auto_translate_razor.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import os
import re
import hashlib
import time
import logging
import requests
import xml.etree.ElementTree as ET
from collections import OrderedDict
import argparse

# === Конфигурация ===
# Директивы для .cshtml
USING_DIRECTIVE  = "@using Microsoft.Extensions.Localization"
INJECT_DIRECTIVE = "@inject IStringLocalizer<CPG_CaptchaWebPentestGame.SharedResource> Localizer"

# Пути
ROOT_DIR      = "."  # корневой каталог проекта
PROJECT_DIR   = os.path.join(ROOT_DIR, "CaptchaWebPentestGame")
RESOURCES_DIR = os.path.join(PROJECT_DIR, "Resources")
BASE_NAME     = "SharedResource"
DEFAULT_RESX  = os.path.join(RESOURCES_DIR, f"{BASE_NAME}.resx")
RU_RESX       = os.path.join(RESOURCES_DIR, f"{BASE_NAME}.ru.resx")

# Языки перевода
SOURCE_LANG = "EN"
TARGET_LANG = "RU"

# DeepL API
DEEPL_URL = os.getenv("DEEPL_URL", "https://api-free.deepl.com/v2/translate")
DEEPL_AUTH_KEY = os.getenv("DEEPL_AUTH_KEY")
if not DEEPL_AUTH_KEY:
    print("ERROR: Не задан ключ DeepL. Установите переменную окружения DEEPL_AUTH_KEY.")
    exit(1)

# === Шаблоны для извлечения ===
TITLE_PATTERN       = re.compile(r'ViewData\["Title"\]\s*=\s*"([^"]+)"')
HTML_TEXT_PATTERN   = re.compile(r'>([\s\S]*?)(?=<)', re.MULTILINE)
PLACEHOLDER_PATTERN = re.compile(r'placeholder="([^"]+)"')

# === Утилита для чтения файла с детектом кодировки ===
def read_file(path):
    """
    Пытается прочитать файл в нескольких кодировках (utf-8, cp1251, latin1).
    При неудаче возвращает содержимое с заменой некорректных байт.
    """
    for enc in ('utf-8', 'cp1251', 'latin1'):
        try:
            with open(path, 'r', encoding=enc) as f:
                return f.read()
        except UnicodeDecodeError:
            continue
    with open(path, 'r', encoding='utf-8', errors='replace') as f:
        return f.read()

# === Функции для DeepL ===
def deepl_translate(texts, source, target, auth_key, batch_size=20, pause=1.0):
    headers = {
        "Content-Type": "application/json",
        "Authorization": f"DeepL-Auth-Key {auth_key}"
    }
    translated = []
    for i in range(0, len(texts), batch_size):
        batch = texts[i:i+batch_size]
        payload = {
            "text": batch,
            "source_lang": source.upper(),
            "target_lang": target.upper(),
            "preserve_formatting": True
        }
        resp = requests.post(DEEPL_URL, headers=headers, json=payload)
        resp.raise_for_status()
        data = resp.json()
        translated.extend([t['text'] for t in data["translations"]])
        time.sleep(pause)
    return translated

# === Сбор строк ===
def extract_strings_from_file(path):
    if not path.endswith('.cshtml'):
        return []
    content = read_file(path)
    texts = []
    # Фильтр: исключаем кодовые конструкции и системные символы
    def should_translate(txt):
        return (txt and
                not txt.startswith('@') and
                not (txt.startswith('if ') or txt.startswith('if(')) and
                '@' not in txt and 'times' not in txt and
                '{' not in txt and '}' not in txt and
                'case ' not in txt and '=>' not in txt and
                'Localizer' not in txt and ';' not in txt and
                'Model.' not in txt)

    # Заголовки страницы
    for m in TITLE_PATTERN.finditer(content):
        txt = m.group(1).strip()
        if should_translate(txt):
            texts.append(txt)
    # HTML-текст между тегами
    for m in HTML_TEXT_PATTERN.finditer(content):
        txt = m.group(1).strip()
        if should_translate(txt):
            texts.append(txt)
    # Placeholder атрибуты
    for m in PLACEHOLDER_PATTERN.finditer(content):
        txt = m.group(1).strip()
        if should_translate(txt):
            texts.append(txt)
    return texts

# === Утилиты ===
def make_key(text):
    h = hashlib.md5(text.encode('utf-8')).hexdigest()[:10]
    prefix = re.sub(r'[^0-9a-zA-Z]+', '_', text).strip('_')[:20]
    return f"{prefix}_{h}"

# Работа с .resx
def load_resx(path):
    mapping = {}
    if os.path.exists(path):
        tree = ET.parse(path)
        for data in tree.findall('.//data'):
            key = data.get('name')
            val_elem = data.find('value')
            if key and val_elem is not None:
                mapping[key] = val_elem.text or ''
    return mapping

def write_resx_file(path, mapping):
    root = ET.Element('root')
    headers = [
        ('resmimetype','text/microsoft-resx'),
        ('version','2.0'),
        ('reader','System.Resources.ResXResourceReader, System.Windows.Forms'),
        ('writer','System.Resources.ResXResourceWriter, System.Windows.Forms')
    ]
    for name, val in headers:
        ET.SubElement(root, 'resheader', name=name).text = val
    for key, val in mapping.items():
        data = ET.SubElement(root, 'data', name=key)
        ET.SubElement(data, 'value').text = val
    os.makedirs(os.path.dirname(path), exist_ok=True)
    ET.ElementTree(root).write(path, encoding='utf-8', xml_declaration=True)
    logging.info(f"Written resources to {path}")

# === Замена в .cshtml ===
def replace_in_file(path, string_to_key):
    content = read_file(path)
    modified = content
    # Добавляем директивы, если отсутствуют
    if USING_DIRECTIVE not in content:
        modified = USING_DIRECTIVE + "\n" + modified
    if INJECT_DIRECTIVE not in content:
        modified = modified.replace(USING_DIRECTIVE, USING_DIRECTIVE + "\n" + INJECT_DIRECTIVE)
    # Замена заголовка
    def repl_title(m):
        txt = m.group(1).strip()
        key = string_to_key.get(txt)
        return f'ViewData["Title"] = Localizer["{key}"]' if key else m.group(0)
    modified = TITLE_PATTERN.sub(repl_title, modified)
    # Замена HTML-текста
    def repl_html(m):
        original = m.group(1).strip()
        key = string_to_key.get(original)
        if key:
            pre = m.group(0)[:m.start(1)-m.start(0)]
            post = m.group(0)[m.end(1)-m.start(0):]
            return f'{pre}@Localizer["{key}"]{post}'
        return m.group(0)
    modified = HTML_TEXT_PATTERN.sub(repl_html, modified)
    # Замена placeholder
    def repl_placeholder(m):
        original = m.group(1).strip()
        key = string_to_key.get(original)
        if key:
            return f'placeholder="@Localizer["{key}"]"'
        return m.group(0)
    modified = PLACEHOLDER_PATTERN.sub(repl_placeholder, modified)
    if modified != content:
        with open(path, 'w', encoding='utf-8') as fw:
            fw.write(modified)
        logging.info(f"Updated file: {path}")

# === Основной процесс ===
def main():
    parser = argparse.ArgumentParser(description='Translate .cshtml files with DeepL and update resources.')
    parser.add_argument('-f', '--file', help='Path to a single .cshtml file to translate')
    args = parser.parse_args()
    logging.basicConfig(level=logging.INFO, format='%(asctime)s %(levelname)s: %(message)s')
    default_map = load_resx(DEFAULT_RESX)
    ru_map      = load_resx(RU_RESX)
    text_to_key = {v: k for k, v in default_map.items()}
    files_to_process = []
    if args.file:
        if os.path.exists(args.file) and args.file.endswith('.cshtml'):
            files_to_process = [args.file]
        else:
            logging.error(f"Файл не найден или не .cshtml: {args.file}")
            return
    else:
        for root, _, files in os.walk(ROOT_DIR):
            for fname in files:
                if fname.endswith('.cshtml'):
                    files_to_process.append(os.path.join(root, fname))
    all_strings = []
    for path in files_to_process:
        all_strings.extend(extract_strings_from_file(path))
    unique_strings = list(OrderedDict.fromkeys(all_strings))
    logging.info(f"Found {len(unique_strings)} unique strings to translate.")
    if not unique_strings:
        logging.warning("No strings found.")
        return
    to_translate = [s for s in unique_strings if s not in text_to_key]
    logging.info(f"{len(to_translate)} new strings to translate.")
    translations = deepl_translate(to_translate, SOURCE_LANG, TARGET_LANG, DEEPL_AUTH_KEY) if to_translate else []
    for orig in unique_strings:
        if orig in text_to_key:
            key = text_to_key[orig]
        else:
            key = make_key(orig)
            default_map[key] = orig
            text_to_key[orig] = key
            idx = to_translate.index(orig)
            ru_map[key] = translations[idx]
    write_resx_file(DEFAULT_RESX, default_map)
    write_resx_file(RU_RESX, ru_map)
    for path in files_to_process:
        replace_in_file(path, text_to_key)
    logging.info("All done.")

if __name__ == "__main__":
    main()