150 lines
3.8 KiB
Python
150 lines
3.8 KiB
Python
#!/usr/bin/env python
|
|
'''
|
|
Utilities to provide various misc functions.
|
|
'''
|
|
|
|
import urllib.request
|
|
from bs4 import BeautifulSoup
|
|
from urllib.parse import (urljoin, urlsplit)
|
|
|
|
|
|
class UrlPool(object):
|
|
'''
|
|
Object to manage a pool of URLs.
|
|
'''
|
|
|
|
def __init__(self):
|
|
self.pool = set()
|
|
|
|
def check_duplicate(self, new_url):
|
|
'''
|
|
Checks if a URL exists in the current pool.
|
|
'''
|
|
if new_url in self.pool:
|
|
return True
|
|
else:
|
|
return False
|
|
|
|
def remove_from_pool(self):
|
|
'''
|
|
Remove a URL from the pool and return it to be crawled.
|
|
'''
|
|
return(self.pool.pop())
|
|
|
|
def add_to_pool(self, url):
|
|
self.pool.add(url)
|
|
|
|
def list_pool(self):
|
|
pool = self.pool
|
|
return pool
|
|
|
|
|
|
class WebPage(object):
|
|
'''
|
|
Object to manage common operations required to return
|
|
the data from each individual page.
|
|
'''
|
|
|
|
headers = {'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:61.0) Gecko/20100101 Firefox/61.0'}
|
|
|
|
def __init__(self, url=None, base_url=None):
|
|
self.url = url
|
|
self.base_url = base_url
|
|
|
|
|
|
def get_source(self):
|
|
'''
|
|
Retrieve a page's source.
|
|
'''
|
|
|
|
request = urllib.request.Request(self.url, headers=self.headers)
|
|
page = urllib.request.urlopen(request, timeout=5)
|
|
self.source = page.read()
|
|
|
|
|
|
def find_links(self):
|
|
'''
|
|
Find all URLs on a page and ensure they are absolute. If they are
|
|
relative then they will be appended to the base URL.
|
|
'''
|
|
hrefs = set()
|
|
|
|
soup = BeautifulSoup(self.source, 'html.parser')
|
|
links = soup.find_all('a', href=True)
|
|
|
|
for link in links:
|
|
if link['href'].startswith('/'):
|
|
hrefs.add(urljoin(self.url, link['href']))
|
|
else:
|
|
hrefs.add(link['href'])
|
|
|
|
self.discovered_hrefs = hrefs
|
|
|
|
|
|
def parse_urls(self):
|
|
'''
|
|
Iterate through the list of discovered URLs and add them to the
|
|
pool if they start with the base URL.
|
|
'''
|
|
self.urls_to_crawl = set()
|
|
|
|
for url in self.discovered_hrefs:
|
|
if url.startswith(self.url):
|
|
sanitised_url = sanitise_url(url=url)
|
|
self.urls_to_crawl.add(sanitised_url)
|
|
|
|
|
|
def list_urls(self):
|
|
'''
|
|
Returns the contents of the
|
|
'''
|
|
|
|
return self.urls_to_crawl
|
|
|
|
|
|
def run(self):
|
|
try:
|
|
self.get_source()
|
|
except Exception as e:
|
|
print(e)
|
|
|
|
try:
|
|
self.find_links()
|
|
except Exception as e:
|
|
print(e)
|
|
|
|
try:
|
|
self.parse_urls()
|
|
except Exception as e:
|
|
print(e)
|
|
|
|
|
|
def sanitise_url(url, base_url=False):
|
|
'''
|
|
If `base_url` is True, we attempt to standardise `url` to ensure it can be
|
|
prepended to relative URLs. If no scheme has been provided then we default
|
|
to http as any sane https-only site should 301 redirect http > https.
|
|
|
|
If `base_url` is False, we sanitise URLs to strip queries and fragments (we
|
|
don't want to scrape in-page anchors etc).
|
|
|
|
Returns a sanitised URL as a string.
|
|
'''
|
|
default_proto = 'http'
|
|
delim = '://'
|
|
|
|
split_url = urlsplit(url)
|
|
|
|
if base_url:
|
|
# This will sanitise the initial url for the initial page crawl.
|
|
if split_url.scheme and split_url.scheme.startswith('http'):
|
|
sanitised_url = "".join([split_url.scheme, delim, split_url.netloc])
|
|
elif (split_url.path and not split_url.scheme and not split_url.netloc):
|
|
sanitised_url = "".join([default_proto, delim, split_url.path])
|
|
else:
|
|
# Sanitise discovered URLs. We already expect them in the format
|
|
# protocol://base_url/path
|
|
sanitised_url = "".join([split_url.scheme, delim, split_url.netloc, split_url.path])
|
|
|
|
return sanitised_url
|