Compare commits

...

6 Commits

Author SHA1 Message Date
CI Bot 1ccec767d1 fix(worker): 修复 dedup fingerprint JSON 序列化失败 - np.float32 转原生 float
CI/CD Pipeline / Frontend Lint (pull_request) Successful in 55s
CI/CD Pipeline / Validate Code Quality And Tests (pull_request) Successful in 1m39s
CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (pull_request) Has been skipped
CI/CD Pipeline / Build Production Runtime Images (pull_request) Has been skipped
CI/CD Pipeline / Staging E2E Tests (pull_request) Has been skipped
CI/CD Pipeline / Deploy Production (pull_request) Has been skipped
CI/CD Pipeline / Staging API Integration Tests (pull_request) Has been skipped
CI/CD Pipeline / Production Browser E2E (pull_request) Has been skipped
CI/CD Pipeline / Integration Tests (pull_request) Successful in 1m8s
根因:cv2.normalize() 返回 np.float32 数组,color_histograms 直接存进 dict 后
SQLAlchemy JSON 序列化时报 'float32 is not JSON serializable'

修复:VideoFingerprint.to_dict() 中统一转 Python 原生类型
- color_histograms: np.float32 → float
- duration: 统一转 float
- resolution: np.int32 → int
2026-07-13 08:41:24 +08:00
xiaoxia 788559ff29 test(render-compare): 灰度对比工具包 + 5个P1修复 (#237)
CI/CD Pipeline / Frontend Lint (push) Successful in 48s
CI/CD Pipeline / Validate Code Quality And Tests (push) Successful in 1m6s
CI/CD Pipeline / Build Production Runtime Images (push) Has been skipped
CI/CD Pipeline / Deploy Production (push) Has been skipped
CI/CD Pipeline / Production Browser E2E (push) Has been skipped
CI/CD Pipeline / Integration Tests (push) Successful in 1m22s
CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (push) Successful in 53m17s
CI/CD Pipeline / Staging E2E Tests (push) Successful in 1m10s
CI/CD Pipeline / Staging API Integration Tests (push) Successful in 3m16s
2026-07-13 07:46:38 +08:00
xiaoxia bdf99bba39 fix(worker): 修复 render_edit_plan 素材下载 + 总时长日志 + video_processing 导入链 (#236)
CI/CD Pipeline / Validate Code Quality And Tests (push) Successful in 4m30s
CI/CD Pipeline / Frontend Lint (push) Successful in 3m22s
CI/CD Pipeline / Build Production Runtime Images (push) Has been skipped
CI/CD Pipeline / Deploy Production (push) Has been skipped
CI/CD Pipeline / Production Browser E2E (push) Has been skipped
CI/CD Pipeline / Integration Tests (push) Successful in 3m13s
CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (push) Successful in 7m31s
CI/CD Pipeline / Staging E2E Tests (push) Successful in 51s
CI/CD Pipeline / Staging API Integration Tests (push) Successful in 3m19s
2026-07-13 01:32:28 +08:00
xiaoxia bfe8bfe2da feat(feature-flag): Redis Feature Flag 灰度发布基础设施
CI/CD Pipeline / Validate Code Quality And Tests (push) Successful in 2m14s
CI/CD Pipeline / Frontend Lint (push) Successful in 2m7s
CI/CD Pipeline / Build Production Runtime Images (push) Has been skipped
CI/CD Pipeline / Deploy Production (push) Has been skipped
CI/CD Pipeline / Production Browser E2E (push) Has been skipped
CI/CD Pipeline / Integration Tests (push) Successful in 2m35s
CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (push) Successful in 7m54s
CI/CD Pipeline / Staging E2E Tests (push) Failing after 2m35s
CI/CD Pipeline / Staging API Integration Tests (push) Successful in 3m21s
Redis Feature Flag 灰度发布基础设施:白名单+百分比切流+全局开关,热更新,内部管理API
2026-07-13 01:11:46 +08:00
xiaoxia fdcf48103e feat(unified-render): Phase 3 - 音频统一混音 + 灰度观测埋点
CI/CD Pipeline / Validate Code Quality And Tests (push) Successful in 1m27s
CI/CD Pipeline / Frontend Lint (push) Successful in 2m4s
CI/CD Pipeline / Build Production Runtime Images (push) Has been skipped
CI/CD Pipeline / Deploy Production (push) Has been skipped
CI/CD Pipeline / Production Browser E2E (push) Has been skipped
CI/CD Pipeline / Integration Tests (push) Successful in 1m44s
CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (push) Successful in 9m39s
CI/CD Pipeline / Staging E2E Tests (push) Has been skipped
CI/CD Pipeline / Staging API Integration Tests (push) Has been skipped
Phase 3 unified render with audio mixing + gray scale observation metrics
2026-07-12 23:00:51 +08:00
xiaoxia 9e87ac05c6 style: fix isort imports + black formatting for render adapter (#234)
CI/CD Pipeline / Validate Code Quality And Tests (push) Successful in 1m55s
CI/CD Pipeline / Frontend Lint (push) Successful in 4m4s
CI/CD Pipeline / Build Production Runtime Images (push) Has been skipped
CI/CD Pipeline / Deploy Production (push) Has been skipped
CI/CD Pipeline / Production Browser E2E (push) Has been skipped
CI/CD Pipeline / Integration Tests (push) Successful in 2m21s
CI/CD Pipeline / Build & Push Staging (Watchtower auto-deploy) (push) Successful in 8m56s
CI/CD Pipeline / Staging E2E Tests (push) Successful in 2m5s
CI/CD Pipeline / Staging API Integration Tests (push) Successful in 4m51s
style: fix isort imports + black formatting for render adapter
2026-07-12 20:57:29 +08:00
22 changed files with 3843 additions and 80 deletions
+5
View File
@@ -8,6 +8,7 @@ from app.api.routes.dashboard import router as dashboard_router
from app.api.routes.duplication import router as duplication_router
from app.api.routes.edit_plans import router as edit_plans_router
from app.api.routes.edit_templates import router as edit_templates_router
from app.api.routes.feature_flags import router as feature_flags_router
from app.api.routes.generated_videos import router as generated_videos_router
from app.api.routes.generation_tasks import router as generation_tasks_router
from app.api.routes.health import router as health_check_router
@@ -151,3 +152,7 @@ api_router.include_router(
prefix="/tts",
tags=["TTS"],
)
api_router.include_router(
feature_flags_router,
tags=["Internal"],
)
+195
View File
@@ -0,0 +1,195 @@
"""Feature Flag 内部管理接口。
通过内部 API Key 鉴权,支持查看和修改 Feature Flag 配置。
主要用于灰度发布期间的动态开关控制。
API:
GET /api/v1/internal/feature-flags - 列出所有 flag
GET /api/v1/internal/feature-flags/{name} - 查看单个 flag
PUT /api/v1/internal/feature-flags/{name} - 设置 flag 配置
DELETE /api/v1/internal/feature-flags/{name} - 删除 flag
鉴权:X-API-Key header,走内部 API Key 验证
"""
from __future__ import annotations
import logging
from typing import Optional
from app.api.routes.auth import _verify_internal_api_key
from app.config import settings
from fastapi import APIRouter, Depends, HTTPException, Query, status
from pydantic import BaseModel, Field
from packages.adapters.redis.feature_flag_store import (
FEATURE_FLAG_REDIS_PREFIX,
FeatureFlagConfig,
RedisFeatureFlagStore,
)
logger = logging.getLogger(__name__)
router = APIRouter(prefix="/internal/feature-flags", tags=["Internal"])
# 允许管理的 flag 白名单(防止误操作其他系统 flag)
ALLOWED_FLAGS = {
"render_engine",
}
def _get_feature_flag_store() -> RedisFeatureFlagStore:
"""获取 Feature Flag 存储实例。"""
return RedisFeatureFlagStore(redis_url=settings.REDIS_URL)
class FeatureFlagUpdateRequest(BaseModel):
"""Feature Flag 更新请求体。"""
enabled: bool = Field(..., description="是否启用")
percentage: int = Field(0, ge=0, le=100, description="灰度百分比 (0-100)")
whitelist: list[str] = Field(default_factory=list, description="白名单列表(如 user_id")
class FeatureFlagResponse(BaseModel):
"""Feature Flag 响应。"""
name: str
enabled: bool
percentage: int
whitelist: list[str]
@classmethod
def from_config(cls, config: FeatureFlagConfig) -> "FeatureFlagResponse":
return cls(
name=config.name,
enabled=config.enabled,
percentage=config.percentage,
whitelist=sorted(config.whitelist),
)
class FeatureFlagCheckResponse(BaseModel):
"""Flag 激活检查响应。"""
name: str
active: bool
identifier: Optional[str] = None
def _validate_flag_name(name: str) -> None:
"""校验 flag 名称是否在允许列表中。"""
if name not in ALLOWED_FLAGS:
raise HTTPException(
status_code=status.HTTP_400_BAD_REQUEST,
detail=f"Unsupported flag: {name}. Allowed: {sorted(ALLOWED_FLAGS)}",
)
@router.get("", response_model=list[FeatureFlagResponse])
async def list_feature_flags(
_: bool = Depends(_verify_internal_api_key),
store: RedisFeatureFlagStore = Depends(_get_feature_flag_store),
):
"""列出所有 Feature Flag。"""
try:
flags = store.list_all()
# 同时返回预定义的 flag(即使未设置也显示默认值)
result = []
for name in sorted(ALLOWED_FLAGS):
config = flags.get(name) or FeatureFlagConfig(name=name, enabled=False)
result.append(FeatureFlagResponse.from_config(config))
# 加上已存在但不在白名单中的 flag(只读展示)
for name, config in flags.items():
if name not in ALLOWED_FLAGS:
result.append(FeatureFlagResponse.from_config(config))
return sorted(result, key=lambda x: x.name)
except Exception as exc:
logger.error("Failed to list feature flags: %s", exc)
raise HTTPException(status_code=500, detail=f"Failed to list flags: {exc}")
@router.get("/{name}", response_model=FeatureFlagResponse)
async def get_feature_flag(
name: str,
_: bool = Depends(_verify_internal_api_key),
store: RedisFeatureFlagStore = Depends(_get_feature_flag_store),
):
"""获取单个 Feature Flag 配置。"""
try:
config = store.get(name)
return FeatureFlagResponse.from_config(config)
except Exception as exc:
logger.error("Failed to get feature flag %s: %s", name, exc)
raise HTTPException(status_code=500, detail=f"Failed to get flag: {exc}")
@router.get("/{name}/check", response_model=FeatureFlagCheckResponse)
async def check_feature_flag(
name: str,
identifier: Optional[str] = Query(None, description="标识符,如 user_id"),
_: bool = Depends(_verify_internal_api_key),
store: RedisFeatureFlagStore = Depends(_get_feature_flag_store),
):
"""检查某个标识符是否命中 Feature Flag。"""
try:
active = store.is_active(name, identifier=identifier)
return FeatureFlagCheckResponse(name=name, active=active, identifier=identifier)
except Exception as exc:
logger.error("Failed to check feature flag %s: %s", name, exc)
raise HTTPException(status_code=500, detail=f"Failed to check flag: {exc}")
@router.put("/{name}", response_model=FeatureFlagResponse)
async def update_feature_flag(
name: str,
request: FeatureFlagUpdateRequest,
_: bool = Depends(_verify_internal_api_key),
store: RedisFeatureFlagStore = Depends(_get_feature_flag_store),
):
"""更新 Feature Flag 配置。
只允许修改 ALLOWED_FLAGS 列表中的 flag。
"""
_validate_flag_name(name)
try:
config = FeatureFlagConfig(
name=name,
enabled=request.enabled,
percentage=request.percentage,
whitelist=set(request.whitelist),
)
store.set(config)
logger.info(
"Feature flag updated: name=%s enabled=%s percentage=%d whitelist=%d",
name,
config.enabled,
config.percentage,
len(config.whitelist),
)
return FeatureFlagResponse.from_config(config)
except Exception as exc:
logger.error("Failed to update feature flag %s: %s", name, exc)
raise HTTPException(status_code=500, detail=f"Failed to update flag: {exc}")
@router.delete("/{name}", status_code=status.HTTP_204_NO_CONTENT)
async def delete_feature_flag(
name: str,
_: bool = Depends(_verify_internal_api_key),
store: RedisFeatureFlagStore = Depends(_get_feature_flag_store),
):
"""删除 Feature Flag。
只允许删除 ALLOWED_FLAGS 列表中的 flag。
"""
_validate_flag_name(name)
try:
deleted = store.delete(name)
logger.info("Feature flag deleted: name=%s deleted=%s", name, deleted)
return None
except Exception as exc:
logger.error("Failed to delete feature flag %s: %s", name, exc)
raise HTTPException(status_code=500, detail=f"Failed to delete flag: {exc}")
+6 -10
View File
@@ -1,21 +1,17 @@
"""
视频处理模块
轻量工具(ffmpeg_utils / oss_helpers / dedup_helpers)顶层直接导出,
无额外依赖。渲染相关组件(UnifiedRenderService / RenderAdapter /
VideoProcessor 等)按需从子模块导入,避免 __init__ 阶段引入
packages / DB 等重依赖。
"""
# 共享工具模块(供 editing_modes / generation / edit_plan_generation 等复用)
# 共享工具模块(零外部依赖,供 editing_modes / generation / edit_plan_generation 等复用)
from . import dedup_helpers, ffmpeg_utils, oss_helpers
from .processor import VideoProcessor, VideoResult
from .unified_render_service import RenderResult, UnifiedRenderService
from .render_adapter import RenderAdapter, RenderAdapterResult
__all__ = [
"VideoProcessor",
"VideoResult",
"ffmpeg_utils",
"oss_helpers",
"dedup_helpers",
"UnifiedRenderService",
"RenderResult",
"RenderAdapter",
"RenderAdapterResult",
]
+7 -3
View File
@@ -93,12 +93,16 @@ class VideoFingerprint:
resolution: tuple[int, int]
def to_dict(self) -> dict:
# 注意:color_histograms 里的值可能是 np.float32(来自 cv2.normalize),
# 直接存进 dict 后 SQLAlchemy JSON 序列化会报 "float32 is not JSON serializable"。
# 这里统一转成 Python 原生 float。
native_histograms = [[float(v) for v in hist] for hist in self.color_histograms]
return {
"md5": self.md5,
"keyframe_phashes": self.keyframe_phashes,
"color_histograms": self.color_histograms,
"duration": self.duration,
"resolution": list(self.resolution),
"color_histograms": native_histograms,
"duration": float(self.duration),
"resolution": [int(self.resolution[0]), int(self.resolution[1])],
}
@@ -85,6 +85,41 @@ def run_ffmpeg(
raise
def probe_has_audio(local_path: str | Path) -> bool:
"""探测文件是否包含音频流。
Args:
local_path: 本地文件路径
Returns:
True 表示有音频流(或探测失败保守返回),False 表示确认无音频流
"""
try:
result = subprocess.run( # nosec B603
[
FFPROBE_BIN,
"-v",
"error",
"-select_streams",
"a:0",
"-show_entries",
"stream=codec_type",
"-of",
"default=noprint_wrappers=1:nokey=1",
str(local_path),
],
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=10,
)
return result.stdout.strip() == "audio"
except Exception:
# 探测失败保守返回 True,让 FFmpeg 自己处理(避免误删音频)
return True
def probe_duration(local_path: str | Path) -> float:
"""用 ffprobe 获取视频时长(秒)。
+23 -13
View File
@@ -20,20 +20,13 @@ from pathlib import Path
from typing import Any, Callable
from sqlalchemy.orm import Session
from video_processing.oss_helpers import download_asset, upload_to_oss
from video_processing.unified_render_service import RenderResult, UnifiedRenderService
from packages.adapters.sqlalchemy_impl.edit_plan_clip_repository import (
SQLAlchemyEditPlanClipRepository,
)
from packages.adapters.sqlalchemy_impl.edit_plan_repository import (
SQLAlchemyEditPlanRepository,
)
from packages.adapters.sqlalchemy_impl.edit_plan_clip_repository import SQLAlchemyEditPlanClipRepository
from packages.adapters.sqlalchemy_impl.edit_plan_repository import SQLAlchemyEditPlanRepository
from packages.domain.edit_plan import EditPlan, EditPlanStatus
from packages.domain.edit_plan_clip import EditPlanClip, EditPlanClipStatus
from video_processing.oss_helpers import download_asset, upload_to_oss
from video_processing.unified_render_service import (
RenderResult,
UnifiedRenderService,
)
logger = logging.getLogger(__name__)
@@ -140,7 +133,7 @@ class RenderAdapter:
)
logger.info(
"开始渲染: plan_id=%s job_id=%s ready_clips=%d",
"开始渲染: plan_id=%s job_id=%s ready_clips=%d engine=unified",
plan_id,
job_id,
len(ready_clips),
@@ -176,6 +169,18 @@ class RenderAdapter:
self._report_progress(progress_cb, 100.0, "渲染完成")
logger.info(
"[render-adapter] render success: plan_id=%s job_id=%s engine=unified "
"duration=%.2fs file_size=%d resolution=%dx%d clip_count=%d",
plan_id,
job_id,
result.duration,
result.file_size,
result.width,
result.height,
len(ready_clips),
)
return RenderAdapterResult(
success=True,
output_url=output_url or "",
@@ -188,7 +193,12 @@ class RenderAdapter:
)
except Exception as exc:
logger.exception("渲染失败: plan_id=%s", plan_id)
logger.exception(
"[render-adapter] render failed: plan_id=%s job_id=%s engine=unified error=%s",
plan_id,
job_id,
str(exc)[:200],
)
return RenderAdapterResult(
success=False,
error_message=str(exc)[:500],
+204
View File
@@ -0,0 +1,204 @@
"""渲染引擎 Feature Flag 解析器。
封装渲染引擎选择逻辑,支持:
- 环境变量作为默认值(RENDER_ENGINE=legacy/unified
- Redis Feature Flag 运行时覆盖(白名单 + 百分比 + 全局开关)
- 定时刷新,支持热更新不重启 worker
使用方式:
resolver = RenderEngineResolver(redis_url="redis://...", default_engine="legacy")
engine = resolver.get_engine(user_id="user123")
# engine: "legacy""unified"
"""
from __future__ import annotations
import logging
import threading
from typing import Optional
from packages.adapters.redis.feature_flag_store import (
FeatureFlagConfig,
FeatureFlagStore,
InMemoryFeatureFlagStore,
RedisFeatureFlagStore,
)
logger = logging.getLogger(__name__)
# Feature Flag 名称常量
FLAG_RENDER_ENGINE = "render_engine"
# 引擎常量
ENGINE_LEGACY = "legacy"
ENGINE_UNIFIED = "unified"
VALID_ENGINES = {ENGINE_LEGACY, ENGINE_UNIFIED}
class RenderEngineResolver:
"""渲染引擎选择器。
判定逻辑(从高到低):
1. Redis flag 白名单匹配 → unified
2. Redis flag 百分比命中 → unified
3. Redis flag 全局开启(100%)→ unified
4. 环境变量默认值 → legacy / unified
当 Redis 不可用时,自动降级到环境变量默认值,不影响业务。
"""
def __init__(
self,
default_engine: str = ENGINE_LEGACY,
redis_url: Optional[str] = None,
refresh_interval: float = 30.0,
store: Optional[FeatureFlagStore] = None,
) -> None:
"""
Args:
default_engine: 环境变量默认的引擎名(legacy / unified
redis_url: Redis 连接 URL,传 None 时使用内存实现(测试用)
refresh_interval: Redis flag 配置刷新间隔(秒)
store: 直接传入 store 实例(测试用,优先级高于 redis_url)
"""
self._default_engine = default_engine.lower() if default_engine else ENGINE_LEGACY
if self._default_engine not in VALID_ENGINES:
logger.warning(
"Invalid default engine '%s', fallback to '%s'",
self._default_engine,
ENGINE_LEGACY,
)
self._default_engine = ENGINE_LEGACY
if store is not None:
self._store = store
elif redis_url:
self._store = RedisFeatureFlagStore(redis_url=redis_url)
else:
self._store = InMemoryFeatureFlagStore()
logger.info("No Redis configured, using in-memory feature flag store")
self._refresh_interval = refresh_interval
self._lock = threading.Lock()
self._cached_config: Optional[FeatureFlagConfig] = None
self._last_refresh: float = 0.0
def _maybe_refresh(self) -> None:
"""惰性刷新配置,超过刷新间隔时从存储重新读取。"""
import time
now = time.time()
if now - self._last_refresh < self._refresh_interval:
return
try:
config = self._store.get(FLAG_RENDER_ENGINE)
with self._lock:
self._cached_config = config
self._last_refresh = now
except Exception as exc:
logger.warning("Failed to refresh render engine flag: %s", exc)
# 刷新失败时保留旧缓存,不中断业务
if self._cached_config is None:
# 首次就读失败,设一个默认值
with self._lock:
self._cached_config = FeatureFlagConfig(name=FLAG_RENDER_ENGINE)
self._last_refresh = now
def _get_config(self) -> FeatureFlagConfig:
"""获取当前 flag 配置(带缓存)。"""
if self._cached_config is None:
self._maybe_refresh()
else:
self._maybe_refresh()
return self._cached_config or FeatureFlagConfig(name=FLAG_RENDER_ENGINE)
def get_engine(self, user_id: Optional[str] = None) -> str:
"""获取当前应该使用的渲染引擎。
Args:
user_id: 用户ID,用于白名单匹配和百分比哈希。
传 None 时只看全局开关。
Returns:
"legacy""unified"
"""
config = self._get_config()
# 全局关闭 → 用默认值
if not config.enabled:
return self._default_engine
# 白名单匹配 / 百分比命中 → unified
if config.is_active(user_id):
return ENGINE_UNIFIED
# 未命中灰度 → 用默认值
return self._default_engine
def should_use_unified(self, user_id: Optional[str] = None) -> bool:
"""便捷方法:是否应该使用统一渲染引擎。"""
return self.get_engine(user_id) == ENGINE_UNIFIED
def force_refresh(self) -> None:
"""强制立即刷新配置(用于管理接口修改后立即生效)。"""
self._last_refresh = 0.0
if isinstance(self._store, RedisFeatureFlagStore):
self._store.invalidate_cache(FLAG_RENDER_ENGINE)
self._maybe_refresh()
def get_config_snapshot(self) -> dict:
"""获取当前配置快照(用于管理接口展示)。"""
config = self._get_config()
return {
"flag_name": FLAG_RENDER_ENGINE,
"default_engine": self._default_engine,
"enabled": config.enabled,
"percentage": config.percentage,
"whitelist": sorted(config.whitelist),
"refresh_interval": self._refresh_interval,
"last_refresh": self._last_refresh,
}
def set_flag(self, config: FeatureFlagConfig) -> None:
"""设置 flag 配置(管理接口用)。"""
config.name = FLAG_RENDER_ENGINE
self._store.set(config)
self.force_refresh()
# 全局单例
_resolver: Optional[RenderEngineResolver] = None
_resolver_lock = threading.Lock()
def get_render_engine_resolver() -> RenderEngineResolver:
"""获取全局单例(基于 worker 配置)。"""
global _resolver
if _resolver is not None:
return _resolver
with _resolver_lock:
if _resolver is not None:
return _resolver
try:
from worker_app.core.config import get_settings
settings = get_settings()
redis_url = getattr(settings, "redis_url", None) or getattr(settings, "broker_url", None)
default = getattr(settings, "render_engine", ENGINE_LEGACY)
_resolver = RenderEngineResolver(
default_engine=default,
redis_url=redis_url,
)
logger.info(
"RenderEngineResolver initialized: default=%s, redis=%s",
default,
bool(redis_url),
)
except Exception as exc:
logger.warning("Failed to init RenderEngineResolver from settings: %s", exc)
_resolver = RenderEngineResolver(default_engine=ENGINE_LEGACY)
return _resolver
@@ -24,6 +24,7 @@ from __future__ import annotations
import logging
import os
import subprocess
import time
from dataclasses import dataclass, field
from pathlib import Path
from typing import Any
@@ -356,6 +357,8 @@ def _resolve_layer_role(clip_type: str, config: dict[str, Any]) -> str:
# main type
if role == "b_roll":
return "broll"
if role == "audio":
return "audio"
return "main"
@@ -415,9 +418,16 @@ class UnifiedRenderService:
1. 视频主渲染(直通或完整链路)
2. 如有 title/subtitle,叠加 ASS 字幕
音频后处理:
1. 主图层音频 concat 拼接
2. 独立音频轨 amix 混入
3. 合并到输出视频
Raises:
ValueError: 没有可渲染的片段时抛出
"""
t_start = time.time()
# 1. 解析 clips → ResolvedClips(跳过无素材的 clip
resolved = self._resolve_clips()
if not resolved:
@@ -432,18 +442,88 @@ class UnifiedRenderService:
# 4. 生成 ASS 字幕文件(如果有 title/subtitle 配置)
ass_path = self._maybe_generate_ass(video_duration)
output_path = self.work_dir / f"rendered_{self.plan.id}.mp4"
# 灰度埋点:开始渲染
layer_roles = [layer.role for layer in layers]
clip_counts = {layer.role: len(layer.clips) for layer in layers}
logger.info(
"[unified-render] start render: plan_id=%s clip_count=%d layers=%s clip_counts=%s",
self.plan.id,
len(resolved),
layer_roles,
clip_counts,
)
# 5. 视频主渲染
if self._can_use_pass_through(layers):
self._render_pass_through(layers, output_path, ass_path=ass_path)
t_video_start = time.time()
video_only_path = self.work_dir / f"rendered_{self.plan.id}_video.mp4"
output_path = self.work_dir / f"rendered_{self.plan.id}.mp4"
is_pass_through = self._can_use_pass_through(layers)
pass_through_has_audio = False
if is_pass_through:
# 直通优化:单clip场景一次FFmpeg同时处理视频+音频,省去提取+合并两次调用
pass_through_has_audio = self._render_pass_through(
layers, output_path, ass_path=ass_path, video_duration=video_duration
)
else:
filter_complex, input_args = self._build_filter_complex(layers, ass_path=ass_path)
self._execute_ffmpeg(filter_complex, input_args, output_path)
self._execute_ffmpeg(filter_complex, input_args, video_only_path)
# 6. 探测输出
t_video_end = time.time()
video_render_ms = int((t_video_end - t_video_start) * 1000)
logger.info(
"[unified-render] video render done: plan_id=%s duration_ms=%d pass_through=%s",
self.plan.id,
video_render_ms,
is_pass_through,
)
# 6. 音频后处理混音(直通场景已合并处理,跳过)
t_audio_start = time.time()
audio_mix_ms = 0
has_audio = False
if is_pass_through:
# 直通场景已在一次调用中完成视频+音频
has_audio = pass_through_has_audio
else:
audio_path = self._mix_audio(layers, video_duration)
t_audio_end = time.time()
audio_mix_ms = int((t_audio_end - t_audio_start) * 1000)
has_audio = audio_path is not None
if has_audio:
logger.info(
"[unified-render] audio mix done: plan_id=%s duration_ms=%d",
self.plan.id,
audio_mix_ms,
)
# 7. 合并音视频
self._merge_audio_video(video_only_path, audio_path, output_path)
else:
# 无音频,直接用无声视频
import shutil
shutil.copy2(video_only_path, output_path)
# 8. 探测输出
duration, file_size, width, height = self._probe_output(output_path)
t_total = int((time.time() - t_start) * 1000)
logger.info(
"[unified-render] render done: plan_id=%s total_ms=%d video_ms=%d audio_ms=%d "
"output_duration=%.2fs output_size=%d resolution=%dx%d has_audio=%s",
self.plan.id,
t_total,
video_render_ms,
audio_mix_ms if has_audio else 0,
duration,
file_size,
width,
height,
has_audio,
)
return RenderResult(
output_path=output_path,
duration=duration,
@@ -470,14 +550,7 @@ class UnifiedRenderService:
if not main_layer or not main_layer.clips:
return 0.0
total = sum(
(
min(c.duration, c.actual_duration)
if c.duration > 0 and c.actual_duration > 0
else (c.duration if c.duration > 0 else c.actual_duration)
)
for c in main_layer.clips
)
total = sum(UnifiedRenderService._clip_effective_duration(c) for c in main_layer.clips)
# 减去转场重叠时间(粗略估算)
n_clips = len(main_layer.clips)
@@ -547,30 +620,36 @@ class UnifiedRenderService:
return True
def _render_pass_through(
self, layers: list[RenderLayer], output_path: Path, *, ass_path: Path | None = None
) -> None:
"""单图层单 clip 直通渲染(使用 -vf 而非 -filter_complex)。
self,
layers: list[RenderLayer],
output_path: Path,
*,
ass_path: Path | None = None,
video_duration: float = 0.0,
) -> bool:
"""单图层单 clip 直通渲染(使用 -vf 而非 -filter_complex),一次性输出带音频的最终视频。
性能优化:避免 filter_complex 的解析和调度开销,
对于一镜到底场景性能提升 ~30%,接近链路A水平。
性能优化:
- 避免 filter_complex 的解析和调度开销,单clip场景性能提升 ~30%
- 视频+音频一次FFmpeg调用完成,省去后续音频提取+音视频合并两次调用
Args:
layers: 图层列表(只有1个图层1个clip)
output_path: 输出文件路径
ass_path: ASS 字幕文件路径,有则叠加字幕
video_duration: 视频总时长(用于截断音频,0表示不额外截断)
Returns:
True 表示输出包含音频(近似判断,实际以输出文件为准)
"""
clip = layers[0].clips[0]
role = layers[0].role
# 构建滤镜链(与 _build_filter_complex 中预处理逻辑一致)
# 构建视频滤镜链(与 _build_filter_complex 中预处理逻辑一致)
filters: list[str] = []
# trim
effective_duration = 0.0
if clip.duration > 0:
effective_duration = min(clip.duration, clip.actual_duration) if clip.actual_duration > 0 else clip.duration
elif clip.actual_duration > 0:
effective_duration = clip.actual_duration
effective_duration = UnifiedRenderService._clip_effective_duration(clip)
if effective_duration > 0:
filters.append(f"trim=duration={effective_duration}")
@@ -592,12 +671,16 @@ class UnifiedRenderService:
# 字幕叠加
if ass_path is not None:
# ASS 文件路径需要转义:Windows 反斜杠转正斜杠,冒号转义
ass_filter_path = str(ass_path).replace("\\", "/").replace(":", "\\:")
filters.append(f"subtitles='{ass_filter_path}'")
vf_str = ",".join(filters)
# 最终输出时长:取 clip 有效时长和 video_duration 的较小值
final_duration = effective_duration
if video_duration > 0 and (final_duration <= 0 or final_duration > video_duration):
final_duration = video_duration
command = [
FFMPEG_BIN,
"-y",
@@ -615,16 +698,27 @@ class UnifiedRenderService:
"yuv420p",
"-movflags",
"+faststart",
"-an", # 直通模式暂不处理音频,音频统一在后续混音阶段处理
str(output_path),
]
# 音频处理:background 通常是图片无音频,跳过;其他编码为 aac
# background 以外的视频素材,默认带音频
has_audio = role != "background"
if has_audio:
command.extend(["-c:a", "aac", "-b:a", "128k"])
# 统一截断时长(同时作用于视频和音频)
if final_duration > 0:
command.extend(["-t", f"{final_duration:.3f}"])
command.append(str(output_path))
logger.info(
"直通渲染: plan_id=%s clip=%s role=%s duration=%.2fs",
"直通渲染: plan_id=%s clip=%s role=%s duration=%.2fs has_audio=%s",
self.plan.id,
clip.clip_id,
role,
effective_duration,
has_audio,
)
try:
run_ffmpeg(command)
@@ -638,6 +732,8 @@ class UnifiedRenderService:
)
raise
return has_audio
# ── 内部方法 ──────────────────────────────────────────────────────────────
def _resolve_clips(self) -> list[ResolvedClip]:
@@ -759,14 +855,7 @@ class UnifiedRenderService:
filters: list[str] = []
# trim — 始终将输出截断到有效时长,防止 xfade offset 与实际时长不匹配
# 有效时长 = min(指定时长, 实际时长);若均未设置则跳过
effective_duration = 0.0
if clip.duration > 0:
effective_duration = (
min(clip.duration, clip.actual_duration) if clip.actual_duration > 0 else clip.duration
)
elif clip.actual_duration > 0:
effective_duration = clip.actual_duration
effective_duration = UnifiedRenderService._clip_effective_duration(clip)
if effective_duration > 0:
filters.append(f"trim=duration={effective_duration}")
@@ -803,14 +892,7 @@ class UnifiedRenderService:
layer_clip_indices = [all_clips.index(c) for c in layer.clips]
layer_labels = [preprocessed_labels[i] for i in layer_clip_indices]
# 使用 trim 后的有效时长,与 Step 1 的 trim=duration 保持一致
layer_durations = []
for i in layer_clip_indices:
c = all_clips[i]
if c.duration > 0:
eff = min(c.duration, c.actual_duration) if c.actual_duration > 0 else c.duration
else:
eff = c.actual_duration if c.actual_duration > 0 else 0.0
layer_durations.append(eff)
layer_durations = [UnifiedRenderService._clip_effective_duration(all_clips[i]) for i in layer_clip_indices]
layer_transitions = [all_clips[i].transition_effect for i in layer_clip_indices]
if len(layer_labels) == 1:
@@ -930,7 +1012,7 @@ class UnifiedRenderService:
raise
def _probe_output(self, output_path: Path) -> tuple[float, int, int, int]:
"""探测输出文件的时长、大小、宽高
"""探测输出文件的时长、大小、宽高.
Returns:
(duration, file_size, width, height)
@@ -943,3 +1025,304 @@ class UnifiedRenderService:
info["width"],
info["height"],
)
# ── 音频后处理 ────────────────────────────────────────────────────────
def _mix_audio(self, layers: list[RenderLayer], video_duration: float) -> Path | None:
"""音频后处理混音.
处理逻辑:
1. 主音频源按优先级查找:main > brollbackground 不参与主音频,通常是图片无音轨)
2. 主图层音频按顺序 concat 拼接
3. 独立音频轨(audio role)用 amix 混入
4. 输出时长截断到 video_duration
5. 无音频流的 clip 会被自动跳过,避免 FFmpeg 引用 [i:a] 失败
Args:
layers: 图层列表
video_duration: 视频总时长(用于截断音频)
Returns:
混音后的音频文件路径,无音频时返回 None
"""
# 按优先级精确查找主音频图层:main > broll
# background 不参与主音频(通常是静态图片,无音轨)
layer_map = {layer.role: layer for layer in layers}
main_layer = None
for role in ("main", "broll"):
if role in layer_map and layer_map[role].clips:
main_layer = layer_map[role]
break
main_clips: list[ResolvedClip] = main_layer.clips if main_layer else []
# 没有主视频图层时兜底:检查 overlay/corner_voice 层是否有带音频的素材
if not main_clips:
for role in ("overlay", "corner_voice"):
if role in layer_map and layer_map[role].clips:
main_clips = layer_map[role].clips
break
# 收集独立音频轨
audio_clips: list[ResolvedClip] = []
if "audio" in layer_map:
audio_clips = layer_map["audio"].clips
# ── 防御:过滤掉无音频流的 clip ──
# 源视频可能没有音频流(如静音视频、纯图片转的视频),直接引用 [i:a] 会导致 FFmpeg 失败
main_clips = [c for c in main_clips if self._clip_has_audio(c)]
audio_clips = [c for c in audio_clips if self._clip_has_audio(c)]
if not main_clips and not audio_clips:
return None
# 构建音频处理命令
output_path = self.work_dir / f"audio_{self.plan.id}.aac"
# 简单场景:只有主图层 + 无独立音频 → 直接从视频提取音频并拼接
if main_clips and not audio_clips:
self._concat_main_audio(main_clips, output_path, video_duration)
return output_path
# 有独立音频轨 → amix 混音
self._mix_with_independent_audio(main_clips, audio_clips, output_path, video_duration)
return output_path
def _concat_main_audio(self, clips: list[ResolvedClip], output_path: Path, video_duration: float) -> None:
"""主图层音频 concat 拼接(对齐链路A行为).
每个 clip 提取音频 → trim → 按顺序 concat。
"""
if len(clips) == 1:
# 单 clip,直接提取音频,截断到 min(clip有效时长, 视频总时长)
clip = clips[0]
effective_duration = self._clip_effective_duration(clip)
# 最终时长:取 clip 有效时长和视频总时长的较小值
# (视频总时长由主图层决定,但单 clip 场景下两者应该一致,仍做保护)
final_duration = effective_duration
if video_duration > 0 and (final_duration <= 0 or final_duration > video_duration):
final_duration = video_duration
command = [
FFMPEG_BIN,
"-y",
"-i",
str(clip.local_path),
"-vn",
"-acodec",
"aac",
"-b:a",
"128k",
]
if final_duration > 0:
command.extend(["-t", f"{final_duration:.3f}"])
command.append(str(output_path))
run_ffmpeg(command)
return
# 多 clip,用 filter_complex concat
input_args: list[str] = []
filter_parts: list[str] = []
for i, clip in enumerate(clips):
input_args.extend(["-i", str(clip.local_path)])
effective_duration = self._clip_effective_duration(clip)
if effective_duration > 0:
filter_parts.append(f"[{i}:a]atrim=0:{effective_duration:.3f},asetpts=PTS-STARTPTS[a{i}]")
else:
filter_parts.append(f"[{i}:a]asetpts=PTS-STARTPTS[a{i}]")
audio_labels = "".join(f"[a{i}]" for i in range(len(clips)))
filter_parts.append(f"{audio_labels}concat=n={len(clips)}:v=0:a=1[outa]")
# 截断到视频总时长
if video_duration > 0:
filter_parts.append(f"[outa]atrim=0:{video_duration:.3f}[final_audio]")
final_label = "final_audio"
else:
final_label = "outa"
filter_complex = ";".join(filter_parts)
command = [
FFMPEG_BIN,
"-y",
*input_args,
"-filter_complex",
filter_complex,
"-map",
f"[{final_label}]",
"-acodec",
"aac",
"-b:a",
"128k",
str(output_path),
]
run_ffmpeg(command)
def _mix_with_independent_audio(
self,
main_clips: list[ResolvedClip],
audio_clips: list[ResolvedClip],
output_path: Path,
video_duration: float,
) -> None:
"""主音频 + 独立音频轨 amix 混音.
Args:
main_clips: 主视频 clips(提取音频后 concat
audio_clips: 独立音频轨 clips
output_path: 输出路径
video_duration: 视频总时长
"""
input_args: list[str] = []
filter_parts: list[str] = []
mix_labels: list[str] = []
input_idx = 0
# 1. 主图层音频 concat
if main_clips:
for clip in main_clips:
input_args.extend(["-i", str(clip.local_path)])
effective_duration = self._clip_effective_duration(clip)
if effective_duration > 0:
filter_parts.append(
f"[{input_idx}:a]atrim=0:{effective_duration:.3f},asetpts=PTS-STARTPTS[ma{input_idx}]"
)
else:
filter_parts.append(f"[{input_idx}:a]asetpts=PTS-STARTPTS[ma{input_idx}]")
input_idx += 1
if len(main_clips) == 1:
mix_labels.append("ma0")
else:
main_labels = "".join(f"[ma{i}]" for i in range(len(main_clips)))
filter_parts.append(f"{main_labels}concat=n={len(main_clips)}:v=0:a=1[main_audio]")
mix_labels.append("main_audio")
# 2. 独立音频轨
for j, clip in enumerate(audio_clips):
input_args.extend(["-i", str(clip.local_path)])
effective_duration = self._clip_effective_duration(clip)
volume = clip.config.get("volume", 1.0) if clip.config else 1.0
label = f"ia{j}"
filters = []
if effective_duration > 0:
filters.append(f"atrim=0:{effective_duration:.3f}")
filters.append("asetpts=PTS-STARTPTS")
if volume != 1.0:
filters.append(f"volume={volume}")
filter_parts.append(f"[{input_idx}:a]{','.join(filters)}[{label}]")
mix_labels.append(label)
input_idx += 1
# 3. amix 混音
mix_inputs = "".join(f"[{label}]" for label in mix_labels)
n_inputs = len(mix_labels)
# normalized=0 保持音量,duration=shortest 取最短
filter_parts.append(f"{mix_inputs}amix=inputs={n_inputs}:duration=longest:normalize=0[mixed_audio]")
# 4. 截断到视频时长
if video_duration > 0:
filter_parts.append(f"[mixed_audio]atrim=0:{video_duration:.3f}[final_audio]")
final_label = "final_audio"
else:
final_label = "mixed_audio"
filter_complex = ";".join(filter_parts)
command = [
FFMPEG_BIN,
"-y",
*input_args,
"-filter_complex",
filter_complex,
"-map",
f"[{final_label}]",
"-acodec",
"aac",
"-b:a",
"128k",
str(output_path),
]
logger.info(
"音频混音: plan_id=%s main_clips=%d audio_clips=%d",
self.plan.id,
len(main_clips),
len(audio_clips),
)
try:
run_ffmpeg(command)
except subprocess.CalledProcessError as e:
logger.error(
"音频混音失败: plan_id=%s exit_code=%d\nfilter_complex:\n%s",
self.plan.id,
e.returncode,
filter_complex[:3000],
)
raise
def _merge_audio_video(self, video_path: Path, audio_path: Path, output_path: Path) -> None:
"""将音频合并到视频中(视频流拷贝,音频直接复用).
Args:
video_path: 无声视频路径
audio_path: 音频文件路径
output_path: 输出文件路径
"""
command = [
FFMPEG_BIN,
"-y",
"-i",
str(video_path),
"-i",
str(audio_path),
"-c:v",
"copy",
"-c:a",
"aac",
"-b:a",
"128k",
"-map",
"0:v:0",
"-map",
"1:a:0",
"-shortest",
"-movflags",
"+faststart",
str(output_path),
]
logger.info("合并音视频: plan_id=%s", self.plan.id)
try:
run_ffmpeg(command)
except subprocess.CalledProcessError as e:
logger.error(
"合并音视频失败: plan_id=%s exit_code=%d",
self.plan.id,
e.returncode,
)
raise
@staticmethod
def _clip_effective_duration(clip: ResolvedClip) -> float:
"""计算 clip 的有效时长."""
if clip.duration > 0:
return min(clip.duration, clip.actual_duration) if clip.actual_duration > 0 else clip.duration
return clip.actual_duration if clip.actual_duration > 0 else 0.0
def _clip_has_audio(self, clip: ResolvedClip) -> bool:
"""探测 clip 是否有音频流(带缓存).
避免同一个 clip 被多次 ffprobe 探测。
"""
if not hasattr(self, "_audio_cache"):
self._audio_cache: dict[str, bool] = {}
key = str(clip.local_path)
if key not in self._audio_cache:
from .ffmpeg_utils import probe_has_audio
self._audio_cache[key] = probe_has_audio(clip.local_path)
return self._audio_cache[key]
+11 -4
View File
@@ -61,10 +61,12 @@ def compose_video(self, job_id: str, **kwargs):
return {"status": "error", "message": "Missing plan_id"}
# 判断使用哪个渲染引擎
from worker_app.core.config import get_settings as get_worker_settings
# 优先级:Redis Feature Flag(白名单 > 百分比) > 环境变量默认
from video_processing.render_engine_resolver import get_render_engine_resolver
worker_settings = get_worker_settings()
engine = (worker_settings.render_engine or "legacy").lower()
resolver = get_render_engine_resolver()
user_id = job.created_by_user_id or None
engine = resolver.get_engine(user_id=user_id)
if engine == "unified":
return _compose_with_unified_engine(self, job_service, job, plan_id, db)
@@ -207,7 +209,12 @@ def _compose_with_unified_engine(task, job_service, job, plan_id: str, db) -> di
}
job_service.complete_job(job_id, result=result_data)
logger.info("视频合成完成(unified): job_id=%s plan_id=%s duration=%.2fs", job_id, plan_id, result.duration)
logger.info(
"视频合成完成(unified): job_id=%s plan_id=%s duration=%.2fs",
job_id,
plan_id,
result.duration,
)
return {"status": "completed", "job_id": job_id, "result": result_data}
@@ -116,6 +116,15 @@ def render_edit_plan(self, plan_id: str) -> dict:
rendered_clip_ids: list[str] = []
failed_clip_ids: list[str] = []
# 预先批量查询所有素材的 storage_keyfile_url
from packages.adapters.sqlalchemy_impl.models import AssetModel
clip_asset_ids = [c.asset_id for c in clips if c.asset_id]
asset_storage_map: dict[str, str] = {}
if clip_asset_ids:
assets = db.query(AssetModel).filter(AssetModel.id.in_(clip_asset_ids)).all()
asset_storage_map = {a.id: a.file_url for a in assets if a.file_url}
for clip in clips:
if not clip.asset_id:
# 没有素材的片段跳过,标记为失败
@@ -129,10 +138,22 @@ def render_edit_plan(self, plan_id: str) -> dict:
rendered_clip_ids.append(clip.id)
continue
storage_key = asset_storage_map.get(clip.asset_id)
if not storage_key:
logger.warning(
"片段素材无 storage_key,跳过: clip_id=%s asset_id=%s",
clip.id,
clip.asset_id,
)
clip.mark_failed()
clip_repo.update(clip)
failed_clip_ids.append(clip.id)
continue
# 下载素材
ext = Path(clip.asset_id).suffix or ".mp4"
ext = Path(storage_key).suffix or ".mp4"
local_path = tmpdir_path / f"clip_{clip.order:04d}{ext}"
if download_asset(clip.asset_id, local_path):
if download_asset(storage_key, local_path):
asset_path_map[clip.asset_id] = local_path
rendered_clip_ids.append(clip.id)
else:
+7 -1
View File
@@ -162,13 +162,15 @@ def _build_plan_and_clips_from_task(
"""
plan = _VirtualPlan(id=task_id, name=f"Generated-{task_id[:8]}")
# 为每个下载路径生成合成 asset_id
# 为每个下载路径生成合成 asset_id,并预探测素材时长
asset_path_map: dict[str, Path] = {}
path_to_asset_id: dict[Path, str] = {}
path_duration: dict[Path, float] = {}
for i, p in enumerate(downloaded_paths):
asset_id = f"gen_{task_id[:8]}_{i:03d}{p.suffix or '.mp4'}"
asset_path_map[asset_id] = p
path_to_asset_id[p] = asset_id
path_duration[p] = probe_duration(p)
clips: list[_VirtualClip] = []
n = len(downloaded_paths)
@@ -184,6 +186,7 @@ def _build_plan_and_clips_from_task(
clip_type=clip_type,
order=i,
asset_id=path_to_asset_id[p],
duration=path_duration[p],
)
)
elif mode == "voice_over":
@@ -196,6 +199,7 @@ def _build_plan_and_clips_from_task(
clip_type="main",
order=i,
asset_id=path_to_asset_id[p],
duration=path_duration[p],
config={"role": "b_roll"},
)
)
@@ -215,6 +219,7 @@ def _build_plan_and_clips_from_task(
clip_type=clip_type,
order=i,
asset_id=path_to_asset_id[p],
duration=path_duration[p],
)
)
else:
@@ -227,6 +232,7 @@ def _build_plan_and_clips_from_task(
clip_type="main",
order=i,
asset_id=path_to_asset_id[p],
duration=path_duration[p],
)
)
+16 -1
View File
@@ -1,3 +1,9 @@
from packages.adapters.redis.feature_flag_store import (
FeatureFlagConfig,
FeatureFlagStore,
InMemoryFeatureFlagStore,
RedisFeatureFlagStore,
)
from packages.adapters.redis.session_store import (
NoopSessionStore,
RedisConfig,
@@ -5,4 +11,13 @@ from packages.adapters.redis.session_store import (
get_session_store,
)
__all__ = ["NoopSessionStore", "RedisConfig", "SessionStore", "get_session_store"]
__all__ = [
"FeatureFlagConfig",
"FeatureFlagStore",
"InMemoryFeatureFlagStore",
"NoopSessionStore",
"RedisConfig",
"RedisFeatureFlagStore",
"SessionStore",
"get_session_store",
]
+259
View File
@@ -0,0 +1,259 @@
"""Feature Flag 存储实现。
支持两种后端:
- RedisFeatureFlagStore:生产环境使用,支持多实例共享、热更新
- InMemoryFeatureFlagStore:测试/开发环境使用,纯内存
支持的 Flag 类型:
- 全局开关(enabled: bool
- 白名单(whitelist: Set[str],如 user_id 列表)
- 百分比切流(percentage: 0-100,基于标识符哈希取模)
判定优先级:白名单 > 百分比 > 全局开关
"""
from __future__ import annotations
import hashlib
import json
import logging
import threading
import time
from abc import ABC, abstractmethod
from dataclasses import dataclass, field
from typing import Optional, Set
logger = logging.getLogger(__name__)
# Redis key 前缀
FEATURE_FLAG_REDIS_PREFIX = "feature_flag:"
@dataclass
class FeatureFlagConfig:
"""单个 Feature Flag 的配置。"""
name: str
enabled: bool = False
percentage: int = 0 # 0-100
whitelist: Set[str] = field(default_factory=set)
def to_dict(self) -> dict:
return {
"name": self.name,
"enabled": self.enabled,
"percentage": self.percentage,
"whitelist": sorted(self.whitelist),
}
@classmethod
def from_dict(cls, data: dict) -> "FeatureFlagConfig":
return cls(
name=data["name"],
enabled=bool(data.get("enabled", False)),
percentage=int(data.get("percentage", 0)),
whitelist=set(data.get("whitelist", [])),
)
def is_active(self, identifier: Optional[str] = None) -> bool:
"""判断当前 flag 是否激活。
判定优先级:
1. 全局关闭 → False
2. 白名单匹配 → True
3. 百分比命中 → True
4. 其他 → False
Args:
identifier: 用于白名单匹配和百分比哈希的标识符(如 user_id)。
传 None 时只看全局开关 + 百分比(百分比用随机值)。
"""
if not self.enabled:
return False
# 白名单:精确匹配
if identifier and identifier in self.whitelist:
return True
# 百分比:0 直接 False100 直接 True
if self.percentage <= 0:
# 没有白名单且百分比为0 → 未启用
return False
if self.percentage >= 100:
return True
# 基于 identifier 做哈希取模,确保同一用户始终落在同一侧
if identifier:
hash_val = int(
hashlib.md5(f"{self.name}:{identifier}".encode("utf-8")).hexdigest(), 16 # nosec B324
) # nosec B324 - 用于哈希取模做百分比切流,非安全用途
return (hash_val % 100) < self.percentage
# 无 identifier 且百分比在 0-100 之间 → 按比例随机(不保证一致性)
import random
return random.randint(0, 99) < self.percentage
class FeatureFlagStore(ABC):
"""Feature Flag 存储抽象接口。"""
@abstractmethod
def get(self, name: str) -> FeatureFlagConfig:
"""获取指定 flag 的配置,不存在则返回默认配置(关闭状态)。"""
...
@abstractmethod
def set(self, config: FeatureFlagConfig) -> None:
"""设置 flag 配置。"""
...
@abstractmethod
def delete(self, name: str) -> bool:
"""删除 flag,返回是否成功删除。"""
...
@abstractmethod
def list_all(self) -> dict[str, FeatureFlagConfig]:
"""列出所有 flag。"""
...
def is_active(self, name: str, identifier: Optional[str] = None) -> bool:
"""便捷方法:判断 flag 是否激活。"""
return self.get(name).is_active(identifier)
class InMemoryFeatureFlagStore(FeatureFlagStore):
"""内存实现,用于测试和本地开发。"""
def __init__(self) -> None:
self._flags: dict[str, FeatureFlagConfig] = {}
self._lock = threading.Lock()
def get(self, name: str) -> FeatureFlagConfig:
with self._lock:
return self._flags.get(name, FeatureFlagConfig(name=name, enabled=False))
def set(self, config: FeatureFlagConfig) -> None:
with self._lock:
self._flags[config.name] = config
def delete(self, name: str) -> bool:
with self._lock:
if name in self._flags:
del self._flags[name]
return True
return False
def list_all(self) -> dict[str, FeatureFlagConfig]:
with self._lock:
return dict(self._flags)
class RedisFeatureFlagStore(FeatureFlagStore):
"""Redis 实现,支持多实例共享配置。
每个 flag 存在一个独立的 Redis hash key 中:
Key: feature_flag:{name}
Fields: enabled, percentage, whitelist(JSON array)
"""
def __init__(self, redis_url: str, key_prefix: str = FEATURE_FLAG_REDIS_PREFIX) -> None:
import redis as redis_lib
self._redis = redis_lib.from_url(redis_url, decode_responses=True)
self._key_prefix = key_prefix
# 本地缓存 + TTL,减少 Redis 调用
self._cache: dict[str, tuple[FeatureFlagConfig, float]] = {}
self._cache_ttl = 5.0 # 秒,默认5秒本地缓存
self._lock = threading.Lock()
def _redis_key(self, name: str) -> str:
return f"{self._key_prefix}{name}"
def _parse_whitelist(self, raw: Optional[str]) -> Set[str]:
if not raw:
return set()
try:
data = json.loads(raw)
return set(data) if isinstance(data, list) else set()
except (json.JSONDecodeError, TypeError):
return set()
def get(self, name: str) -> FeatureFlagConfig:
now = time.time()
# 先查本地缓存
with self._lock:
cached = self._cache.get(name)
if cached and now - cached[1] < self._cache_ttl:
return cached[0]
# 从 Redis 读取
try:
key = self._redis_key(name)
data = self._redis.hgetall(key)
if not data:
config = FeatureFlagConfig(name=name, enabled=False)
else:
config = FeatureFlagConfig(
name=name,
enabled=(data.get("enabled", "0") in ("1", "true", "True")),
percentage=int(data.get("percentage", 0)),
whitelist=self._parse_whitelist(data.get("whitelist")),
)
# 写入本地缓存
with self._lock:
self._cache[name] = (config, now)
return config
except Exception as exc:
logger.warning("Failed to get feature flag %s from Redis: %s", name, exc)
# Redis 不可用时返回默认值(关闭),不影响业务
return FeatureFlagConfig(name=name, enabled=False)
def set(self, config: FeatureFlagConfig) -> None:
key = self._redis_key(config.name)
self._redis.hset(
key,
mapping={
"enabled": "1" if config.enabled else "0",
"percentage": str(config.percentage),
"whitelist": json.dumps(sorted(config.whitelist), ensure_ascii=False),
},
)
# 失效本地缓存
with self._lock:
self._cache.pop(config.name, None)
def delete(self, name: str) -> bool:
key = self._redis_key(name)
result = self._redis.delete(key)
with self._lock:
self._cache.pop(name, None)
return bool(result)
def list_all(self) -> dict[str, FeatureFlagConfig]:
pattern = f"{self._key_prefix}*"
result: dict[str, FeatureFlagConfig] = {}
try:
cursor = 0
while True:
cursor, keys = self._redis.scan(cursor=cursor, match=pattern, count=100)
for key in keys:
name = key[len(self._key_prefix) :]
result[name] = self.get(name)
if cursor == 0:
break
except Exception as exc:
logger.warning("Failed to list feature flags from Redis: %s", exc)
return result
def invalidate_cache(self, name: Optional[str] = None) -> None:
"""手动失效本地缓存。"""
with self._lock:
if name:
self._cache.pop(name, None)
else:
self._cache.clear()
+136
View File
@@ -0,0 +1,136 @@
# 灰度对比测试工具
用于统一渲染引擎灰度发布期间的新旧引擎对比验证。
## 能力
- **像素对比**:基于 FFmpeg SSIM + PSNR 双指标,评估视频画质差异
- **音频对比**:基于差值音频 RMS,评估音频波形差异
- **批量对比**10个预设场景覆盖 P0/P1/P2 优先级
- **HTML 报告**:可视化对比结果,包含画质、音频、性能三维度
- **两种切换方式**:支持 engine 参数直传 或 Feature Flag 白名单切换
## 目录结构
```
tests/render_compare/
├── __init__.py # 包导出
├── README.md # 本文档
├── video_diff.py # 视频像素对比(SSIM + PSNR
├── audio_diff.py # 音频对比(差值 RMS)
├── scenarios.py # 预定义对比场景(10个)
└── runner.py # 批量对比执行器 + HTML 报告生成
```
## 快速开始
### 环境要求
- FFmpeg 4.4+(需带 ssim 和 psnr 滤镜)
- Python 3.10+
- httpxAPI 调用)
### 配置环境变量
```bash
export STAGING_API_URL=https://api.staging.example.com
export STAGING_API_KEY=your_api_key
export STAGING_INTERNAL_API_KEY=your_internal_key # 可选,Feature Flag 模式需要
```
### 运行对比
```bash
# 运行所有 P0 场景(最核心的5个)
python -m tests.render_compare.runner --priority P0 --output ./report/
# 运行 P0 + P1 场景
python -m tests.render_compare.runner --priority P1 --output ./report/
# 只跑指定场景
python -m tests.render_compare.runner --scenarios simple_pass_through,subtitle_rendering
# 使用 Feature Flag 方式切换引擎(需要 internal key
python -m tests.render_compare.runner --priority P0 --flag-mode
# 自定义阈值
python -m tests.render_compare.runner --priority P0 --ssim-threshold 0.95 --psnr-threshold 30
```
## 对比场景
| ID | 名称 | 优先级 | 验证点 |
|----|------|--------|--------|
| simple_pass_through | 简单直通 | P0 | 直通优化路径正确性 |
| multi_clip_transition | 多clip转场 | P0 | 转场效果 + concat |
| subtitle_rendering | 字幕渲染 | P0 | ASS字幕渲染 |
| independent_audio_track | 独立音频轨 | P0 | 音频混音(amix |
| no_audio_video | 无音轨视频 | P0 | 无音轨防御逻辑 |
| picture_in_picture | 画中画 | P1 | overlay 图层 |
| multi_layer_mix | 多图层混合 | P1 | 多图层复杂场景 |
| image_background | 图片背景 | P1 | background 层 + 无音频 |
| long_video_stress | 长视频压力 | P2 | 多clip性能 |
| vertical_portrait | 竖屏9:16 | P2 | scale 策略(铺满裁剪) |
## 验收标准(建议)
### 视频质量
- **平均 SSIM >= 0.90**:通过(有微小差异但视觉可接受)
- **平均 SSIM >= 0.95**:优秀(视觉几乎无差异)
- **平均 PSNR >= 25 dB**:通过
- **分辨率一致 + 时长差 < 0.1s**:通过
### 音频质量
- **相似度 >= 0.85**:通过
- **采样率/声道数一致**:通过
### 性能
- **平均性能差异在 ±10% 以内**:可接受
- **直通场景新引擎更快**(预期 +30%)
## API 约定
Runner 默认假设渲染 API 支持以下接口:
### 提交任务
```
POST /api/v1/render/compose
Authorization: Bearer {api_key}
Body: { ...plan_payload, "engine": "legacy" | "unified" }
Response: { "task_id": "xxx" }
```
### 查询状态
```
GET /api/v1/tasks/{task_id}
Response: { "status": "completed", "output_url": "...", "duration_sec": 5.2 }
```
### Feature Flagflag-mode
```
PUT /api/v1/internal/feature-flags/render_engine
X-API-Key: {internal_key}
Body: { "enabled": true, "percentage": 100 }
```
如果你的 API 接口不同,请修改 `StagingAPI` 类中的对应方法。
## 故障排查
### 对比失败定位指南
1. **像素差异大(SSIM < 0.90**
- 检查分辨率是否一致
- 检查帧率是否一致
-`save_diff_frame` 生成差异帧可视化
- 检查转场效果(slideup/slidedown 是新引擎独有)
2. **音频不一致**
- 检查音频编码参数(码率、采样率)
- 检查主音频源优先级(main > broll
- 用 ffprobe 对比两视频音频流参数
3. **渲染失败**
- 检查日志:`[unified-render] render failed`
- 检查素材是否完整下载
- 检查 FFmpeg 命令是否正确
+26
View File
@@ -0,0 +1,26 @@
"""灰度对比测试工具包.
用于新旧渲染引擎的批量对比测试,包含:
- video_diff: 视频像素对比(SSIM + PSNR
- audio_diff: 音频对比(差值 RMS
- scenarios: 预定义对比场景
- runner: 批量对比执行器 + HTML 报告
"""
from .audio_diff import AudioDiffResult, compute_audio_diff, extract_audio, probe_duration, probe_has_audio
from .scenarios import SCENARIOS, CompareScenario, get_scenarios_by_priority
from .video_diff import VideoDiffResult, compute_video_diff, save_diff_frame
__all__ = [
"VideoDiffResult",
"compute_video_diff",
"save_diff_frame",
"AudioDiffResult",
"compute_audio_diff",
"extract_audio",
"probe_has_audio",
"probe_duration",
"SCENARIOS",
"CompareScenario",
"get_scenarios_by_priority",
]
+322
View File
@@ -0,0 +1,322 @@
"""音频对比工具 — 基于 FFmpeg 的音频质量对比.
使用以下指标评估两段音频的相似度:
1. 波形差异(RMS 差值)
2. 频谱相似度(FFT 分帧比较)
3. 时长差异
对比方式:
- 直接对两个音频做 `ametadata=select='gt(scene\\,0.3)'` 过于复杂
- 简化方案:用 `amerge` + `astats` 计算差值音频的 RMS
更精确的方案(已实现):
- 将两轨音频做差(amix=0:weights='1 -1' → 实际上用 pan 更简单)
- 对差值音频做 astats,获取差值的 RMS、峰值等指标
"""
from __future__ import annotations
import json
import re
import shutil
import subprocess # nosec B404
from dataclasses import asdict, dataclass
from pathlib import Path
from typing import Any
FFMPEG_BIN: str = shutil.which("ffmpeg") or "ffmpeg"
FFPROBE_BIN: str = shutil.which("ffprobe") or "ffprobe"
@dataclass
class AudioDiffResult:
"""音频对比结果."""
audio_a: str
audio_b: str
duration_a: float
duration_b: float
duration_diff: float
sample_rate_match: bool
channels_match: bool
diff_rms_db: float # 差值音频的 RMS(dB,越低越相似)
diff_peak_db: float # 差值音频的峰值(dB,越低越相似)
similarity_score: float # 综合相似度评分 [0, 1],1 = 完全一致
passed: bool
def to_dict(self) -> dict[str, Any]:
return asdict(self)
def probe_duration(file_path: str) -> float:
"""探测文件时长(秒),失败返回 0."""
try:
result = subprocess.run( # nosec B603
[
FFPROBE_BIN,
"-v",
"error",
"-show_entries",
"format=duration",
"-of",
"default=noprint_wrappers=1:nokey=1",
str(file_path),
],
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=10,
)
return round(float(result.stdout.strip()), 3)
except Exception:
return 0.0
def probe_has_audio(file_path: str | Path) -> bool:
"""探测文件是否包含音频流."""
try:
result = subprocess.run( # nosec B603
[
FFPROBE_BIN,
"-v",
"error",
"-select_streams",
"a:0",
"-show_entries",
"stream=codec_type",
"-of",
"default=noprint_wrappers=1:nokey=1",
str(file_path),
],
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=10,
)
return result.stdout.strip() == "audio"
except Exception:
return False # 探测失败保守返回 False,避免误判有音频
def compute_audio_diff(
audio_a: str | Path,
audio_b: str | Path,
*,
similarity_threshold: float = 0.90,
duration_tolerance: float = 0.1,
) -> AudioDiffResult:
"""计算两段音频的差异.
方案:用 pan 滤镜将两轨相减,对差值音频做 astats 分析。
Args:
audio_a: 音频A(基线)
audio_b: 音频B(对比)
similarity_threshold: 相似度合格阈值
duration_tolerance: 时长容忍度(秒)
Returns:
AudioDiffResult 对比结果
"""
dur_a = probe_duration(str(audio_a))
dur_b = probe_duration(str(audio_b))
duration_diff = abs(dur_a - dur_b)
# 获取音频元信息
info_a = _probe_audio_info(str(audio_a))
info_b = _probe_audio_info(str(audio_b))
sample_rate_match = info_a["sample_rate"] == info_b["sample_rate"]
channels_match = info_a["channels"] == info_b["channels"]
# 相减后分析差值
# 取较短时长做对比
min_dur = min(dur_a, dur_b)
if min_dur <= 0:
return AudioDiffResult(
audio_a=str(audio_a),
audio_b=str(audio_b),
duration_a=dur_a,
duration_b=dur_b,
duration_diff=duration_diff,
sample_rate_match=sample_rate_match,
channels_match=channels_match,
diff_rms_db=-999.0,
diff_peak_db=-999.0,
similarity_score=0.0,
passed=False,
)
# 做差值音频:a - b
# 注意:amix 会自动按输入数归一化音量(除以N),
# 所以 a + (-1)*b 经过 amix=inputs=2 后整体音量会减半(-6dB)。
# 加 volume=2 补偿回来,确保差值 RMS 反映真实差异幅度。
command = [
FFMPEG_BIN,
"-i",
str(audio_a),
"-i",
str(audio_b),
"-filter_complex",
# 第2轨反相 → amix混合 → volume=2补偿amix的自动缩放
"[1:a]volume=-1[inv];[0:a][inv]amix=inputs=2:duration=shortest:dropout_transition=0,volume=2[diff]",
"-map",
"[diff]",
"-f",
"null",
"-af",
"astats=metadata=1:reset=0",
"-",
]
try:
result = subprocess.run( # nosec B603
command,
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=120,
)
stderr = result.stderr or ""
except subprocess.CalledProcessError as e:
# 如果音频格式不兼容,返回失败
return AudioDiffResult(
audio_a=str(audio_a),
audio_b=str(audio_b),
duration_a=dur_a,
duration_b=dur_b,
duration_diff=duration_diff,
sample_rate_match=sample_rate_match,
channels_match=channels_match,
diff_rms_db=999.0,
diff_peak_db=999.0,
similarity_score=0.0,
passed=False,
)
diff_rms_db, diff_peak_db = _parse_astats(stderr)
# 相似度评分:基于差值 RMS
# 差值 RMS -60dB → 相似度 ~1.0(几乎无声差)
# 差值 RMS -20dB → 相似度 ~0.5(有明显差异)
# 差值 RMS 0dB → 相似度 ~0.0(完全相反)
if diff_rms_db <= -60:
similarity_score = 1.0
elif diff_rms_db >= 0:
similarity_score = 0.0
else:
# 线性映射:-60dB → 1.0, 0dB → 0.0
similarity_score = max(0.0, min(1.0, 1.0 + diff_rms_db / 60.0))
passed = (
duration_diff <= duration_tolerance
and sample_rate_match
and channels_match
and similarity_score >= similarity_threshold
)
return AudioDiffResult(
audio_a=str(audio_a),
audio_b=str(audio_b),
duration_a=round(dur_a, 3),
duration_b=round(dur_b, 3),
duration_diff=round(duration_diff, 3),
sample_rate_match=sample_rate_match,
channels_match=channels_match,
diff_rms_db=round(diff_rms_db, 2),
diff_peak_db=round(diff_peak_db, 2),
similarity_score=round(similarity_score, 4),
passed=passed,
)
def _probe_audio_info(file_path: str) -> dict[str, int]:
"""探测音频元信息."""
try:
result = subprocess.run( # nosec B603
[
FFPROBE_BIN,
"-v",
"error",
"-select_streams",
"a:0",
"-show_entries",
"stream=sample_rate,channels",
"-of",
"json",
file_path,
],
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=10,
)
info = json.loads(result.stdout)
stream = info.get("streams", [{}])[0]
return {
"sample_rate": int(stream.get("sample_rate", 44100)),
"channels": int(stream.get("channels", 2)),
}
except Exception:
return {"sample_rate": 0, "channels": 0}
def _parse_astats(stderr: str) -> tuple[float, float]:
"""从 astats 输出中解析 RMS 和峰值.
astats 输出格式(在 stderr 中):
[Parsed_astats_1 @ 0x...] Channel: 1
[Parsed_astats_1 @ 0x...] ...
[Parsed_astats_1 @ 0x...] Overall
[Parsed_astats_1 @ 0x...] DC offset: 0.000000
[Parsed_astats_1 @ 0x...] Min level: -0.123456
[Parsed_astats_1 @ 0x...] Max level: 0.789012
[Parsed_astats_1 @ 0x...] Peak level dB: -2.01
[Parsed_astats_1 @ 0x...] RMS level dB: -10.56
...
"""
lines = stderr.split("\n")
rms_db = -999.0
peak_db = -999.0
for line in lines:
# 找 Overall 部分的统计(双声道时取整体值)
rms_match = re.search(r"RMS level dB:\s*(-?\d+\.?\d*)", line)
peak_match = re.search(r"Peak level dB:\s*(-?\d+\.?\d*)", line)
if rms_match:
rms_db = float(rms_match.group(1))
if peak_match:
peak_db = float(peak_match.group(1))
return rms_db, peak_db
def extract_audio(video_path: str | Path, output_path: str | Path) -> Path:
"""从视频中提取音频(AAC 格式).
Args:
video_path: 视频文件路径
output_path: 输出音频路径
Returns:
输出音频文件路径
"""
command = [
FFMPEG_BIN,
"-y",
"-i",
str(video_path),
"-vn",
"-acodec",
"aac",
"-b:a",
"128k",
str(output_path),
]
subprocess.run(command, check=True, capture_output=True, timeout=120) # nosec B603
return Path(output_path)
+628
View File
@@ -0,0 +1,628 @@
"""灰度对比测试 Runner — 新旧引擎批量对比 + 报告生成.
使用方法:
# 配置环境变量
export STAGING_API_URL=https://api.staging.example.com
export STAGING_API_KEY=your_key
# 运行全部 P0 场景
python -m tests.render_compare.runner --priority P0 --output ./report/
# 只跑指定场景
python -m tests.render_compare.runner --scenario simple_pass_through,subtitle_rendering
对比流程:
1. 对每个场景,分别提交到 legacy 和 unified 引擎(通过 Feature Flag 白名单/百分比控制)
- 方式A:通过内部 API 临时切换 flag(需要 admin key
- 方式B:提交任务时指定 engine 参数(如果 API 支持)
2. 等待任务完成,下载输出视频
3. 像素对比(SSIM + PSNR+ 音频对比(差值RMS)
4. 生成 HTML 对比报告
注意:默认假设 API 支持 `engine` 参数来指定渲染引擎。
如果不支持,需要先通过内部 API 切换 Feature Flag,然后提交任务。
"""
from __future__ import annotations
import argparse
import json
import os
import sys
import time
from dataclasses import dataclass, field
from datetime import datetime
from pathlib import Path
from typing import Any
import httpx
# 确保项目根目录在 path 中
sys.path.insert(0, str(Path(__file__).parent.parent.parent))
from .audio_diff import AudioDiffResult, compute_audio_diff
from .scenarios import SCENARIOS, CompareScenario, get_scenarios_by_priority
from .video_diff import VideoDiffResult, compute_video_diff
@dataclass
class ScenarioResult:
"""单个场景的对比结果."""
scenario: CompareScenario
legacy_task_id: str = ""
unified_task_id: str = ""
legacy_video_path: str = ""
unified_video_path: str = ""
legacy_duration_sec: float = 0.0
unified_duration_sec: float = 0.0
video_diff: VideoDiffResult | None = None
audio_diff: AudioDiffResult | None = None
legacy_success: bool = False
unified_success: bool = False
error: str = ""
@property
def passed(self) -> bool:
if not (self.legacy_success and self.unified_success):
return False
if self.video_diff and not self.video_diff.passed:
return False
if self.audio_diff and not self.audio_diff.passed:
return False
return True
class StagingAPI:
"""Staging 环境 API 客户端."""
def __init__(self, base_url: str, api_key: str, internal_api_key: str = ""):
self.base_url = base_url.rstrip("/")
self.api_key = api_key
self.internal_api_key = internal_api_key
self.client = httpx.Client(timeout=30.0)
def _headers(self, internal: bool = False) -> dict[str, str]:
headers = {"Authorization": f"Bearer {self.api_key}"}
if internal and self.internal_api_key:
headers["X-API-Key"] = self.internal_api_key
return headers
def submit_render_task(self, plan_payload: dict[str, Any], engine: str = "") -> str:
"""提交渲染任务,返回 task_id.
Args:
plan_payload: EditPlan payload
engine: 可选,指定引擎("legacy" / "unified"
Returns:
task_id
"""
url = f"{self.base_url}/api/v1/render/compose"
payload = dict(plan_payload)
if engine:
payload["engine"] = engine
resp = self.client.post(url, json=payload, headers=self._headers())
resp.raise_for_status()
data = resp.json()
return data.get("task_id") or data.get("id", "")
def get_task_status(self, task_id: str) -> dict[str, Any]:
"""获取任务状态."""
url = f"{self.base_url}/api/v1/tasks/{task_id}"
resp = self.client.get(url, headers=self._headers())
resp.raise_for_status()
return resp.json()
def wait_for_task(self, task_id: str, timeout: float = 300.0, poll_interval: float = 3.0) -> dict[str, Any]:
"""等待任务完成.
Returns:
最终任务状态
Raises:
TimeoutError: 超时
"""
start = time.time()
while time.time() - start < timeout:
status = self.get_task_status(task_id)
state = status.get("status", "")
if state in ("completed", "success", "done", "failed", "error"):
return status
time.sleep(poll_interval)
raise TimeoutError(f"Task {task_id} timed out after {timeout}s")
def set_feature_flag(self, flag_name: str, enabled: bool, percentage: int = 0, whitelist: list[str] | None = None):
"""通过内部 API 设置 Feature Flag.
用于不支持 engine 参数的场景,切换全局灰度比例。
"""
if not self.internal_api_key:
raise ValueError("internal_api_key is required for feature flag operations")
url = f"{self.base_url}/api/v1/internal/feature-flags/{flag_name}"
body: dict[str, Any] = {"enabled": enabled, "percentage": percentage}
if whitelist is not None:
body["whitelist"] = whitelist
resp = self.client.put(url, json=body, headers=self._headers(internal=True))
resp.raise_for_status()
return resp.json()
def get_feature_flag(self, flag_name: str) -> dict[str, Any]:
"""获取 Feature Flag 配置."""
if not self.internal_api_key:
raise ValueError("internal_api_key is required")
url = f"{self.base_url}/api/v1/internal/feature-flags/{flag_name}"
resp = self.client.get(url, headers=self._headers(internal=True))
resp.raise_for_status()
return resp.json()
def download_video(self, video_url: str, output_path: str | Path) -> Path:
"""下载视频文件."""
output_path = Path(output_path)
output_path.parent.mkdir(parents=True, exist_ok=True)
with self.client.stream("GET", video_url, timeout=60.0) as resp:
resp.raise_for_status()
with open(output_path, "wb") as f:
for chunk in resp.iter_bytes():
f.write(chunk)
return output_path
class CompareRunner:
"""新旧引擎对比 Runner."""
# 全局默认阈值(唯一真实来源,所有入口统一引用)
DEFAULT_SSIM_THRESHOLD: float = 0.95
DEFAULT_PSNR_THRESHOLD: float = 28.0
DEFAULT_AUDIO_SIMILARITY_THRESHOLD: float = 0.90
DEFAULT_DURATION_TOLERANCE: float = 0.1
DEFAULT_TASK_TIMEOUT: float = 300.0
def __init__(
self,
api: StagingAPI,
output_dir: Path,
*,
ssim_threshold: float | None = None,
psnr_threshold: float | None = None,
audio_similarity_threshold: float | None = None,
task_timeout: float | None = None,
flag_mode: bool = False, # 是否使用 Feature Flag 方式切换引擎
duration_tolerance: float | None = None,
):
self.api = api
self.output_dir = output_dir
self.ssim_threshold = ssim_threshold if ssim_threshold is not None else self.DEFAULT_SSIM_THRESHOLD
self.psnr_threshold = psnr_threshold if psnr_threshold is not None else self.DEFAULT_PSNR_THRESHOLD
self.audio_similarity_threshold = (
audio_similarity_threshold
if audio_similarity_threshold is not None
else self.DEFAULT_AUDIO_SIMILARITY_THRESHOLD
)
self.duration_tolerance = (
duration_tolerance if duration_tolerance is not None else self.DEFAULT_DURATION_TOLERANCE
)
self.task_timeout = task_timeout if task_timeout is not None else self.DEFAULT_TASK_TIMEOUT
self.flag_mode = flag_mode
self.results: list[ScenarioResult] = []
# flag_mode 下保存原始配置,测试结束后恢复(防污染线上)
self._original_flag_config: dict[str, Any] | None = None
def run_scenario(self, scenario: CompareScenario) -> ScenarioResult:
"""运行单个场景对比."""
print(f"\n{'='*60}")
print(f"[{scenario.priority}] {scenario.id}: {scenario.name}")
print(f" {scenario.description}")
result = ScenarioResult(scenario=scenario)
scenario_dir = self.output_dir / scenario.id
scenario_dir.mkdir(parents=True, exist_ok=True)
try:
# 1. 提交两个引擎的任务
legacy_task_id = self._submit_with_engine(scenario, "legacy")
unified_task_id = self._submit_with_engine(scenario, "unified")
result.legacy_task_id = legacy_task_id
result.unified_task_id = unified_task_id
print(f" legacy task: {legacy_task_id}")
print(f" unified task: {unified_task_id}")
# 2. 等待完成
print(" waiting for legacy...", end="", flush=True)
legacy_status = self.api.wait_for_task(legacy_task_id, timeout=self.task_timeout)
result.legacy_success = legacy_status.get("status") in ("completed", "success", "done")
legacy_video_url = legacy_status.get("output_url", "") or legacy_status.get("video_url", "")
print(f" {'' if result.legacy_success else ''} ({legacy_status.get('duration_sec', '?')}s)")
print(" waiting for unified...", end="", flush=True)
unified_status = self.api.wait_for_task(unified_task_id, timeout=self.task_timeout)
result.unified_success = unified_status.get("status") in ("completed", "success", "done")
unified_video_url = unified_status.get("output_url", "") or unified_status.get("video_url", "")
print(f" {'' if result.unified_success else ''} ({unified_status.get('duration_sec', '?')}s)")
result.legacy_duration_sec = float(legacy_status.get("duration_sec", 0))
result.unified_duration_sec = float(unified_status.get("duration_sec", 0))
if not (result.legacy_success and result.unified_success):
result.error = f"Legacy success={result.legacy_success}, Unified success={result.unified_success}"
print(" ⚠️ 任务未全部成功,跳过对比")
return result
# 3. 下载视频
print(" downloading...", end="", flush=True)
legacy_path = self.api.download_video(legacy_video_url, scenario_dir / "legacy.mp4")
unified_path = self.api.download_video(unified_video_url, scenario_dir / "unified.mp4")
result.legacy_video_path = str(legacy_path)
result.unified_video_path = str(unified_path)
print("")
# 4. 像素对比
print(" computing video diff...", end="", flush=True)
result.video_diff = compute_video_diff(
legacy_path,
unified_path,
ssim_threshold=self.ssim_threshold,
psnr_threshold=self.psnr_threshold,
duration_tolerance=self.duration_tolerance,
)
print(
f" SSIM={result.video_diff.avg_ssim:.4f} PSNR={result.video_diff.avg_psnr:.2f}dB {'' if result.video_diff.passed else ''}"
)
# 5. 音频对比(仅当都有音频时)
from .audio_diff import probe_has_audio
legacy_has_audio = probe_has_audio(legacy_path)
unified_has_audio = probe_has_audio(unified_path)
if legacy_has_audio and unified_has_audio:
print(" computing audio diff...", end="", flush=True)
result.audio_diff = compute_audio_diff(
legacy_path,
unified_path,
similarity_threshold=self.audio_similarity_threshold,
)
print(
f" similarity={result.audio_diff.similarity_score:.4f} {'' if result.audio_diff.passed else ''}"
)
elif legacy_has_audio != unified_has_audio:
result.error = f"音频不一致: legacy_has_audio={legacy_has_audio}, unified_has_audio={unified_has_audio}"
print(f" ⚠️ 音频不一致: legacy={legacy_has_audio}, unified={unified_has_audio}")
else:
print(" audio: both silent (skip)")
except Exception as e:
result.error = str(e)
print(f" ❌ 错误: {e}")
self.results.append(result)
return result
def _submit_with_engine(self, scenario: CompareScenario, engine: str) -> str:
"""提交指定引擎的任务.
如果 flag_mode=True,通过 Feature Flag 切换,否则通过 engine 参数。
"""
if self.flag_mode:
# 先设置 flag(用白名单方式,确保只有当前测试用户命中)
percentage = 0 if engine == "legacy" else 100
self.api.set_feature_flag("render_engine", enabled=True, percentage=percentage)
time.sleep(1) # 给 worker 一点时间刷新配置
return self.api.submit_render_task(scenario.plan_payload)
else:
return self.api.submit_render_task(scenario.plan_payload, engine=engine)
def run_all(self, scenarios: list[CompareScenario]) -> list[ScenarioResult]:
"""运行所有场景.
flag_mode=True 时,测试开始前保存原始 Feature Flag 配置,
结束后(无论成功失败)自动恢复,避免污染线上环境。
"""
print(f"\n灰度对比测试开始 - {len(scenarios)} 个场景")
print(f"输出目录: {self.output_dir}")
print(f"视频阈值: SSIM>={self.ssim_threshold}, PSNR>={self.psnr_threshold}dB")
print(f"音频阈值: similarity>={self.audio_similarity_threshold}")
# flag_mode:保存原始配置,测试结束后恢复(防污染)
if self.flag_mode:
try:
self._original_flag_config = self.api.get_feature_flag("render_engine")
print(f" [flag_mode] 已保存原始配置: {self._original_flag_config}")
except Exception as e:
print(f" ⚠️ [flag_mode] 保存原始配置失败: {e}")
print(" 为避免污染线上,将中止测试。请检查 internal_api_key 配置。")
return self.results
try:
for i, scenario in enumerate(scenarios):
print(f"\n进度: {i+1}/{len(scenarios)}")
self.run_scenario(scenario)
finally:
# 始终恢复原始 flag 配置
if self.flag_mode and self._original_flag_config:
try:
orig = self._original_flag_config
self.api.set_feature_flag(
"render_engine",
enabled=orig.get("enabled", False),
percentage=orig.get("percentage", 0),
whitelist=orig.get("whitelist"),
)
print("\n[flag_mode] ✅ 已恢复原始 Feature Flag 配置")
except Exception as e:
print(f"\n[flag_mode] ❌ 恢复 Feature Flag 失败: {e}")
print(" 请手动检查并恢复 render_engine flag 配置!")
return self.results
def summary(self) -> dict[str, Any]:
"""生成汇总统计."""
total = len(self.results)
passed = sum(1 for r in self.results if r.passed)
failed = total - passed
# 性能对比
perf_diffs = []
for r in self.results:
if r.legacy_success and r.unified_success and r.legacy_duration_sec > 0:
diff_pct = (r.unified_duration_sec - r.legacy_duration_sec) / r.legacy_duration_sec * 100
perf_diffs.append(diff_pct)
avg_perf_diff = sum(perf_diffs) / len(perf_diffs) if perf_diffs else 0.0
return {
"total": total,
"passed": passed,
"failed": failed,
"pass_rate": f"{passed/total*100:.1f}%" if total > 0 else "0%",
"avg_perf_diff_pct": round(avg_perf_diff, 2),
"scenarios": [self._result_to_dict(r) for r in self.results],
"timestamp": datetime.now().isoformat(),
"ssim_threshold": self.ssim_threshold,
"psnr_threshold": self.psnr_threshold,
"audio_threshold": self.audio_similarity_threshold,
}
def _result_to_dict(self, r: ScenarioResult) -> dict[str, Any]:
return {
"id": r.scenario.id,
"name": r.scenario.name,
"priority": r.scenario.priority,
"passed": r.passed,
"legacy_success": r.legacy_success,
"unified_success": r.unified_success,
"legacy_duration_sec": r.legacy_duration_sec,
"unified_duration_sec": r.unified_duration_sec,
"video_diff": r.video_diff.to_dict() if r.video_diff else None,
"audio_diff": r.audio_diff.to_dict() if r.audio_diff else None,
"error": r.error,
}
def generate_html_report(summary: dict[str, Any], output_path: Path):
"""生成 HTML 对比报告."""
scenarios = summary["scenarios"]
# 按通过/失败分组
passed_list = [s for s in scenarios if s["passed"]]
failed_list = [s for s in scenarios if not s["passed"]]
# 构建场景卡片
scenario_cards = ""
for s in scenarios:
status_class = "pass" if s["passed"] else "fail"
status_text = "✅ 通过" if s["passed"] else "❌ 失败"
vdiff = s.get("video_diff") or {}
adiff = s.get("audio_diff") or {}
video_info = ""
if vdiff:
video_info = f"""
<div class="metric-row">
<span>SSIM:</span>
<span class="{'good' if vdiff.get('avg_ssim', 0) >= 0.95 else 'warn'}">{vdiff.get('avg_ssim', 0):.4f}</span>
</div>
<div class="metric-row">
<span>PSNR:</span>
<span>{vdiff.get('avg_psnr', 0):.2f} dB</span>
</div>
<div class="metric-row">
<span>时长差:</span>
<span>{vdiff.get('duration_diff', 0):.3f}s</span>
</div>
"""
audio_info = ""
if adiff:
audio_info = f"""
<div class="metric-row">
<span>音频相似度:</span>
<span class="{'good' if adiff.get('similarity_score', 0) >= 0.9 else 'warn'}">{adiff.get('similarity_score', 0):.4f}</span>
</div>
<div class="metric-row">
<span>差值 RMS:</span>
<span>{adiff.get('diff_rms_db', 0):.2f} dB</span>
</div>
"""
perf_info = ""
if s["legacy_duration_sec"] and s["unified_duration_sec"]:
diff = s["unified_duration_sec"] - s["legacy_duration_sec"]
pct = diff / s["legacy_duration_sec"] * 100 if s["legacy_duration_sec"] else 0
trend = "🔴" if pct > 10 else ("🟡" if pct > 0 else "🟢")
perf_info = f"""
<div class="perf-row">
<span>Legacy: {s['legacy_duration_sec']:.2f}s</span>
<span>Unified: {s['unified_duration_sec']:.2f}s</span>
<span>{trend} {pct:+.1f}%</span>
</div>
"""
error_info = f'<div class="error-box">{s["error"]}</div>' if s["error"] else ""
scenario_cards += f"""
<div class="card {status_class}">
<div class="card-header">
<span class="badge">{s['priority']}</span>
<span class="scenario-name">{s['name']}</span>
<span class="status {status_class}">{status_text}</span>
</div>
<div class="card-body">
<div class="grid-2">
<div>
<h4>视频质量</h4>
{video_info or '<p class="muted">无数据</p>'}
</div>
<div>
<h4>音频质量</h4>
{audio_info or '<p class="muted">无音频或跳过</p>'}
</div>
</div>
<div>
<h4>性能对比</h4>
{perf_info or '<p class="muted">无数据</p>'}
</div>
{error_info}
</div>
</div>
"""
html = f"""<!DOCTYPE html>
<html lang="zh-CN">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>统一渲染引擎灰度对比报告</title>
<style>
* {{ box-sizing: border-box; margin: 0; padding: 0; }}
body {{ font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, sans-serif; background: #f5f5f5; color: #333; padding: 20px; }}
.container {{ max-width: 1200px; margin: 0 auto; }}
h1 {{ margin-bottom: 20px; font-size: 24px; }}
.summary {{ background: white; border-radius: 12px; padding: 24px; margin-bottom: 24px; display: flex; gap: 32px; flex-wrap: wrap; }}
.summary-item {{ text-align: center; }}
.summary-item .value {{ font-size: 32px; font-weight: bold; margin-bottom: 4px; }}
.summary-item .label {{ color: #666; font-size: 14px; }}
.pass .value {{ color: #10b981; }}
.fail .value {{ color: #ef4444; }}
.card {{ background: white; border-radius: 12px; margin-bottom: 16px; overflow: hidden; border-left: 4px solid #10b981; }}
.card.fail {{ border-left-color: #ef4444; }}
.card-header {{ padding: 16px 20px; background: #fafafa; display: flex; align-items: center; gap: 12px; border-bottom: 1px solid #eee; }}
.badge {{ background: #e5e7eb; color: #374151; padding: 2px 8px; border-radius: 4px; font-size: 12px; font-weight: 600; }}
.scenario-name {{ flex: 1; font-weight: 600; }}
.status {{ font-weight: 600; }}
.status.pass {{ color: #10b981; }}
.status.fail {{ color: #ef4444; }}
.card-body {{ padding: 20px; }}
.grid-2 {{ display: grid; grid-template-columns: 1fr 1fr; gap: 24px; margin-bottom: 16px; }}
h4 {{ margin-bottom: 12px; color: #374151; font-size: 14px; }}
.metric-row {{ display: flex; justify-content: space-between; padding: 6px 0; font-size: 14px; }}
.metric-row .good {{ color: #10b981; font-weight: 600; }}
.metric-row .warn {{ color: #f59e0b; font-weight: 600; }}
.perf-row {{ display: flex; gap: 24px; padding: 8px 0; font-size: 14px; background: #f9fafb; padding: 12px; border-radius: 8px; }}
.error-box {{ background: #fef2f2; color: #dc2626; padding: 12px; border-radius: 8px; margin-top: 12px; font-size: 13px; }}
.muted {{ color: #9ca3af; font-size: 14px; }}
.timestamp {{ text-align: center; color: #9ca3af; font-size: 12px; margin-top: 24px; }}
</style>
</head>
<body>
<div class="container">
<h1>🎬 统一渲染引擎灰度对比报告</h1>
<div class="summary">
<div class="summary-item">
<div class="value">{summary['total']}</div>
<div class="label">总场景数</div>
</div>
<div class="summary-item pass">
<div class="value">{summary['passed']}</div>
<div class="label">通过</div>
</div>
<div class="summary-item fail">
<div class="value">{summary['failed']}</div>
<div class="label">失败</div>
</div>
<div class="summary-item">
<div class="value">{summary['pass_rate']}</div>
<div class="label">通过率</div>
</div>
<div class="summary-item">
<div class="value {'good' if summary['avg_perf_diff_pct'] <= 0 else 'warn'}" style="font-size: 24px; color: {'#10b981' if summary['avg_perf_diff_pct'] <= 0 else '#f59e0b'}">{summary['avg_perf_diff_pct']:+.1f}%</div>
<div class="label">平均性能差异</div>
</div>
</div>
{scenario_cards}
<div class="timestamp">生成时间: {summary['timestamp']}</div>
</div>
</body>
</html>"""
output_path.parent.mkdir(parents=True, exist_ok=True)
output_path.write_text(html, encoding="utf-8")
return output_path
def main():
parser = argparse.ArgumentParser(description="统一渲染引擎灰度对比测试")
parser.add_argument("--priority", default="P0", choices=["P0", "P1", "P2"], help="最低优先级")
parser.add_argument("--scenarios", default="", help="指定场景ID,逗号分隔")
parser.add_argument("--output", default="./gray_compare_report", help="输出目录")
parser.add_argument("--ssim-threshold", type=float, default=None, help="SSIM阈值(默认0.95")
parser.add_argument("--psnr-threshold", type=float, default=None, help="PSNR阈值(dB)(默认28.0")
parser.add_argument("--audio-threshold", type=float, default=None, help="音频相似度阈值(默认0.90")
parser.add_argument("--flag-mode", action="store_true", help="使用Feature Flag方式切换引擎")
parser.add_argument("--task-timeout", type=float, default=300.0, help="单任务超时时间(秒)")
args = parser.parse_args()
base_url = os.environ.get("STAGING_API_URL", "")
api_key = os.environ.get("STAGING_API_KEY", "")
internal_key = os.environ.get("STAGING_INTERNAL_API_KEY", "")
if not base_url or not api_key:
print("❌ 请设置环境变量 STAGING_API_URL 和 STAGING_API_KEY")
sys.exit(1)
# 选择场景
if args.scenarios:
scenario_ids = [s.strip() for s in args.scenarios.split(",")]
selected = [s for s in SCENARIOS if s.id in scenario_ids]
if not selected:
print(f"❌ 未找到匹配的场景: {scenario_ids}")
print(f"可用场景: {[s.id for s in SCENARIOS]}")
sys.exit(1)
else:
selected = get_scenarios_by_priority(args.priority)
output_dir = Path(args.output).resolve()
output_dir.mkdir(parents=True, exist_ok=True)
api = StagingAPI(base_url, api_key, internal_key)
runner = CompareRunner(
api,
output_dir,
ssim_threshold=args.ssim_threshold,
psnr_threshold=args.psnr_threshold,
audio_similarity_threshold=args.audio_threshold,
flag_mode=args.flag_mode,
task_timeout=args.task_timeout,
)
runner.run_all(selected)
# 生成报告
summary = runner.summary()
# JSON 报告
json_path = output_dir / "report.json"
json_path.write_text(json.dumps(summary, indent=2, ensure_ascii=False), encoding="utf-8")
# HTML 报告
html_path = output_dir / "report.html"
generate_html_report(summary, html_path)
print(f"\n{'='*60}")
print(f"对比完成: {summary['passed']}/{summary['total']} 通过 ({summary['pass_rate']})")
print(f"报告: {html_path}")
print(f"JSON: {json_path}")
if __name__ == "__main__":
main()
+257
View File
@@ -0,0 +1,257 @@
"""灰度对比测试场景定义 — 覆盖典型渲染场景.
每个场景对应一个 EditPlan,用于新旧引擎对比。
覆盖场景:
1. 简单直通(单clip无特效)
2. 多clip转场(fade + slide
3. 画中画(main + overlay
4. 字幕渲染(ASS字幕)
5. 独立音频轨(主视频 + BGM
6. 多图层混合(main + broll + overlay + audio
7. 背景图片 + 主视频(图片背景无音频)
8. 无音频视频(纯画面,验证无音轨防御)
9. 长视频(10+ clip,压力测试)
10. 分辨率非标(竖屏9:16,验证scale策略)
"""
from __future__ import annotations
from dataclasses import dataclass, field
from typing import Any
@dataclass
class CompareScenario:
"""对比测试场景."""
id: str
name: str
description: str
priority: str # P0 / P1 / P2
plan_payload: dict[str, Any] # EditPlan JSON payload(提交给 API 的数据)
expected: dict[str, Any] = field(default_factory=dict) # 预期结果
SCENARIOS: list[CompareScenario] = [
CompareScenario(
id="simple_pass_through",
name="简单直通",
description="单主clip,无转场无特效,验证直通优化路径",
priority="P0",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 5.0,
"order": 0,
}
],
},
),
CompareScenario(
id="multi_clip_transition",
name="多clip转场",
description="3个clipfade + slideleft 转场",
priority="P0",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 3.0,
"order": 0,
"transition_effect": "cut",
},
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 3.0,
"order": 1,
"transition_effect": "fade",
},
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 3.0,
"order": 2,
"transition_effect": "slideleft",
},
],
},
),
CompareScenario(
id="picture_in_picture",
name="画中画",
description="主视频 + 角落小窗(corner_voice",
priority="P1",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{"clip_type": "main", "asset_id": "sample_5s.mp4", "duration": 5.0, "order": 0},
{"clip_type": "corner_voice", "asset_id": "sample_5s.mp4", "duration": 5.0, "order": 0},
],
},
),
CompareScenario(
id="subtitle_rendering",
name="字幕渲染",
description="主视频 + ASS字幕",
priority="P0",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 5.0,
"order": 0,
"config": {"subtitles": [{"text": "测试字幕 Test Subtitle", "start_time": 0, "end_time": 5.0}]},
}
],
},
),
CompareScenario(
id="independent_audio_track",
name="独立音频轨",
description="主视频(带音频)+ 独立BGM轨,验证音频混音",
priority="P0",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{"clip_type": "main", "asset_id": "sample_5s.mp4", "duration": 5.0, "order": 0},
{
"clip_type": "main",
"asset_id": "sample_bgm.mp3",
"duration": 5.0,
"order": 0,
"config": {"role": "audio", "volume": 0.5},
},
],
},
),
CompareScenario(
id="multi_layer_mix",
name="多图层混合",
description="main + broll + overlay + audio 四图层",
priority="P1",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 4.0,
"order": 0,
"transition_effect": "fade",
},
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 4.0,
"order": 1,
"transition_effect": "slideup",
},
{"clip_type": "broll", "asset_id": "sample_broll.mp4", "duration": 8.0, "order": 0},
{"clip_type": "overlay", "asset_id": "sample_overlay.png", "duration": 8.0, "order": 0},
{
"clip_type": "main",
"asset_id": "sample_bgm.mp3",
"duration": 8.0,
"order": 0,
"config": {"role": "audio", "volume": 0.3},
},
],
},
),
CompareScenario(
id="image_background",
name="图片背景",
description="background图片层 + 主视频,验证背景层无音频",
priority="P1",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{"clip_type": "background", "asset_id": "sample_bg.jpg", "duration": 5.0, "order": 0},
{"clip_type": "main", "asset_id": "sample_5s.mp4", "duration": 5.0, "order": 0},
],
},
),
CompareScenario(
id="no_audio_video",
name="无音轨视频",
description="源视频无音频流,验证无音轨防御逻辑",
priority="P0",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{"clip_type": "main", "asset_id": "sample_silent_5s.mp4", "duration": 5.0, "order": 0},
],
},
),
CompareScenario(
id="long_video_stress",
name="长视频压力",
description="10个clip + 多种转场,性能压力测试",
priority="P2",
plan_payload={
"width": 1280,
"height": 720,
"fps": 25,
"clips": [
{
"clip_type": "main",
"asset_id": "sample_5s.mp4",
"duration": 3.0,
"order": i,
"transition_effect": ["cut", "fade", "slideleft", "slidedown", "dissolve"][i % 5],
}
for i in range(10)
],
},
),
CompareScenario(
id="vertical_portrait",
name="竖屏9:16",
description="竖屏分辨率,验证scale策略(铺满裁剪)",
priority="P2",
plan_payload={
"width": 720,
"height": 1280,
"fps": 25,
"clips": [
{"clip_type": "main", "asset_id": "sample_5s.mp4", "duration": 5.0, "order": 0},
],
},
),
]
def get_scenarios_by_priority(min_priority: str = "P2") -> list[CompareScenario]:
"""按优先级过滤场景.
P0 包含 P0
P1 包含 P0 + P1
P2 包含全部
"""
priority_order = {"P0": 0, "P1": 1, "P2": 2}
threshold = priority_order.get(min_priority, 2)
return [s for s in SCENARIOS if priority_order.get(s.priority, 2) <= threshold]
+283
View File
@@ -0,0 +1,283 @@
"""视频对比工具 — 基于 FFmpeg 的像素级质量对比.
使用 SSIM + PSNR 双指标评估两个视频的相似度:
- SSIM (Structural Similarity): 结构相似性,范围 [0, 1],越接近 1 越相似
- PSNR (Peak Signal-to-Noise Ratio): 峰值信噪比,单位 dB,越高越好
灰度验收标准:
- 平均 SSIM >= 0.95 → 视觉上几乎无差异(P0 场景必达)
- 最低 SSIM >= 0.90 → 最严重帧差异可接受
- 平均 PSNR >= 28dB → 质量达标
"""
from __future__ import annotations
import json
import re
import shutil
import subprocess # nosec B404
from dataclasses import asdict, dataclass
from pathlib import Path
from typing import Any
FFMPEG_BIN: str = shutil.which("ffmpeg") or "ffmpeg"
FFPROBE_BIN: str = shutil.which("ffprobe") or "ffprobe"
@dataclass
class VideoDiffResult:
"""视频对比结果."""
video_a: str
video_b: str
width: int
height: int
duration_a: float
duration_b: float
avg_ssim: float
min_ssim: float
avg_psnr: float # dB
min_psnr: float
frame_count: int
duration_diff: float # 时长差(秒)
resolution_match: bool
passed: bool # 是否通过阈值
def to_dict(self) -> dict[str, Any]:
return asdict(self)
def probe_video_info(video_path: str) -> dict[str, Any]:
"""获取视频信息(宽、高、时长、fps."""
try:
result = subprocess.run( # nosec B603
[
FFPROBE_BIN,
"-v",
"error",
"-select_streams",
"v:0",
"-show_entries",
"stream=width,height,r_frame_rate,duration",
"-show_entries",
"format=duration",
"-of",
"json",
video_path,
],
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=10,
)
info = json.loads(result.stdout)
stream = info.get("streams", [{}])[0]
fmt = info.get("format", {})
width = int(stream.get("width", 1280))
height = int(stream.get("height", 720))
fps_str = stream.get("r_frame_rate", "25/1")
if "/" in fps_str:
num, den = fps_str.split("/")
fps = float(num) / float(den) if float(den) > 0 else 25.0
else:
fps = float(fps_str) if fps_str else 25.0
duration = float(fmt.get("duration", 0)) or float(stream.get("duration", 0))
return {"width": width, "height": height, "duration": duration, "fps": round(fps, 2)}
except Exception:
return {"width": 1280, "height": 720, "duration": 0.0, "fps": 25.0}
def compute_video_diff(
video_a: str | Path,
video_b: str | Path,
*,
ssim_threshold: float = 0.95,
psnr_threshold: float = 28.0,
duration_tolerance: float = 0.1,
) -> VideoDiffResult:
"""计算两个视频的像素差异.
使用 FFmpeg ssim + psnr 滤镜一次性计算两个指标。
Args:
video_a: 视频A路径(基线)
video_b: 视频B路径(对比)
ssim_threshold: SSIM 合格阈值(默认 0.90
psnr_threshold: PSNR 合格阈值(默认 25dB
duration_tolerance: 时长容忍度(秒,默认 0.1s)
Returns:
VideoDiffResult 对比结果
Raises:
subprocess.CalledProcessError: FFmpeg 执行失败
"""
info_a = probe_video_info(str(video_a))
info_b = probe_video_info(str(video_b))
duration_diff = abs(info_a["duration"] - info_b["duration"])
resolution_match = info_a["width"] == info_b["width"] and info_a["height"] == info_b["height"]
# ssim 和 psnr 的 stats_file 都输出到 stdout
# 用行格式区分:SSIM 行含 "All:"PSNR 行含 "psnr_avg:"
command = [
FFMPEG_BIN,
"-i",
str(video_a),
"-i",
str(video_b),
"-lavfi",
"[0:v][1:v]ssim=stats_file=-[out1];[0:v][1:v]psnr=stats_file=-[out2]",
"-f",
"null",
"-",
]
result = subprocess.run( # nosec B603
command,
check=True,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True,
timeout=300,
)
# 逐帧统计在 stdoutstats_file=-),汇总日志在 stderr
stats_stdout = result.stdout or ""
avg_ssim, min_ssim = _parse_ssim_stats(stats_stdout)
avg_psnr, min_psnr = _parse_psnr_stats(stats_stdout)
frame_count = _count_frames(result.stderr or "")
passed = (
resolution_match
and duration_diff <= duration_tolerance
and avg_ssim >= ssim_threshold
and avg_psnr >= psnr_threshold
)
return VideoDiffResult(
video_a=str(video_a),
video_b=str(video_b),
width=info_a["width"],
height=info_a["height"],
duration_a=round(info_a["duration"], 3),
duration_b=round(info_b["duration"], 3),
avg_ssim=round(avg_ssim, 6),
min_ssim=round(min_ssim, 6),
avg_psnr=round(avg_psnr, 3),
min_psnr=round(min_psnr, 3),
frame_count=frame_count,
duration_diff=round(duration_diff, 3),
resolution_match=resolution_match,
passed=passed,
)
def _parse_ssim_stats(stats_output: str) -> tuple[float, float]:
"""从 SSIM stats_file 输出中解析逐帧 SSIM.
FFmpeg ssim 滤镜 stats_file 输出格式(每行一帧):
n:1 Y:0.987654 U:0.991234 V:0.990000 All:0.989000 (19.585642)
n:2 Y:0.986543 U:0.990123 V:0.988888 All:0.987654 (19.123456)
...
Returns:
(avg_ssim, min_ssim)
"""
ssim_values: list[float] = []
for line in stats_output.split("\n"):
# 匹配 stats_file 格式:n:数字 ... All:数字
if not line.startswith("n:"):
continue
match = re.search(r"All:(\d+\.\d+)", line)
if match:
ssim_values.append(float(match.group(1)))
if not ssim_values:
return 0.0, 0.0
avg_ssim = sum(ssim_values) / len(ssim_values)
min_ssim = min(ssim_values)
return avg_ssim, min_ssim
def _parse_psnr_stats(stats_output: str) -> tuple[float, float]:
"""从 PSNR stats_file 输出中解析逐帧 PSNR.
FFmpeg psnr 滤镜 stats_file 输出格式(每行一帧):
n:1 mse_avg:100.23 mse_y:150.12 mse_u:50.34 mse_v:80.56 psnr_avg:28.12 psnr_y:26.34 psnr_u:31.12 psnr_v:29.08
n:2 ...
Returns:
(avg_psnr, min_psnr) — avg_psnr 是逐帧 psnr_avg 的均值,min_psnr 是逐帧最小值
"""
psnr_values: list[float] = []
for line in stats_output.split("\n"):
if not line.startswith("n:"):
continue
match = re.search(r"psnr_avg:(\d+\.\d+)", line)
if match:
psnr_values.append(float(match.group(1)))
if not psnr_values:
return 0.0, 0.0
avg_psnr = sum(psnr_values) / len(psnr_values)
min_psnr = min(psnr_values)
return avg_psnr, min_psnr
def _count_frames(stderr: str) -> int:
"""从 FFmpeg 输出中统计帧数."""
match = re.search(r"frame=\s*(\d+)", stderr)
return int(match.group(1)) if match else 0
def save_diff_frame(
video_a: str | Path,
video_b: str | Path,
output_path: str | Path,
*,
timestamp: float = 1.0,
) -> Path:
"""生成差异帧可视化图(红绿色差).
使用 blend 滤镜生成差异可视化图,差异越大越亮。
Args:
video_a: 视频A
video_b: 视频B
output_path: 输出图片路径
timestamp: 截取的时间点(秒)
Returns:
输出图片路径
"""
command = [
FFMPEG_BIN,
"-y",
"-ss",
str(timestamp),
"-i",
str(video_a),
"-ss",
str(timestamp),
"-i",
str(video_b),
"-lavfi",
"[0:v][1:v]blend=all_mode=difference,eq=contrast=5:brightness=0.5[diff]",
"-map",
"[diff]",
"-vframes",
"1",
str(output_path),
]
subprocess.run(command, check=True, capture_output=True, timeout=60) # nosec B603
return Path(output_path)
+424
View File
@@ -0,0 +1,424 @@
"""Feature Flag 单元测试。
测试 FeatureFlagConfig、InMemoryFeatureFlagStore、RenderEngineResolver 的核心逻辑。
"""
from __future__ import annotations
import time
from unittest.mock import MagicMock, patch
import pytest
from packages.adapters.redis.feature_flag_store import (
FeatureFlagConfig,
InMemoryFeatureFlagStore,
)
# ── FeatureFlagConfig 测试 ──────────────────────────────────────────────────
class TestFeatureFlagConfig:
"""FeatureFlagConfig 核心逻辑测试。"""
def test_default_disabled(self):
"""默认配置为关闭状态。"""
config = FeatureFlagConfig(name="test_flag")
assert config.enabled is False
assert config.percentage == 0
assert config.whitelist == set()
assert config.is_active() is False
assert config.is_active("user1") is False
def test_global_enabled_100_percent(self):
"""100% + 启用 = 全部命中。"""
config = FeatureFlagConfig(name="test_flag", enabled=True, percentage=100)
assert config.is_active() is True
assert config.is_active("user1") is True
assert config.is_active("any_user") is True
def test_global_enabled_0_percent_no_whitelist(self):
"""启用但 0% 且无白名单 = 不命中。"""
config = FeatureFlagConfig(name="test_flag", enabled=True, percentage=0)
assert config.is_active() is False
assert config.is_active("user1") is False
def test_whitelist_takes_priority(self):
"""白名单优先级高于百分比。"""
config = FeatureFlagConfig(
name="test_flag",
enabled=True,
percentage=0,
whitelist={"user1", "user2"},
)
assert config.is_active("user1") is True
assert config.is_active("user2") is True
assert config.is_active("user3") is False
def test_whitelist_with_percentage(self):
"""白名单用户即使百分比为0也命中,非白名单按百分比。"""
config = FeatureFlagConfig(
name="test_flag",
enabled=True,
percentage=100, # 100% 所有人命中
whitelist={"user1"},
)
assert config.is_active("user1") is True
assert config.is_active("user999") is True # 100% 命中
def test_percentage_consistency_same_user(self):
"""同一用户多次调用结果一致(哈希确定性)。"""
config = FeatureFlagConfig(name="test_flag", enabled=True, percentage=50)
results = [config.is_active("user_fixed") for _ in range(100)]
assert all(r == results[0] for r in results)
def test_percentage_different_users_distributed(self):
"""不同用户分布大致符合百分比(统计检验,宽松阈值)。"""
config = FeatureFlagConfig(name="test_flag", enabled=True, percentage=50)
active_count = sum(1 for i in range(1000) if config.is_active(f"user_{i}"))
# 50% 上下浮动 10% 都算合理
assert 400 <= active_count <= 600, f"Expected ~500, got {active_count}"
def test_percentage_boundary_0_and_100(self):
"""0% 和 100% 的边界情况。"""
config_0 = FeatureFlagConfig(name="test", enabled=True, percentage=0)
config_100 = FeatureFlagConfig(name="test", enabled=True, percentage=100)
for i in range(100):
assert config_0.is_active(f"user_{i}") is False
assert config_100.is_active(f"user_{i}") is True
def test_disabled_ignores_all_other_settings(self):
"""关闭时忽略白名单和百分比。"""
config = FeatureFlagConfig(
name="test_flag",
enabled=False,
percentage=100,
whitelist={"user1"},
)
assert config.is_active("user1") is False
assert config.is_active() is False
def test_none_identifier_with_percentage(self):
"""无 identifier 时按随机比例(0% 和 100% 是确定的)。"""
config_0 = FeatureFlagConfig(name="test", enabled=True, percentage=0)
config_100 = FeatureFlagConfig(name="test", enabled=True, percentage=100)
assert config_0.is_active(None) is False
assert config_100.is_active(None) is True
def test_to_dict_and_from_dict(self):
"""序列化和反序列化对称。"""
original = FeatureFlagConfig(
name="test_flag",
enabled=True,
percentage=30,
whitelist={"user_a", "user_b", "user_c"},
)
data = original.to_dict()
restored = FeatureFlagConfig.from_dict(data)
assert restored.name == original.name
assert restored.enabled == original.enabled
assert restored.percentage == original.percentage
assert restored.whitelist == original.whitelist
def test_from_dict_with_missing_fields(self):
"""from_dict 缺失字段时使用默认值。"""
config = FeatureFlagConfig.from_dict({"name": "minimal"})
assert config.name == "minimal"
assert config.enabled is False
assert config.percentage == 0
assert config.whitelist == set()
# ── InMemoryFeatureFlagStore 测试 ───────────────────────────────────────────
class TestInMemoryFeatureFlagStore:
"""内存存储实现测试。"""
def test_get_nonexistent_returns_default(self):
"""获取不存在的 flag 返回默认配置(关闭)。"""
store = InMemoryFeatureFlagStore()
config = store.get("nonexistent")
assert config.name == "nonexistent"
assert config.enabled is False
def test_set_and_get(self):
"""设置后可以读取。"""
store = InMemoryFeatureFlagStore()
config = FeatureFlagConfig(name="test", enabled=True, percentage=50, whitelist={"u1"})
store.set(config)
got = store.get("test")
assert got.enabled is True
assert got.percentage == 50
assert got.whitelist == {"u1"}
def test_delete_existing(self):
"""删除存在的 flag 返回 True。"""
store = InMemoryFeatureFlagStore()
store.set(FeatureFlagConfig(name="test", enabled=True))
assert store.delete("test") is True
assert store.get("test").enabled is False
def test_delete_nonexistent(self):
"""删除不存在的 flag 返回 False。"""
store = InMemoryFeatureFlagStore()
assert store.delete("nonexistent") is False
def test_list_all(self):
"""列出所有 flag。"""
store = InMemoryFeatureFlagStore()
store.set(FeatureFlagConfig(name="flag_a", enabled=True))
store.set(FeatureFlagConfig(name="flag_b", percentage=10))
all_flags = store.list_all()
assert len(all_flags) == 2
assert "flag_a" in all_flags
assert "flag_b" in all_flags
assert all_flags["flag_a"].enabled is True
def test_is_active_convenience(self):
"""is_active 便捷方法。"""
store = InMemoryFeatureFlagStore()
store.set(FeatureFlagConfig(name="render", enabled=True, percentage=0, whitelist={"vip_user"}))
assert store.is_active("render", "vip_user") is True
assert store.is_active("render", "normal_user") is False
assert store.is_active("nonexistent") is False
# ── RenderEngineResolver 测试 ───────────────────────────────────────────────
class TestRenderEngineResolver:
"""渲染引擎选择器测试。"""
def test_default_legacy_when_flag_disabled(self):
"""flag 关闭时使用默认引擎(legacy)。"""
store = InMemoryFeatureFlagStore()
resolver = self._make_resolver(store=store, default="legacy")
assert resolver.get_engine() == "legacy"
assert resolver.get_engine("user1") == "legacy"
def test_default_unified_when_flag_disabled(self):
"""flag 关闭但默认值是 unified 时返回 unified。"""
store = InMemoryFeatureFlagStore()
resolver = self._make_resolver(store=store, default="unified")
assert resolver.get_engine() == "unified"
def test_whitelist_user_uses_unified(self):
"""白名单用户走新引擎。"""
store = InMemoryFeatureFlagStore()
store.set(
FeatureFlagConfig(
name="render_engine",
enabled=True,
percentage=0,
whitelist={"beta_tester"},
)
)
resolver = self._make_resolver(store=store, default="legacy")
assert resolver.get_engine("beta_tester") == "unified"
assert resolver.get_engine("normal_user") == "legacy"
def test_100_percent_all_unified(self):
"""100% 时所有用户走新引擎。"""
store = InMemoryFeatureFlagStore()
store.set(FeatureFlagConfig(name="render_engine", enabled=True, percentage=100))
resolver = self._make_resolver(store=store, default="legacy")
for i in range(50):
assert resolver.get_engine(f"user_{i}") == "unified"
def test_invalid_default_engine_fallback(self):
"""无效默认值回退到 legacy。"""
store = InMemoryFeatureFlagStore()
resolver = self._make_resolver(store=store, default="invalid_value")
assert resolver.get_engine() == "legacy"
def test_should_use_unified_helper(self):
"""should_use_unified 便捷方法。"""
store = InMemoryFeatureFlagStore()
store.set(
FeatureFlagConfig(
name="render_engine",
enabled=True,
percentage=0,
whitelist={"user_a"},
)
)
resolver = self._make_resolver(store=store)
assert resolver.should_use_unified("user_a") is True
assert resolver.should_use_unified("user_b") is False
def test_config_snapshot(self):
"""配置快照。"""
store = InMemoryFeatureFlagStore()
store.set(
FeatureFlagConfig(
name="render_engine",
enabled=True,
percentage=30,
whitelist={"u1", "u2"},
)
)
resolver = self._make_resolver(store=store)
snapshot = resolver.get_config_snapshot()
assert snapshot["flag_name"] == "render_engine"
assert snapshot["enabled"] is True
assert snapshot["percentage"] == 30
assert snapshot["whitelist"] == ["u1", "u2"]
def test_set_flag_updates_config(self):
"""通过 set_flag 修改后立即生效。"""
store = InMemoryFeatureFlagStore()
resolver = self._make_resolver(store=store, default="legacy")
# 初始:关闭
assert resolver.get_engine("user1") == "legacy"
# 开启 100%
resolver.set_flag(FeatureFlagConfig(name="render_engine", enabled=True, percentage=100))
assert resolver.get_engine("user1") == "unified"
# 关闭
resolver.set_flag(FeatureFlagConfig(name="render_engine", enabled=False))
assert resolver.get_engine("user1") == "legacy"
def test_force_refresh(self):
"""强制刷新不报错。"""
store = InMemoryFeatureFlagStore()
resolver = self._make_resolver(store=store)
resolver.force_refresh() # 不抛异常即可
def test_does_not_affect_in_flight_tasks(self):
"""
热更新不影响在途任务验证:
任务开始时确定引擎,中途配置变更不改变当前任务的引擎选择。
(这是通过"每次调用 get_engine 时读取当前配置"来保证的,
任务开始时调用一次拿到结果,之后不再变化)
"""
store = InMemoryFeatureFlagStore()
store.set(FeatureFlagConfig(name="render_engine", enabled=True, percentage=100))
resolver = self._make_resolver(store=store, default="legacy")
# 模拟任务开始时获取引擎
engine_at_start = resolver.get_engine("user1")
assert engine_at_start == "unified"
# 任务进行中关闭 flag
store.set(FeatureFlagConfig(name="render_engine", enabled=False))
resolver.force_refresh()
# 在途任务持有的 engine_at_start 仍然是 unified(不随配置变化)
assert engine_at_start == "unified"
# 新任务会拿到 legacy
assert resolver.get_engine("user1") == "legacy"
# ── 辅助方法 ──
@staticmethod
def _make_resolver(store=None, default="legacy"):
from apps.worker.video_processing.render_engine_resolver import (
RenderEngineResolver,
)
return RenderEngineResolver(
default_engine=default,
store=store or InMemoryFeatureFlagStore(),
refresh_interval=9999, # 测试时禁用自动刷新
)
# ── RedisFeatureFlagStore 降级测试(无 Redis 环境) ───────────────────────
class TestRedisStoreDegradation:
"""Redis 不可用时的降级行为测试。"""
def test_get_returns_default_when_redis_unavailable(self):
"""Redis 连接失败时返回默认关闭配置,不抛异常。"""
import importlib
from packages.adapters.redis import feature_flag_store as ff_module
# 模拟 redis 模块不存在的场景不好做,这里直接测试异常捕获逻辑
store = ff_module.RedisFeatureFlagStore.__new__(ff_module.RedisFeatureFlagStore)
store._redis = MagicMock()
store._redis.hgetall.side_effect = ConnectionError("Redis down")
store._key_prefix = ff_module.FEATURE_FLAG_REDIS_PREFIX
store._cache = {}
store._cache_ttl = 5.0
import threading
store._lock = threading.Lock()
config = store.get("render_engine")
assert config.enabled is False
assert config.name == "render_engine"
def test_list_all_returns_empty_on_redis_error(self):
"""Redis 错误时 list_all 返回空字典。"""
import importlib
from packages.adapters.redis import feature_flag_store as ff_module
store = ff_module.RedisFeatureFlagStore.__new__(ff_module.RedisFeatureFlagStore)
store._redis = MagicMock()
store._redis.scan.side_effect = ConnectionError("Redis down")
store._key_prefix = ff_module.FEATURE_FLAG_REDIS_PREFIX
store._cache = {}
store._cache_ttl = 5.0
import threading
store._lock = threading.Lock()
result = store.list_all()
assert result == {}
class TestRedisStoreListAll:
"""RedisFeatureFlagStore list_all 正常路径测试。"""
def _make_store(self):
from packages.adapters.redis import feature_flag_store as ff_module
store = ff_module.RedisFeatureFlagStore.__new__(ff_module.RedisFeatureFlagStore)
store._redis = MagicMock()
store._key_prefix = ff_module.FEATURE_FLAG_REDIS_PREFIX
store._cache = {}
store._cache_ttl = 5.0
import threading
store._lock = threading.Lock()
return store
def test_list_all_scan_with_match_param(self):
"""list_all 调用 redis.scan 时使用正确的 match 参数名。"""
store = self._make_store()
prefix = store._key_prefix
# 模拟 scan 返回 2 个 key,分 2 次游标
store._redis.scan.side_effect = [
(10, [f"{prefix}render_engine", f"{prefix}other_flag"]),
(0, []),
]
# 模拟 hgetall 返回配置
store._redis.hgetall.return_value = {
b"enabled": b"true",
b"percentage": b"50",
b"whitelist": b'["user1","user2"]',
}
result = store.list_all()
# 验证 scan 被调用了 2 次(游标遍历)
assert store._redis.scan.call_count == 2
# 验证参数名是 match(不是 match_pattern
first_call_kwargs = store._redis.scan.call_args_list[0][1]
assert "match" in first_call_kwargs
assert "match_pattern" not in first_call_kwargs
assert first_call_kwargs["match"] == f"{prefix}*"
# 验证返回了 2 个 flag
assert len(result) == 2
assert "render_engine" in result
assert "other_flag" in result
+2 -1
View File
@@ -13,7 +13,6 @@ from unittest.mock import MagicMock, patch
import pytest
from video_processing.render_adapter import RenderAdapter, RenderAdapterResult
# ── Fixtures ──────────────────────────────────────────────────────────────────
@@ -243,6 +242,7 @@ class TestRenderPlan:
@patch("video_processing.render_adapter.download_asset")
def test_successful_render(self, mock_download, mock_render_cls, mock_upload, tmp_path):
"""完整渲染流程成功。"""
# 素材下载成功
def _fake_download(asset_id, local_path):
local_path.parent.mkdir(parents=True, exist_ok=True)
@@ -295,6 +295,7 @@ class TestRenderPlan:
@patch("video_processing.render_adapter.download_asset")
def test_progress_callback(self, mock_download, tmp_path):
"""进度回调被正确触发。"""
def _fake_download(asset_id, local_path):
local_path.parent.mkdir(parents=True, exist_ok=True)
local_path.write_bytes(b"fake data")
+546
View File
@@ -573,6 +573,8 @@ class TestPassThrough:
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_render_pass_through") as mock_pass,
patch.object(svc, "_execute_ffmpeg") as mock_exec,
patch.object(svc, "_mix_audio", return_value=None),
patch("shutil.copy2"),
patch.object(svc, "_probe_output", return_value=(5.0, 1024, 1280, 720)),
):
result = svc.render()
@@ -598,6 +600,8 @@ class TestPassThrough:
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_render_pass_through") as mock_pass,
patch.object(svc, "_execute_ffmpeg") as mock_exec,
patch.object(svc, "_mix_audio", return_value=None),
patch("shutil.copy2"),
patch.object(svc, "_probe_output", return_value=(5.5, 2048, 1280, 720)),
):
result = svc.render()
@@ -816,6 +820,8 @@ class TestRender:
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_render_pass_through") as mock_pass,
patch.object(svc, "_mix_audio", return_value=None),
patch("shutil.copy2"),
patch.object(svc, "_probe_output", return_value=(5.0, 1024, 1280, 720)),
):
result = svc.render()
@@ -843,6 +849,8 @@ class TestRender:
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_execute_ffmpeg") as mock_exec,
patch.object(svc, "_mix_audio", return_value=None),
patch("shutil.copy2"),
patch.object(svc, "_probe_output", return_value=(5.5, 2048, 1280, 720)),
):
result = svc.render()
@@ -853,3 +861,541 @@ class TestRender:
assert result.width == 1280
assert result.height == 720
mock_exec.assert_called_once()
# ── 测试音频后处理 ──────────────────────────────────────────────────────────
class TestAudioMixing:
"""测试音频后处理混音功能。"""
def test_clip_effective_duration_with_both(self):
"""指定时长和实际时长都有时取较小值。"""
clip = ResolvedClip(
clip_id="c1",
asset_id="a1",
local_path=Path("/tmp/c1.mp4"),
clip_type="main",
order=0,
duration=3.0,
actual_duration=5.0,
)
assert UnifiedRenderService._clip_effective_duration(clip) == 3.0
def test_clip_effective_duration_only_actual(self):
"""只有实际时长时用实际时长。"""
clip = ResolvedClip(
clip_id="c1",
asset_id="a1",
local_path=Path("/tmp/c1.mp4"),
clip_type="main",
order=0,
duration=0.0,
actual_duration=5.0,
)
assert UnifiedRenderService._clip_effective_duration(clip) == 5.0
def test_clip_effective_duration_only_specified(self):
"""只有指定时长时用指定时长。"""
clip = ResolvedClip(
clip_id="c1",
asset_id="a1",
local_path=Path("/tmp/c1.mp4"),
clip_type="main",
order=0,
duration=3.0,
actual_duration=0.0,
)
assert UnifiedRenderService._clip_effective_duration(clip) == 3.0
def test_clip_effective_duration_zero(self):
"""都没有时返回0。"""
clip = ResolvedClip(
clip_id="c1",
asset_id="a1",
local_path=Path("/tmp/c1.mp4"),
clip_type="main",
order=0,
duration=0.0,
actual_duration=0.0,
)
assert UnifiedRenderService._clip_effective_duration(clip) == 0.0
def test_mix_audio_single_main_clip(self):
"""单主clip时直接提取音频。"""
clips = [_make_clip("c1", "main", order=0, duration=5.0)]
asset_paths = {"asset_c1.mp4": Path("/tmp/asset_c1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
assert result.name == "audio_plan_001.aac"
mock_run.assert_called_once()
# 验证命令包含 -vn(无视频)和 aac 编码
cmd = mock_run.call_args[0][0]
assert "-vn" in cmd
assert "aac" in cmd
def test_mix_audio_multi_main_clips(self):
"""多主clip时用concat拼接音频。"""
clips = [
_make_clip("c1", "main", order=0, duration=3.0),
_make_clip("c2", "main", order=1, duration=2.0),
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_c2.mp4": Path("/tmp/asset_c2.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 4.5)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
# 验证有 filter_complex 和 concat
assert "-filter_complex" in cmd
cmd_str = " ".join(cmd)
assert "concat=n=2:v=0:a=1" in cmd_str
def test_mix_audio_with_independent_audio_track(self):
"""有独立音频轨时用amix混音。"""
clips = [
_make_clip("c1", "main", order=0, duration=5.0),
_make_clip(
"bgm1",
"main",
order=0,
duration=5.0,
config={"role": "audio", "volume": 0.5},
),
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_bgm1.mp4": Path("/tmp/asset_bgm1.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
cmd_str = " ".join(cmd)
assert "amix" in cmd_str
assert "volume=0.5" in cmd_str
def test_mix_audio_no_audio_returns_none(self):
"""没有音频素材时返回None。"""
# 构造一个没有音频的场景(比如纯文字)
clips = [_make_clip("t1", "title", order=0, duration=3.0)]
clips[0].asset_id = "" # 无素材
asset_paths: dict[str, Path] = {}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=3.0),
):
# 没有素材的clip会被跳过,layers为空
resolved = svc._resolve_clips()
layers = svc._group_clips_into_layers(resolved)
result = svc._mix_audio(layers, 3.0)
assert result is None
def test_mix_audio_background_not_used_as_main(self):
"""background 图层不参与主音频,main 优先级更高。"""
clips = [
_make_clip("bg1", "background", order=0, duration=5.0),
_make_clip("c1", "main", order=0, duration=5.0),
]
asset_paths = {
"asset_bg1.mp4": Path("/tmp/asset_bg1.mp4"),
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
# 验证主音频源是 main 的 c1,不是 background 的 bg1
# 单 main clip 走直接提取路径,输入文件应该只有 c1
cmd_str = " ".join(cmd)
assert "asset_c1.mp4" in cmd_str
assert "asset_bg1.mp4" not in cmd_str
def test_mix_audio_main_priority_over_broll(self):
"""main 图层优先级高于 broll。"""
clips = [
_make_clip("b1", "b_roll", order=0, duration=5.0),
_make_clip("c1", "main", order=0, duration=5.0),
]
asset_paths = {
"asset_b1.mp4": Path("/tmp/asset_b1.mp4"),
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
cmd_str = " ".join(cmd)
# 主音频源应该是 main 的 c1,不是 broll 的 b1
assert "asset_c1.mp4" in cmd_str
assert "asset_b1.mp4" not in cmd_str
def test_mix_audio_broll_used_when_no_main(self):
"""没有 main 时,broll 作为主音频源。"""
clips = [_make_clip("b1", "b_roll", order=0, duration=5.0)]
asset_paths = {"asset_b1.mp4": Path("/tmp/asset_b1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
assert "-vn" in cmd
assert "asset_b1.mp4" in " ".join(cmd)
def test_mix_audio_single_clip_truncated_to_video_duration(self):
"""单clip音频截断到 video_durationvideo_duration < clip有效时长)。"""
clips = [_make_clip("c1", "main", order=0, duration=10.0)]
asset_paths = {"asset_c1.mp4": Path("/tmp/asset_c1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=10.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
# video_duration 只有 3.0,小于 clip 的 10.0
result = svc._mix_audio(layers, 3.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
# 验证 -t 参数是 3.0 不是 10.0
t_index = cmd.index("-t")
assert t_index >= 0
t_value = float(cmd[t_index + 1])
assert t_value == 3.0
def test_merge_audio_video(self):
"""合并音视频命令正确。"""
svc = _make_service([], {})
video_path = Path("/tmp/video.mp4")
audio_path = Path("/tmp/audio.aac")
output_path = Path("/tmp/output.mp4")
with patch("video_processing.unified_render_service.run_ffmpeg") as mock_run:
svc._merge_audio_video(video_path, audio_path, output_path)
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
assert "-c:v" in cmd
assert "copy" in cmd
assert "-map" in cmd
assert "-shortest" in cmd
def test_render_calls_audio_mixing(self):
"""多clip完整render流程会调用音频混音(非直通路径)。"""
clips = [
_make_clip("c1", "main", order=0, duration=3.0),
_make_clip("c2", "main", order=1, duration=2.0),
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_c2.mp4": Path("/tmp/asset_c2.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_execute_ffmpeg"),
patch.object(svc, "_mix_audio", return_value=Path("/tmp/audio.aac")) as mock_mix,
patch.object(svc, "_merge_audio_video") as mock_merge,
patch.object(svc, "_probe_output", return_value=(5.0, 1024, 1280, 720)),
):
result = svc.render()
mock_mix.assert_called_once()
mock_merge.assert_called_once()
assert result.duration == 5.0
def test_render_without_audio_copies_video(self):
"""多clip无音频时走copy路径(非直通路径)。"""
clips = [
_make_clip("c1", "main", order=0, duration=3.0),
_make_clip("c2", "main", order=1, duration=2.0),
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_c2.mp4": Path("/tmp/asset_c2.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_execute_ffmpeg"),
patch.object(svc, "_mix_audio", return_value=None),
patch("shutil.copy2") as mock_copy,
patch.object(svc, "_probe_output", return_value=(5.0, 1024, 1280, 720)),
):
result = svc.render()
mock_copy.assert_called_once()
assert result.duration == 5.0
def test_render_pass_through_skips_audio_mix(self):
"""直通场景下视频+音频一次完成,跳过 _mix_audio 和 _merge_audio_video。"""
clips = [_make_clip("c1", "main", order=0, duration=5.0)]
asset_paths = {"asset_c1.mp4": Path("/tmp/asset_c1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch.object(svc, "_render_pass_through", return_value=True) as mock_pt,
patch.object(svc, "_mix_audio") as mock_mix,
patch.object(svc, "_merge_audio_video") as mock_merge,
patch("shutil.copy2") as mock_copy,
patch.object(svc, "_probe_output", return_value=(5.0, 1024, 1280, 720)),
):
result = svc.render()
# 直通场景调用了 _render_pass_through,跳过了 _mix_audio / _merge / copy
mock_pt.assert_called_once()
mock_mix.assert_not_called()
mock_merge.assert_not_called()
mock_copy.assert_not_called()
assert result.duration == 5.0
def test_pass_through_main_has_aac_audio(self):
"""直通main/broll场景输出带aac音频。"""
clips = [_make_clip("c1", "main", order=0, duration=5.0)]
asset_paths = {"asset_c1.mp4": Path("/tmp/asset_c1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._render_pass_through(layers, Path("/tmp/out.mp4"), video_duration=5.0)
assert result is True # main 类型返回有音频
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
assert "-an" not in cmd # 不再是无声
assert "aac" in cmd # 有aac音频编码
assert "-b:a" in cmd
def test_pass_through_background_no_audio(self):
"""直通background场景不带音频(图片素材)。"""
clips = [_make_clip("bg1", "background", order=0, duration=5.0)]
asset_paths = {"asset_bg1.mp4": Path("/tmp/asset_bg1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._render_pass_through(layers, Path("/tmp/out.mp4"))
assert result is False # background 返回无音频
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
assert "aac" not in cmd # 没有音频编码参数
# ── 无音轨视频防御测试 ──
def test_mix_audio_main_no_audio_stream_returns_none(self):
"""主图层clip无音频流且无独立音频轨时,返回None(不报错)。"""
clips = [_make_clip("c1", "main", order=0, duration=5.0)]
asset_paths = {"asset_c1.mp4": Path("/tmp/asset_c1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.ffmpeg_utils.probe_has_audio", return_value=False),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is None
# 没有音频流时不应调用 FFmpeg
mock_run.assert_not_called()
def test_mix_audio_partial_clips_no_audio_filtered(self):
"""部分主图层clip无音频流时,过滤掉无音轨的,剩余有音频的正常concat。"""
clips = [
_make_clip("c1", "main", order=0, duration=3.0), # 无音频
_make_clip("c2", "main", order=1, duration=2.0), # 有音频
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_c2.mp4": Path("/tmp/asset_c2.mp4"),
}
svc = _make_service(clips, asset_paths)
# 模拟:c1 无音频,c2 有音频
def fake_has_audio(path):
return "c2" in str(path)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.ffmpeg_utils.probe_has_audio", side_effect=fake_has_audio),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
cmd_str = " ".join(cmd)
# 只剩 1 个有效音频 clip,走单clip路径(-vn),不走 filter_complex concat
assert "-vn" in cmd
assert "concat=n=2" not in cmd_str
def test_mix_audio_all_main_no_audio_but_independent_track(self):
"""主图层全部无音频,但有独立音频轨时,正常走amix混音。"""
clips = [
_make_clip("c1", "main", order=0, duration=5.0), # 无音频
_make_clip(
"bgm1",
"main",
order=0,
duration=5.0,
config={"role": "audio", "volume": 0.5},
), # 独立音频轨(有音频)
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_bgm1.mp4": Path("/tmp/asset_bgm1.mp4"),
}
svc = _make_service(clips, asset_paths)
def fake_has_audio(path):
return "bgm" in str(path)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.ffmpeg_utils.probe_has_audio", side_effect=fake_has_audio),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is not None
mock_run.assert_called_once()
cmd = mock_run.call_args[0][0]
cmd_str = " ".join(cmd)
# 只有独立音频轨参与混音,amix 输入数=1
assert "amix=inputs=1" in cmd_str
def test_mix_audio_both_no_audio_returns_none(self):
"""主图层和独立音频轨都无音频时,返回None。"""
clips = [
_make_clip("c1", "main", order=0, duration=5.0),
_make_clip(
"bgm1",
"main",
order=0,
duration=5.0,
config={"role": "audio"},
),
]
asset_paths = {
"asset_c1.mp4": Path("/tmp/asset_c1.mp4"),
"asset_bgm1.mp4": Path("/tmp/asset_bgm1.mp4"),
}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
patch("video_processing.ffmpeg_utils.probe_has_audio", return_value=False),
patch("video_processing.unified_render_service.run_ffmpeg") as mock_run,
):
layers = svc._group_clips_into_layers(svc._resolve_clips())
result = svc._mix_audio(layers, 5.0)
assert result is None
mock_run.assert_not_called()
def test_clip_has_audio_cache(self):
"""_clip_has_audio 带缓存,同一clip只探测一次。"""
clips = [_make_clip("c1", "main", order=0, duration=5.0)]
asset_paths = {"asset_c1.mp4": Path("/tmp/asset_c1.mp4")}
svc = _make_service(clips, asset_paths)
with (
_patch_path_exists(),
patch("video_processing.unified_render_service.probe_duration", return_value=5.0),
):
resolved = svc._resolve_clips()
clip = resolved[0]
with patch("video_processing.ffmpeg_utils.probe_has_audio", return_value=True) as mock_probe:
# 调用 3 次
r1 = svc._clip_has_audio(clip)
r2 = svc._clip_has_audio(clip)
r3 = svc._clip_has_audio(clip)
assert r1 is True and r2 is True and r3 is True
# 实际只探测了 1 次
assert mock_probe.call_count == 1