academic-refchecker 1.2.45__py3-none-any.whl → 1.2.46__py3-none-any.whl

__version__.py

@@ -1,3 +1,3 @@
 """Version information for RefChecker."""
 
-__version__ = "1.2.45"
+__version__ = "1.2.46"

{academic_refchecker-1.2.45.dist-info → academic_refchecker-1.2.46.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: academic-refchecker
-Version: 1.2.45
+Version: 1.2.46
 Summary: A comprehensive tool for validating reference accuracy in academic papers
 Author-email: Mark Russinovich <markrussinovich@hotmail.com>
 License-Expression: MIT

{academic_refchecker-1.2.45.dist-info → academic_refchecker-1.2.46.dist-info}/RECORD

@@ -1,5 +1,5 @@
-__version__.py,sha256=8vdrigO4-YfHufQMfh_RQ9NlN5btmqndss2dAOLxa1Q,65
-academic_refchecker-1.2.45.dist-info/licenses/LICENSE,sha256=Kwrx3fePVCeEFDCZvCW4OuoTNBiSoYbpGBI6qzGhWF0,1067
+__version__.py,sha256=PuEdWOtQEDRwfL2m1mk5h6WKAoScQu-kbHu9VkBS764,65
+academic_refchecker-1.2.46.dist-info/licenses/LICENSE,sha256=Kwrx3fePVCeEFDCZvCW4OuoTNBiSoYbpGBI6qzGhWF0,1067
 checkers/__init__.py,sha256=T0PAHTFt6UiGvn-WGoJU8CdhXNmf6zaHmcGVoWHhmJQ,533
 checkers/crossref.py,sha256=cLYmSzE8ehJ5sNko_R3fEiGBGiPH5_HxLhFM-pCfDRM,20378
 checkers/enhanced_hybrid_checker.py,sha256=6yf5tV4jLSVzjX1xR_kQq0NOgQIst-z_WmkiqqMc8hQ,23469
@@ -7,8 +7,8 @@ checkers/github_checker.py,sha256=GoepG4aRRUqAomkM4HgOSNf20BPxQgocZEpsk0ZTZZU,14
 checkers/local_semantic_scholar.py,sha256=DgGMjmR_w_saz2UTMatEhfXbhUED9tUmDG3vlZAIzc4,20428
 checkers/openalex.py,sha256=Fbc7iscZzmXjAZxH32PDX2r2Nwo9b5Ku-Sh1Ut9KpLA,19550
 checkers/openreview_checker.py,sha256=EYRcVu7sZzrFjqdOEKp72vd0S3sXSSjTgI9jsAVpdD0,20524
-checkers/semantic_scholar.py,sha256=0LcVahf3twyHqaD7bQ2eJiTyg-AQ9NGvVohb9nqaHdA,34884
-checkers/webpage_checker.py,sha256=woY8mNgZ4Lr9Ug53CN-Xo_2P62BTpR2u_FZyUPgTEuA,21833
+checkers/semantic_scholar.py,sha256=99KOHLAiYs31nSdx-gcMR_TWIlV8G4juNL0bmV4AoUs,34768
+checkers/webpage_checker.py,sha256=Ivzhu0xcpeWZcCTlAt9C06Lfsz5gKEiSko9gJ7EyQEw,22324
 config/__init__.py,sha256=r7sONsX2-ITviUJRU1KEz76uAuTRqZlzU-TVkvFRGYY,15
 config/logging.conf,sha256=r1tP0ApLHtlz7rV-oKS1MVO7oXJOgahbZFTtYmKnf9U,687
 config/settings.py,sha256=-vODFoXbWbGPUElpmchE5zbCj_n4Vtxr8HU1hQDFp_c,6164
@@ -26,7 +26,7 @@ scripts/start_vllm_server.py,sha256=ZepWp2y2cKFW0Kgsoima2RbmF02fTU29UFcLLpsBhFU,
 services/__init__.py,sha256=jGi9S74Msak3YR-C4Qb68VU7HB4oLaX9o1rlVAFpOFI,187
 services/pdf_processor.py,sha256=vu_JnhFGZY6jFVbDbPvG-mlQojvB-3Dzc8_946KVV2E,9427
 utils/__init__.py,sha256=1RrGoIIn1_gVzxd56b6a7HeAS-wu7uDP-nxLbR3fJ-8,1199
-utils/arxiv_utils.py,sha256=Y8sDJgDwHxp1L33BkQoDumIl0Pkp-BuYZb1PwWYsmak,18251
+utils/arxiv_utils.py,sha256=MxyD3Q0EzrmE0xORMJw8wdVtZ4Fp-ux_cn6jLMQimV8,18168
 utils/author_utils.py,sha256=DLTo1xsxef2wxoe4s_MWrh36maj4fgnvFlsDLpDE-qQ,5507
 utils/biblatex_parser.py,sha256=OkHXQcjiBrEDuhBfEk0RtmAYxufu5lAxAjb8__DzMjI,25537
 utils/bibliography_utils.py,sha256=mpmdAklzAs1CT3gqrOcjujGhouL95OuliCx0LE9Pg90,11705
@@ -36,11 +36,11 @@ utils/db_utils.py,sha256=_wSupfBlm0ILFvntQTvoj7tLDCbrYPRQrp9NDvphF_E,6281
 utils/doi_utils.py,sha256=qAf2slTLkgBwwnkYsWMkwMSDf6q8wNqd2YwNxhQudEc,3995
 utils/error_utils.py,sha256=wrLQhVJIIWUzcTSOLVvrvCAtMRNXt_1YYL5OXYnYf0A,11085
 utils/mock_objects.py,sha256=QxU-UXyHSY27IZYN8Sb8ei0JtNkpGSdMXoErrRLHXvE,6437
-utils/text_utils.py,sha256=T67Y-HSNokj-mOcdCtOcULNviBxyaG9xTjRd_l9titI,210088
+utils/text_utils.py,sha256=gFI-qu6g-9Lo1s3w1OjgBZ9SvdPufL1mMg-05l0BwD0,210269
 utils/unicode_utils.py,sha256=-WBKarXO756p7fd7gCeNsMag4ztDNURwFX5IVniOtwY,10366
 utils/url_utils.py,sha256=HdxIO8QvciP6Jp8Wd4sTSrS8JQrOMwgM7pxdUC8RJb4,9176
-academic_refchecker-1.2.45.dist-info/METADATA,sha256=mY4M9FRaDKcyS5yOFvR3X0Y0bj47_YmZeMayvrrpS38,22576
-academic_refchecker-1.2.45.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-academic_refchecker-1.2.45.dist-info/entry_points.txt,sha256=WdI89tYkIfz-M628PiboOfOLzTBWZAqvlF29qCVCkek,61
-academic_refchecker-1.2.45.dist-info/top_level.txt,sha256=6RlcQEA0kHb7-ndbKMFMZnYnJQVohgsU6BBkbEvJvEs,69
-academic_refchecker-1.2.45.dist-info/RECORD,,
+academic_refchecker-1.2.46.dist-info/METADATA,sha256=lP_pMHS9uI4hhXEsox_yGlSnT6pSl1_6W58CtzZEbDM,22576
+academic_refchecker-1.2.46.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+academic_refchecker-1.2.46.dist-info/entry_points.txt,sha256=WdI89tYkIfz-M628PiboOfOLzTBWZAqvlF29qCVCkek,61
+academic_refchecker-1.2.46.dist-info/top_level.txt,sha256=6RlcQEA0kHb7-ndbKMFMZnYnJQVohgsU6BBkbEvJvEs,69
+academic_refchecker-1.2.46.dist-info/RECORD,,

checkers/semantic_scholar.py

@@ -543,49 +543,50 @@ class NonArxivReferenceChecker:
         elif paper_venue and not isinstance(paper_venue, str):
             paper_venue = str(paper_venue)
         
+        # Check venue mismatches
         if cited_venue and paper_venue:
             # Use the utility function to check if venues are substantially different
             if are_venues_substantially_different(cited_venue, paper_venue):
                 from utils.error_utils import create_venue_warning
                 errors.append(create_venue_warning(cited_venue, paper_venue))
         elif not cited_venue and paper_venue:
-            # Check if this is an arXiv paper first
-            external_ids = paper_data.get('externalIds', {})
-            arxiv_id = external_ids.get('ArXiv') if external_ids else None
-            
-            if arxiv_id:
-                # For arXiv papers, suggest including the arXiv URL instead of venue
-                arxiv_url = f"https://arxiv.org/abs/{arxiv_id}"
-                
-                # Check if the reference already includes this ArXiv URL or equivalent DOI
-                reference_url = reference.get('url', '')
-                
-                # Check for direct arXiv URL match
-                has_arxiv_url = arxiv_url in reference_url
-                
-                # Also check for arXiv DOI URL (e.g., https://doi.org/10.48550/arxiv.2505.11595)
-                arxiv_doi_url = f"https://doi.org/10.48550/arxiv.{arxiv_id}"
-                has_arxiv_doi = arxiv_doi_url.lower() in reference_url.lower()
-                
-                if not (has_arxiv_url or has_arxiv_doi):
+            # Original reference has the venue in raw text but not parsed correctly
+            raw_text = reference.get('raw_text', '')
+            if raw_text and '#' in raw_text:
+                # Check if venue might be in the raw text format (author#title#venue#year#url)
+                parts = raw_text.split('#')
+                if len(parts) >= 3 and parts[2].strip():
+                    # Venue is present in raw text but missing from parsed reference
                     errors.append({
                         'warning_type': 'venue',
-                        'warning_details': f"Reference should include arXiv URL: {arxiv_url}",
-                        'ref_url_correct': arxiv_url
+                        'warning_details': f"Venue missing: should include '{paper_venue}'",
+                        'ref_venue_correct': paper_venue
                     })
-            else:
-                # Original reference has the venue in raw text but not parsed correctly
-                raw_text = reference.get('raw_text', '')
-                if raw_text and '#' in raw_text:
-                    # Check if venue might be in the raw text format (author#title#venue#year#url)
-                    parts = raw_text.split('#')
-                    if len(parts) >= 3 and parts[2].strip():
-                        # Venue is present in raw text but missing from parsed reference
-                        errors.append({
-                            'warning_type': 'venue',
-                            'warning_details': f"Venue missing: should include '{paper_venue}'",
-                            'ref_venue_correct': paper_venue
-                        })
+
+        # Always check for missing arXiv URLs when paper has arXiv ID
+        external_ids = paper_data.get('externalIds', {})
+        arxiv_id = external_ids.get('ArXiv') if external_ids else None
+        
+        if arxiv_id:
+            # For arXiv papers, check if reference includes the arXiv URL
+            arxiv_url = f"https://arxiv.org/abs/{arxiv_id}"
+            
+            # Check if the reference already includes this ArXiv URL or equivalent DOI
+            reference_url = reference.get('url', '')
+            
+            # Check for direct arXiv URL match
+            has_arxiv_url = arxiv_url in reference_url
+            
+            # Also check for arXiv DOI URL (e.g., https://doi.org/10.48550/arxiv.2505.11595)
+            arxiv_doi_url = f"https://doi.org/10.48550/arxiv.{arxiv_id}"
+            has_arxiv_doi = arxiv_doi_url.lower() in reference_url.lower()
+            
+            if not (has_arxiv_url or has_arxiv_doi):
+                errors.append({
+                    'warning_type': 'url',
+                    'warning_details': f"Reference could include arXiv URL: {arxiv_url}",
+                    'ref_url_correct': arxiv_url
+                })
 
         # Verify DOI
         paper_doi = None

checkers/webpage_checker.py

@@ -71,7 +71,8 @@ class WebPageChecker:
         doc_indicators = [
             'docs', 'documentation', 'readthedocs.io', 'help', 'guide', 'tutorial',
             'reference', 'manual', 'wiki', 'blog', 'api', 'developer', 'platform',
-            'index', 'research', 'news', 'insights', 'whitepaper', 'brief', 'develop'
+            'index', 'research', 'news', 'insights', 'whitepaper', 'brief', 'develop',
+            'posts'  # For blog posts and forum posts like LessWrong
         ]
         
         return any(indicator in url.lower() for indicator in doc_indicators) or self._is_likely_webpage(url)
@@ -84,7 +85,8 @@ class WebPageChecker:
         doc_domains = [
             'pytorch.org', 'tensorflow.org', 'readthedocs.io', 'onnxruntime.ai',
             'deepspeed.ai', 'huggingface.co', 'openai.com', 'microsoft.com',
-            'google.com', 'nvidia.com', 'intel.com', 'langchain.com'
+            'google.com', 'nvidia.com', 'intel.com', 'langchain.com',
+            'lesswrong.com'  # LessWrong rationality and AI safety blog platform
         ]
         
         return any(domain in parsed.netloc for domain in doc_domains)
@@ -395,6 +397,14 @@ class WebPageChecker:
         organization = site_info.get('organization', '').lower()
         domain = site_info.get('domain', '').lower()
         
+        # Accept generic web resource terms - these are valid for any web URL
+        generic_web_terms = [
+            'web resource', 'web site', 'website', 'online resource', 
+            'online', 'web', 'internet resource', 'web page', 'webpage'
+        ]
+        if cited_lower in generic_web_terms:
+            return True
+        
         # Direct matches
         if cited_lower in organization or organization in cited_lower:
             return True

utils/arxiv_utils.py

@@ -392,32 +392,29 @@ def get_bibtex_content(paper):
         logger.debug(f"Detected ArXiv paper {arxiv_id}, checking for structured bibliography")
         tex_content, bib_content, bbl_content = download_arxiv_source(arxiv_id)
         
-        # Choose between .bib and .bbl files based on content richness
-        # Prioritize .bbl if it has more references than filtered .bib, otherwise prefer .bib
+        # Choose between .bib and .bbl files - .bbl files take priority when they contain entries
+        # .bbl files are processed biblatex output that reflects exactly what was cited
         if bib_content and bbl_content:
-            # Count entries in both
+            # Count entries in both for logging
             bib_entry_count = len(re.findall(r'@\w+\s*\{', bib_content))
             bbl_entry_count = len(re.findall(r'\\bibitem\[', bbl_content))
             
-            # If we have LaTeX content, get filtered BibTeX count
-            filtered_bib_count = bib_entry_count
-            filtered_content = bib_content
-            if tex_content:
-                cited_keys = extract_cited_keys_from_tex({}, tex_content)
-                if cited_keys:
-                    logger.debug(f"Found {len(cited_keys)} cited keys, filtering BibTeX")
-                    filtered_content = filter_bibtex_by_citations(bib_content, {}, tex_content)
-                    filtered_bib_count = len(re.findall(r'@\w+\s*\{', filtered_content))
-            
-            logger.debug(f"Bibliography comparison: .bbl has {bbl_entry_count} entries, filtered .bib has {filtered_bib_count} entries")
+            logger.debug(f"Bibliography comparison: .bbl has {bbl_entry_count} entries, .bib has {bib_entry_count} entries")
             
-            # Prioritize .bbl if it has significantly more entries
-            if bbl_entry_count > filtered_bib_count * 1.5:  # 50% more entries threshold
-                logger.info(f"Using .bbl files from ArXiv source")
+            # Only use .bbl if it actually contains bibliography entries
+            if bbl_entry_count > 0:
+                logger.info(f"Using .bbl files from ArXiv source (biblatex takes priority over bibtex)")
                 return bbl_content
             else:
-                logger.info(f"Using filtered .bib files")
-                return filtered_content
+                logger.info(f"Using .bib files from ArXiv source (.bbl file is empty)")
+                # If we have LaTeX content, filter BibTeX by cited keys
+                if tex_content:
+                    cited_keys = extract_cited_keys_from_tex({}, tex_content)
+                    if cited_keys:
+                        logger.debug(f"Found {len(cited_keys)} cited keys, filtering BibTeX")
+                        filtered_content = filter_bibtex_by_citations(bib_content, {}, tex_content)
+                        return filtered_content
+                return bib_content
                     
         elif bib_content:
             logger.info(f"Found .bib files in ArXiv source for {arxiv_id}")

utils/text_utils.py

@@ -2111,7 +2111,9 @@ def compare_authors(cited_authors: list, correct_authors: list, normalize_func=N
         # Allow minor flexibility (1 author difference) but not more
         if abs(len(cleaned_cited) - len(correct_names)) > 1:
             from utils.error_utils import format_author_count_mismatch
-            error_msg = format_author_count_mismatch(len(cleaned_cited), len(correct_names), cleaned_cited, correct_names)
+            # Convert cited names to display format (First Last) before showing in error
+            display_cited = [format_author_for_display(author) for author in cleaned_cited]
+            error_msg = format_author_count_mismatch(len(cleaned_cited), len(correct_names), display_cited, correct_names)
             return False, error_msg
         
         # Use the shorter list for comparison