""" Report Generation Module URL mapping helpers, markdown report generation, and PDF conversion for preclinical synthesis reports. """ import os import re from datetime import datetime from typing import List, Dict # --------------------------------------------------------------------------- # URL mapping helpers for hyperlinked reports # --------------------------------------------------------------------------- def build_url_maps(results: List[Dict]): """Build PMID->URL and DOI->URL lookup maps from search results.""" pmid_url = {} doi_url = {} for r in results: pmid = str(r.get("pmid", "")).strip() doi = str(r.get("doi", "")).strip() url = str(r.get("url", "")).strip() if pmid and url: pmid_url[pmid] = url if doi and url: doi_url[doi] = url return pmid_url, doi_url def hyperlink_pmid(pmid: str, pmid_url_map: dict) -> str: """Return markdown hyperlink for a PMID.""" pmid = str(pmid).strip() if not pmid: return "" url = pmid_url_map.get(pmid, "") if not url and pmid.isdigit(): url = f"https://pubmed.ncbi.nlm.nih.gov/{pmid}/" if url: return f"[PMID: {pmid}]({url})" return f"PMID: {pmid}" def hyperlink_pmids_in_text(text: str, pmid_url_map: dict) -> str: """Replace bare [PMID] patterns in text with hyperlinked markdown.""" def _replace_match(match): pmid = match.group(1) url = pmid_url_map.get(pmid, "") if not url and pmid.isdigit(): url = f"https://pubmed.ncbi.nlm.nih.gov/{pmid}/" if url: return f"[PMID: {pmid}]({url})" return match.group(0) return re.sub(r'\[(\d{5,})\]', _replace_match, text) # --------------------------------------------------------------------------- # Markdown report generation # --------------------------------------------------------------------------- def generate_markdown_report( results: List[Dict], experiments: List[Dict], synthesis: Dict, target: str, disease: str, output_file: str ) -> None: """Generate structured markdown synthesis report.""" n_papers = synthesis["total_papers"] type_bd = synthesis["experiment_type_breakdown"] # Year range years = [e.get("publication_date", "")[:4] for e in experiments] years = [y for y in years if y.isdigit()] year_range = f"{min(years)}-{max(years)}" if years else "N/A" md = f"""# Preclinical Literature Extraction Report | | | |---|---| | **Target** | {target} | | **Disease** | {disease} | | **Date Generated** | {datetime.now().strftime('%Y-%m-%d')} | | **Total Papers** | {n_papers} | | **Date Range** | {year_range} | --- ## Summary This report extracts and synthesizes preclinical experiment details from {n_papers} papers studying **{target}** in **{disease}**. ### Experiment Type Breakdown | Type | Count | Percentage | |------|-------|------------| | In vitro only | {type_bd.get('in_vitro', 0)} | {type_bd.get('in_vitro', 0)/max(n_papers,1)*100:.1f}% | | In vivo only | {type_bd.get('in_vivo', 0)} | {type_bd.get('in_vivo', 0)/max(n_papers,1)*100:.1f}% | | Both | {type_bd.get('both', 0)} | {type_bd.get('both', 0)/max(n_papers,1)*100:.1f}% | | Unclassified | {type_bd.get('unclassified', 0)} | {type_bd.get('unclassified', 0)/max(n_papers,1)*100:.1f}% | """ # Build URL maps for hyperlinking papers pmid_url_map, doi_url_map = build_url_maps(results) # Insert narrative synthesis section if available narrative = synthesis.get("narrative", {}) narrative_md = narrative.get("narrative_markdown", "") if narrative_md: # Hyperlink bare PMID references in narrative text narrative_md = hyperlink_pmids_in_text(narrative_md, pmid_url_map) md += "\n---\n\n" md += narrative_md md += """--- ## In Vitro Experiments ### Most Common Cell Lines """ cell_lines = synthesis.get("cell_line_frequency", {}) if cell_lines: md += "| Rank | Cell Line | Papers |\n|------|-----------|--------|\n" for i, (cl, count) in enumerate(list(cell_lines.items())[:15], 1): md += f"| {i} | {cl} | {count} |\n" else: md += "_No cell lines detected._\n" md += "\n### Most Common Assay Types\n\n" assays = synthesis.get("assay_frequency", {}) if assays: md += "| Rank | Assay Type | Papers |\n|------|-----------|--------|\n" for i, (assay, count) in enumerate(list(assays.items())[:15], 1): md += f"| {i} | {assay} | {count} |\n" else: md += "_No assays detected._\n" md += "\n---\n\n## In Vivo Experiments\n\n### Animal Model Types\n\n" models = synthesis.get("animal_model_frequency", {}) if models: md += "| Rank | Model Type | Papers |\n|------|-----------|--------|\n" for i, (model, count) in enumerate(list(models.items())[:10], 1): md += f"| {i} | {model} | {count} |\n" else: md += "_No animal models detected._\n" md += "\n### Endpoints Measured\n\n" endpoints = synthesis.get("endpoint_frequency", {}) if endpoints: md += "| Rank | Endpoint | Papers |\n|------|----------|--------|\n" for i, (ep, count) in enumerate(list(endpoints.items())[:10], 1): md += f"| {i} | {ep} | {count} |\n" else: md += "_No endpoints detected._\n" md += "\n---\n\n## Papers with Both In Vitro and In Vivo Data\n\n" both_papers = [e for e in experiments if e["experiment_type"] == "both"] if both_papers: n_both = len(both_papers) verb = "reports" if n_both == 1 else "report" noun = "paper" if n_both == 1 else "papers" md += f"{n_both} {noun} {verb} both in vitro and in vivo experiments:\n\n" for i, e in enumerate(both_papers[:20], 1): pmid = str(e.get('pmid', '')).strip() paper_url = pmid_url_map.get(pmid, '') if not paper_url: doi = str(e.get('doi', '')).strip() paper_url = doi_url_map.get(doi, '') if paper_url: md += f"{i}. **[{e['title']}]({paper_url})**\n" else: md += f"{i}. **{e['title']}**\n" author_line = f" - {e['authors'][:60]}... ({e['publication_date'][:4]})" if pmid: author_line += f" \u2014 {hyperlink_pmid(pmid, pmid_url_map)}" md += f"{author_line}\n" if e["cell_lines"]: md += f" - Cell lines: {e['cell_lines']}\n" if e["assays"]: md += f" - Assays: {e['assays']}\n" if e["animal_models"]: md += f" - Models: {e['animal_models']}\n" if e["endpoints"]: md += f" - Endpoints: {e['endpoints']}\n" md += "\n" else: md += "_No papers with both experiment types found._\n" md += "\n---\n\n## Publication Timeline\n\n" year_dist = synthesis.get("year_distribution", {}) if year_dist: md += "| Year | Papers |\n|------|--------|\n" for year, count in sorted(year_dist.items()): md += f"| {year} | {count} |\n" md += "\n---\n\n## Per-Paper Extraction Details\n\n" md += "See `experiment_extraction.csv` for full per-paper details.\n\n" md += "---\n\n" md += "## Full-Text Insights\n\n" md += "_Pending: Read full text for top papers and replace this section " md += "with per-paper findings. See references/full-text-enrichment-guide.md._\n\n" # References section with all papers hyperlinked md += "---\n\n## References\n\n" for i, r in enumerate(results, 1): pmid = str(r.get('pmid', '')).strip() doi = str(r.get('doi', '')).strip() url = str(r.get('url', '')).strip() title = r.get('title', 'Untitled') authors = r.get('authors', 'Unknown') journal = r.get('journal', '') ref = f"{i}. {authors}. " if url: ref += f"[{title}]({url})" else: ref += title if journal: ref += f". *{journal}*" ref += "." identifiers = [] if pmid: identifiers.append(hyperlink_pmid(pmid, pmid_url_map)) if doi: identifiers.append(f"[DOI: {doi}](https://doi.org/{doi})") if identifiers: ref += " " + " | ".join(identifiers) md += f"{ref}\n\n" md += "---\n\n" md += "_Report generated by literature-preclinical skill._\n" with open(output_file, "w", encoding="utf-8") as f: f.write(md) # --------------------------------------------------------------------------- # PDF generation # --------------------------------------------------------------------------- def generate_pdf_report(md_path: str, pdf_path: str) -> bool: """Convert markdown report to PDF. Tries weasyprint first (best quality), falls back to pandoc. Returns True if PDF was generated successfully. """ # Strategy 1: markdown + weasyprint (pip install markdown weasyprint) try: import markdown as md_lib from weasyprint import HTML with open(md_path, "r", encoding="utf-8") as f: md_content = f.read() html_body = md_lib.markdown( md_content, extensions=["tables", "fenced_code"], ) # Make all links open in a new window/tab html_body = re.sub(r'\n\n\n\n\n" f"{html_body}\n" "\n" ) HTML(string=styled_html).write_pdf(pdf_path) return True except ImportError: pass except Exception as e: print(f" weasyprint error: {e}") # Strategy 2: pandoc (system tool) try: import shutil import subprocess if shutil.which("pandoc"): # Use xelatex for Unicode support (alpha, beta, etc. in abstracts) pdf_engine = "xelatex" if shutil.which("xelatex") else "pdflatex" result = subprocess.run( ["pandoc", md_path, "-o", pdf_path, f"--pdf-engine={pdf_engine}", "-V", "geometry:margin=2.5cm", "-V", "colorlinks=true", "-V", "linkcolor=blue"], capture_output=True, text=True, timeout=60, ) if result.returncode == 0: return True print(f" pandoc error: {result.stderr[:200]}") except Exception as e: print(f" pandoc error: {e}") return False