k+ entries would cause the tool to hang.
+            # Use .bbl if: (1) .bbl has entries AND (2) .bib has >10x more entries than .bbl OR >1000 entries
+            excessive_bib = bib_entry_count > 1000 or (bbl_entry_count > 0 and bib_entry_count > bbl_entry_count * 10)
+            
+            if bbl_entry_count > 0 and excessive_bib:
+                logger.info(f"Using .bbl files from ArXiv source (.bib has {bib_entry_count} entries which is excessive, .bbl has {bbl_entry_count})")
+                return bbl_content
+            elif uses_bibtex and bib_entry_count > 0 and not excessive_bib:
                 logger.info(f"Using .bib files from ArXiv source (main TeX uses \\bibliography{{...}})")
                 return bib_content
             elif bbl_entry_count > 0: