import pandas as pd
import os
import re
# 需要处理的原始 csv 名字（不带后缀）
csv_list = [
    'dialogs', 'factions', 'game_menus', 'game_strings', 'info_pages',
    'item_kinds', 'item_modifiers', 'parties', 'party_templates', 'quests',
    'quick_strings', 'skills', 'skins', 'troops', 'ui'
]

def space_for_game(text: str) -> str:
    text = text.replace(' ', '')
    tokens = re.findall(r'[A-Za-z0-9{}_%$@]+|.', text)
    return ' '.join(tokens)

def process_csv(path: str):
    """
    从 ./path.csv 中提取文本，按行分割：
    left|right  ->  id = left, text = right
    并把 text 列保存到 extract/path.csv，供翻译使用。
    """
    csv_path = f'./{path}.csv'
    rows = []

    # 新建 extract 文件夹
    os.makedirs('extract', exist_ok=True)

    # 读取原始 csv，每行只按第一个 | 分割
    with open(csv_path, 'r', encoding='utf-8', errors='ignore') as f:
        for line in f:
            line = line.rstrip('\n')
            left, right = (line.split('|', 1) + [''])[:2]
            rows.append([left, right])

    df = pd.DataFrame(rows, columns=['id', 'text'])

    # 如果你想去掉 text 中所有空格，可以打开下面这一行
    # df['text'] = df['text'].str.replace(' ', '')

    out_path = f'extract/{os.path.basename(path)}.csv'
    df['text'].to_csv(out_path, index=False, header=False, encoding='utf-8')

    print(f'[提取完成] {csv_path} -> {out_path}')


def combine_csv():
    # 新建 combine 文件夹
    os.makedirs('combine', exist_ok=True)

    for name in csv_list:
        original_path = f'./{name}.csv'
        extract_path = f'./extract/{name}.csv'
        out_path = f'./combine/{name}.csv'

        if not os.path.exists(original_path):
            print(f'[跳过] 找不到原始文件：{original_path}')
            continue

        if not os.path.exists(extract_path):
            print(f'[跳过] 找不到翻译文件：{extract_path}')
            continue

        # 1) 读取原始文件，拿到所有 id（左边部分）
        ids = []
        with open(original_path, 'r', encoding='utf-8', errors='ignore') as f:
            for line in f:
                line = line.rstrip('\n')
                left, _ = (line.split('|', 1) + [''])[:2]
                ids.append(left)

        # 2) 读取翻译后的 text（每行一条）
        #    这里假设你只改了内容，没有增加/减少行
        df_trans = pd.read_csv(extract_path, header=None, encoding='utf-8')
        # 防止多列情况，这里取最后一列
        translated_texts = df_trans.iloc[:, -1].astype(str).tolist()

        # 3) 对齐长度（防止不小心少翻 / 多翻了几行）
        n_ids = len(ids)
        n_txt = len(translated_texts)
        n = min(n_ids, n_txt)

        if n_ids != n_txt:
            print(
                f'[警告] {name}: 原始行数 = {n_ids}, 翻译行数 = {n_txt}，'
                f'将只使用前 {n} 行进行合并'
            )

        # 4) 写出新的 csv：id|translated_text
        with open(out_path, 'w', encoding='utf-8') as f_out:
            for i in range(n):
                text_spaced = space_for_game(translated_texts[i])
                new_line = f"{ids[i]}|{text_spaced}"
                f_out.write(new_line + '\n')

        print(f'[合并完成] {original_path} + {extract_path} -> {out_path}')


def main():
    while True:
        print('请选择想要执行的操作：')
        print('A：从原始 csv 提取 text 到 extract/')
        print('B：用翻译后的 text 替换并生成 combine/ 下的新 csv')
        print('Q：退出程序')
        choice = input('输入 A 或 B 或 Q：').strip().upper()

        if choice == 'A':
            for csv_name in csv_list:
                process_csv(csv_name)
        elif choice == 'B':
            combine_csv()
        elif choice == 'Q':
            break
        else:
            print('输入错误，请重新运行并输入 A 或 B 或 Q')


if __name__ == '__main__':
    main()