Simple Web Spider in Python

From XPUB & Lens-Based wiki

A web spider that starts at a given URL, and follows links

Opening an network connection with urllib2

import urllib2

request = urllib2.Request("http://www.volkskrant.nl/")
f=urllib2.urlopen(request)

print f.geturl()
print f.info()
print f.read()

Some sites require that you set the "User-Agent" header.

import urllib2

request = urllib2.Request("http://www.volkskrant.nl/")
request.add_header("User-Agent", "Mozilla/5.0 (X11; U; Linux x86_64; fr; rv:1.9.1.5) Gecko/20091109 Ubuntu/9.10 (karmic) Firefox/3.5.5")
f=urllib2.urlopen(request)

Get the URL of all the links in a page and jump to a random page

import random
import urllib2, urlparse, html5lib, lxml
from lxml.cssselect import CSSSelector

targets = ['http://www.volkskrant.nl/']

while True:
    target = random.choice(targets)
    print '*** '+target+' ***'
    request = urllib2.Request(target)
    request.add_header("User-Agent", "Mozilla/5.0 (X11; U; Linux x86_64; fr; rv:1.9.1.5) Gecko/20091109 Ubuntu/9.10 (karmic) Firefox/3.5.5")
    f=urllib2.urlopen(request)

    parser = html5lib.HTMLParser(tree=html5lib.treebuilders.getTreeBuilder("lxml"), namespaceHTMLElements=False)
    page = parser.parse(f)

    targets = []

    for links in CSSSelector('a[href]')(page):
        href = urlparse.urljoin(f.geturl(), links.attrib['href'])
        if href.split(':')[0] == 'http':
            targets.append(href)

Get the URL of all the links in a page and jump to a random page AND BE SMART

import random
import urllib2, urlparse, html5lib, lxml
from lxml.cssselect import CSSSelector

targets = ['http://www.volkskrant.nl/']

while True:
    target = random.choice(targets)
    print '*** '+target+' ***'

    # try to open URL ...
    try:
        request = urllib2.Request(target)
        request.add_header("User-Agent", "Mozilla/5.0 (X11; U; Linux x86_64; fr; rv:1.9.1.5) Gecko/20091109 Ubuntu/9.10 (karmic) Firefox/3.5.5")
        f=urllib2.urlopen(request)

        # Is it really something that I can parse? srsly?
        try:
            parser = html5lib.HTMLParser(tree=html5lib.treebuilders.getTreeBuilder("lxml"), namespaceHTMLElements=False)
            page = parser.parse(f)
        except ValueError, err:
            print "Value Error:", err, target

        links = []
        if CSSSelector('a[href]')(page): # Any links for me?
            for link in CSSSelector('a[href]')(page):
                href = urlparse.urljoin(f.geturl(), link.attrib['href'])
                if href.split(':')[0] == 'http': # No js links
                    links.append(href)
        if links: # Anything left?
            targets = links


    # ... catch HTTP and URL errors
    except urllib2.HTTPError, err:
        print "HTTP Error:",err.code , target
        print "trying other URL"
    except urllib2.URLError, err:
        print "URL Error:",err.reason , target
        print "trying other URL"

Same as above and grab all the pictures found on each page AND BE SMARTER

import random
import urllib2, urlparse, html5lib, lxml
from lxml.cssselect import CSSSelector

useragent = "Mozilla/5.001 (windows; U; NT4.0; en-US; rv:1.0) Gecko/25250101"
targets = ['http://www.volkskrant.nl/']

while True:
    target = random.choice(targets)
    print '*** '+target+' ***'

    # try to open URL ...
    try:
        request = urllib2.Request(target, None, {'User-Agent': useragent})
        f=urllib2.urlopen(request)

        # Is it really something that I can parse? srsly?
        try:
            parser = html5lib.HTMLParser(tree=html5lib.treebuilders.getTreeBuilder("lxml"), namespaceHTMLElements=False)
            page = parser.parse(f)

            links = []
            if CSSSelector('a[href]')(page): # Any links for me?
                for link in CSSSelector('a[href]')(page):
                    href = urlparse.urljoin(f.geturl(), link.attrib['href'])
                    if href.split(':')[0] == 'http' and href != target: # No useless links
                        links.append(href)
            if links: # Anything left?
                targets = links
                for elt in CSSSelector('img[src]')(page):
                    href = urlparse.urljoin(f.geturl(), elt.attrib['src'])
                    request = urllib2.Request(href, None, {'User-Agent': useragent})
                    remotefile = urllib2.urlopen(request)
                    print 'downloading ' + href
                    localfile = open('dump/'+href.split('/')[-1], "wb")
                    localfile.write(remotefile.read())
                    localfile.close()

        except IOError:
            print "Ooops"

        except ValueError:
            print "Ooops"

        except AssertionError:
            print "Ooops"

    # ... catch HTTP and URL errors
    except urllib2.HTTPError, err:
        print "HTTP Error:",err.code , target
        print "trying other URL"
    except urllib2.URLError, err:
        print "URL Error:",err.reason , target
        print "trying other URL"