Git-Mirrors/brozzler
1
0
Fork 0
mirror of https://github.com/internetarchive/brozzler.git synced 2025-02-24 16:49:56 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity

small ruff formatting pass

Browse Source
This commit is contained in:
Gretchen Miller 2024-09-13 13:38:21 -07:00
parent c722549697
commit 77e6b9ed52
1 changed files with 14 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
brozzler/worker.py
View File

@ -248,8 +248,13 @@ class BrozzlerWorker:
if not self._needs_browsing(page_headers): if not self._needs_browsing(page_headers):
self.logger.info("needs fetch: %s", page) self.logger.info("needs fetch: %s", page)
if self._is_video_type(page_headers) and site.video_capture != "ENABLE_VIDEO_CAPTURE": if (
self.logger.info("skipping video content: video capture disabled for site") self._is_video_type(page_headers)
and site.video_capture != "ENABLE_VIDEO_CAPTURE"
):
self.logger.info(
"skipping video content: video capture disabled for site"
)
else: else:
self._fetch_url(site, page=page) self._fetch_url(site, page=page)
else: else:
@ -305,10 +310,15 @@ class BrozzlerWorker:
return {} return {}
def _needs_browsing(self, page_headers) -> bool: def _needs_browsing(self, page_headers) -> bool:
return not bool("content-type" in page_headers and "html" not in page_headers["content-type"]) return not bool(
"content-type" in page_headers
and "html" not in page_headers["content-type"]
)
def _is_video_type(self, page_headers) -> bool: def _is_video_type(self, page_headers) -> bool:
return ("content-type" in page_headers and "video" in page_headers["content-type"]) return (
"content-type" in page_headers and "video" in page_headers["content-type"]
)
def _browse_page(self, browser, site, page, on_screenshot=None, on_request=None): def _browse_page(self, browser, site, page, on_screenshot=None, on_request=None):
def _on_screenshot(screenshot_jpeg): def _on_screenshot(screenshot_jpeg):