[Add] browser-use and main.py

This commit is contained in:
tv0924@icloud.com 2025-05-18 21:57:54 +09:00
commit 96914d44ac
221 changed files with 30952 additions and 1 deletions

View file

@ -0,0 +1,39 @@
import httpx
import pytest
from browser_use.browser.browser import Browser, BrowserConfig
@pytest.mark.asyncio
async def test_browser_close_doesnt_affect_external_httpx_clients():
"""
Test that Browser.close() doesn't close HTTPX clients created outside the Browser instance.
This test demonstrates the issue where Browser.close() is closing all HTTPX clients.
"""
# Create an external HTTPX client that should remain open
external_client = httpx.AsyncClient()
# Create a Browser instance
browser = Browser(config=BrowserConfig(headless=True))
# Close the browser (which should trigger cleanup_httpx_clients)
await browser.close()
# Check if the external client is still usable
try:
# If the client is closed, this will raise RuntimeError
# Using a simple HEAD request to a reliable URL
await external_client.head('https://www.example.com', timeout=2.0)
client_is_closed = False
except RuntimeError as e:
# If we get "Cannot send a request, as the client has been closed"
client_is_closed = 'client has been closed' in str(e)
except Exception:
# Any other exception means the client is not closed but request failed
client_is_closed = False
finally:
# Always clean up our test client properly
await external_client.aclose()
# Our external client should not be closed by browser.close()
assert not client_is_closed, 'External HTTPX client was incorrectly closed by Browser.close()'

View file

@ -0,0 +1,36 @@
import asyncio
import base64
import pytest
from browser_use.browser.browser import Browser, BrowserConfig
async def test_take_full_page_screenshot():
browser = Browser(config=BrowserConfig(headless=False, disable_security=True))
try:
async with await browser.new_context() as context:
page = await context.get_current_page()
# Go to a test page
await page.goto('https://example.com')
await asyncio.sleep(3)
# Take full page screenshot
screenshot_b64 = await context.take_screenshot(full_page=True)
await asyncio.sleep(3)
# Verify screenshot is not empty and is valid base64
assert screenshot_b64 is not None
assert isinstance(screenshot_b64, str)
assert len(screenshot_b64) > 0
# Test we can decode the base64 string
try:
base64.b64decode(screenshot_b64)
except Exception as e:
pytest.fail(f'Failed to decode base64 screenshot: {str(e)}')
finally:
await browser.close()
if __name__ == '__main__':
asyncio.run(test_take_full_page_screenshot())

View file

@ -0,0 +1,96 @@
import asyncio
import json
import anyio
import pytest
from browser_use.browser.browser import Browser, BrowserConfig
from browser_use.dom.views import DOMBaseNode, DOMElementNode, DOMTextNode
from browser_use.utils import time_execution_sync
class ElementTreeSerializer:
@staticmethod
def dom_element_node_to_json(element_tree: DOMElementNode) -> dict:
def node_to_dict(node: DOMBaseNode) -> dict:
if isinstance(node, DOMTextNode):
return {'type': 'text', 'text': node.text}
elif isinstance(node, DOMElementNode):
return {
'type': 'element',
'tag_name': node.tag_name,
'attributes': node.attributes,
'highlight_index': node.highlight_index,
'children': [node_to_dict(child) for child in node.children],
}
return {}
return node_to_dict(element_tree)
# run with: pytest browser_use/browser/tests/test_clicks.py
@pytest.mark.asyncio
async def test_highlight_elements():
browser = Browser(config=BrowserConfig(headless=False, disable_security=True))
async with await browser.new_context() as context:
page = await context.get_current_page()
# await page.goto('https://immobilienscout24.de')
# await page.goto('https://help.sap.com/docs/sap-ai-core/sap-ai-core-service-guide/service-plans')
# await page.goto('https://google.com/search?q=elon+musk')
# await page.goto('https://kayak.com')
# await page.goto('https://www.w3schools.com/tags/tryit.asp?filename=tryhtml_iframe')
# await page.goto('https://dictionary.cambridge.org')
# await page.goto('https://github.com')
await page.goto('https://huggingface.co/')
await asyncio.sleep(1)
while True:
try:
# await asyncio.sleep(10)
state = await context.get_state(True)
async with await anyio.open_file('./tmp/page.json', 'w') as f:
await f.write(
json.dumps(
ElementTreeSerializer.dom_element_node_to_json(state.element_tree),
indent=1,
)
)
# await time_execution_sync('highlight_selector_map_elements')(
# browser.highlight_selector_map_elements
# )(state.selector_map)
# Find and print duplicate XPaths
xpath_counts = {}
if not state.selector_map:
continue
for selector in state.selector_map.values():
xpath = selector.xpath
if xpath in xpath_counts:
xpath_counts[xpath] += 1
else:
xpath_counts[xpath] = 1
print('\nDuplicate XPaths found:')
for xpath, count in xpath_counts.items():
if count > 1:
print(f'XPath: {xpath}')
print(f'Count: {count}\n')
print(list(state.selector_map.keys()), 'Selector map keys')
print(state.element_tree.clickable_elements_to_string())
action = input('Select next action: ')
await time_execution_sync('remove_highlight_elements')(context.remove_highlights)()
node_element = state.selector_map[int(action)]
# check if index of selector map are the same as index of items in dom_items
await context._click_element_node(node_element)
except Exception as e:
print(e)