#!/usr/bin/env python3
"""
合并所有Sage模块的i18n内容到 wwwroot/i18n

从 load_path.py 中提取sage引用的所有模块，
读取每个模块的 i18n/{lang}/msg.txt，
合并输出到 wwwroot/i18n/{lang}/i18n.json。

冲突策略：后加载的模块覆盖先加载的，sage自身覆盖所有。
"""

import os
import re
import json
import sys
from pathlib import Path

# --- 配置 ---
SAGE_DIR = Path(__file__).resolve().parent       # /d/ymq/repos/sage
REPOS_DIR = SAGE_DIR.parent                        # /d/ymq/repos
I18N_OUTPUT_DIR = SAGE_DIR / 'wwwroot' / 'i18n'    # 输出目标
LANGS = ['zh', 'en', 'jp', 'ko']                   # 支持的语言


def parse_load_path(load_path_file):
    """从 load_path.py 中提取模块名列表

    load_path.py 中每行格式: /module_name/rest/of/path ...
    只提取行首的第一个路径段作为模块名。
    """
    modules = set()
    # 已知的非模块路径前缀（系统路径、资源目录等）
    skip_prefixes = {
        'conf', 'logs', 'files', 'script', 'skills', 'plugins',
        'py3', 'bin', 'initln', 'docs', 'agents', 'hermes_work',
        'bricks', 'imgs', 'i18n', 'public', 'api', 'shell',
    }

    with open(load_path_file, 'r', encoding='utf-8') as f:
        for line in f:
            line = line.strip()
            if not line or line.startswith('#'):
                continue
            # 匹配行首的 /module_name/
            m = re.match(r'^/\s*([a-zA-Z_][a-zA-Z0-9_]*)\s*/', line)
            if not m:
                continue
            mod = m.group(1)
            if mod not in skip_prefixes:
                modules.add(mod)

    # sage 自身始终排在前面（优先级最低，因为 load_path 中 sage 是根路径）
    modules.discard('sage')
    result = ['sage'] + sorted(modules)
    return result


def find_module_i18n_dir(module_name):
    """查找模块的 i18n 目录"""
    candidate = REPOS_DIR / module_name / 'i18n'
    if candidate.is_dir():
        return candidate
    return None


def parse_msg_txt(filepath):
    """解析 msg.txt，格式: key: value（每行一个）"""
    result = {}
    if not filepath.is_file():
        return result
    
    with open(filepath, 'r', encoding='utf-8') as f:
        for line in f:
            line = line.rstrip('\n').rstrip('\r')
            if not line:
                continue
            # 格式: key: value
            idx = line.find(': ')
            if idx > 0:
                key = line[:idx]
                value = line[idx+2:]
                result[key] = value
            else:
                # 可能只有 key 没有 value，或者格式不对，跳过
                idx = line.find(':')
                if idx > 0:
                    key = line[:idx]
                    value = line[idx+1:]
                    result[key] = value
    return result


def merge_i18n(modules):
    """合并所有模块的 i18n，返回 {lang: {key: value}}"""
    merged = {lang: {} for lang in LANGS}
    
    for module_name in modules:
        i18n_dir = find_module_i18n_dir(module_name)
        if not i18n_dir:
            print(f"  [跳过] {module_name}: 未找到i18n目录")
            continue
        
        for lang in LANGS:
            msg_file = i18n_dir / lang / 'msg.txt'
            if msg_file.is_file():
                translations = parse_msg_txt(msg_file)
                count = len(translations)
                before = len(merged[lang])
                merged[lang].update(translations)
                after = len(merged[lang])
                new_keys = after - before
                if new_keys > 0 or count > 0:
                    print(f"  {module_name}/{lang}: {count}条, 新增{new_keys}个唯一key")
            else:
                print(f"  [警告] {module_name}/{lang}: msg.txt 不存在")
    
    return merged


def write_output(merged):
    """将合并结果写入 wwwroot/i18n/{lang}/i18n.json 和 msg.txt"""
    for lang in LANGS:
        lang_dir = I18N_OUTPUT_DIR / lang
        lang_dir.mkdir(parents=True, exist_ok=True)
        
        data = merged[lang]
        
        # 写入 i18n.json（前端读取此文件）
        json_file = lang_dir / 'i18n.json'
        with open(json_file, 'w', encoding='utf-8') as f:
            json.dump(data, f, ensure_ascii=False, indent='\t')
        
        # 同时写入 msg.txt（人类可读格式）
        msg_file = lang_dir / 'msg.txt'
        with open(msg_file, 'w', encoding='utf-8') as f:
            for key, value in data.items():
                f.write(f"{key}: {value}\n")
        
        print(f"  写入: {json_file} ({len(data)} 条)")
        print(f"  写入: {msg_file} ({len(data)} 条)")


def main():
    load_path_file = SAGE_DIR / 'load_path.py'
    if not load_path_file.is_file():
        print(f"错误: 找不到 {load_path_file}")
        sys.exit(1)
    
    print(f"解析 load_path.py: {load_path_file}")
    modules = parse_load_path(load_path_file)
    print(f"发现 {len(modules)} 个模块:")
    for m in modules:
        has_i18n = (REPOS_DIR / m / 'i18n').is_dir()
        status = '✓' if has_i18n else '✗'
        print(f"  {status} {m}")
    
    print(f"\n开始合并 i18n...")
    merged = merge_i18n(modules)
    
    print(f"\n合并结果:")
    for lang in LANGS:
        print(f"  {lang}: {len(merged[lang])} 个唯一key")
    
    print(f"\n写入 wwwroot/i18n...")
    write_output(merged)
    
    print(f"\n完成!")


if __name__ == '__main__':
    main()