TeamAAS-2.5/Tools/LocalizationExtractor/Program.cs

using System.Text;
using System.Text.Json;
using System.Text.Json.Nodes;
using System.Text.RegularExpressions;

// ============================================================================
// 多语言原文提取器：扫描源码中所有 Translate 调用点，把"中文原文"合并进
// Localization\zh-CN.json，保证界面上用到的文本不会漏配翻译。
//
// 扫描的调用点：
//   1. XAML  标记扩展   {loc:Translate '原文'} / {loc:Translate "原文"} / Key='原文'
//   2. C#   方法调用    xxx.Translate("原文") / xxx.Translate("原文", "ctx")
//                          （含 @"..." 逐字字符串）
//
// 合并策略（只增不删，防止误删运行时动态文本）：
//   - 新原文      → 追加进 translations（value = 原文）
//   - 已有原文    → 保留不动
//   - 疑似废弃    → 保留，仅打印提醒（人工到 LocalizationTool 里确认后再删）
//
// 另外校验其他语言包（en-US.json 等）相对 zh-CN 缺失的 key，打印清单供翻译。
// ============================================================================

var root = args.Length > 0 ? Path.GetFullPath(args[0]) : FindRepoRoot();
if (!Directory.Exists(root))
{
    Console.Error.WriteLine($"[LocalizationExtractor] 仓库根目录不存在: {root}");
    return 1;
}

// 语言包母本：ExE\win-x64\Runtime\Localization（程序运行时加载的目录，翻译在此维护）。
// 仓库根 Localization\ 仅是历史快照，仅作兜底。
var localizationDir = Path.Combine(root, "ExE", "win-x64", "Runtime", "Localization");
if (!Directory.Exists(localizationDir))
    localizationDir = Path.Combine(root, "Localization");
var zhFile = Path.Combine(localizationDir, "zh-CN.json");
if (!File.Exists(zhFile))
{
    Console.Error.WriteLine($"[LocalizationExtractor] 找不到 {zhFile}");
    return 1;
}

// 默认扫描除工具/输出目录外的所有源码工程目录
var excluded = new HashSet<string>(StringComparer.OrdinalIgnoreCase)
    { "bin", "obj", ".vs", "packages", "lib", "ExE", "EXE", "Tools", ".git" };
var scanDirs = Directory.EnumerateDirectories(root)
    .Where(d => !excluded.Contains(Path.GetFileName(d)))
    .ToList();

var originals = new SortedSet<string>(StringComparer.Ordinal);

foreach (var dir in scanDirs)
{
    foreach (var file in Directory.EnumerateFiles(dir, "*", SearchOption.AllDirectories))
    {
        var ext = Path.GetExtension(file).ToLowerInvariant();
        if (ext != ".xaml" && ext != ".cs") continue;
        if (IsUnderExcludedDir(file)) continue;

        string content;
        try { content = File.ReadAllText(file); }
        catch { continue; } // 文件被占用等，跳过即可

        if (ext == ".xaml") ExtractXaml(content, originals);
        else ExtractCSharp(content, originals);
    }
}

bool IsUnderExcludedDir(string file)
{
    var rel = Path.GetRelativePath(root, file);
    return rel.Split(Path.DirectorySeparatorChar)
        .Any(seg => excluded.Contains(seg));
}

// ----------------------------------------------------------------------------
// 读取 zh-CN.json 并合并
// ----------------------------------------------------------------------------
var jsonOptions = new JsonSerializerOptions
{
    WriteIndented = true,
    Encoder = System.Text.Encodings.Web.JavaScriptEncoder.UnsafeRelaxedJsonEscaping,
};

var node = JsonNode.Parse(File.ReadAllText(zhFile))!.AsObject();
var translations = node["translations"]!.AsObject();

var added = new List<string>();
foreach (var text in originals)
    if (!translations.ContainsKey(text))
    {
        translations[text] = text; // 中文包 value = 原文
        added.Add(text);
    }

var removedCandidates = translations.Select(t => t.Key).Where(k => !originals.Contains(k)).ToList();

if (added.Count > 0)
    File.WriteAllText(zhFile, node.ToJsonString(jsonOptions), new UTF8Encoding(false));

// ----------------------------------------------------------------------------
// 其他语言包缺失校验（只报告，不改写——翻译走 LocalizationTool / AI 流程）
// ----------------------------------------------------------------------------
var zhKeys = translations.Select(t => t.Key).ToHashSet(StringComparer.Ordinal);
var missingReport = new StringBuilder();
foreach (var packFile in Directory.EnumerateFiles(localizationDir, "*.json"))
{
    var name = Path.GetFileName(packFile);
    if (name.Equals("zh-CN.json", StringComparison.OrdinalIgnoreCase)) continue;

    JsonObject pack;
    try { pack = JsonNode.Parse(File.ReadAllText(packFile))!["translations"]!.AsObject(); }
    catch { continue; }

    var packKeys = pack.Select(p => p.Key).ToHashSet(StringComparer.Ordinal);
    var missing = zhKeys.Where(k => !packKeys.Contains(k)).ToList();
    if (missing.Count > 0)
        missingReport.AppendLine($"  {name} 缺失 {missing.Count} 条: {string.Join(" | ", missing.Take(10))}{(missing.Count > 10 ? " ..." : "")}");
}

// ----------------------------------------------------------------------------
// 汇总输出
// ----------------------------------------------------------------------------
Console.WriteLine($"[LocalizationExtractor] 扫描 {scanDirs.Count} 个工程目录，共提取原文 {originals.Count} 条");
if (added.Count > 0)
{
    Console.WriteLine($"[LocalizationExtractor] zh-CN.json 新增 {added.Count} 条");
}
if (removedCandidates.Count > 0)
{
    Console.WriteLine($"[LocalizationExtractor] 提示: {removedCandidates.Count} 条原文未在源码中找到（已保留，确认废弃请手动删除）:");
    foreach (var k in removedCandidates.Take(10))
        Console.WriteLine($"    - {k}");
    if (removedCandidates.Count > 10) Console.WriteLine("    ...");
}
if (missingReport.Length > 0)
{
    Console.WriteLine("[LocalizationExtractor] 其他语言包缺失词条:");
    Console.Write(missingReport);
}

return 0;

// ----------------------------------------------------------------------------
// 提取逻辑
// ----------------------------------------------------------------------------
static void ExtractXaml(string content, SortedSet<string> into)
{
    // {loc:Translate '原文'} 或 {loc:Translate "原文"}
    foreach (Match m in Regex.Matches(content, @"loc:Translate\s+(['""])(?<text>(?:(?!\1).)*)\1"))
        into.Add(UnescapeXml(m.Groups["text"].Value));

    // {loc:Translate Key='原文'} 属性写法
    foreach (Match m in Regex.Matches(content, @"loc:Translate\s+Key=(['""])(?<text>(?:(?!\1).)*)\1"))
        into.Add(UnescapeXml(m.Groups["text"].Value));
}

static void ExtractCSharp(string content, SortedSet<string> into)
{
    // 普通字符串：Translate("原文"...（含 xxx.Translate 与裸调用）
    foreach (Match m in Regex.Matches(content, @"(?:^|[^.\w])Translate\(\s*""(?<text>(?:[^""\\]|\\.)*)"""))
        into.Add(UnescapeCSharp(m.Groups["text"].Value));

    // 逐字字符串：Translate(@"原文...
    foreach (Match m in Regex.Matches(content, @"(?:^|[^.\w])Translate\(\s*@""(?<text>(?:[^""]|"""")*)"""))
        into.Add(m.Groups["text"].Value.Replace("\"\"", "\""));

    // L10n.T("原文") / L10n.F("模板{0}"...) — C# 界面文案入口
    foreach (Match m in Regex.Matches(content, @"L10n\.(?:T|F)\(\s*""(?<text>(?:[^""\\]|\\.)*)"""))
        into.Add(UnescapeCSharp(m.Groups["text"].Value));

    foreach (Match m in Regex.Matches(content, @"L10n\.(?:T|F)\(\s*@""(?<text>(?:[^""]|"""")*)"""))
        into.Add(m.Groups["text"].Value.Replace("\"\"", "\""));
}

static string UnescapeXml(string s) => s
    .Replace("&lt;", "<").Replace("&gt;", ">")
    .Replace("&quot;", "\"").Replace("&apos;", "'")
    .Replace("&amp;", "&");

static string UnescapeCSharp(string s) => Regex.Replace(s, @"\\(?:u(?<u>[0-9a-fA-F]{4})|x(?<x>[0-9a-fA-F]{2,4})|(?<c>.))",
    m =>
    {
        if (m.Groups["u"].Success) return ((char)Convert.ToInt32(m.Groups["u"].Value, 16)).ToString();
        if (m.Groups["x"].Success) return ((char)Convert.ToInt32(m.Groups["x"].Value, 16)).ToString();
        return m.Groups["c"].Value switch
        {
            "n" => "\n", "t" => "\t", "r" => "\r", "0" => "\0",
            "a" => "\a", "b" => "\b", "f" => "\f", "v" => "\v",
            var c => c, // \\ \" \' 等
        };
    });

static string FindRepoRoot()
{
    // 从当前目录向上找包含 TeamAAS.sln 的目录
    var dir = new DirectoryInfo(Environment.CurrentDirectory);
    while (dir != null && !File.Exists(Path.Combine(dir.FullName, "TeamAAS.sln")))
        dir = dir.Parent;
    return dir?.FullName ?? Environment.CurrentDirectory;
}