Archives/langchain
2
0
Fork 1
mirror of https://github.com/hwchase17/langchain synced 2024-11-08 07:10:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

refactor: modify method visibility in url_playwright

Browse Source
This commit is contained in:
Youngwook Kim 2023-08-09 10:59:22 +09:00
parent 224263aa24
commit ef7f4aea32
1 changed files with 4 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
libs/langchain/langchain/document_loaders/url_playwright.py
View File

@ -48,7 +48,7 @@ class PlaywrightURLLoader(BaseLoader):
self.headless = headless
self.remove_selectors = remove_selectors
def sync_evaluate(self, page, browser, response):
def _sync_evaluate(self, page, browser, response):
"""Process a page and return the text content synchronously.
Args:
@ -72,7 +72,7 @@ class PlaywrightURLLoader(BaseLoader):
text = "\n\n".join([str(el) for el in elements])
return text
async def async_evaluate(self, page, browser, response):
async def _async_evaluate(self, page, browser, response):
"""Process a page and return the text content asynchronously.
Args:
@ -112,7 +112,7 @@ class PlaywrightURLLoader(BaseLoader):
try:
page = browser.new_page()
response = page.goto(url)
text = self.sync_evaluate(page, browser, response)
text = self._sync_evaluate(page, browser, response)
metadata = {"source": url}
docs.append(Document(page_content=text, metadata=metadata))
except Exception as e:
@ -142,7 +142,7 @@ class PlaywrightURLLoader(BaseLoader):
try:
page = await browser.new_page()
response = await page.goto(url)
text = await self.async_evaluate(page, browser, response)
text = await self._async_evaluate(page, browser, response)
metadata = {"source": url}
docs.append(Document(page_content=text, metadata=metadata))
except Exception as e: