File size: 2,553 Bytes
39a78f4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
#!/usr/bin/env python3
import json, os, re, subprocess, threading, time
from concurrent.futures import ThreadPoolExecutor

ROOT = os.environ.get("UNSC_CORPUS_ROOT", ".")
OUT = f"{ROOT}/publication/data/fulltext.jsonl"
FAMILIES = ["resolutions", "presidential_statements", "verbatim_debates", "documents"]
_lock = threading.Lock()


def safe(s):
    return re.sub(r"[^A-Za-z0-9.]+", "_", s).strip("_")


def find_pdf(rec, fam_of):
    base = safe(rec.get("symbol") or rec["record_id"])
    for fam in fam_of:
        p = f"{ROOT}/{fam}/pdfs/{base}-EN.pdf"
        if os.path.exists(p) and os.path.getsize(p) > 1000:
            return p, fam
    return None, None


def extract(path):
    try:
        out = subprocess.run(["pdftotext", "-q", path, "-"], capture_output=True,
                             timeout=120).stdout.decode("utf-8", "ignore")
        return out
    except Exception:
        return ""


def main():
    merged, fam_of = {}, {}
    for fam in FAMILIES:
        p = f"{ROOT}/{fam}/metadata/records.jsonl"
        if not os.path.exists(p):
            continue
        for line in open(p):
            r = json.loads(line)
            rid = r.get("record_id")
            if not rid:
                continue
            fam_of.setdefault(rid, []).append(fam)
            merged.setdefault(rid, r)
    items = list(merged.items())
    print(f"{len(items)} unique records to process", flush=True)

    fh = open(OUT, "w")
    n = [0]; withtext = [0]; notext = [0]; nopdf = [0]

    def one(item):
        rid, r = item
        path, fam = find_pdf(r, fam_of[rid])
        if not path:
            with _lock:
                nopdf[0] += 1; n[0] += 1
            return
        txt = extract(path).strip()
        rec = {"record_id": rid, "symbol": r.get("symbol"), "family": fam,
               "date": r.get("date"), "n_chars": len(txt), "text": txt}
        with _lock:
            fh.write(json.dumps(rec, ensure_ascii=False) + "\n")
            n[0] += 1
            if len(txt) >= 50: withtext[0] += 1
            else: notext[0] += 1
            if n[0] % 5000 == 0:
                fh.flush()
                print(f"  {n[0]}/{len(items)} (text {withtext[0]}, "
                      f"empty {notext[0]}, no-pdf {nopdf[0]})", flush=True)

    with ThreadPoolExecutor(max_workers=10) as ex:
        list(ex.map(one, items))
    fh.close()
    print(f"DONE: {n[0]} records | with-text {withtext[0]} | "
          f"empty(image-only) {notext[0]} | no-pdf {nopdf[0]} -> {OUT}", flush=True)


if __name__ == "__main__":
    main()