ayrisdev/crawl4ai
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'main' of https://github.com/unclecode/crawl4ai

Browse Source
This commit is contained in:
UncleCode
2024-11-22 19:56:56 +08:00
parent a5decaa7cf 3439f7886d
commit d7a112fefe
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
crawl4ai/async_webcrawler.py
View File

@@ -477,7 +477,7 @@ class AsyncWebCrawler:
word_count_threshold=word_count_threshold, word_count_threshold=word_count_threshold,
css_selector=css_selector, css_selector=css_selector,
only_text=kwargs.pop("only_text", False), only_text=kwargs.pop("only_text", False),
image_description_min_word_threshold=kwargs.get( image_description_min_word_threshold=kwargs.pop(
"image_description_min_word_threshold", IMAGE_DESCRIPTION_MIN_WORD_THRESHOLD "image_description_min_word_threshold", IMAGE_DESCRIPTION_MIN_WORD_THRESHOLD
), ),
content_filter = content_filter, content_filter = content_filter,

2
crawl4ai/crawler_strategy.py
View File

@@ -283,7 +283,7 @@ class LocalSeleniumCrawlerStrategy(CrawlerStrategy):
print(f"[LOG] ✅ Crawled {url} successfully!") print(f"[LOG] ✅ Crawled {url} successfully!")
return html return html
except InvalidArgumentException: except InvalidArgumentException as e:
if not hasattr(e, 'msg'): if not hasattr(e, 'msg'):
e.msg = sanitize_input_encode(str(e)) e.msg = sanitize_input_encode(str(e))
raise InvalidArgumentException(f"Failed to crawl {url}: {e.msg}") raise InvalidArgumentException(f"Failed to crawl {url}: {e.msg}")