def parse_links(markup):
"""Parse to add target, nofollow, and include missing TLDs"""
# Add missing top-level domains to linkfify.
# Workaround from https://github.com/mozilla/bleach/issues/519
tlds = linkifier.TLDS
tlds.append(u'chat')
tlds.append(u'cloud')
tlds.append(u'community')
tlds.append(u'fund')
tlds.append(u'to')
tlds.append(u'today')
improved_url_re = build_url_re(tlds=tlds)
def set_target(attrs, new=False):
p = urllib.parse.urlparse(attrs[(None, 'href')])
# TODO: get URL from request to figure out if it's an internal link.
if p.netloc not in ['blender.community']:
attrs[(None, 'target')] = '_blank'
attrs[(None, 'rel')] = 'nofollow'
attrs[(None, 'class')] = 'is-external'
return attrs
skip_tags = ['code', 'pre']
linker = linkifier.Linker(url_re=improved_url_re, callbacks=[set_target], skip_tags=skip_tags)
markup = linker.linkify(markup)
# markup = parse_phabricator_tasks(markup)
markup = clean_url_path(markup)
return markup
def test_custom_url_re(self):
url_re = build_url_re(["example"])
expected = '<p><a href="https://domain.example" rel="nofollow">https://domain.example</a></p>'
actual = markdown(
"https://domain.example",
extensions=[LinkifyExtension(linker_options={"url_re": url_re})],
)
self.assertEqual(expected, actual)
ALLOWED_ATTRIBUTES = {
'a': ['href', 'title', 'class'],
'abbr': ['title'],
'acronym': ['title'],
'table': ['width'],
'td': ['width', 'align'],
'div': ['class'],
'p': ['class'],
'span': ['class', 'title'],
# Update doc/user/markdown.rst if you change this!
}
ALLOWED_PROTOCOLS = ['http', 'https', 'mailto', 'tel']
URL_RE = build_url_re(tlds=sorted(tld_set, key=len, reverse=True))
EMAIL_RE = build_email_re(tlds=sorted(tld_set, key=len, reverse=True))
def safelink_callback(attrs, new=False):
"""
Makes sure that all links to a different domain are passed through a redirection handler
to ensure there's no passing of referers with secrets inside them.
"""
url = attrs.get((None, 'href'), '/')
if not url_has_allowed_host_and_scheme(url, allowed_hosts=None) and not url.startswith('mailto:') and not url.startswith('tel:'):
signer = signing.Signer(salt='safe-redirect')
attrs[None, 'href'] = reverse('redirect') + '?url=' + urllib.parse.quote(signer.sign(url))
attrs[None, 'target'] = '_blank'
attrs[None, 'rel'] = 'noopener'
ALLOWED_ATTRIBUTES = {
'a': ['href', 'title', 'class'],
'abbr': ['title'],
'acronym': ['title'],
'table': ['width'],
'td': ['width', 'align'],
'div': ['class'],
'p': ['class'],
'span': ['class', 'title'],
# Update doc/user/markdown.rst if you change this!
}
ALLOWED_PROTOCOLS = ['http', 'https', 'mailto', 'tel']
URL_RE = SimpleLazyObject(
lambda: build_url_re(tlds=sorted(tld_set, key=len, reverse=True)))
EMAIL_RE = SimpleLazyObject(
lambda: build_email_re(tlds=sorted(tld_set, key=len, reverse=True)))
def safelink_callback(attrs, new=False):
"""
Makes sure that all links to a different domain are passed through a redirection handler
to ensure there's no passing of referers with secrets inside them.
"""
url = attrs.get((None, 'href'), '/')
if not url_has_allowed_host_and_scheme(
url, allowed_hosts=None) and not url.startswith(
'mailto:') and not url.startswith('tel:'):
signer = signing.Signer(salt='safe-redirect')
print(linker.linkify('abc models.py def'))
# abc models.py def
print('\n' + '-' * 100 + '\n')
linker = Linker(skip_tags=['pre'])
print(linker.linkify('a b c http://example.com d e f'))
# a b c <a href="http://example.com" rel="nofollow">http://example.com</a> d e f
print(linker.linkify('<pre>http://example.com</pre>'))
# <pre>http://example.com</pre>
print('\n' + '-' * 100 + '\n')
only_fish_tld_url_re = build_url_re(tlds=['fish'])
linker = Linker(url_re=only_fish_tld_url_re)
print(linker.linkify('com TLD does not link https://example.com'))
# com TLD does not link https://example.com
print(linker.linkify('fish TLD links https://example.fish'))
# fish TLD links <a href="https://example.fish" rel="nofollow">https://example.fish</a>
print('\n' + '-' * 100 + '\n')
only_https_url_re = build_url_re(protocols=['https'])
linker = Linker(url_re=only_https_url_re)
print(linker.linkify('gopher does not link gopher://example.link'))
# gopher does not link gopher://example.link
ALLOWED_ATTRIBUTES = {
'a': ['href', 'title', 'class'],
'abbr': ['title'],
'acronym': ['title'],
'table': ['width'],
'td': ['width', 'align'],
'div': ['class'],
'p': ['class'],
'span': ['class', 'title'],
# Update doc/user/markdown.rst if you change this!
}
ALLOWED_PROTOCOLS = ['http', 'https', 'mailto', 'tel']
URL_RE = build_url_re(tlds=tld_set)
EMAIL_RE = build_email_re(tlds=tld_set)
def safelink_callback(attrs, new=False):
url = attrs.get((None, 'href'), '/')
if not url_has_allowed_host_and_scheme(
url, allowed_hosts=None) and not url.startswith(
'mailto:') and not url.startswith('tel:'):
signer = signing.Signer(salt='safe-redirect')
attrs[None,
'href'] = reverse('redirect') + '?url=' + urllib.parse.quote(
signer.sign(url))
attrs[None, 'target'] = '_blank'
attrs[None, 'rel'] = 'noopener'
