#!/usr/bin/env python3
"""Probe PIT/PM drug image key assumptions.

This script is intentionally read-only. It does not require production credentials in the repo;
wire `query_rows()` to pyodbc/pymssql in your environment.
"""
from __future__ import annotations
import base64
import hashlib
import json
import re
import urllib.request
import urllib.error
from dataclasses import dataclass


def classify_key(value: str) -> dict:
    value = value or ''
    out = {'value': value, 'length': len(value), 'kind': 'unknown'}
    if re.fullmatch(r'\d{13}', value):
        out['kind'] = 'plain_13_digit_dik_code'
        return out
    try:
        raw = base64.b64decode(value, validate=True)
        out['base64_bytes'] = len(raw)
        out['base64_hex_prefix'] = raw[:8].hex()
        if len(value) == 44 and len(raw) == 32:
            out['kind'] = 'base64_32byte_digest_like_key'
        else:
            out['kind'] = 'base64_other'
    except Exception:
        out['kind'] = 'non_base64_text'
    return out


def sha256_b64_candidates(plain: str) -> dict:
    return {
        enc: base64.b64encode(hashlib.sha256(plain.encode(enc)).digest()).decode()
        for enc in ['utf-8', 'cp949', 'utf-16le']
    }


def probe_url(url: str, referer: str | None = None, timeout: int = 10) -> dict:
    headers = {'User-Agent': 'Mozilla/5.0'}
    if referer:
        headers['Referer'] = referer
    try:
        req = urllib.request.Request(url, headers=headers)
        with urllib.request.urlopen(req, timeout=timeout) as r:
            head = r.read(32)
            return {
                'url': url,
                'ok': True,
                'status': r.status,
                'content_type': r.headers.get('content-type'),
                'content_length': r.headers.get('content-length'),
                'head_hex': head.hex(),
                'is_jpeg': head.startswith(b'\xff\xd8\xff'),
            }
    except Exception as e:
        return {'url': url, 'ok': False, 'error': f'{type(e).__name__}: {e}'}


def extract_zoom_image_candidates(dik_code: str) -> dict:
    url = f'https://www.health.kr/drug_info/sb/zoom.asp?drug_code={dik_code}'
    result = {'zoom_url': url, 'ok': False, 'candidates': []}
    try:
        req = urllib.request.Request(url, headers={'User-Agent': 'Mozilla/5.0'})
        with urllib.request.urlopen(req, timeout=15) as r:
            data = r.read(300000)
        # health.kr pages are usually euc-kr/cp949-ish.
        text = data.decode('euc-kr', errors='replace')
        result['ok'] = True
        imgs = re.findall(r'''(?:src|myImg\.src)\s*=\s*['\"]([^'\"]+)''', text, flags=re.I)
        paths = re.findall(r'''https?://[^'\"<>\s]+(?:sb_photo|big3|marked3)[^'\"<>\s]+''', text, flags=re.I)
        result['candidates'] = sorted(set(imgs + paths))
    except Exception as e:
        result['error'] = f'{type(e).__name__}: {e}'
    return result


def demo():
    # Replace these with real rows from PM_DRUG..CD_GOODS and PM_IMAGE..DrugImg.
    plain_dik = '2015012700008'
    image_key = 'xiY4jUZTAZB0p8JpHEa+Zt9DEcOVQFq9sgtAxSG7GXY='
    print(json.dumps({
        'plain_dik': classify_key(plain_dik),
        'image_key': classify_key(image_key),
        'sha256_plain_candidates': sha256_b64_candidates(plain_dik),
        'direct_big3': probe_url(f'http://pharm.or.kr/images/sb_photo/big3/{plain_dik}.jpg'),
        'zoom_candidates': extract_zoom_image_candidates(plain_dik),
    }, ensure_ascii=False, indent=2))


if __name__ == '__main__':
    demo()