backup/pywb
1
0
Fork 0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-15 08:04:49 +01:00
Code Issues Projects Releases Activity

rewrite deprefix: improve query deprefix to also test url-encoded params, closes #119

Browse Source
This commit is contained in:
Ilya Kreymer 2015-07-06 18:19:01 -07:00
parent d2e17bf505
commit 93d49ae24b
2 changed files with 14 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
pywb/rewrite/test/test_url_rewriter.py
View File

@ -114,6 +114,14 @@
>>> do_deprefix('http://example.com/file.html?param=http://localhost:8080/pywb/https%3A//example.com/filename.html&other=value&a=b&param2=http://localhost:8080/pywb/http://test.example.com', '/pywb/', 'http://localhost:8080/pywb/')
'http://example.com/file.html?param=https://example.com/filename.html&other=value&a=b&param2=http://test.example.com'
# urlencoded
>>> do_deprefix('http://example.com/file.html?foo=bar&url=' + urllib.quote_plus('http://localhost:8080/pywb/http://example.com/filename.html') + '&foo2=bar2', '/pywb/', 'http://localhost:8080/pywb/')
'http://example.com/file.html?foo=bar&url=http://example.com/filename.html&foo2=bar2'
# with extra path
>>> do_deprefix('http://example.com/file.html?foo=bar&url=' + urllib.quote_plus('http://localhost:8080/pywb/extra/path/http://example.com/filename.html') + '&foo2=bar2', '/pywb/', 'http://localhost:8080/pywb/')
'http://example.com/file.html?foo=bar&url=http://example.com/filename.html&foo2=bar2'
# HttpsUrlRewriter tests
>>> httpsrewriter = HttpsUrlRewriter('http://example.com/', None)
>>> httpsrewriter.rewrite('https://example.com/abc')

8
pywb/rewrite/wburl.py
View File

@ -243,6 +243,7 @@ class WbUrl(BaseWbUrl):
self.timestamp = res[0]
self.mod = res[1]
self.url = res[2]
if self.timestamp:
self.type = self.REPLAY
else:
@ -256,8 +257,11 @@ class WbUrl(BaseWbUrl):
def deprefix_url(self, prefix):
rex_query = '=' + re.escape(prefix) + '([0-9])*([\w]{2}_)?/?'
new_url = re.sub(rex_query, '=', self.url)
self.url = new_url
self.url = re.sub(rex_query, '=', self.url)
rex_query = '=(' + urllib.quote_plus(prefix) + '.*?)((?:https?%3A)?%2F%2F[^&]+)'
self.url = re.sub(rex_query, '=\\2', self.url)
return self.url
def get_url(self, url=None):