"""Why did two PDFs score ~0? Count unreadable (cid:) words and italic-only heading fonts."""
import sys, pdfplumber
d = sys.argv[1]
for pid in ["2210.03629", "2312.10997"]:
    with pdfplumber.open(f"{d}/{pid}.pdf") as pdf:
        ws = [w for p in pdf.pages for w in p.extract_words(extra_attrs=["fontname"])]
    cid = sum("(cid:" in w["text"] for w in ws)
    ital = sum("ital" in w["fontname"].lower() for w in ws)
    print(f"{pid}: words={len(ws)} cid_words={cid} ({cid/len(ws):.1%}) italic_words={ital} ({ital/len(ws):.1%})")