PDF Reader MCP Server

""" Academic section detection for research papers """ import re from typing import Dict, List, Any, Optional from ..core.pdf_processor import PDFProcessor class SectionDetector: """Detects and extracts academic paper sections""" # Common academic section patterns SECTION_PATTERNS = { "abstract": [ r"^ABSTRACT\s*$", r"^Abstract\s*$", r"^\d+\.\s*ABSTRACT", r"^\d+\.\s*Abstract" ], "introduction": [ r"^INTRODUCTION\s*$", r"^Introduction\s*$", r"^\d+\.\s*INTRODUCTION", r"^\d+\.\s*Introduction", r"^1\.\s*Introduction" ], "methods": [ r"^METHODS?\s*$", r"^Methods?\s*$", r"^METHODOLOGY\s*$", r"^Methodology\s*$", r"^\d+\.\s*METHODS?", r"^\d+\.\s*Methods?", r"^\d+\.\s*METHODOLOGY", r"^\d+\.\s*Methodology" ], "results": [ r"^RESULTS?\s*$", r"^Results?\s*$", r"^FINDINGS\s*$", r"^Findings\s*$", r"^\d+\.\s*RESULTS?", r"^\d+\.\s*Results?", r"^\d+\.\s*FINDINGS", r"^\d+\.\s*Findings" ], "discussion": [ r"^DISCUSSION\s*$", r"^Discussion\s*$", r"^\d+\.\s*DISCUSSION", r"^\d+\.\s*Discussion" ], "conclusion": [ r"^CONCLUSION\s*$", r"^Conclusion\s*$", r"^CONCLUSIONS\s*$", r"^Conclusions\s*$", r"^\d+\.\s*CONCLUSION", r"^\d+\.\s*Conclusion", r"^\d+\.\s*CONCLUSIONS", r"^\d+\.\s*Conclusions" ], "references": [ r"^REFERENCES\s*$", r"^References\s*$", r"^BIBLIOGRAPHY\s*$", r"^Bibliography\s*$", r"^\d+\.\s*REFERENCES", r"^\d+\.\s*References" ] } @staticmethod async def detect_sections(pdf_path: str) -> Dict[str, Any]: """Detect academic sections in the PDF""" text = await PDFProcessor.extract_raw_text(pdf_path) lines = text.split('\n') sections = {} current_section = None section_content = [] for i, line in enumerate(lines): line = line.strip() if not line: continue # Check if line matches any section pattern detected_section = SectionDetector._match_section_pattern(line) if detected_section: # Save previous section if current_section and section_content: sections[current_section] = { "content": '\n'.join(section_content).strip(), "line_start": i - len(section_content), "line_end": i - 1, "word_count": len(' '.join(section_content).split()) } # Start new section current_section = detected_section section_content = [] else: # Add to current section if current_section: section_content.append(line) # Save last section if current_section and section_content: sections[current_section] = { "content": '\n'.join(section_content).strip(), "line_start": len(lines) - len(section_content), "line_end": len(lines) - 1, "word_count": len(' '.join(section_content).split()) } return { "sections": sections, "sections_found": list(sections.keys()), "total_sections": len(sections) } @staticmethod def _match_section_pattern(line: str) -> Optional[str]: """Check if line matches any section pattern""" for section_name, patterns in SectionDetector.SECTION_PATTERNS.items(): for pattern in patterns: if re.match(pattern, line, re.IGNORECASE): return section_name return None @staticmethod async def extract_abstract(pdf_path: str) -> Dict[str, Any]: """Extract abstract specifically""" sections = await SectionDetector.detect_sections(pdf_path) if "abstract" in sections["sections"]: abstract_data = sections["sections"]["abstract"] return { "abstract": abstract_data["content"], "word_count": abstract_data["word_count"], "found": True } # Fallback: look for abstract in first few paragraphs text = await PDFProcessor.extract_raw_text(pdf_path) paragraphs = text.split('\n\n')[:5] # First 5 paragraphs for para in paragraphs: if len(para.split()) > 50 and len(para.split()) < 300: # Abstract length if any(word in para.lower() for word in ['study', 'research', 'analysis', 'investigation']): return { "abstract": para.strip(), "word_count": len(para.split()), "found": True, "method": "heuristic" } return {"abstract": "", "found": False} @staticmethod async def extract_key_sections(pdf_path: str) -> Dict[str, str]: """Extract key academic sections for agent understanding""" sections_data = await SectionDetector.detect_sections(pdf_path) sections = sections_data["sections"] key_sections = {} # Priority sections for agents priority_sections = ["abstract", "introduction", "methods", "results", "conclusion"] for section in priority_sections: if section in sections: content = sections[section]["content"] # Truncate very long sections for agent consumption if len(content.split()) > 500: words = content.split()[:500] content = ' '.join(words) + "... [truncated]" key_sections[section] = content return key_sections @staticmethod async def get_section_summary(pdf_path: str) -> Dict[str, Any]: """Get a summary of detected sections""" sections_data = await SectionDetector.detect_sections(pdf_path) sections = sections_data["sections"] summary = { "has_abstract": "abstract" in sections, "has_introduction": "introduction" in sections, "has_methods": "methods" in sections, "has_results": "results" in sections, "has_discussion": "discussion" in sections, "has_conclusion": "conclusion" in sections, "has_references": "references" in sections, "total_sections": len(sections), "estimated_structure": "academic_paper" if len(sections) >= 4 else "other_document" } # Calculate section word counts section_stats = {} for section_name, section_data in sections.items(): section_stats[section_name] = { "word_count": section_data["word_count"], "percentage": round(section_data["word_count"] / sum(s["word_count"] for s in sections.values()) * 100, 1) } summary["section_statistics"] = section_stats return summary