CalvinBackup/anonymous_github
1
0
Fork 0
mirror of https://github.com/tdurieux/anonymous_github.git synced 2026-02-13 02:42:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

Anonymize whole URLs (#19)

Browse Source
This commit is contained in:
Remi Rampin
2019-03-12 12:22:38 -04:00
committed by Thomas Durieux
parent c4d88758a4
commit b2f44417d3
1 changed files with 14 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
server.py
View File

@@ -93,7 +93,7 @@ class Anonymous_Github:
application.jinja_env.add_extension('jinja2.ext.do')
@application.template_filter('remove_terms', )
def remove_terms(content, repository_configuration, word_boundaries=True):
def remove_terms(content, repository_configuration, word_boundaries=True, whole_urls=True):
"""
remove the blacklisted terms from the content
:param content: the content to anonymize
@@ -108,9 +108,20 @@ class Anonymous_Github:
content = re.compile(repo, re.IGNORECASE).sub("%s/repository/%s" % (self.public_url, repository_configuration["id"]), content)
for term in repository_configuration['terms']:
if word_boundaries:
content = re.compile(r'\b%s\b' % term, re.IGNORECASE).sub("XXX", content)
regex = re.compile(r'\b%s\b' % term, re.IGNORECASE)
else:
content = re.compile(term, re.IGNORECASE).sub("XXX", content)
regex = re.compile(term, re.IGNORECASE)
if whole_urls:
def sub_url(m):
if regex.search(m.group(0)):
return 'XXX'
return m.group(0)
url_regex = re.compile('\\b((https?|ftp|file)://)[-A-Za-z0-9+&@#/%?=~_|!:,.;]+[-A-Za-z0-9+&@#/%=~_|]\\b')
content = url_regex.sub(sub_url, content)
content = regex.sub("XXX", content)
return content
@application.template_filter('file_render', )