def crawl_voice_of_nigeria(crawler, out, urlprefix):
assert urlprefix.startswith('/'), urlprefix
assert urlprefix.endswith('/'), urlprefix
site = urljoin('http://von.gov.ng/', urlprefix)
for url in sorted(find_wordpress_urls(crawler, site)):
doc = crawler.fetch(url)
if doc.status != 200:
continue
html = doc.content.decode('utf-8')
title = re.search(r'<h1[^>]*>(.+?)</h1>', html, re.DOTALL).group(1)
pubdate = re.search(r'<meta itemprop="dateModified" content="(.+?)"',
html)
if pubdate is None: # only a few pages with little content
continue
pubdate = cleantext(pubdate.group(1))
content = re.split('<p[^>]*>', html, 1)[1].split('<footer', 1)[0]
content = content.replace('\n', ' ').replace('</p>', '\n')
paras = [title] + content.splitlines()
paras = filter(None, [cleantext(p) for p in paras])
if not paras:
continue
out.write('# Location: %s\n' % url)
out.write('# Genre: News\n')
if pubdate:
out.write('# Publication-Date: %s\n' % pubdate)
out.write('\n'.join(paras) + '\n')
def find_wordpress_urls(crawler, site):
urls = set()
mainpage = crawler.fetch(site).content.decode('utf-8')
for category in re.findall(r'/(category/[^/"]+/)">', mainpage):
caturl = urljoin(site, category)
catdoc = crawler.fetch(caturl)
assert catdoc.status == 200, (catdoc.status, caturl)
pages = [int(n) for n in re.findall(r'/page/(\d)+/', catdoc.content)]
for page in range(1, 1 + max([0] + pages)):
pgurl = urljoin(caturl, 'page/%d/' % page) if page > 1 else caturl
pgdoc = crawler.fetch(pgurl)
assert pgdoc.status == 200, (pgdoc.status, pgurl)
pgcontent = pgdoc.content.decode('utf-8')
for url in re.findall(r'"(%s[^/"]+/)"' % site, pgcontent):
if not url.endswith('/feed/'):
urls.add(url)
return urls
def find_wordpress_urls(crawler, site):
urls = set()
mainpage = crawler.fetch_content(site)
for category in re.findall(r'/(category/[^/"]+/)">', mainpage):
caturl = urljoin(site, category)
catdoc = crawler.fetch(caturl)
assert catdoc.status == 200, (catdoc.status, caturl)
pages = [int(n) for n in re.findall(r'/page/(\d)+/', catdoc.content)]
for page in range(1, 1 + max([0] + pages)):
pgurl = urljoin(caturl, 'page/%d/' % page) if page > 1 else caturl
pgdoc = crawler.fetch(pgurl)
if pgdoc.status != 200:
print('Error %3d: %s' % (pgdoc.status, pgurl))
continue
pgcontent = pgdoc.content.decode('utf-8')
for url in re.findall(r'"(%s[^"]+)"' % site, pgcontent):
url = replace_html_entities(url.split('#')[0])
if url.find('/category/') < 0 and not url.endswith('/feed/'):
urls.add(url)
return urls
def crawl_voice_of_america(self, out, host, ignore_ascii=False):
site = 'https://%s' % host
sitemap = self.fetch_sitemap(urljoin(site, 'sitemap.xml'))
for url in sorted(sitemap.keys()):
doc = self.fetch(url)
if doc.status != 200:
continue
try:
html = doc.content.decode('utf-8')
except UnicodeDecodeError:
continue
title = re.search(r'<title>(.+?)</title>', html)
title = title.group(1) if title else ''
pubdate = re.search(
r'<div class="published">\s*<span class="date"\s*>\s*'
r'<time datetime="(.+?)"', html)
pubdate = cleantext(pubdate.group(1)) if pubdate else ''
if pubdate.startswith('1900'):
pubdate = ''
description = re.search(
r'<meta name="description" content="(.+?)"', html)
description = description.group(1) if description else ''
if description == title:
description = ''
paragraphs = [title, description]
if html.find('<div class="intro content-offset">') > 0:
intro = html.split('<div class="intro content-offset">', 1)[1]
intro = intro.split('</div')[0]
intro = intro.replace('</p>', '\n').replace('</P>', '\n')
paragraphs.extend(intro.splitlines())
if html.find('<div class="wsw">') > 0:
content = html.split('<div class="wsw">', 1)[1]
content = content.split('<div')[0]
content = content.replace('</p>', '\n').replace('</P>', '\n')
paragraphs.extend(content.splitlines())
paragraphs = filter(None, [cleantext(p) for p in paragraphs])
paragraphs = [p for p in paragraphs if not p.startswith('VOA')]
if ignore_ascii:
paragraphs = [
p for p in paragraphs
if not (ord(p[0]) >= 0x30 and ord(p[0]) <= 0xff)
]
if len(paragraphs) > 0:
out.write('# Location: %s\n' % url)
out.write('# Genre: News\n')
if pubdate:
out.write('# Publication-Date: %s\n' % pubdate)
out.write('\n'.join(paragraphs) + '\n')
def urlallowed(url):
if CONFIG['skip-robots-txt']:
return True
protocol, domain = urlparse.urlparse(url)[:2]
for bd in ROBOTS_TXT_BLACKLIST_DOMAINS:
if re.match(bd, domain):
return True
for d in ['sourceforge', 'berlios', 'github.com']:
if d in domain:
return True
if protocol == 'ftp':
return True
baseurl = '%s://%s' % (protocol, domain)
robotsurl = urlparse.urljoin(baseurl, 'robots.txt')
if rpcache.has_key(baseurl):
rp = rpcache[baseurl]
else:
from socket import setdefaulttimeout, getdefaulttimeout
timeout = getdefaulttimeout()
setdefaulttimeout(5)
rp = robotparser.RobotFileParser()
rp.set_url(robotsurl)
try:
rp.read()
rpcache[baseurl] = rp
except:
rp = None
setdefaulttimeout(timeout)
return rp.can_fetch(CONFIG['user-agent'], url) if rp else False
def urlallowed(url):
if CONFIG['skip-robots-txt']:
return True
protocol, domain = urlparse.urlparse(url)[:2]
for bd in ROBOTS_TXT_BLACKLIST_DOMAINS:
if re.match(bd, domain):
return True
for d in ['sourceforge', 'berlios', 'github.com']:
if d in domain:
return True
if protocol == 'ftp':
return True
baseurl = '%s://%s' % (protocol, domain)
robotsurl = urlparse.urljoin(baseurl, 'robots.txt')
if baseurl in rpcache:
rp = rpcache[baseurl]
else:
from socket import setdefaulttimeout, getdefaulttimeout
timeout = getdefaulttimeout()
setdefaulttimeout(5)
rp = robotparser.RobotFileParser()
rp.set_url(robotsurl)
try:
rp.read()
rpcache[baseurl] = rp
except:
rp = None
setdefaulttimeout(timeout)
return rp.can_fetch(CONFIG['user-agent'], url) if rp else True
