修改浏览器方案为同步方案

tool/browser.py

@@ -4,18 +4,21 @@
 - open_page(url, wait_ms): 打开页面，返回 page_id
 - query(page_id, selector): 选中节点并提取文字、图片 URL 与属性
 
-实现上内部常驻一个 Browser；每次 open_page 创建独立 BrowserContext 和 Page，
-query 完成后自动释放对应资源。
+实现上内部常驻一个 Browser；每次 open_page 创建独立 BrowserContext 和 Page。
+Playwright 只在一个后台线程里运行，但外层 API 保持同步。
 """
 
 from __future__ import annotations
 
-import asyncio
+import atexit
+import queue
 import threading
 import uuid
+from concurrent.futures import Future
 from dataclasses import dataclass
+from typing import Any, Callable
 
-from playwright.async_api import Browser, BrowserContext, Page, async_playwright
+from playwright.sync_api import Browser, BrowserContext, Page, sync_playwright
 
 
 @dataclass
@@ -35,41 +38,59 @@ class _PageSession:
 
 
 class _BrowserRuntime:
-    """在后台事件循环中管理 Playwright 生命周期。"""
+    """在单独线程中管理 Playwright 生命周期。"""
 
     def __init__(self) -> None:
         self._thread: threading.Thread | None = None
-        self._loop: asyncio.AbstractEventLoop | None = None
         self._ready = threading.Event()
+        self._closed = threading.Event()
+        self._job_queue: queue.Queue[
+            tuple[Callable[..., Any], tuple[Any, ...], dict[str, Any], Future[Any]]
+        ] = queue.Queue()
         self._lock = threading.RLock()
         self._playwright = None
         self._browser: Browser | None = None
         self._pages: dict[str, _PageSession] = {}
+        atexit.register(self.close)
 
     def _thread_main(self) -> None:
-        loop = asyncio.new_event_loop()
-        asyncio.set_event_loop(loop)
-        self._loop = loop
+        with sync_playwright() as playwright:
+            self._playwright = playwright
+            self._browser = playwright.chromium.launch(headless=True)
             self._ready.set()
-        loop.run_forever()
 
-        pending = asyncio.all_tasks(loop)
-        for task in pending:
-            task.cancel()
+            while True:
+                job = self._job_queue.get()
+                if job is None:
+                    break
 
-        if pending:
-            loop.run_until_complete(asyncio.gather(*pending, return_exceptions=True))
+                fn, args, kwargs, fut = job
+                try:
+                    result = fn(*args, **kwargs)
+                except BaseException as exc:  # pragma: no cover - propagate failures
+                    fut.set_exception(exc)
+                else:
+                    fut.set_result(result)
+
+            for session in self._pages.values():
+                try:
+                    session.page.close()
+                except Exception:
+                    pass
+                try:
+                    session.context.close()
+                except Exception:
+                    pass
+            self._pages.clear()
 
             if self._browser is not None:
-            loop.run_until_complete(self._browser.close())
+                try:
+                    self._browser.close()
+                finally:
                     self._browser = None
 
-        if self._playwright is not None:
-            loop.run_until_complete(self._playwright.stop())
             self._playwright = None
-
-        loop.run_until_complete(loop.shutdown_asyncgens())
-        loop.close()
+            self._closed.set()
 
     def _ensure_thread(self) -> None:
         with self._lock:
@@ -84,42 +105,45 @@ class _BrowserRuntime:
             self._thread.start()
             self._ready.wait()
 
-    def _run(self, coro):
+    def _run(self, fn: Callable[..., Any], *args: Any, **kwargs: Any) -> Any:
         self._ensure_thread()
-        assert self._loop is not None
-        future = asyncio.run_coroutine_threadsafe(coro, self._loop)
-        return future.result()
+        fut: Future[Any] = Future()
+        self._job_queue.put((fn, args, kwargs, fut))
+        return fut.result()
 
-    async def _ensure_browser(self) -> Browser:
-        if self._browser is not None:
+    def close(self) -> None:
+        with self._lock:
+            if self._closed.is_set():
+                return
+            if self._thread is None:
+                return
+            self._job_queue.put(None)
+            self._closed.wait(timeout=10)
+
+    def _ensure_browser(self) -> Browser:
+        if self._browser is None:
+            raise RuntimeError("Browser runtime is not ready.")
         return self._browser
 
-        if self._playwright is None:
-            self._playwright = await async_playwright().start()
-
-        self._browser = await self._playwright.chromium.launch(headless=True)
-        return self._browser
-
-    async def _open_page(self, url: str, wait_ms: int) -> str:
-        browser = await self._ensure_browser()
-        context = await browser.new_context()
-        page = await context.new_page()
-        await page.goto(url, wait_until="domcontentloaded", timeout=30000)
+    def _open_page_impl(self, url: str, wait_ms: int) -> str:
+        browser = self._ensure_browser()
+        context = browser.new_context()
+        page = context.new_page()
+        page.goto(url, wait_until="domcontentloaded", timeout=30000)
         if wait_ms > 0:
-            await page.wait_for_timeout(wait_ms)
+            page.wait_for_timeout(wait_ms)
 
         page_id = uuid.uuid4().hex
         self._pages[page_id] = _PageSession(context=context, page=page)
         return page_id
 
-    async def _query(self, page_id: str, selector: str) -> list[NodeSnapshot]:
+    def _query_impl(self, page_id: str, selector: str) -> list[NodeSnapshot]:
         session = self._pages.get(page_id)
         if session is None:
             raise KeyError(f"Unknown page_id: {page_id}")
 
-        try:
         locator = session.page.locator(selector)
-            payload = await locator.evaluate_all(
+        payload = locator.evaluate_all(
             r"""(elements) => elements.map((el) => {
                 const attrs = {};
                 for (const attr of Array.from(el.attributes || [])) {
@@ -223,10 +247,6 @@ class _BrowserRuntime:
             )
             for item in payload
         ]
-        finally:
-            await session.page.close()
-            await session.context.close()
-            self._pages.pop(page_id, None)
 
 
 _runtime = _BrowserRuntime()
@@ -234,12 +254,12 @@ _runtime = _BrowserRuntime()
 
 def open_page(url: str, wait_ms: int = 3000) -> str:
     """打开页面并返回 page_id。"""
-    return _runtime._run(_runtime._open_page(url, wait_ms))
+    return _runtime._run(_runtime._open_page_impl, url, wait_ms)
 
 
 def query(page_id: str, selector: str) -> list[NodeSnapshot]:
     """按选择器抓取节点，提取文本、图片 URL 与属性。"""
-    items = _runtime._run(_runtime._query(page_id, selector))
+    items = _runtime._run(_runtime._query_impl, page_id, selector)
     seen_urls: set[str] = set()
 
     for item in items: