feat: support any yt-dlp site + readable download filenames

frontend/src/components/UrlInput.tsx

@@ -72,14 +72,14 @@ export function UrlInput({ onSubmit, disabled }: UrlInputProps) {
       return;
     }
 
-    const youtubePattern = /^https?:\/\/(www\.)?(youtube\.com\/watch\?v=|youtu\.be\/)/;
-    if (!youtubePattern.test(url)) {
+    const urlPattern = /^https?:\/\/.+/;
+    if (!urlPattern.test(url)) {
       setError(t('error.invalid_url'));
       return;
     }
 
     const request: JobCreateRequest = {
-      youtube_url: url,
+      source_url: url,
     };
     if (startSeconds !== undefined) request.start_time = startSeconds;
     if (endSeconds !== undefined) request.end_time = endSeconds;

frontend/src/i18n/en.json

@@ -1,7 +1,7 @@
 {
   "app": {
     "title": "BilingualSub",
-    "subtitle": "YouTube Bilingual Subtitle Generator",
+    "subtitle": "Bilingual Subtitle Generator for YouTube, X, TikTok & More",
     "processing_title": "Processing Video",
     "processing_desc": "Please wait while we transcribe and translate.",
     "error_title": "Something went wrong",
@@ -17,8 +17,7 @@
     "download_audio": "Download Audio"
   },
   "form": {
-    "url_placeholder": "Paste YouTube video URL",
-    "paste_placeholder": "Paste YouTube URL...",
+    "paste_placeholder": "Paste video URL (YouTube / X / Twitter / TikTok…)",
     "source_lang": "Source Language",
     "target_lang": "Target Language",
     "label_translate": "Translate",
@@ -33,7 +32,7 @@
     "enable_range": "Set Range",
     "disable_range": "Disable Range",
     "clear_range": "Clear",
-    "inputModeUrl": "YouTube URL",
+    "inputModeUrl": "Video URL",
     "inputModeFile": "Upload File",
     "filePlaceholder": "Choose a video or audio file",
     "fileSelected": "Selected: {{filename}}"
@@ -66,7 +65,7 @@
     "unsupported": "Your browser does not support video playback"
   },
   "error": {
-    "invalid_url": "Please enter a valid YouTube URL",
+    "invalid_url": "Please enter a valid video URL",
     "job_not_found": "Job not found",
     "download_failed": "Video download failed",
     "transcription_failed": "Transcription failed",

frontend/src/i18n/zh-TW.json

@@ -1,7 +1,7 @@
 {
   "app": {
     "title": "BilingualSub",
-    "subtitle": "YouTube 雙語字幕自動生成工具",
+    "subtitle": "多平台雙語字幕自動生成工具",
     "processing_title": "影片處理中",
     "processing_desc": "請稍候，我們正在進行語音辨識與翻譯。",
     "error_title": "發生錯誤",
@@ -17,8 +17,7 @@
     "download_audio": "下載音訊"
   },
   "form": {
-    "url_placeholder": "貼上 YouTube 影片網址",
-    "paste_placeholder": "貼上 YouTube 網址...",
+    "paste_placeholder": "貼上影片網址（YouTube / X / Twitter / TikTok…）",
     "source_lang": "原始語言",
     "target_lang": "翻譯語言",
     "label_translate": "翻譯語言",
@@ -33,7 +32,7 @@
     "enable_range": "設定範圍",
     "disable_range": "取消範圍",
     "clear_range": "清除",
-    "inputModeUrl": "YouTube 網址",
+    "inputModeUrl": "影片網址",
     "inputModeFile": "上傳檔案",
     "filePlaceholder": "選擇影片或音訊檔案",
     "fileSelected": "已選擇：{{filename}}"
@@ -66,7 +65,7 @@
     "unsupported": "您的瀏覽器不支援影片播放"
   },
   "error": {
-    "invalid_url": "請輸入有效的 YouTube 網址",
+    "invalid_url": "請輸入有效的影片網址",
     "job_not_found": "找不到此任務",
     "download_failed": "影片下載失敗",
     "transcription_failed": "語音辨識失敗",

frontend/src/types.ts

@@ -1,7 +1,7 @@
 import type { FileType, JobStatus } from './constants';
 
 export interface JobCreateRequest {
-  youtube_url: string;
+  source_url: string;
   source_lang?: string;
   target_lang?: string;
   start_time?: number; // seconds

src/bilingualsub/api/jobs.py

@@ -29,7 +29,7 @@ class Job:
     """Represents a subtitle generation job."""
 
     id: str
-    youtube_url: str = ""
+    source_url: str = ""
     source_lang: str = ""
     target_lang: str = ""
     local_video_path: Path | None = None
@@ -59,7 +59,7 @@ def __init__(self) -> None:
 
     def create_job(
         self,
-        youtube_url: str = "",
+        source_url: str = "",
         source_lang: str = "",
         target_lang: str = "",
         start_time: float | None = None,
@@ -70,15 +70,15 @@ def create_job(
         job_id = uuid.uuid4().hex[:12]
         job = Job(
             id=job_id,
-            youtube_url=youtube_url,
+            source_url=source_url,
             source_lang=source_lang,
             target_lang=target_lang,
             local_video_path=local_video_path,
             start_time=start_time,
             end_time=end_time,
         )
         self._jobs[job_id] = job
-        logger.info("job_created", job_id=job_id, youtube_url=youtube_url)
+        logger.info("job_created", job_id=job_id, source_url=source_url)
         return job
 
     def get_job(self, job_id: str) -> Job | None:

src/bilingualsub/api/pipeline.py

@@ -23,7 +23,7 @@
     TranscriptionError,
     TranslationError,
     VideoMetadata,
-    download_youtube_video,
+    download_video,
     merge_subtitles,
     transcribe_audio,
     translate_subtitle,
@@ -236,8 +236,8 @@ def _put_event() -> None:
             loop.call_soon_threadsafe(_put_event)
 
     metadata = await asyncio.to_thread(
-        download_youtube_video,
-        job.youtube_url,
+        download_video,
+        job.source_url,
         video_path,
         on_progress=_on_download_progress,
         start_time=job.start_time,
@@ -385,112 +385,6 @@ async def run_subtitle(job: Job) -> None:
         )
 
 
-async def run_pipeline(job: Job) -> None:
-    """Execute the full subtitle generation pipeline for a job.
-
-    Steps: download -> transcribe -> translate -> merge/serialize -> burn.
-    All blocking core calls are wrapped in asyncio.to_thread().
-    Progress events are sent to job.event_queue at each step.
-    """
-    log = logger.bind(job_id=job.id)
-    work_dir = Path(tempfile.mkdtemp(prefix=f"bilingualsub_{job.id}_"))
-
-    try:
-        # --- Step 1: Download or use local file ---
-        video_path, metadata = await _acquire_video(job, work_dir, log)
-        # Trimming is now handled during download via start_time/end_time
-
-        # --- Step 1.5: Extract audio ---
-        audio_path = await _extract_audio_step(job, video_path, work_dir, log)
-
-        # --- Step 2: Transcribe ---
-        _send_progress(
-            job, JobStatus.TRANSCRIBING, 20.0, "transcribe", "Transcribing audio"
-        )
-        t0 = time.monotonic()
-        original_sub = await asyncio.to_thread(
-            transcribe_audio, audio_path, language=job.source_lang
-        )
-        log.info(
-            "step_done",
-            step="transcribe",
-            duration_ms=int((time.monotonic() - t0) * 1000),
-        )
-
-        # --- Step 3: Translate ---
-        _send_progress(
-            job, JobStatus.TRANSLATING, 50.0, "translate", "Translating subtitles"
-        )
-        t0 = time.monotonic()
-
-        _on_translate_progress = _make_translate_progress_cb(job)
-        _on_rate_limit = _make_rate_limit_cb(job)
-
-        translated_sub = await asyncio.to_thread(
-            translate_subtitle,
-            original_sub,
-            source_lang=job.source_lang,
-            target_lang=job.target_lang,
-            video_title=metadata.title,
-            video_description=metadata.description,
-            on_progress=_on_translate_progress,
-            on_rate_limit=_on_rate_limit,
-        )
-        log.info(
-            "step_done",
-            step="translate",
-            duration_ms=int((time.monotonic() - t0) * 1000),
-        )
-
-        # --- Step 4: Merge & Serialize ---
-        _send_progress(
-            job, JobStatus.MERGING, 70.0, "merge", "Merging bilingual subtitles"
-        )
-        t0 = time.monotonic()
-
-        merged_entries = await asyncio.to_thread(
-            merge_subtitles, original_sub.entries, translated_sub.entries
-        )
-        merged_sub = Subtitle(entries=merged_entries)
-
-        srt_content = serialize_srt(merged_sub)
-        srt_path = work_dir / "subtitle.srt"
-        srt_path.write_text(srt_content, encoding="utf-8")
-        job.output_files[FileType.SRT] = srt_path
-
-        ass_content = serialize_bilingual_ass(
-            original_sub,
-            translated_sub,
-            video_width=metadata.width,
-            video_height=metadata.height,
-        )
-        ass_path = work_dir / "subtitle.ass"
-        ass_path.write_text(ass_content, encoding="utf-8")
-        job.output_files[FileType.ASS] = ass_path
-
-        log.info(
-            "step_done", step="merge", duration_ms=int((time.monotonic() - t0) * 1000)
-        )
-
-        # --- Step 5: Save source video & complete ---
-        job.output_files[FileType.SOURCE_VIDEO] = video_path
-        _send_complete(job)
-        log.info("pipeline_complete", job_id=job.id)
-
-    except PipelineError:
-        raise
-    except Exception as exc:
-        pipeline_err = _to_pipeline_error(exc)
-        _send_error(
-            job, pipeline_err.code, pipeline_err.message, pipeline_err.detail or ""
-        )
-        log.error(
-            "pipeline_failed",
-            error_code=pipeline_err.code,
-            error=str(exc),
-        )
-
-
 async def run_burn(job: Job, srt_content: str) -> None:
     """Burn user-edited SRT into the source video."""
     log = logger.bind(job_id=job.id)

src/bilingualsub/api/routes.py

@@ -4,9 +4,10 @@
 
 import asyncio
 import json
+import re
 import tempfile
 from pathlib import Path
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING, Any, NamedTuple
 
 import structlog
 from fastapi import APIRouter, Form, Request, UploadFile
@@ -53,6 +54,52 @@
     ".webm",
 }
 
+_DEFAULT_FILENAME = "video"
+_SUFFIX_ORIGINAL = "(original)"
+_LANG_SEPARATOR = "_to_"
+
+
+class _FileMeta(NamedTuple):
+    ext: str
+    media_type: str
+
+
+_FILE_META: dict[FileType, _FileMeta] = {
+    FileType.SRT: _FileMeta("srt", "text/plain"),
+    FileType.ASS: _FileMeta("ass", "text/plain"),
+    FileType.VIDEO: _FileMeta("mp4", "video/mp4"),
+    FileType.AUDIO: _FileMeta("mp3", "audio/mpeg"),
+    FileType.SOURCE_VIDEO: _FileMeta("mp4", "video/mp4"),
+}
+
+# Windows-reserved + POSIX control characters (NTFS/FAT32/ext4 safe)
+_FILENAME_BAD_CHARS = re.compile(r'[<>:"/\\|?*\x00-\x1f]')
+
+_MAX_UPLOAD_BYTES = 500 * 1024 * 1024  # 500 MB
+
+
+def _sanitize_filename(name: str) -> str:
+    """Remove filesystem-unsafe characters and truncate to 120 chars."""
+    cleaned = _FILENAME_BAD_CHARS.sub("", name).strip(" .")
+    truncated = (cleaned or _DEFAULT_FILENAME)[:120]
+    return truncated.rstrip(" .") or _DEFAULT_FILENAME
+
+
+def _build_download_filename(job: Job, file_type: FileType) -> str:
+    """Build a human-readable download filename for the given job and file type."""
+    base = _sanitize_filename(job.video_title or _DEFAULT_FILENAME)
+    original_only = file_type in (FileType.SOURCE_VIDEO, FileType.AUDIO)
+    same_lang = (
+        not job.source_lang or not job.target_lang or job.source_lang == job.target_lang
+    )
+    suffix = (
+        _SUFFIX_ORIGINAL
+        if original_only or same_lang
+        else f"({job.source_lang}{_LANG_SEPARATOR}{job.target_lang})"
+    )
+    ext = _FILE_META[file_type].ext
+    return f"{base} {suffix}.{ext}"
+
 
 def _get_job_manager(request: Request) -> JobManager:
     """Get the JobManager from app state."""
@@ -83,7 +130,7 @@ async def create_job(body: JobCreateRequest, request: Request) -> JobCreateRespo
     """Create a new subtitle generation job."""
     manager = _get_job_manager(request)
     job = manager.create_job(
-        youtube_url=str(body.youtube_url),
+        source_url=str(body.source_url),
         source_lang=body.source_lang,
         target_lang=body.target_lang,
         start_time=body.start_time,
@@ -104,7 +151,6 @@ async def create_job_from_upload(
     request: Request,
 ) -> JobCreateResponse:
     """Create a subtitle generation job from an uploaded file."""
-    # Validate file extension
     filename = file.filename or ""
     suffix = Path(filename).suffix.lower()
     if suffix not in _ALLOWED_UPLOAD_EXTENSIONS:
@@ -113,11 +159,9 @@ async def create_job_from_upload(
             detail=f"Allowed: {', '.join(sorted(_ALLOWED_UPLOAD_EXTENSIONS))}",
         )
 
-    # Sanitize filename to prevent path traversal
     safe_name = Path(filename).name or f"upload{suffix}"
 
-    # Save uploaded file to temp directory with size limit
-    max_size = 500 * 1024 * 1024  # 500 MB
+    max_size = _MAX_UPLOAD_BYTES
     tmp_dir = Path(tempfile.mkdtemp(prefix="bilingualsub_upload_"))
     saved_path = tmp_dir / safe_name
     bytes_written = 0
@@ -212,25 +256,10 @@ async def download_file(job_id: str, file_type: str, request: Request) -> FileRe
             detail="Job may not have completed this step",
         )
 
-    # Set appropriate media type and filename
-    media_types = {
-        FileType.SRT: "text/plain",
-        FileType.ASS: "text/plain",
-        FileType.VIDEO: "video/mp4",
-        FileType.AUDIO: "audio/mpeg",
-        FileType.SOURCE_VIDEO: "video/mp4",
-    }
-    extensions = {
-        FileType.SRT: "srt",
-        FileType.ASS: "ass",
-        FileType.VIDEO: "mp4",
-        FileType.AUDIO: "mp3",
-        FileType.SOURCE_VIDEO: "mp4",
-    }
     return FileResponse(
         path=path,
-        media_type=media_types[ft],
-        filename=f"bilingualsub.{extensions[ft]}",
+        media_type=_FILE_META[ft].media_type,
+        filename=_build_download_filename(job, ft),
     )