fix(worker+api): resolve batch title race + auto-queue pending tasks (#2098)

Two batch-rendering bugs found from staging logs (6-video batch 17:30-17:43):

Bug A (P1) — title_config race when same plan renders concurrently:
  _sync_task_config_to_plan wrote each task's title_config/bgm/resolution
  to the shared plan.config row, then RenderAdapter.render_plan read back
  from plan.config during rendering. If two tasks sharing the same plan
  (e.g. batch retry, preview+render, retry+new) ran concurrently, Task B's
  write could overwrite Task A's title before Task A's ffmpeg read it,
  producing videos with the wrong title text/style.

  Fix: stop mutating plan.config in the worker render path. Introduce
  task_config_override threading through render_plan / _do_render /
  UnifiedRenderService; URS reads config via _effective_config() which
  does a deep-copy plan.config merged with per-task override (title/bgm/
  export). _build_task_config_override builds the override dict (with
  field-name normalization: font_size→size, font_color→color) from
  task_info, and _download_voice_for_task handles voiceover download
  independently. plan.config is left as-authored by the API writeback;
  concurrent tasks no longer race on it.

  Files:
  - apps/worker/video_processing/unified_render_service.py: add
    override_config param + _cfg_section/_effective_config helpers;
    replace all render-time self.plan.config reads (title/subtitle/bgm/
    export/TTS) with _effective_config().
  - apps/worker/video_processing/render_adapter.py: render_plan/_do_render/
    _prepare_bgm accept task_config_override, merge into bgm/export reads,
    forward to URS.
  - apps/worker/worker_app/tasks/generation.py: replace
    _sync_task_config_to_plan (which wrote plan.config) with
    _build_task_config_override + _download_voice_for_task; pass
    task_config_override to render_plan.

Bug B (P0) — USER_PENDING_LIMIT=3 returns HTTP 429, blocks batch submit:
  Pre-check in create_generation_task rejected the whole batch with 429
  USER_QUEUE_FULL once user had ≥3 pending tasks; UI showed 'wait ~4 min'
  and prevented any more submissions. Users expect to submit a batch of
  6 and have them queue naturally (worker concurrency=2).

  Fix:
  - USER_PENDING_LIMIT 3→20 (soft cap for abuse protection, supports
    typical batch sizes of 6-10 with headroom).
  - Remove user-level 429 rejection from pre-check, retry endpoint,
    single-task confirm endpoint, and safe_enqueue (user path now logs
    a warning and continues to enqueue). Global GLOBAL_PENDING_LIMIT=20
    is retained as a hard 503 system-busy guard.
  - safe_enqueue post-enqueue check: user-over only logs, does not
    fail the task or raise.
  - Batch loop UserPendingLimitExceeded except branch now treats it as
    a successful enqueue (should not trigger in practice).

  Files:
  - apps/api/app/core/task_enqueue.py: limit 3→20; user checks log-only.
  - apps/api/app/api/routes/generation_tasks.py: remove user pre-check
    429; soften retry/single/batch except branches.

Stacked on PR #2095/#2097 which already fixed title baseline/width-scaling
and default-constant alignment with CPU video_filter_builder.

Tests:
- tests/unit/test_gpu_direct_pipeline.py: 30 passed
- tests/unit/test_render_adapter_pure.py: 18 passed
- ruff check/format clean; py_compile clean

apps/worker/video_processing/unified_render_service.py

@@ -158,6 +158,7 @@ class UnifiedRenderService:
         bgm_path: str | None = None,  # BGM 本地文件路径
         voiceover_audio_path: str | None = None,  # 配音素材库音频本地路径
         clip_has_text: list[bool] | None = None,  # 源视频片段是否有文字（来自 atom_clip.ai_tags.has_text）
+        override_config: dict | None = None,  # Bug A: task 级 config 覆盖（title/bgm/export/subtitle），防并发竞态
     ):
         self.plan = plan
         self.clips = clips
@@ -170,6 +171,9 @@ class UnifiedRenderService:
         self.asr_service = asr_service
         self.bgm_path = bgm_path
         self.voiceover_audio_path = voiceover_audio_path
+        # Bug A: task 级 config override（深拷贝），优先级高于 plan.config；
+        # 避免同 plan 多任务并发渲染时 _sync_task_config_to_plan 写 plan.config["title"] 互相覆盖。
+        self._override_config = dict(override_config) if isinstance(override_config, dict) else {}
         # #1970：片段级文字检测（顺序与非 audio 的源视频片段一致）；None 表示无可靠检测，保守不翻转
         self._clip_has_text = clip_has_text
         self._transition_engine = TransitionEngine(default_duration=transition_duration)
@@ -180,6 +184,28 @@ class UnifiedRenderService:
         self._micro_plan_cache: Any = None
         self._micro_plan_loaded = False
 
+    def _cfg_section(self, section: str) -> dict:
+        """读取单个配置段：override_config 优先于 plan.config（Bug A 防并发竞态）。"""
+        base = dict((self.plan.config or {}).get(section, {}) or {})
+        override = self._override_config.get(section)
+        if isinstance(override, dict) and override:
+            base.update(override)  # 浅合并，保留 base 中未被覆盖字段
+        return base
+
+    def _effective_config(self) -> dict:
+        """读取完整 config：override_config 顶层段覆盖 plan.config（Bug A 防并发竞态）。"""
+        import copy
+
+        full = copy.deepcopy(self.plan.config or {})
+        for k, v in self._override_config.items():
+            if isinstance(v, dict):
+                sec = dict(full.get(k, {}) or {})
+                sec.update(v)
+                full[k] = sec
+            else:
+                full[k] = v
+        return full
+
     # ── #1970 PR2 智能降重：片段级微变换 ───────────────────────────────────
     def _dedup_enabled(self) -> bool:
         """读取 plan.config.dedup_enabled，缺省视为 True（向后兼容）。"""
@@ -431,7 +457,7 @@ class UnifiedRenderService:
             has_audio = pass_through_has_audio
             # 直通模式下也支持 BGM 混音：提取音频 → 混 BGM → 合并回视频
             if self.bgm_path and pass_through_has_audio:
-                config = self.plan.config or {}
+                config = self._effective_config()
                 bgm_config = config.get("bgm", {}) or {}
                 if bgm_config.get("enabled", False):
                     ctx = RenderContext(work_dir=self.work_dir, plan_id=self.plan.id)
@@ -471,7 +497,7 @@ class UnifiedRenderService:
                             "[unified-render] pass-through BGM mix failed, skipping: plan_id=%s", self.plan.id
                         )
         else:
-            config = self.plan.config or {}
+            config = self._effective_config()
             bgm_config = config.get("bgm", {}) or {}
             if not isinstance(bgm_config, dict):
                 bgm_config = {}
@@ -696,7 +722,7 @@ class UnifiedRenderService:
         Returns:
             ASS 文件路径，没有字幕时返回 None
         """
-        config = self.plan.config or {}
+        config = self._effective_config()
         # #1901 统一读 "title"，兼容老数据 "title_config"
         title_cfg = config.get("title", {}) or {}
         if not isinstance(title_cfg, dict) or not (title_cfg.get("text") or "").strip():
@@ -881,7 +907,7 @@ class UnifiedRenderService:
         Returns:
             是否成功添加了配音音轨
         """
-        config = self.plan.config or {}
+        config = self._effective_config()
         tts_cfg = config.get("tts", {}) or {}
         if not isinstance(tts_cfg, dict):
             tts_cfg = {}
@@ -2275,7 +2301,7 @@ class UnifiedRenderService:
         try:
             from video_processing import gpu_direct_pipeline as gdp
 
-            cfg = self.plan.config or {}
+            cfg = self._effective_config()
             video_layer = next(_lyr for _lyr in layers if _lyr.role not in ("audio",))
             video_clips = [c for c in video_layer.clips if c.clip_type != "audio"]