import fitz, re, os, glob, sys
d = "/Users/liuxinyuan/Desktop/Hermes输出-工作类/资料/文献与专利/2025-2027/companies/raw"
for f in sorted(glob.glob(d + "/ar_*.pdf")):
    try:
        doc = fitz.open(f)
    except Exception as e:
        print("ERR", f, e); continue
    txt = "\n".join(p.get_text() for p in doc)
    out = f.replace(".pdf", ".txt")
    open(out, "w").write(txt)
    print("==", os.path.basename(f), len(doc), "pages", len(txt), "chars")
    for kw in ["分子诊断", "分行业", "营业收入构成"]:
        print("   ", kw, len(re.findall(kw, txt)))