#!/usr/bin/env python3
"""Extract translatable strings (text-domain gdpr-content-blocker) from the plugin PHP."""
import re, sys, os, glob

FUNCS = r"(?:esc_html__|esc_html_e|esc_attr__|esc_attr_e|esc_attr_e|__|_e|esc_attr_e)"
# capture first single-quoted arg
PAT = re.compile(FUNCS + r"\s*\(\s*'((?:[^'\\]|\\.)*)'\s*,\s*'gdpr-content-blocker'")

def unescape(s):
    return s.replace("\\'", "'").replace('\\\\', '\\')

base = sys.argv[1]
seen = []
seenset = set()
for path in glob.glob(os.path.join(base, '**', '*.php'), recursive=True):
    with open(path, encoding='utf-8') as f:
        txt = f.read()
    for m in PAT.finditer(txt):
        s = unescape(m.group(1))
        if s not in seenset:
            seenset.add(s)
            seen.append(s)

out = sys.argv[2] if len(sys.argv) > 2 else 'strings.txt'
with open(out, 'w', encoding='utf-8') as f:
    f.write(f"# {len(seen)} unique strings\n")
    for s in seen:
        f.write("MSGID\t" + s + "\n")
print(f"{len(seen)} strings written to {out}")