Spaces:
No application file
No application file
File size: 20,174 Bytes
4de915b |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 |
# This file has the code which we tried for multiple tries to get element screenshot working.
# It is kept here for reference but not currently used.
# class ElementScreenshotParams(BaseModel):
# selector: str = Field(..., description="A CSS selector for the element to screenshot.")
# @tools.action(
# description="Take a screenshot of an element at the given CSS selector and return status",
# param_model=ElementScreenshotParams,
# )
# async def element_screenshot(params: ElementScreenshotParams, browser_session: BrowserSession) -> ActionResult:
# # This is a dummy implementation
# print('element_screenshot ---------------')
# print(params)
# return ActionResult(extracted_content=json.dumps({
# "success": True,
# "selector": params.selector,
# }))
# @tools.action(
# description="""
# A robust tool to capture screenshots of web elements.
# Args:
# params (ElementScreenshotParams): Parameters containing selectors, filename, highlight, padding, scroll_if_needed, fallback_to_full_page.
# browser_session (BrowserSession): The active browser session.
# Returns:
# ActionResult: Contains extracted_content, long_term_memory, vision_content, or error.
# """,
# param_model=ElementScreenshotParams,
# )
# async def element_screenshot(params: ElementScreenshotParams, browser_session: BrowserSession) -> ActionResult:
# page = await browser_session.get_current_page()
# try:
# session_base = getattr(browser_session, 'file_system_path', None)
# if session_base:
# base_path = session_base
# else:
# base_path = os.getcwd()
# parsed_url = urlparse(await page.get_url())
# website_name = parsed_url.netloc.replace('www.', '').replace('.', '_').replace(':', '_')
# timestamp = int(time.time())
# screenshot_dir = os.path.join(base_path, "tempImgs", f"{website_name}-{timestamp}")
# os.makedirs(screenshot_dir, exist_ok=True)
# output_path = os.path.join(screenshot_dir, params.filename)
# except Exception as e:
# print(e)
# output_path = os.path.join(os.path.abspath('.'), params.filename)
# element = None
# used_selector = None
# error_messages = []
# for selector in params.selectors:
# try:
# element = await page.query_selector(selector)
# if element:
# used_selector = selector
# break
# except Exception as e:
# error_messages.append(f"{selector}: {str(e)}")
# if not element:
# return ActionResult(error=f"Could not find any element using selectors: {params.selectors}. Errors: {'; '.join(error_messages)}")
# try:
# clip_obj = dict(element).get('clip')
# if not clip_obj or clip_obj.get('x') is None:
# return ActionResult(error=f"Could not get element bounds for selector '{used_selector}'")
# try:
# await page.screenshot(
# path=output_path,
# clip={
# 'x': clip_obj['x'] - params.padding,
# 'y': clip_obj['y'] - params.padding,
# 'width': clip_obj['width'] + (2 * params.padding),
# 'height': clip_obj['height'] + (2 * params.padding)
# }
# )
# except Exception as e:
# return ActionResult(error=f"Screenshot failed for selector '{used_selector}': {str(e)}")
# success_msg = f"Element screenshot saved at: {output_path} (selector: '{used_selector}')"
# return ActionResult(
# extracted_content=success_msg,
# include_in_memory=True,
# long_term_memory=f"Element screenshot taken: {used_selector} -> {output_path}",
# vision_content=[{"type": "image", "path": output_path}]
# )
# except PlaywrightTimeoutError:
# return ActionResult(error=f"Element screenshot failed: Timeout waiting for element '{used_selector}' to be visible or stable.")
# except Exception as e:
# return ActionResult(error=f"Element screenshot failed for selector '{used_selector}': {str(e)}")
# @tools.action(
# description="""
# Finds and returns a web page element based on a natural language description.
# Captures a base64 screenshot of the element and returns it along with the backend_node_id.
# Args:
# params (FindElementByPromptParams): Parameters containing the query string.
# browser_session (BrowserSession): The active browser session.
# Returns:
# ActionResult: Result containing element_base64, backend_node_id, and screenshot_path.
# """,
# param_model=FindElementByPromptParams,
# )
# async def find_element_by_prompt(params: FindElementByPromptParams, browser_session: BrowserSession) -> ActionResult:
# page = await browser_session.get_current_page()
# try:
# # CRITICAL: Refresh DOM tree BEFORE finding element
# print('🔄 Loading initial DOM tree...')
# await page.dom_service.get_dom_tree(target_id=page._target_id)
# print('✅ Initial DOM tree loaded ---------')
# print(params.query)
# # Find element using prompt
# element = await page.must_get_element_by_prompt(params.query, llm=get_model("browser_agent_openrouter:google/gemini-2.5-flash"))
# print(f'✅ Element found: {element}')
# # Refresh DOM tree again before taking screenshot
# print('🔄 Refreshing DOM tree for screenshot...')
# await page.dom_service.get_dom_tree(target_id=page._target_id)
# print('✅ DOM tree refreshed')
# backend_node_id = element._backend_node_id
# # Create fresh element reference and ensure DOM is ready before screenshot
# try:
# print("🔄 Preparing for screenshot operation...")
# print(f"Backend node ID: {element._backend_node_id}")
# await page.dom_service.get_dom_tree(target_id=page._target_id)
# await page._ensure_session() # Make sure session is active
# fresh_element = await page.get_element(backend_node_id=element._backend_node_id)
# print("✅ Got fresh element reference ",fresh_element)
# # Ensure we have a valid session and document
# session_id = await page._ensure_session()
# # Get a fresh document first
# print("🔄 Getting fresh document...")
# doc_result = await page._client.send.DOM.getDocument(
# session_id=session_id
# )
# root_node = doc_result['root']
# print("✅ Got fresh document root")
# # Now get our element's full node info
# print("🔄 Describing node...")
# node_result = await page._client.send.DOM.describeNode(
# params={'backendNodeId': element._backend_node_id},
# session_id=session_id
# )
# print("✅ Got node description")
# # Take screenshot with fresh context
# print("📸 Taking screenshot with fresh context...")
# element_base64 = await fresh_element.screenshot()
# print("✅ Got screenshot data, ", element_base64[:30], "...")
# if not element_base64:
# raise ValueError("Screenshot returned empty data")
# print('✅ Screenshot captured successfully')
# except Exception as screenshot_error:
# print(f"⚠️ Screenshot error: {type(screenshot_error).__name__}: {screenshot_error}")
# raise
# # Get a fresh reference for basic info
# # print('🔄 Refreshing DOM tree for basic info...')
# # await page.dom_service.get_dom_tree(target_id=page._target_id)
# # basic_info_element = await page.get_element(backend_node_id=element.backend_node_id)
# # basic_info = await basic_info_element.get_basic_info()
# # backend_node_id = basic_info.get('backendNodeId')
# # print(f"✅ Element info: backend_node_id={backend_node_id}, nodeName={basic_info.get('nodeName')}")
# # # Try to get bounding box with another fresh reference
# # bbox = None
# # try:
# # print('🔄 Refreshing DOM tree for bounding box...')
# # await page.dom_service.get_dom_tree(target_id=page._target_id)
# # bbox_element = await page.get_element(backend_node_id=backend_node_id)
# # bbox = await bbox_element.get_bounding_box()
# # print(f"✅ Element bounding box: {bbox}")
# # except Exception as bbox_error:
# # print(f"⚠️ Could not get bounding box: {bbox_error}")
# # # Bbox is optional, we can continue without it
# # Save the base64 screenshot to tempImgs folder
# try:
# session_base = getattr(browser_session, 'file_system_path', None)
# base_path = session_base if session_base else os.getcwd()
# screenshot_dir = os.path.join(base_path, "tempImgs", "finding")
# os.makedirs(screenshot_dir, exist_ok=True)
# # Save base64 to file
# screenshot_filename = f"element_{backend_node_id}.png"
# screenshot_path = os.path.join(screenshot_dir, screenshot_filename)
# # Decode base64 and save as PNG file
# try:
# import base64
# screenshot_data = base64.b64decode(element_base64)
# with open(screenshot_path, 'wb') as f:
# f.write(screenshot_data)
# print(f"Screenshot saved to: {screenshot_path}")
# except Exception as save_error:
# print(f"Error saving screenshot: {save_error}")
# screenshot_path = None
# except Exception as e:
# screenshot_path = None
# result_data = {
# "element_base64": element_base64,
# "backend_node_id": backend_node_id,
# "screenshot_path": screenshot_path,
# }
# # Wrap the dictionary in an ActionResult
# return ActionResult(extracted_content=json.dumps(result_data))
# except Exception as e:
# error_data = {
# "element_base64": None,
# "backend_node_id": None,
# "screenshot_path": None,
# "reason": f"llm_error: {e}"
# }
# # Also wrap errors in an ActionResult
# return ActionResult(error=json.dumps(error_data))
# @tools.action(
# description="""
# Adds or removes a visual highlight (red border) around a specified element.
# Args:
# params (HighlightElementParams): Parameters containing selector and remove flag.
# browser_session (BrowserSession): The active browser session.
# Returns:
# dict: Result containing ok, selector, error.
# """,
# param_model=HighlightElementParams,
# )
# async def highlight_element(params: HighlightElementParams, browser_session: BrowserSession) -> dict:
# page = await browser_session.get_current_page()
# selector = params.selector
# remove = params.remove
# try:
# element = await page.query_selector(selector)
# return {"ok": True, "selector": selector}
# except Exception as e:
# return {"ok": False, "selector": selector, "error": str(e)}
# @tools.action(
# description="""
# Gets the position and dimensions of an element on the page.
# Args:
# params (GetBoundingBoxParams): Parameters containing selector.
# browser_session (BrowserSession): The active browser session.
# Returns:
# dict: Element's bounding box containing x, y, width, height, error.
# """,
# param_model=GetBoundingBoxParams,
# )
# async def get_bounding_box(params: GetBoundingBoxParams, browser_session: BrowserSession) -> dict:
# page = await browser_session.get_current_page()
# selector = params.selector
# js = """
# (sel) => {
# const el = document.querySelector(sel);
# if (!el) return { x: null, y: null, width: null, height: null };
# const r = el.getBoundingClientRect();
# return { x: r.x, y: r.y, width: r.width, height: r.height };
# }
# """
# try:
# return await page.evaluate(js, selector)
# except Exception as e:
# return {"error": str(e)}
# @tools.action(
# description="""
# Takes a screenshot of a specific region of the webpage using provided coordinates.
# Args:
# params (ElementScreenshotClipParams): Parameters containing clip and filename.
# browser_session (BrowserSession, optional): The active browser session.
# Returns:
# dict: Result containing ok, path, clip, error.
# """,
# param_model=ElementScreenshotClipParams,
# )
# async def element_screenshot_clip(params: ElementScreenshotClipParams, browser_session: BrowserSession = None) -> dict:
# if browser_session is None:
# return {"ok": False, "error": "No browser session provided"}
# page = await browser_session.get_current_page()
# try:
# # Set up the output path
# session_base = getattr(browser_session, 'file_system_path', None)
# base_path = session_base if session_base else os.getcwd()
# # Create unique subdirectory for this session's screenshots
# parsed_url = urlparse(await page.get_url())
# website_name = parsed_url.netloc.replace('www.', '').replace('.', '_').replace(':', '_')
# timestamp = int(time.time())
# screenshot_dir = os.path.join(base_path, "tempImgs", f"{website_name}-{timestamp}")
# os.makedirs(screenshot_dir, exist_ok=True)
# output_path = os.path.join(screenshot_dir, params.filename)
# # Take the screenshot
# await page.screenshot(path=output_path, clip=clip)
# return {
# "ok": True,
# "path": output_path,
# "clip": clip
# }
# except Exception as e:
# return {"ok": False, "error": str(e)}
# @tools.action(
# description="""
# Verifies that a screenshot corresponds to the element found by a natural language query.
# Args:
# params (VerifyElementVisualParams): Parameters containing query, screenshot_path, tolerance.
# browser_session (BrowserSession): The active browser session.
# Returns:
# dict: Verification result containing verified, selector, bbox, screenshot, reason.
# """,
# param_model=VerifyElementVisualParams,
# )
# async def verify_element_visual(params: VerifyElementVisualParams, browser_session: BrowserSession) -> dict:
# found = await find_element_by_prompt(query, browser_session)
# selector = found.get('selector')
# if not selector:
# return {"verified": False, "reason": f"Could not find element for query: {query}"}
# bbox = await get_bounding_box(selector, browser_session)
# if not bbox or bbox.get('width') is None:
# return {"verified": False, "reason": "Could not get element bounds"}
# try:
# img = Image.open(screenshot_path)
# w, h = img.size
# except Exception as e:
# return {"verified": False, "reason": f"Could not load screenshot: {e}"}
# bw = int(round(bbox['width']))
# bh = int(round(bbox['height']))
# if abs(bw - w) <= tolerance and abs(bh - h) <= tolerance:
# return {
# "verified": True,
# "selector": selector,
# "bbox": bbox,
# "screenshot": screenshot_path
# }
# else:
# return {
# "verified": False,
# "reason": f"Size mismatch: element {bw}x{bh} vs screenshot {w}x{h}"
# }
# class FindByAccessibilityParams(BaseModel):
# query: str = Field(..., description="Natural language description of the element to find (e.g., 'main navigation menu', 'login button').")
# @tools.action(
# description="""
# Finds elements using accessibility tree for more accurate semantic matching.
# This approach mimics how screen readers interpret the page, making it more reliable
# for finding semantically meaningful elements.
# Args:
# params (FindByAccessibilityParams): Parameters containing the search query
# browser_session (BrowserSession): The active browser session
# Returns:
# ActionResult: Contains element details including role, name, and selector
# """,
# param_model=FindByAccessibilityParams
# )
# async def find_by_accessibility(params: FindByAccessibilityParams, browser_session: BrowserSession) -> ActionResult:
# page = await browser_session.get_current_page()
# try:
# # Get accessibility snapshot
# print("🔄 Getting accessibility snapshot...")
# accessibility_tree = await page.accessibility.snapshot()
# if not accessibility_tree:
# return ActionResult(error="Failed to get accessibility snapshot")
# print("✅ Got accessibility snapshot")
# # Use LLM to analyze the tree
# print("🧠 Analyzing with LLM...")
# llm = get_model("browser_agent_openrouter:google/gemini-2.5-flash")
# prompt = f"""
# Given this accessibility tree:
# {json.dumps(accessibility_tree, indent=2)}
# Find the element that best matches this description: "{params.query}"
# Return a JSON object with these fields:
# - role: The ARIA role of the element
# - name: The accessible name
# - selector: A CSS selector that uniquely identifies this element
# - confidence: Number between 0-100 indicating match confidence
# - reasoning: Brief explanation of why this element matches
# Example:
# {{
# "role": "button",
# "name": "Sign up",
# "selector": "#signup-button",
# "confidence": 95,
# "reasoning": "Exact match for button with 'Sign up' text"
# }}
# """
# analysis = await llm(prompt)
# print("✅ LLM analysis complete")
# try:
# result = json.loads(analysis)
# except json.JSONDecodeError:
# return ActionResult(error="Failed to parse LLM response as JSON")
# # Verify the element exists
# print("🔍 Verifying element...")
# element = None
# if result.get("selector"):
# try:
# element = await page.query_selector(result["selector"])
# except Exception as e:
# print(f"⚠️ Selector verification failed: {e}")
# if not element and result.get("role") and result.get("name"):
# try:
# element = await page.get_by_role(result["role"], name=result["name"]).element_handle()
# except Exception as e:
# print(f"⚠️ Role/name verification failed: {e}")
# if not element:
# return ActionResult(error="Could not verify element existence")
# print("✅ Element verified")
# # Just return the element if found
# return ActionResult(
# extracted_content=json.dumps({
# "element": {
# "role": result.get("role"),
# "name": result.get("name"),
# "selector": result.get("selector")
# }
# }),
# include_in_memory=True
# )
# except Exception as e:
# return ActionResult(
# error=f"Accessibility search failed: {str(e)}"
# )
|