This commit is contained in:
@@ -1,13 +1,13 @@
|
|||||||
from fastapi import APIRouter, HTTPException, Header, Response
|
from fastapi import APIRouter, HTTPException, Header, Response
|
||||||
from typing import Optional
|
from typing import Optional, List, Literal, Union, Annotated
|
||||||
from urllib.parse import unquote
|
from urllib.parse import unquote
|
||||||
from app.config import API_KEY
|
from app.config import API_KEY
|
||||||
from app.database import get_cached_data, save_to_cache
|
from app.database import get_cached_data, save_to_cache
|
||||||
from pydantic import BaseModel
|
from pydantic import BaseModel, Field
|
||||||
from app.services.browser import (
|
from app.services.browser import (
|
||||||
visit_url_service,
|
visit_url_service,
|
||||||
screenshot_url_service,
|
screenshot_url_service,
|
||||||
click_selector_service,
|
interactions_service,
|
||||||
extract_seo_service,
|
extract_seo_service,
|
||||||
extract_meta_tags_service,
|
extract_meta_tags_service,
|
||||||
capture_outgoing_calls_service,
|
capture_outgoing_calls_service,
|
||||||
@@ -16,11 +16,27 @@ from app.services.browser import (
|
|||||||
|
|
||||||
router = APIRouter()
|
router = APIRouter()
|
||||||
|
|
||||||
class ClickRequest(BaseModel):
|
class ClickInteraction(BaseModel):
|
||||||
url: str
|
action: Literal["click"]
|
||||||
selector: str
|
selector: str
|
||||||
# Default behavior: return screenshot after clicking.
|
|
||||||
returnType: Optional[str] = "screenshot"
|
|
||||||
|
class TypeInteraction(BaseModel):
|
||||||
|
action: Literal["type"]
|
||||||
|
selector: str
|
||||||
|
text: str
|
||||||
|
|
||||||
|
|
||||||
|
Interaction = Annotated[
|
||||||
|
Union[ClickInteraction, TypeInteraction],
|
||||||
|
Field(discriminator="action"),
|
||||||
|
]
|
||||||
|
|
||||||
|
|
||||||
|
class InteractionsRequest(BaseModel):
|
||||||
|
url: str
|
||||||
|
returnType: Optional[Literal["screenshot", "html"]] = "screenshot"
|
||||||
|
interactions: List[Interaction]
|
||||||
|
|
||||||
@router.get("/")
|
@router.get("/")
|
||||||
async def visit_url(url: str, skipCache: bool = False, x_api_key: Optional[str] = Header(None)):
|
async def visit_url(url: str, skipCache: bool = False, x_api_key: Optional[str] = Header(None)):
|
||||||
@@ -74,21 +90,19 @@ async def screenshot_url(url: str, fullPage: bool = True, x_api_key: Optional[st
|
|||||||
print(f"Error taking screenshot for URL {decoded_url}: {e}")
|
print(f"Error taking screenshot for URL {decoded_url}: {e}")
|
||||||
raise HTTPException(status_code=500, detail=str(e))
|
raise HTTPException(status_code=500, detail=str(e))
|
||||||
|
|
||||||
@router.post("/click")
|
@router.post("/interactions")
|
||||||
async def click_selector(payload: ClickRequest, x_api_key: Optional[str] = Header(None)):
|
async def run_interactions(payload: InteractionsRequest, x_api_key: Optional[str] = Header(None)):
|
||||||
if not x_api_key or x_api_key != API_KEY:
|
if not x_api_key or x_api_key != API_KEY:
|
||||||
raise HTTPException(status_code=401, detail="Invalid API key")
|
raise HTTPException(status_code=401, detail="Invalid API key")
|
||||||
|
|
||||||
decoded_url = unquote(payload.url)
|
decoded_url = unquote(payload.url)
|
||||||
return_type = (payload.returnType or "screenshot").lower()
|
return_type = (payload.returnType or "screenshot").lower()
|
||||||
if return_type not in ("screenshot", "html"):
|
|
||||||
raise HTTPException(status_code=400, detail="Invalid returnType. Use 'screenshot' or 'html'.")
|
|
||||||
|
|
||||||
try:
|
try:
|
||||||
result = await click_selector_service(decoded_url, payload.selector, return_type)
|
result = await interactions_service(decoded_url, payload.interactions, return_type)
|
||||||
|
|
||||||
if isinstance(result, dict) and result.get("status") == "error":
|
if isinstance(result, dict) and result.get("status") == "error":
|
||||||
raise HTTPException(status_code=500, detail=result.get("error", "Click failed"))
|
raise HTTPException(status_code=500, detail=result.get("error", "Interaction failed"))
|
||||||
|
|
||||||
if return_type == "html":
|
if return_type == "html":
|
||||||
if not isinstance(result, str):
|
if not isinstance(result, str):
|
||||||
@@ -101,7 +115,7 @@ async def click_selector(payload: ClickRequest, x_api_key: Optional[str] = Heade
|
|||||||
except HTTPException:
|
except HTTPException:
|
||||||
raise
|
raise
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
print(f"Error clicking selector on URL {decoded_url}: {e}")
|
print(f"Error running interactions on URL {decoded_url}: {e}")
|
||||||
raise HTTPException(status_code=500, detail=str(e))
|
raise HTTPException(status_code=500, detail=str(e))
|
||||||
|
|
||||||
@router.get("/seo")
|
@router.get("/seo")
|
||||||
|
|||||||
@@ -175,6 +175,85 @@ async def click_selector_service(decoded_url: str, selector: str, return_type: s
|
|||||||
|
|
||||||
return await safe_browser_operation(decoded_url, click_operation)
|
return await safe_browser_operation(decoded_url, click_operation)
|
||||||
|
|
||||||
|
async def interactions_service(decoded_url: str, interactions, return_type: str):
|
||||||
|
"""Open `decoded_url`, execute click/type interactions in order, then return screenshot or HTML."""
|
||||||
|
print(f"Running interactions on {decoded_url} (return_type={return_type})")
|
||||||
|
|
||||||
|
async def interactions_operation(page):
|
||||||
|
try:
|
||||||
|
# Navigate and wait for initial DOM.
|
||||||
|
await page.goto(decoded_url, wait_until="domcontentloaded", timeout=30000)
|
||||||
|
await page.wait_for_load_state("load", timeout=30000)
|
||||||
|
|
||||||
|
# Ensure the page is fully loaded (helps with JS-driven UIs).
|
||||||
|
await page.wait_for_function("document.readyState === 'complete'", timeout=30000)
|
||||||
|
|
||||||
|
# Best-effort cookie/banner handling (helps avoid click interception).
|
||||||
|
cookie_buttons = [
|
||||||
|
"Accept all",
|
||||||
|
"Accepteer",
|
||||||
|
"Accepteren",
|
||||||
|
"Accept",
|
||||||
|
]
|
||||||
|
for button in cookie_buttons:
|
||||||
|
try:
|
||||||
|
element = await page.query_selector(f'text="{button}"')
|
||||||
|
if element:
|
||||||
|
await element.click()
|
||||||
|
await page.wait_for_timeout(1000)
|
||||||
|
break
|
||||||
|
except Exception:
|
||||||
|
pass
|
||||||
|
|
||||||
|
for interaction in interactions:
|
||||||
|
# Support both Pydantic model instances and raw dicts.
|
||||||
|
if isinstance(interaction, dict):
|
||||||
|
action = interaction.get("action")
|
||||||
|
selector = interaction.get("selector")
|
||||||
|
text = interaction.get("text")
|
||||||
|
else:
|
||||||
|
action = getattr(interaction, "action", None)
|
||||||
|
selector = getattr(interaction, "selector", None)
|
||||||
|
text = getattr(interaction, "text", None)
|
||||||
|
|
||||||
|
if not action or not selector:
|
||||||
|
raise ValueError("Interaction must contain action and selector")
|
||||||
|
|
||||||
|
if action == "click":
|
||||||
|
await page.wait_for_selector(selector, timeout=30000)
|
||||||
|
locator = page.locator(selector)
|
||||||
|
await locator.scroll_into_view_if_needed(timeout=30000)
|
||||||
|
await locator.click(force=True, timeout=30000)
|
||||||
|
|
||||||
|
elif action == "type":
|
||||||
|
if text is None:
|
||||||
|
raise ValueError("Type interaction must contain text")
|
||||||
|
await page.wait_for_selector(selector, timeout=30000)
|
||||||
|
locator = page.locator(selector)
|
||||||
|
await locator.scroll_into_view_if_needed(timeout=30000)
|
||||||
|
await locator.fill(text)
|
||||||
|
|
||||||
|
else:
|
||||||
|
raise ValueError(f"Unknown interaction action: {action}")
|
||||||
|
|
||||||
|
# Let the UI settle after each action.
|
||||||
|
await page.wait_for_timeout(1000)
|
||||||
|
try:
|
||||||
|
await page.wait_for_load_state("networkidle", timeout=10000)
|
||||||
|
except Exception:
|
||||||
|
pass
|
||||||
|
|
||||||
|
if return_type == "html":
|
||||||
|
return await page.content()
|
||||||
|
|
||||||
|
return await page.screenshot(full_page=True, type="png")
|
||||||
|
|
||||||
|
except Exception as e:
|
||||||
|
print(f"Error during interactions capture: {e}")
|
||||||
|
return {"status": "error", "url": decoded_url, "error": str(e)}
|
||||||
|
|
||||||
|
return await safe_browser_operation(decoded_url, interactions_operation)
|
||||||
|
|
||||||
async def extract_seo_service(decoded_url):
|
async def extract_seo_service(decoded_url):
|
||||||
"""Service function to extract SEO information from a website"""
|
"""Service function to extract SEO information from a website"""
|
||||||
print(f"Extracting SEO from: {decoded_url}")
|
print(f"Extracting SEO from: {decoded_url}")
|
||||||
|
|||||||
Reference in New Issue
Block a user