Skip to content

Automate browser based workflows with AI

update scraped_page references when it's refreshed (#1151)

Shuchang Zhengc531f9bc5d02
diff --git a/skyvern/webeye/scraper/scraper.py b/skyvern/webeye/scraper/scraper.py
index 91c633b4c..3a2f9d41a 100644
--- a/skyvern/webeye/scraper/scraper.py
+++ b/skyvern/webeye/scraper/scraper.py
@@ -242,13 +242,25 @@ class ScrapedPage(BaseModel):
raise UnknownElementTreeFormat(fmt=fmt)
async def refresh(self, with_screenshot: bool = True) -> Self:
- return await scrape_website(
+ refreshed_page = await scrape_website(
browser_state=self._browser_state,
url=self.url,
cleanup_element_tree=self._clean_up_func,
scrape_exclude=self._scrape_exclude,
with_screenshot=with_screenshot,
)
+ self.elements = refreshed_page.elements
+ self.id_to_css_dict = refreshed_page.id_to_css_dict
+ self.id_to_element_dict = refreshed_page.id_to_element_dict
+ self.id_to_frame_dict = refreshed_page.id_to_frame_dict
+ self.id_to_element_hash = refreshed_page.id_to_element_hash
+ self.hash_to_element_ids = refreshed_page.hash_to_element_ids
+ self.element_tree = refreshed_page.element_tree
+ self.element_tree_trimmed = refreshed_page.element_tree_trimmed
+ self.screenshots = refreshed_page.screenshots or self.screenshots
+ self.html = refreshed_page.html
+ self.extracted_text = refreshed_page.extracted_text
+ return self
async def scrape_website(