hhhge/NeteaseUrl
Archived
1
0 Fork 0
Code Issues Actions Packages Projects Releases Wiki Activity

refactor: worker 子系统全面异步化，拆掉同步转异步的桥接层

Browse Source 
拆掉「同步调异步 / 异步桥同步」两类垫片的第二半（HTTP 线 B2 见前序提交）。
worker 侧此前是「起线程跑同步函数、函数体内部再用 asyncio.to_thread 把活塞给
另一条循环」——线程与协程互等，取消/暂停只能靠轮询。本轮把 worker 自身变成
协程，由 _start_task 直接挂到应用事件循环上，桥接层随之失去存在理由。

改动（按子系统）：
- music_downloader：downloader / _aria2 / _concurrency / cover / _http /
  __main__ 全链路协程化。per-file 锁改 asyncio.Lock；计数闸改
  「active 计数 + await asyncio.sleep(0.5)」的可取消等待（原阻塞式 with 会让
  已取消任务仍要等全局名额腾出，卡死并拖后腿退款）；aria2p 的同步 RPC 仍经
  to_thread，轮询改 asyncio.sleep；新增 core.http_client.iter_chunks，把**流式
  读取期**的 aiohttp.ClientError 收敛成 UpstreamNetworkError（请求建立阶段早已
  有这条路径，读取阶段此前没有，断流会绕过调用方的 UpstreamError 分类）。
  删除 get_http_session / _SESSION_LOCAL。
  models._RetryFromScratch 不再继承 requests.RequestException 而继承 Exception：
  它是「续传起点失效、残片已清、从头重下」，既不是确定性失败
  (DownloadException)，也不是残片已保留的网络失败(UpstreamError)，必须独立
  成类且在 except 链中先于二者捕获。
- music_api：podcast.py 全部函数协程化（直链解析内部改
  await NeteaseAPI.get_song_url）；compat.py 删除——16 个死掉的同步门面早已
  清空，文件只剩空壳。
- services：_taskcore / _zipcore / batch / downloads / podcast / tasks 协程化。
  threading.Thread → loop.create_task（任务句柄字段 worker_thread →
  worker_task），ThreadPoolExecutor → asyncio.gather + asyncio.Semaphore，
  终态函数与结算回调全部改 await；监控线程里「worker 已死」判定改 task.done()。
  删除 services/streaming.py（同步流垫片，唯一消费者已迁到
  core.http_client.open）。zipfile 实例仍非线程安全，故「下载」并行、
  「入包」严格串行。
- web：路由层 130+ 处 to_thread 拆除——被包住的目标自己已是协程，再包一层只是
  把协程丢进线程里跑一圈。tasks_registry.start_task 直接 create_task。
  web/app.py 关闭段去掉 async_bridge.shutdown()（不再有第二条约循环）。
  web/routers/download.py 的播客端点补上 3 处漏掉的 await
  （extract_podcast_ref / get_radio_detail）——漏掉时不报错、只静默返回协程对象，
  靠 pyright 抓出来后修正。

留在 to_thread 上的都是**正当的阻塞卸载**：同步 SQLite 落库与账本结算、
mutagen 全量重写标签、zipfile 整段拷贝、shutil.rmtree、aria2p 的同步 RPC、
PBKDF2 的 10 万次迭代。ruff 对残留的本地元数据操作（exists/stat/os.replace/
往 .part 顺序写分块）按「微秒级、卸载的调度开销大于操作本身」放行，pyproject
逐条注明理由。

取消/暂停仍由 threading.Event 承载：置位方是请求线程/监控线程，读取方是 worker
协程，只有「置位」与「轮询读取」两种操作，既不需要跨循环唤醒也不会卡住事件
循环——比 asyncio.Event 更省心（ASYNC110 同理按此豁免）。

测试：10 个文件的替身边界同步迁移——网络替身从 core.sync_http 的
get_http_session 换成 downloader/服务模块的 _http.open + **真实** iter_chunks
（顺带钉住读取期异常收敛）；`_run` 改为在独立线程里 asyncio.run(worker(...))；
aria2 的假时钟从 time.sleep 换成 _POLL_INTERVAL=0；monitor 的线程句柄替身换成
done() 可控的轻量替身（asyncio.Task 的构造需要运行中的事件循环，而监控循环跑在
常驻线程里）；podcast / 封面 / 结算包装等替身一律改 async。

门禁：ruff 全过 / pyright 0-0 / pytest 383 项全绿（0 失败）。

下一步 B2：auth_system 12 个模块改 aiosqlite 背板、web 层剩余 to_thread 收尾；
B3：删 core/sync_http.py、core/async_bridge.py、core/http_errors.py（现仅
sync_http 自用）。
This commit is contained in:
snowy-yang committed 2026-09-28 19:36:56 +08:00
1 parent 65adfc9089
commit 69a50c2c31
35 files changed
+1062 -1005
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

tests/test_zip_workers.py
+31 -19
View File
@@ -16,6 +16,7 @@ ZIP 内容、账本守恒、暂停/恢复/取消语义。断言只依赖「调
账本口径：一律用 ``reserve_and_open`` 真预扣（``ledger_open`` 只记账不扣款，
会让「退款」凭空给余额加次数，守恒断言就失去意义）。
"""
import asyncio
import threading
import time
import zipfile
@@ -59,7 +60,7 @@ class _FakeDownloader:
self.calls: list[int] = []
self._lock = threading.Lock()
def get_music_info(self, mid: Any, quality: str, cookies: Any) -> MusicInfo:
async def get_music_info(self, mid: Any, quality: str, cookies: Any) -> MusicInfo:
return MusicInfo(
id=int(mid), name=f'歌{mid}', artists='歌手A', album='专辑A',
pic_url='', duration=200, track_number=1, download_url='',
@@ -67,7 +68,7 @@ class _FakeDownloader:
lyric='[00:01.00]第一行\n[00:02.00]第二行',
)
def download_music_file(self, mid: Any, quality: str, **kw: Any) -> DownloadResult:
async def download_music_file(self, mid: Any, quality: str, **kw: Any) -> DownloadResult:
with self._lock:
self.calls.append(int(mid))
seq = len(self.calls)
@@ -199,7 +200,11 @@ def _register_podcast(sandbox, pids: list[int], *, reserved: int, user_id: int)
def _run(worker, task_id: str) -> threading.Thread:
th = threading.Thread(target=worker, args=(task_id,), daemon=True)
"""把 worker 协程跑在独立线程的私有事件循环里（等价于生产的任务循环）。"""
def _target() -> None:
asyncio.run(worker(task_id))
th = threading.Thread(target=_target, daemon=True)
th.start()
return th
@@ -346,8 +351,9 @@ def test_batch_zip_cancel_at_publish_checkpoint_delivers(sandbox, monkeypatch):
holder: dict[str, Any] = {}
real_settle = batch_svc._settle_batch_billing
def settle_then_cancel(task_id: str, ok_count: int, final_status: str = 'settled') -> int:
refunded = real_settle(task_id, ok_count, final_status)
async def settle_then_cancel(task_id: str, ok_count: int,
final_status: str = 'settled') -> int:
refunded = await real_settle(task_id, ok_count, final_status)
holder['task']['cancel'].set() # 结算已完成、发布前一刻取消
return refunded
@@ -371,24 +377,27 @@ def test_batch_zip_cancel_at_publish_checkpoint_delivers(sandbox, monkeypatch):
@pytest.fixture()
def podcast_stub(monkeypatch, sandbox):
"""把播客上游与流式下载替换为本地确定性实现。"""
def fake_detail(pid: Any, cookies: Any) -> dict[str, Any]:
async def fake_detail(pid: Any, cookies: Any) -> dict[str, Any]:
n = int(pid)
return {'id': n, 'name': f'单集{n}', 'main_song_id': 1,
'serial_num': n, 'cover_url': '', 'songs': []}
def fake_resolve(prog: dict[str, Any], quality: str, cookies: Any) -> dict[str, Any]:
async def fake_resolve(prog: dict[str, Any], quality: str, cookies: Any) -> dict[str, Any]:
return {'url': f"http://example.invalid/{prog['id']}.mp3", 'size': 2048, 'type': 'mp3'}
def fake_stream(url: str, target: Path, total_size: int, cancel_ev: Any,
task_id: str = '', progress_label: str = '') -> int:
async def fake_stream(url: str, target: Path, total_size: int, cancel_ev: Any,
task_id: str = '', progress_label: str = '') -> int:
if cancel_ev is not None and cancel_ev.is_set():
raise DownloadCanceled('任务已取消')
target.write_bytes(b'\xff\xfb\x90\x00' + b'\x00' * 128)
return target.stat().st_size
async def _no_cover(_u: str) -> None:
return None
monkeypatch.setattr(podcast_svc, '_podcast_resolve_audio', fake_resolve)
monkeypatch.setattr(podcast_svc, '_podcast_stream_download', fake_stream)
monkeypatch.setattr(podcast_svc, 'fetch_cover_bytes', lambda _u: None)
monkeypatch.setattr(podcast_svc, 'fetch_cover_bytes', _no_cover)
monkeypatch.setattr(podcast_svc.podcast, 'get_program_detail', fake_detail)
return sandbox
@@ -418,7 +427,7 @@ def test_podcast_zip_pause_then_resume(podcast_stub, monkeypatch):
holder: dict[str, Any] = {}
fired = {'n': 0}
def _pause_once(url, target, total_size, cancel_ev, task_id='', progress_label=''):
async def _pause_once(url, target, total_size, cancel_ev, task_id='', progress_label=''):
target.write_bytes(b'\xff\xfb\x90\x00' + b'\x00' * 128)
fired['n'] += 1
if fired['n'] == 1: # 只在第一集触发——否则恢复后立刻又被暂停
@@ -447,7 +456,7 @@ def test_podcast_zip_all_fail_refunds(podcast_stub, monkeypatch):
uid = _uid()
before = _balance()
def _boom(prog, quality, cookies):
async def _boom(prog, quality, cookies):
raise RuntimeError('上游不可用')
monkeypatch.setattr(podcast_svc, '_podcast_resolve_audio', _boom)
@@ -470,9 +479,9 @@ def test_podcast_zip_all_fail_refunds(podcast_stub, monkeypatch):
class _SameNameDownloader(_FakeDownloader):
"""所有歌曲同名同歌手、且都带封面——用于验证包内三条目共享序号后缀。"""
def get_music_info(self, mid: Any, quality: str, cookies: Any) -> MusicInfo:
return replace(super().get_music_info(mid, quality, cookies),
name='同名歌', artists='同歌手',
async def get_music_info(self, mid: Any, quality: str, cookies: Any) -> MusicInfo:
base = await super().get_music_info(mid, quality, cookies)
return replace(base, name='同名歌', artists='同歌手',
pic_url='http://example.invalid/cover.jpg')
@@ -485,7 +494,10 @@ def test_batch_zip_dedups_three_arcnames_with_shared_suffix(sandbox, monkeypatch
"""
uid = _uid()
monkeypatch.setattr(api_service, 'downloader', _SameNameDownloader(sandbox['cache']))
monkeypatch.setattr(batch_svc, 'fetch_cover_bytes', lambda _u: b'\xff\xd8\xff\xe0')
async def _cover(_u: str) -> bytes:
return b'\xff\xd8\xff\xe0'
monkeypatch.setattr(batch_svc, 'fetch_cover_bytes', _cover)
tid, task = _register_batch(sandbox, [701, 702], reserved=2, user_id=uid)
th = _run(batch_svc._batch_zip_worker, tid)
@@ -528,7 +540,7 @@ def test_podcast_all_fail_message_and_ledger_status(podcast_stub, monkeypatch):
uid = _uid()
before = _balance()
def _boom(prog, quality, cookies):
async def _boom(prog, quality, cookies):
raise RuntimeError('上游不可用')
monkeypatch.setattr(podcast_svc, '_podcast_resolve_audio', _boom)
@@ -557,8 +569,8 @@ def test_podcast_cleans_temp_files_on_episode_failure(podcast_stub, monkeypatch)
uid = _uid()
before = _balance()
def _flaky(url: str, target: Path, total_size: int, cancel_ev: Any,
task_id: str = '', progress_label: str = '') -> int:
async def _flaky(url: str, target: Path, total_size: int, cancel_ev: Any,
task_id: str = '', progress_label: str = '') -> int:
if '/102.' in url:
target.write_bytes(b'\x00' * 64) # 正文残留
target.with_name(target.name + '.part').write_bytes(b'\x00' * 32) # 残片