import json import re def clean_val(v): if not v: return "" v = v.strip() # remove backticks if v.startswith("`") and v.endswith("`"): v = v[1:-1].strip() # remove bold if v.startswith("**") and v.endswith("**"): v = v[2:-2].strip() # remove quotes if (v.startswith('"') and v.endswith('"')) or (v.startswith("'") and v.endswith("'")): v = v[1:-1].strip() return v def is_req(req_str): req_str = req_str.lower() return "req" in req_str and "opt" not in req_str def parse_md(): with open("web/docs/FASTAPI_BACKEND_API_DOCUMENTATION.md", "r") as f: md_text = f.read() sections = re.split(r"\n(?=###\s+)", md_text) md_items = [] for s in sections[1:]: lines = s.strip().split("\n") header = lines[0].replace("###", "").strip() url_m = re.search(r"\*\*URL:\*\*\s*`([^`]+)`", s) url = url_m.group(1).strip() if url_m else "" subs_m = re.search(r"\*\(Sub-endpoints:\s*([^\)]+)\)\*", s) subs = subs_m.group(1).strip() if subs_m else "" meth_m = re.search(r"\*\*Method:\*\*\s*`?([A-Za-z0-9_\/ ]+)`?", s) method = meth_m.group(1).strip() if meth_m else "" # parse table table_lines = [] in_table = False for l in lines: ls = l.strip() if ls.startswith("|") and ls.endswith("|"): in_table = True table_lines.append(ls) elif in_table: break params = [] if len(table_lines) >= 3: raw_headers = [c.strip() for c in table_lines[0].strip("|").split("|")] for l in table_lines[2:]: cols = [c.strip() for c in l.strip("|").split("|")] if len(cols) == len(raw_headers): row = dict(zip(raw_headers, cols)) field = clean_val(row.get("Field", "")) ty = clean_val(row.get("Type", "")) req_raw = row.get("Required", "") req = is_req(req_raw) df = clean_val(row.get("Default", "")) ds = row.get("Description", "") params.append({ "field": field, "type": ty, "required": req, "required_raw": req_raw, "default": df, "description": ds }) md_items.append({ "header": header, "url": url, "subs": subs, "method": method, "params": params, "raw_text": s }) return md_items def parse_nexus(): with open("web/docs/data.json", "r") as f: nexus_data = json.load(f) nexus_apis = [] for s_idx, sec in enumerate(nexus_data["sections"]): for api in sec.get("eps", []): nexus_apis.append({ "sec_title": sec.get("title"), "api_id": api.get("id"), "api_key": api.get("api_key"), "title": api.get("title"), "api_ip_link": api.get("api_ip_link"), "full_url": api.get("full_url"), "path": api.get("path"), "method": api.get("method"), "params": api.get("params", []), "endpoints": api.get("endpoints", []) }) return nexus_apis if __name__ == "__main__": md_items = parse_md() nexus_apis = parse_nexus() print(f"MD sections: {len(md_items)}, Nexus APIs: {len(nexus_apis)}") import re def run_comparison(): md_items = parse_md() nexus_apis = parse_nexus() # Map by direct full_url, endpoint full_url, api_ip_link, or path url_to_nexus = {} path_to_nexus = {} for a in nexus_apis: if a.get("full_url"): url_to_nexus[a["full_url"].rstrip("/")] = a if a.get("api_ip_link"): url_to_nexus[a["api_ip_link"].rstrip("/")] = a if a.get("path"): path_to_nexus[a["path"].rstrip("/")] = a for se in a.get("endpoints", []): if se.get("full_url"): url_to_nexus[se["full_url"].rstrip("/")] = a if se.get("path"): path_to_nexus[se["path"].rstrip("/")] = a results = [] for idx, md in enumerate(md_items): m_url = md["url"].rstrip("/") matched = None # 1. Direct url match if m_url in url_to_nexus: matched = url_to_nexus[m_url] else: # 2. Check if m_url ends with any known path for p, a in path_to_nexus.items(): if m_url.endswith(p): matched = a break # 3. Path suffix match or partial url if not matched: for u, a in url_to_nexus.items(): if u.endswith(m_url) or m_url.endswith(u): matched = a break # 4. Check special cases if not matched: # check header keyword match hdr_lower = md["header"].lower() for a in nexus_apis: key_lower = a.get("api_key", "").lower() title_lower = a.get("title", "").lower() if "health" in hdr_lower and "health" in title_lower: matched = a break elif "trendy" in hdr_lower and "trendy" in key_lower: matched = a break elif "hair_segmentation_yolo_sam" in m_url and "hair_segmentation" in key_lower: matched = a break elif "caricature_effects" in m_url and "caricature_effects" in key_lower: matched = a break results.append((md, matched)) return results if __name__ == "__main__": results = run_comparison() matched_cnt = sum(1 for m, n in results if n is not None) print(f"Total MD items: {len(results)}, Matched: {matched_cnt}") for idx, (m, n) in enumerate(results): t_str = n["title"] if n else "NO MATCH" print(f"[{idx+1:2d}] {m['header'][:40]:40s} -> {t_str}")