569 lines
25 KiB
Python
569 lines
25 KiB
Python
from app.utils.browser_utils import safe_browser_operation
|
||
from app.utils.http_utils import fetch_url
|
||
import asyncio
|
||
from bs4 import BeautifulSoup
|
||
|
||
async def visit_url_service(decoded_url):
|
||
"""Service function to visit a URL and get its content"""
|
||
print(f"Visiting URL: {decoded_url}")
|
||
|
||
# Define the operation to perform with the browser
|
||
async def visit_operation(page):
|
||
try:
|
||
response = await page.goto(decoded_url, waitUntil='networkidle2', timeout=30000)
|
||
if not response:
|
||
print(f"Warning: No response object returned for {decoded_url}")
|
||
|
||
# Get page content
|
||
content = await page.content()
|
||
return {"status": "success", "content": content}
|
||
except Exception as e:
|
||
print(f"Error during page navigation: {e}")
|
||
# Try to get content anyway
|
||
try:
|
||
content = await page.content()
|
||
return {"status": "partial", "content": content, "error": str(e)}
|
||
except:
|
||
raise Exception(f"Failed to get page content: {str(e)}")
|
||
|
||
# Perform the operation
|
||
return await safe_browser_operation(decoded_url, visit_operation)
|
||
|
||
async def extract_seo_service(decoded_url):
|
||
"""Service function to extract SEO information from a website"""
|
||
print(f"Extracting SEO from: {decoded_url}")
|
||
|
||
# Define the operation to perform with the browser
|
||
async def seo_operation(page):
|
||
try:
|
||
response = await page.goto(decoded_url, waitUntil='networkidle2', timeout=30000)
|
||
|
||
# Extract SEO information
|
||
seo_data = await page.evaluate('''() => {
|
||
const data = {
|
||
title: document.title || '',
|
||
description: '',
|
||
canonical: '',
|
||
h1: [],
|
||
h2: [],
|
||
images: 0,
|
||
links: 0
|
||
};
|
||
|
||
// Get meta description
|
||
const metaDescription = document.querySelector('meta[name="description"]');
|
||
if (metaDescription) {
|
||
data.description = metaDescription.getAttribute('content') || '';
|
||
}
|
||
|
||
// Get canonical link
|
||
const canonicalLink = document.querySelector('link[rel="canonical"]');
|
||
if (canonicalLink) {
|
||
data.canonical = canonicalLink.getAttribute('href') || '';
|
||
}
|
||
|
||
// Get h1 tags
|
||
document.querySelectorAll('h1').forEach(h1 => {
|
||
const text = h1.innerText.trim();
|
||
if (text) data.h1.push(text);
|
||
});
|
||
|
||
// Get h2 tags
|
||
document.querySelectorAll('h2').forEach(h2 => {
|
||
const text = h2.innerText.trim();
|
||
if (text) data.h2.push(text);
|
||
});
|
||
|
||
// Count images
|
||
data.images = document.querySelectorAll('img').length;
|
||
|
||
// Count links
|
||
data.links = document.querySelectorAll('a').length;
|
||
|
||
return data;
|
||
}''')
|
||
|
||
result = {
|
||
"status": "success",
|
||
"url": decoded_url,
|
||
"seo": seo_data
|
||
}
|
||
|
||
return result
|
||
|
||
except Exception as e:
|
||
print(f"Error during SEO extraction: {e}")
|
||
return {"status": "error", "url": decoded_url, "error": str(e)}
|
||
|
||
# Perform the operation
|
||
return await safe_browser_operation(decoded_url, seo_operation)
|
||
|
||
async def extract_meta_tags_service(decoded_url):
|
||
"""Service function to extract meta tags from a website"""
|
||
print(f"Extracting meta tags from: {decoded_url}")
|
||
|
||
try:
|
||
# Fetch the page content using plain HTTP request
|
||
content, status = await fetch_url(decoded_url)
|
||
|
||
# Parse the HTML
|
||
soup = BeautifulSoup(content, 'html.parser')
|
||
|
||
# Extract all meta tags
|
||
meta_tags = []
|
||
for meta in soup.find_all('meta'):
|
||
attributes = {}
|
||
for attr in meta.attrs:
|
||
attributes[attr] = meta[attr]
|
||
meta_tags.append(attributes)
|
||
|
||
# Extract Open Graph tags
|
||
og_tags = {}
|
||
for meta in soup.find_all('meta', property=lambda x: x and x.startswith('og:')):
|
||
og_tags[meta.get('property')] = meta.get('content')
|
||
|
||
# Extract Twitter card tags
|
||
twitter_tags = {}
|
||
for meta in soup.find_all('meta', attrs={'name': lambda x: x and x.startswith('twitter:')}):
|
||
twitter_tags[meta.get('name')] = meta.get('content')
|
||
|
||
# Get page title
|
||
title = soup.title.string if soup.title else ''
|
||
|
||
result = {
|
||
"status": "success",
|
||
"url": decoded_url,
|
||
"meta_tags": meta_tags,
|
||
"open_graph": og_tags,
|
||
"twitter_card": twitter_tags,
|
||
"title": title
|
||
}
|
||
|
||
return result
|
||
|
||
except Exception as e:
|
||
print(f"Error during meta tag extraction: {e}")
|
||
return {"status": "error", "url": decoded_url, "error": str(e)}
|
||
|
||
async def detect_pagination_service(decoded_url):
|
||
"""Service function to detect pagination on a website"""
|
||
print(f"Detecting pagination on: {decoded_url}")
|
||
|
||
# Define the operation to perform with the browser
|
||
async def pagination_operation(page):
|
||
try:
|
||
# Navigate to the URL
|
||
await page.goto(decoded_url, waitUntil='networkidle2', timeout=30000)
|
||
original_url = page.url
|
||
|
||
# Check for common pagination indicators
|
||
pagination_data = await page.evaluate('''() => {
|
||
const data = {
|
||
hasPagination: false,
|
||
paginationType: null,
|
||
paginationElements: [],
|
||
detectedParameter: null,
|
||
lastPageNumber: null
|
||
};
|
||
|
||
// Look for numbered pagination links (1, 2, 3...)
|
||
const numberedLinks = Array.from(document.querySelectorAll('a, button, span'))
|
||
.filter(el => {
|
||
const text = el.innerText.trim();
|
||
|
||
// check for data-page attribute
|
||
const dataPage = el.getAttribute('data-page');
|
||
if (dataPage) {
|
||
return /^[0-9]+$/.test(dataPage);
|
||
}
|
||
|
||
return /^[0-9]+$/.test(text) &&
|
||
(el.tagName === 'A' || el.onclick ||
|
||
el.closest('button, [role="button"]'));
|
||
});
|
||
|
||
console.log(numberedLinks);
|
||
|
||
// Look for next/prev buttons
|
||
const nextButtons = Array.from(document.querySelectorAll('a, button, [role="button"]'))
|
||
.filter(el => {
|
||
const text = el.innerText.trim().toLowerCase();
|
||
const ariaLabel = el.getAttribute('aria-label')?.toLowerCase() || '';
|
||
const hasNextIcon = el.querySelector('i.fa-chevron-right, i.fa-arrow-right, svg[class*="arrow"], svg[class*="next"]');
|
||
|
||
return text.includes('next') ||
|
||
text.includes('›') ||
|
||
text.includes('»') ||
|
||
text.includes('→') ||
|
||
ariaLabel.includes('next') ||
|
||
hasNextIcon;
|
||
});
|
||
|
||
// Check for pagination containers
|
||
const paginationContainers = Array.from(document.querySelectorAll(
|
||
'.pagination, [class*="pagination"], [class*="pager"], nav[aria-label*="pagination"], [role="navigation"]'
|
||
));
|
||
|
||
// Collect all potential pagination elements
|
||
if (numberedLinks.length > 0) {
|
||
data.hasPagination = true;
|
||
data.paginationType = 'numbered';
|
||
|
||
// Get href attributes or other identifiers from numbered links
|
||
data.paginationElements = numberedLinks.slice(0, 5).map(el => {
|
||
return {
|
||
text: el.innerText.trim(),
|
||
dataPage: el.getAttribute('data-page'),
|
||
href: el.tagName === 'A' ? el.href : null,
|
||
classes: el.className,
|
||
id: el.id
|
||
};
|
||
});
|
||
|
||
// Try to find the last page number
|
||
const pageNumbers = numberedLinks
|
||
.map(el => parseInt(el.innerText.replace(/[^\d]/g, '')))
|
||
.filter(num => !isNaN(num));
|
||
|
||
if (pageNumbers.length > 0) {
|
||
data.lastPageNumber = Math.max(...pageNumbers);
|
||
}
|
||
|
||
// Also look for a "last page" element that might have text like "Last" or "»"
|
||
const lastPageElement = Array.from(document.querySelectorAll('a, button'))
|
||
.find(el => {
|
||
const text = el.innerText.trim().toLowerCase();
|
||
const ariaLabel = el.getAttribute('aria-label')?.toLowerCase() || '';
|
||
return text.includes('last') ||
|
||
text === '»' ||
|
||
ariaLabel.includes('last page');
|
||
});
|
||
|
||
if (lastPageElement && lastPageElement.href) {
|
||
// Try to extract page number from the URL
|
||
try {
|
||
const url = new URL(lastPageElement.href);
|
||
// Check common pagination parameters
|
||
['page', 'p', 'pg'].forEach(param => {
|
||
if (url.searchParams.has(param)) {
|
||
const value = parseInt(url.searchParams.get(param));
|
||
if (!isNaN(value) && (data.lastPageNumber === null || value > data.lastPageNumber)) {
|
||
data.lastPageNumber = value;
|
||
}
|
||
}
|
||
});
|
||
|
||
// Check for path-based pagination (like /page/10)
|
||
const pathMatch = url.pathname.match(/\/(?:page|p)\/(\d+)/i);
|
||
if (pathMatch && pathMatch[1]) {
|
||
const value = parseInt(pathMatch[1]);
|
||
if (!isNaN(value) && (data.lastPageNumber === null || value > data.lastPageNumber)) {
|
||
data.lastPageNumber = value;
|
||
}
|
||
}
|
||
} catch (e) {
|
||
console.error("Error parsing last page URL:", e);
|
||
}
|
||
}
|
||
} else if (nextButtons.length > 0) {
|
||
data.hasPagination = true;
|
||
data.paginationType = 'next-prev';
|
||
|
||
// Get information about next buttons
|
||
data.paginationElements = nextButtons.slice(0, 3).map(el => {
|
||
return {
|
||
text: el.innerText.trim(),
|
||
href: el.tagName === 'A' ? el.href : null,
|
||
classes: el.className,
|
||
id: el.id
|
||
};
|
||
});
|
||
}
|
||
|
||
// Check for URL parameters that might indicate pagination
|
||
const currentUrl = window.location.href;
|
||
const urlParams = new URL(currentUrl).searchParams;
|
||
|
||
// Common pagination parameters
|
||
const paginationParams = ['page', 'p', 'pg', 'offset', 'o', 'from', 'start', 'limit'];
|
||
|
||
for (const param of paginationParams) {
|
||
if (urlParams.has(param)) {
|
||
data.detectedParameter = {
|
||
name: param,
|
||
value: urlParams.get(param)
|
||
};
|
||
break;
|
||
}
|
||
}
|
||
|
||
return data;
|
||
}''')
|
||
|
||
# If pagination is detected, try to navigate to the next page by clicking
|
||
next_page_url = None
|
||
pagination_parameter = None
|
||
url_template = None
|
||
step_size = None
|
||
|
||
if pagination_data['hasPagination']:
|
||
print("Pagination detected, attempting to click on a pagination element")
|
||
|
||
# Always try to click on a pagination element, regardless of type
|
||
clicked = False
|
||
|
||
# First try to click on a numbered link (preferably "2" if we're on page 1)
|
||
try:
|
||
clicked = await page.evaluate('''() => {
|
||
// First try to find and click on a "2" link or button
|
||
const page2Elements = Array.from(document.querySelectorAll('a[href], button, [role="button"]'))
|
||
.filter(el => {
|
||
// Check for text content "2"
|
||
if (el.innerText.trim() === '2') {
|
||
return true;
|
||
}
|
||
|
||
// Check for href with page=2 or similar (for anchor elements)
|
||
if (el.tagName === 'A' && el.href) {
|
||
try {
|
||
const url = new URL(el.href, window.location.origin);
|
||
const paginationParams = ['page', 'p', 'pg', 'offset', 'o', 'from', 'start', 'limit',
|
||
'currentpage', 'pagenum', 'pageNumber', 'paged'];
|
||
|
||
for (const param of paginationParams) {
|
||
if (url.searchParams.has(param) && url.searchParams.get(param) === '2') {
|
||
return true;
|
||
}
|
||
}
|
||
|
||
// Check for path-based pagination like /page/2/
|
||
const pathMatch = url.pathname.match(/\/(page|p)\/2\/?$/i);
|
||
if (pathMatch) {
|
||
return true;
|
||
}
|
||
} catch (e) {}
|
||
}
|
||
|
||
// Check for data attributes that might indicate pagination
|
||
if (el.getAttribute('data-page') === '2' ||
|
||
el.getAttribute('data-pagenumber') === '2' ||
|
||
el.getAttribute('data-page-number') === '2') {
|
||
return true;
|
||
}
|
||
|
||
return false;
|
||
});
|
||
|
||
if (page2Elements.length > 0) {
|
||
console.log("Clicking on page 2 element");
|
||
page2Elements[0].click();
|
||
return true;
|
||
}
|
||
|
||
// If no "2" link found, try any numbered link or button
|
||
const numberedElements = Array.from(document.querySelectorAll('a[href], button, [role="button"]'))
|
||
.filter(el => /^\d+$/.test(el.innerText.trim()));
|
||
|
||
if (numberedElements.length > 0) {
|
||
// Sort by number and get the second one (likely page 2)
|
||
const sorted = numberedElements.sort((a, b) => {
|
||
return parseInt(a.innerText.trim()) - parseInt(b.innerText.trim());
|
||
});
|
||
|
||
// Get the second element if available (page 2), otherwise the first one
|
||
const elementToClick = sorted.length > 1 ? sorted[1] : sorted[0];
|
||
console.log("Clicking on numbered element: " + elementToClick.innerText);
|
||
elementToClick.click();
|
||
return true;
|
||
}
|
||
|
||
// If no numbered links, try next button
|
||
const nextTexts = ['next', '›', '»', '→'];
|
||
const nextElements = Array.from(document.querySelectorAll('a, button, [role="button"]'))
|
||
.filter(el => {
|
||
const text = el.textContent.trim().toLowerCase();
|
||
const ariaLabel = el.getAttribute('aria-label')?.toLowerCase() || '';
|
||
return nextTexts.some(t => text.includes(t)) ||
|
||
ariaLabel.includes('next') ||
|
||
el.querySelector('i.fa-chevron-right, i.fa-arrow-right, svg[class*="arrow"], svg[class*="next"]');
|
||
});
|
||
|
||
if (nextElements.length > 0) {
|
||
console.log("Clicking on next button");
|
||
nextElements[0].click();
|
||
return true;
|
||
}
|
||
|
||
return false;
|
||
}''')
|
||
|
||
if clicked:
|
||
print("Successfully clicked on pagination element")
|
||
# Wait for navigation to complete
|
||
await asyncio.sleep(1)
|
||
next_page_url = page.url
|
||
else:
|
||
print("No clickable pagination element found")
|
||
|
||
except Exception as e:
|
||
print(f"Error clicking on pagination element: {e}")
|
||
|
||
# If we successfully navigated to the next page, analyze the URL difference
|
||
if next_page_url:
|
||
print('Searching for pagination parameter')
|
||
# Parse both URLs
|
||
original_parsed = await page.evaluate(f'''(originalUrl) => {{
|
||
const original = new URL(originalUrl);
|
||
const current = new URL(window.location.href);
|
||
|
||
// Check for differences in query parameters
|
||
let paramDiff = null;
|
||
|
||
// Common pagination parameters to check
|
||
const paginationParams = ['page', 'p', 'pg', 'offset', 'o', 'from', 'start', 'limit', 'currentPage', 'current_page', 'currentpage', 'pagenum', 'pageNumber', 'paged'];
|
||
|
||
for (const param of paginationParams) {{
|
||
const originalValue = original.searchParams.get(param);
|
||
const currentValue = current.searchParams.get(param);
|
||
|
||
if (originalValue !== currentValue && currentValue !== null) {{
|
||
paramDiff = {{
|
||
name: param,
|
||
originalValue: originalValue,
|
||
currentValue: currentValue
|
||
}};
|
||
break;
|
||
}}
|
||
}}
|
||
|
||
// Check for path differences (like /page/1 vs /page/2)
|
||
const originalPath = original.pathname;
|
||
const currentPath = current.pathname;
|
||
|
||
let pathDiff = null;
|
||
if (originalPath !== currentPath) {{
|
||
const originalSegments = originalPath.split('/').filter(s => s);
|
||
const currentSegments = currentPath.split('/').filter(s => s);
|
||
|
||
// Find the segment that changed
|
||
if (originalSegments.length === currentSegments.length) {{
|
||
for (let i = 0; i < originalSegments.length; i++) {{
|
||
if (originalSegments[i] !== currentSegments[i]) {{
|
||
// Check if the difference is numeric
|
||
if (!isNaN(originalSegments[i]) && !isNaN(currentSegments[i])) {{
|
||
pathDiff = {{
|
||
index: i,
|
||
originalValue: originalSegments[i],
|
||
currentValue: currentSegments[i]
|
||
}};
|
||
}}
|
||
}}
|
||
}}
|
||
}}
|
||
}}
|
||
|
||
return {{
|
||
paramDiff,
|
||
pathDiff,
|
||
originalUrl: originalUrl,
|
||
currentUrl: window.location.href
|
||
}};
|
||
}}''', original_url)
|
||
|
||
# Determine the pagination parameter and create URL template
|
||
print(original_parsed)
|
||
if original_parsed['paramDiff']:
|
||
param_name = original_parsed['paramDiff']['name']
|
||
pagination_parameter = {
|
||
'type': 'query',
|
||
'name': param_name,
|
||
'value': original_parsed['paramDiff']['currentValue']
|
||
}
|
||
|
||
# --- STEP SIZE DETECTION FOR QUERY PARAM ---
|
||
orig_val = original_parsed['paramDiff']['originalValue'] if original_parsed['paramDiff']['originalValue'] is not None else 0
|
||
curr_val = original_parsed['paramDiff']['currentValue']
|
||
try:
|
||
if orig_val is not None and curr_val is not None:
|
||
orig_num = int(orig_val)
|
||
curr_num = int(curr_val)
|
||
step_size = abs(curr_num - orig_num)
|
||
except Exception:
|
||
step_size = None
|
||
|
||
# Create URL template for query parameter
|
||
url_obj = await page.evaluate(f'''(url, paramName) => {{
|
||
const urlObj = new URL(url);
|
||
urlObj.searchParams.set(paramName, "{{PAGE_NUMBER}}");
|
||
return urlObj.toString();
|
||
}}''', original_url, param_name)
|
||
|
||
url_template = url_obj
|
||
|
||
elif original_parsed['pathDiff']:
|
||
path_index = original_parsed['pathDiff']['index']
|
||
pagination_parameter = {
|
||
'type': 'path',
|
||
'index': path_index,
|
||
'value': original_parsed['pathDiff']['currentValue']
|
||
}
|
||
|
||
# --- STEP SIZE DETECTION FOR PATH PARAM ---
|
||
orig_val = original_parsed['pathDiff']['originalValue']
|
||
curr_val = original_parsed['pathDiff']['currentValue']
|
||
try:
|
||
if orig_val is not None and curr_val is not None:
|
||
orig_num = int(orig_val)
|
||
curr_num = int(curr_val)
|
||
step_size = abs(curr_num - orig_num)
|
||
except Exception:
|
||
step_size = None
|
||
|
||
# Create URL template for path parameter
|
||
url_template = await page.evaluate(f'''(url, pathIndex) => {{
|
||
const urlObj = new URL(url);
|
||
const pathSegments = urlObj.pathname.split('/').filter(s => s);
|
||
pathSegments[pathIndex] = "{{PAGE_NUMBER}}";
|
||
urlObj.pathname = '/' + pathSegments.join('/');
|
||
return urlObj.toString();
|
||
}}''', original_url, path_index)
|
||
|
||
if url_template:
|
||
# Decode URL-encoded characters in the template
|
||
import urllib.parse
|
||
url_template = urllib.parse.unquote(url_template)
|
||
|
||
# If we couldn't determine the URL template from navigation, try to infer it
|
||
if not url_template and pagination_data['detectedParameter']:
|
||
param_name = pagination_data['detectedParameter']['name']
|
||
url_template = await page.evaluate(f'''(url, paramName) => {{
|
||
const urlObj = new URL(url);
|
||
urlObj.searchParams.set(paramName, "{{PAGE_NUMBER}}");
|
||
return urlObj.toString();
|
||
}}''', original_url, param_name)
|
||
|
||
# Return the pagination detection results with a simplified structure
|
||
result = {
|
||
"status": "success",
|
||
"hasPagination": pagination_data['hasPagination'],
|
||
"urlTemplate": url_template,
|
||
"lastPage": pagination_data['lastPageNumber'],
|
||
"stepSize": step_size if step_size is not None and step_size >= 5 else 1
|
||
}
|
||
|
||
return result
|
||
|
||
except Exception as e:
|
||
print(f"Error during pagination detection: {e}")
|
||
return {
|
||
"status": "error",
|
||
"hasPagination": False,
|
||
"urlTemplate": None,
|
||
"lastPage": None,
|
||
"stepSize": None,
|
||
"error": str(e)
|
||
}
|
||
|
||
# Perform the operation
|
||
return await safe_browser_operation(decoded_url, pagination_operation)
|