def test_fix_relative():
assert fix_relative_urls('https://example.org',
'page.html') == 'https://example.org/page.html'
assert fix_relative_urls(
'http://example.org',
'//example.org/page.html') == 'http://example.org/page.html'
assert fix_relative_urls('https://example.org',
'./page.html') == 'https://example.org/page.html'
assert fix_relative_urls('https://example.org',
'/page.html') == 'https://example.org/page.html'
# fixing partial URLs
assert fix_relative_urls(
'https://example.org',
'https://example.org/test.html') == 'https://example.org/test.html'
assert fix_relative_urls('https://example.org',
'/test.html') == 'https://example.org/test.html'
assert fix_relative_urls(
'https://example.org',
'//example.org/test.html') == 'https://example.org/test.html'
assert fix_relative_urls(
'http://example.org',
'//example.org/test.html') == 'http://example.org/test.html'
assert fix_relative_urls('https://example.org',
'test.html') == 'https://example.org/test.html'
assert fix_relative_urls(
'https://example.org',
'../../test.html') == 'https://example.org/test.html'
def refresh_detection(htmlstring, homepage):
"Check if there could be a redirection by meta-refresh tag."
if '"refresh"' in htmlstring or '"REFRESH"' in htmlstring:
try:
html_tree = load_html(htmlstring)
# test meta-refresh redirection
# https://stackoverflow.com/questions/2318446/how-to-follow-meta-refreshes-in-python
attr = html_tree.xpath(
'//meta[@http-equiv="refresh"]/@content|//meta[@http-equiv="REFRESH"]/@content'
)[0]
_, text = attr.split(';')
text = text.strip().lower()
if text.startswith('url=') or text.startswith('URL='):
url2 = text[4:]
if not url2.startswith('http'):
# Relative URL, adapt
_, base_url = get_hostinfo(url2)
url2 = fix_relative_urls(base_url, url2)
# second fetch
newhtmlstring = fetch_url(url2)
if newhtmlstring is None:
logging.warning('failed redirect: %s', url2)
return None, None
#else:
htmlstring, homepage = newhtmlstring, url2
logging.info('successful redirect: %s', url2)
except (IndexError, etree.ParserError, etree.XMLSyntaxError,
etree.XPathEvalError) as err:
logging.info('no redirect found: %s %s', homepage, err)
return htmlstring, homepage
def handle_link(link, sitemapurl, domainname, baseurl, target_lang):
'''Examine a link and determine if it's valid and if it leads to
a sitemap or a web page.'''
state = '0'
# safety net: recursivity
if link == sitemapurl:
return link, state
# fix and check
link = fix_relative_urls(baseurl, link)
# clean and normalize
link = clean_url(link, target_lang)
if link is not None:
if lang_filter(link, target_lang) is True:
newdomain = extract_domain(link)
if newdomain is not None:
# don't take links from another domain and make an exception for main platforms
if newdomain != domainname and not WHITELISTED_PLATFORMS.search(newdomain):
LOGGER.warning('Diverging domain names: %s %s', domainname, newdomain)
else:
if DETECT_SITEMAP_LINK.search(link):
state = 'sitemap'
else:
state = 'link'
else:
LOGGER.error("Couldn't extract domain: %s", link)
return link, state
def determine_feed(htmlstring, baseurl, reference):
'''Try to extract the feed URL from the home page.
Adapted from http://www.aaronsw.com/2002/feedfinder/'''
# parse the page to look for feeds
tree = load_html(htmlstring)
# safeguard
if tree is None:
LOGGER.debug('Invalid HTML/Feed page: %s', baseurl)
return []
feed_urls = []
for linkelem in tree.xpath('//link[@rel="alternate"]'):
# discard elements without links
if 'href' not in linkelem.attrib:
continue
# most common case
if 'type' in linkelem.attrib and linkelem.get('type') in FEED_TYPES:
feed_urls.append(linkelem.get('href'))
# websites like geo.de
elif 'atom' in linkelem.get('href') or 'rss' in linkelem.get('href'):
feed_urls.append(linkelem.get('href'))
# backup
if not feed_urls:
for linkelem in tree.xpath('//a[@href]'):
if linkelem.get('href')[-4:].lower() in ('.rss', '.rdf', '.xml'):
feed_urls.append(linkelem.get('href'))
elif linkelem.get('href')[-5:].lower() == '.atom':
feed_urls.append(linkelem.get('href'))
elif 'atom' in linkelem.get('href') or 'rss' in linkelem.get(
'href'):
feed_urls.append(linkelem.get('href'))
# refine
output_urls = []
for link in sorted(set(feed_urls)):
link = fix_relative_urls(baseurl, link)
link = clean_url(link)
if link == reference or validate_url(link)[0] is False:
continue
if BLACKLIST.search(link):
continue
output_urls.append(link)
# log result
LOGGER.debug('Feed URLs found: %s of which %s valid', len(feed_urls),
len(output_urls))
return output_urls
def handle_link_list(linklist, domainname, baseurl, target_lang=None):
'''Examine links to determine if they are valid and
lead to a web page'''
output_links = []
# sort and uniq
for item in sorted(set(linklist)):
# fix and check
link = fix_relative_urls(baseurl, item)
# control output for validity
checked = check_url(link, language=target_lang)
if checked is not None:
output_links.append(checked[0])
if checked[1] != domainname:
LOGGER.warning('Diverging domain names: %s %s', domainname,
checked[1])
# Feedburner/Google feeds
elif 'feedburner' in item or 'feedproxy' in item:
output_links.append(item)
return output_links
def extract_robots_sitemaps(robotstxt, baseurl):
'Read a robots.txt file and find sitemap links.'
# sanity check on length (cause: redirections)
if robotstxt is None or len(robotstxt) > 10000:
return []
sitemapurls = []
# source: https://github.com/python/cpython/blob/3.8/Lib/urllib/robotparser.py
for line in robotstxt.splitlines():
# remove optional comment and strip line
i = line.find('#')
if i >= 0:
line = line[:i]
line = line.strip()
if not line:
continue
line = line.split(':', 1)
if len(line) == 2:
line[0] = line[0].strip().lower()
if line[0] == "sitemap":
# urllib.parse.unquote(line[1].strip())
candidate = fix_relative_urls(baseurl, line[1].strip())
sitemapurls.append(candidate)
LOGGER.debug('%s sitemaps found in robots.txt', len(sitemapurls))
return sitemapurls
