Merge pull request 'Fix #104: DeepSearch fails with Playwright async context manager error ('__aexit__' missing)' (#124) from typosaurus/ticket-104 into master

Browse Source 
Reviewed-on: #124
This commit is contained in:
typosaurus 2026-07-26 00:36:04 +02:00
commit 2620ecc0f1
1 changed files with 114 additions and 111 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
devplacepy/services/jobs/deepsearch/crawl.py
View File

@ -22,6 +22,8 @@ from .pdf import MAX_PDF_BYTES, extract_pdf_text, is_pdf
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
_pw_lock = asyncio.Lock()
RSEARCH_URL = "https://rsearch.app.molodetz.nl" RSEARCH_URL = "https://rsearch.app.molodetz.nl"
RSEARCH_TIMEOUT_SECONDS = 45.0 RSEARCH_TIMEOUT_SECONDS = 45.0
FETCH_TIMEOUT_SECONDS = 20.0 FETCH_TIMEOUT_SECONDS = 20.0
@ -172,13 +174,7 @@ async def _render_with_playwright(
) -> tuple[str, str, int, list[tuple[str, str]]]: ) -> tuple[str, str, int, list[tuple[str, str]]]:
from playwright.async_api import async_playwright from playwright.async_api import async_playwright
own_browser = browser is None async def _render(browser) -> tuple[str, str, int, list[tuple[str, str]]]:
if own_browser:
pw = await async_playwright().__aenter__()
browser = await pw.chromium.launch(
headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
)
try:
context = await browser.new_context(user_agent=USER_AGENT) context = await browser.new_context(user_agent=USER_AGENT)
page = await context.new_page() page = await context.new_page()
response = await page.goto(url, wait_until="load", timeout=30000) response = await page.goto(url, wait_until="load", timeout=30000)
@ -189,10 +185,18 @@ async def _render_with_playwright(
await context.close() await context.close()
extracted = extract_html(content, base_url=url) extracted = extract_html(content, base_url=url)
return extracted.title, extracted.text, status, extracted.links return extracted.title, extracted.text, status, extracted.links
if browser is None:
async with async_playwright() as pw:
browser = await pw.chromium.launch(
headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
)
try:
return await _render(browser)
finally: finally:
if own_browser:
await browser.close() await browser.close()
await pw.__aexit__(None, None, None) else:
return await _render(browser)
async def fetch_page(url: str, depth: int, browser=None) -> CrawledPage | None: async def fetch_page(url: str, depth: int, browser=None) -> CrawledPage | None:
@ -242,8 +246,12 @@ async def fetch_page(url: str, depth: int, browser=None) -> CrawledPage | None:
except (LookupError, ValueError) as exc: except (LookupError, ValueError) as exc:
logger.info("deepsearch decode failed for %s: %s", url, exc) logger.info("deepsearch decode failed for %s: %s", url, exc)
if len(text) < MIN_PAGE_CHARS: if len(text) < MIN_PAGE_CHARS:
async with _pw_lock:
try: try:
r_title, r_text, r_status, r_links = await _render_with_playwright(url, browser) r_title, r_text, r_status, r_links = await _render_with_playwright(url, browser)
except Exception as exc:
logger.info("deepsearch render failed for %s: %s", url, exc)
r_title, r_text, r_status, r_links = "", "", 0, []
if len(r_text) > len(text): if len(r_text) > len(text):
title, text, status, source, links = ( title, text, status, source, links = (
r_title or title, r_title or title,
@ -252,8 +260,6 @@ async def fetch_page(url: str, depth: int, browser=None) -> CrawledPage | None:
"playwright", "playwright",
r_links, r_links,
) )
except Exception as exc:
logger.info("deepsearch render failed for %s: %s", url, exc)
if len(text) < MIN_PAGE_CHARS: if len(text) < MIN_PAGE_CHARS:
return None return None
return CrawledPage( return CrawledPage(
@ -296,10 +302,9 @@ async def crawl(
total = min(len(level_candidates), max_pages) total = min(len(level_candidates), max_pages)
cancelled = False cancelled = False
pw = None
browser = None browser = None
async with async_playwright() as pw:
try: try:
pw = await async_playwright().__aenter__()
browser = await pw.chromium.launch( browser = await pw.chromium.launch(
headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"] headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
) )
@ -392,6 +397,4 @@ async def crawl(
finally: finally:
if browser is not None: if browser is not None:
await browser.close() await browser.close()
if pw is not None:
await pw.__aexit__(None, None, None)
return outcome return outcome