Create content_analysis.py

content_analysis.py

@@ -0,0 +1,210 @@
+# content_analysis.py
+import re
+from typing import List, Dict, Any
+from collections import Counter
+import language_tool_python
+import traceback
+
+# Import utility from text_utils
+from text_utils import convert_markdown_to_plain_text
+
+def check_text_presence(full_text: str, search_terms: List[str]) -> Dict[str, bool]:
+    return {term: term.lower() in full_text.lower() for term in search_terms}
+
+def label_authors(full_text: str) -> str:
+    author_line_regex = r"^(?:.*\n)(.*?)(?:\n\n)"
+    match = re.search(author_line_regex, full_text, re.MULTILINE)
+    if match:
+        authors = match.group(1).strip()
+        return full_text.replace(authors, f"Authors: {authors}")
+    return full_text
+
+def check_metadata(plain_text: str) -> Dict[str, Any]:
+    return {
+        "author_email": bool(re.search(r'\b[\w.-]+?@\w+?\.\w+?\b', plain_text)),
+        "list_of_authors": bool(re.search(r'Authors?:', plain_text, re.IGNORECASE)),
+        "keywords_list": bool(re.search(r'Keywords?:', plain_text, re.IGNORECASE)),
+        "word_count": len(plain_text.split()) or "Missing"
+    }
+
+def check_disclosures(plain_text: str) -> Dict[str, bool]:
+    search_terms = [
+        "conflict of interest statement",
+        "ethics statement",
+        "funding statement",
+        "data access statement"
+    ]
+    results = check_text_presence(plain_text, search_terms)
+    has_author_contribution = ("author contribution statement" in plain_text.lower() or
+                               "author contributions statement" in plain_text.lower())
+    results["author contribution statement"] = has_author_contribution
+    return results
+
+def check_figures_and_tables(plain_text: str) -> Dict[str, bool]:
+    return {
+        "figures_with_citations": bool(re.search(r'Figure \d+.*?citation', plain_text, re.IGNORECASE)),
+        "figures_legends": bool(re.search(r'Figure \d+.*?legend', plain_text, re.IGNORECASE)),
+        "tables_legends": bool(re.search(r'Table \d+.*?legend', plain_text, re.IGNORECASE))
+    }
+
+def check_references_summary(plain_text: str) -> Dict[str, Any]:
+    abstract_candidate = plain_text[:2000] 
+    return {
+        "old_references": bool(re.search(r'\b19[0-9]{2}\b', plain_text)), 
+        "citations_in_abstract": bool(re.search(r'\[\d+\]', abstract_candidate, re.IGNORECASE)) or \
+                                 bool(re.search(r'\bcit(?:ation|ed)\b', abstract_candidate, re.IGNORECASE)),
+        "reference_count": len(re.findall(r'\[\d+(?:,\s*\d+)*\]', plain_text)), 
+        "self_citations": bool(re.search(r'Self-citation', plain_text, re.IGNORECASE)) 
+    }
+
+def check_structure(plain_text: str) -> Dict[str, bool]:
+    text_lower = plain_text.lower()
+    return {
+        "imrad_structure": all(section.lower() in text_lower for section in ["introduction", "method", "result", "discussion"]),
+        "abstract_structure": "structured abstract" in text_lower
+    }
+
+def check_language_issues_and_regex(markdown_text_from_pdf: str) -> Dict[str, Any]:
+    if not markdown_text_from_pdf.strip():
+        return {"total_issues": 0, "issues_list": [], "text_used_for_analysis": ""}
+
+    plain_text_from_markdown = convert_markdown_to_plain_text(markdown_text_from_pdf)
+    text_for_analysis = plain_text_from_markdown.replace('\n', ' ')
+    text_for_analysis = re.sub(r'\s+', ' ', text_for_analysis).strip()
+
+    if not text_for_analysis:
+        return {"total_issues": 0, "issues_list": [], "text_used_for_analysis": ""}
+
+    text_for_analysis_lower = text_for_analysis.lower()
+    
+    abstract_match = re.search(r'\babstract\b', text_for_analysis_lower)
+    content_start_index = abstract_match.start() if abstract_match else 0
+    if abstract_match: print(f"Found 'abstract' at index {content_start_index}")
+    else: print(f"Did not find 'abstract', starting language analysis from index 0")
+
+    references_match = re.search(r'\breferences\b', text_for_analysis_lower)
+    bibliography_match = re.search(r'\bbibliography\b', text_for_analysis_lower)
+    content_end_index = len(text_for_analysis)
+
+    if references_match and bibliography_match:
+        content_end_index = min(references_match.start(), bibliography_match.start())
+        print(f"Found 'references' at {references_match.start()} and 'bibliography' at {bibliography_match.start()}. Using {content_end_index} as end boundary.")
+    elif references_match:
+        content_end_index = references_match.start()
+        print(f"Found 'references' at {content_end_index}. Using it as end boundary.")
+    elif bibliography_match:
+        content_end_index = bibliography_match.start()
+        print(f"Found 'bibliography' at {content_end_index}. Using it as end boundary.")
+    else:
+        print(f"Did not find 'references' or 'bibliography'. Language analysis up to end of text (index {content_end_index}).")
+
+    if content_start_index >= content_end_index:
+        print(f"Warning: Content start index ({content_start_index}) is not before content end index ({content_end_index}). No language issues will be reported from this range.")
+    
+    tool = None
+    processed_issues: List[Dict[str, Any]] = []
+    try:
+        tool = language_tool_python.LanguageTool('en-US') 
+        raw_lt_matches = tool.check(text_for_analysis)
+        
+        lt_issues_in_range = 0
+        for idx, match in enumerate(raw_lt_matches):
+            if match.ruleId == "EN_SPLIT_WORDS_HYPHEN": continue
+            if not (content_start_index <= match.offset < content_end_index): continue
+            lt_issues_in_range +=1
+            context_str = text_for_analysis[match.offset : match.offset + match.errorLength]
+            processed_issues.append({
+                '_internal_id': f"lt_{idx}", 'ruleId': match.ruleId, 'message': match.message,
+                'context_text': context_str, 'offset_in_text': match.offset, 'error_length': match.errorLength,
+                'replacements_suggestion': match.replacements[:3] if match.replacements else [],
+                'category_name': match.category, 'is_mapped_to_pdf': False,
+                'pdf_coordinates_list': [], 'mapped_page_number': -1
+            })
+        print(f"LanguageTool found {len(raw_lt_matches)} raw issues, {lt_issues_in_range} issues within defined content range.")
+        
+        regex_pattern = r'\b(\w+)\[(\d+)\]'
+        regex_matches = list(re.finditer(regex_pattern, text_for_analysis))
+        
+        regex_issues_in_range = 0
+        for reg_idx, match in enumerate(regex_matches):
+            if not (content_start_index <= match.start() < content_end_index): continue
+            regex_issues_in_range += 1
+            word = match.group(1); number = match.group(2)
+            processed_issues.append({
+                '_internal_id': f"regex_{reg_idx}", 'ruleId': "SPACE_BEFORE_BRACKET",
+                'message': f"Missing space before '[' in '{word}[{number}]'. Should be '{word} [{number}]'.",
+                'context_text': text_for_analysis[match.start():match.end()],
+                'offset_in_text': match.start(), 'error_length': match.end() - match.start(),
+                'replacements_suggestion': [f"{word} [{number}]"], 'category_name': "Formatting",
+                'is_mapped_to_pdf': False, 'pdf_coordinates_list': [], 'mapped_page_number': -1
+            })
+        print(f"Regex check found {len(regex_matches)} raw matches, {regex_issues_in_range} issues within defined content range.")
+        
+        return {
+            "total_issues": len(processed_issues), "issues_list": processed_issues,
+            "text_used_for_analysis": text_for_analysis 
+        }
+    except Exception as e:
+        print(f"Error in check_language_issues_and_regex: {e}")
+        traceback.print_exc()
+        return {"error": str(e), "total_issues": 0, "issues_list": [], "text_used_for_analysis": text_for_analysis}
+    finally:
+        if tool: tool.close()
+
+def check_figure_order(plain_text: str) -> Dict[str, Any]:
+    figure_pattern = r'(?:Fig(?:ure)?\.?|Figure)\s*(\d+)'
+    figure_references_str = re.findall(figure_pattern, plain_text, re.IGNORECASE)
+    
+    valid_figure_numbers_int = [int(num_str) for num_str in figure_references_str if num_str.isdigit()]
+    
+    unique_sorted_figures = sorted(list(set(valid_figure_numbers_int)))
+    is_sequential = all(unique_sorted_figures[i] + 1 == unique_sorted_figures[i+1] for i in range(len(unique_sorted_figures)-1))
+
+    missing_figures = []
+    if unique_sorted_figures:
+        expected_figures = set(range(1, max(unique_sorted_figures) + 1))
+        missing_figures = sorted(list(expected_figures - set(unique_sorted_figures)))
+    
+    counts = Counter(valid_figure_numbers_int)
+    duplicate_refs = [num for num, count in counts.items() if count > 1]
+    
+    return {
+        "sequential_order_of_unique_figures": is_sequential, 
+        "figure_count_unique": len(unique_sorted_figures),
+        "missing_figures_in_sequence_to_max": missing_figures, 
+        "figure_order_as_encountered": valid_figure_numbers_int,
+        "duplicate_references_to_same_figure_number": duplicate_refs
+    }
+
+def check_reference_order(plain_text: str) -> Dict[str, Any]:
+    reference_pattern = r'\[(\d+)\]' 
+    references_str = re.findall(reference_pattern, plain_text)
+    ref_numbers_int = [int(ref) for ref in references_str if ref.isdigit()]
+    
+    max_ref_val = 0
+    out_of_order_details = []
+    
+    if ref_numbers_int:
+        max_ref_val = max(ref_numbers_int)
+        current_max_seen_in_text = 0
+        for i, ref in enumerate(ref_numbers_int):
+            if ref < current_max_seen_in_text : 
+                 out_of_order_details.append({
+                     "position_in_text_occurrences": i + 1, "value": ref,
+                     "previous_max_value_seen": current_max_seen_in_text,
+                     "message": f"Reference [{ref}] appeared after a higher reference [{current_max_seen_in_text}] was already cited."
+                 })
+            current_max_seen_in_text = max(current_max_seen_in_text, ref)
+
+    all_expected_refs_up_to_max = set(range(1, max_ref_val + 1)) if max_ref_val > 0 else set()
+    used_refs_set = set(ref_numbers_int)
+    missing_refs_in_sequence_to_max = sorted(list(all_expected_refs_up_to_max - used_refs_set))
+    
+    is_ordered_in_text = all(ref_numbers_int[i] <= ref_numbers_int[i+1] for i in range(len(ref_numbers_int)-1))
+
+    return {
+        "max_reference_number_cited": max_ref_val,
+        "out_of_order_citations_details": out_of_order_details, 
+        "missing_references_up_to_max_cited": missing_refs_in_sequence_to_max,
+        "is_citation_order_non_decreasing_in_text": is_ordered_in_text
+    }