#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""OCR teks caption video TS Law (4K) pakai easyocr — untuk memastikan isi/topik video."""
import glob
import os
import sys

import easyocr

DIR = os.environ.get("FRAMEDIR", "frames1s")
reader = easyocr.Reader(["en"], gpu=False, verbose=False)
for f in sorted(glob.glob(os.path.join(DIR, "*.jpg"))):
    res = reader.readtext(f, detail=1, paragraph=False)
    teks = " | ".join(t for _, t, c in res if c >= 0.3)
    print(f"{os.path.basename(f)}: {teks}")