backup/pywb
1
0
Fork 0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-24 06:59:52 +01:00
Code Issues Projects Releases Activity

js regex rewrite: scheme-rel rewrite must be preceded by a quote no semicolon, to avoid rewriting ;//comment; as url

Browse Source 
add rewrite tests
This commit is contained in:
Ilya Kreymer 2015-05-09 17:58:04 -07:00
parent 8eee469c93
commit 0f5d32e80e
2 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
pywb/rewrite/regex_rewriters.py
View File

@ -113,7 +113,7 @@ class JSLinkRewriterMixin(object):
""" """
#JS_HTTPX = r'(?:(?:(?<=["\';])https?:)|(?<=["\']))\\{0,4}/\\{0,4}/[A-Za-z0-9:_@.-]+.*(?=["\s\';&\\])' #JS_HTTPX = r'(?:(?:(?<=["\';])https?:)|(?<=["\']))\\{0,4}/\\{0,4}/[A-Za-z0-9:_@.-]+.*(?=["\s\';&\\])'
#JS_HTTPX = r'(?<=["\';])(?:https?:)?\\{0,4}/\\{0,4}/[A-Za-z0-9:_@.\-/\\?&#]+(?=["\';&\\])' #JS_HTTPX = r'(?<=["\';])(?:https?:)?\\{0,4}/\\{0,4}/[A-Za-z0-9:_@.\-/\\?&#]+(?=["\';&\\])'
JS_HTTPX = r'(?<=["\';])(?:https?:)?\\{0,4}/\\{0,4}/[A-Za-z0-9:_@.-][^"\s\';&\\]*(?=["\';&\\])' JS_HTTPX = r'(?:(?<=["\';])https?:|(?<=["\']))\\{0,4}/\\{0,4}/[A-Za-z0-9:_@.-][^"\s\';&\\]*(?=["\';&\\])'
def __init__(self, rewriter, rules=[]): def __init__(self, rewriter, rules=[]):
rules = rules + [ rules = rules + [

5
pywb/rewrite/test/test_regex_rewriters.py
View File

@ -92,6 +92,9 @@ r"""
>>> _test_js('cool_Location = "//example.com/abc.html" //comment') >>> _test_js('cool_Location = "//example.com/abc.html" //comment')
'cool_Location = "/web/20131010/http://example.com/abc.html" //comment' 'cool_Location = "/web/20131010/http://example.com/abc.html" //comment'
>>> _test_js('A = B;//C + D;')
'A = B;//C + D;'
# document.cookie test # document.cookie test
>>> _test_js('document.cookie = "a=b; Path=/"') >>> _test_js('document.cookie = "a=b; Path=/"')
'document.WB_wombat_cookie = "a=b; Path=/"' 'document.WB_wombat_cookie = "a=b; Path=/"'
@ -103,6 +106,8 @@ r"""
>>> _test_js('"http:\/\/sub-site.example.com\/path-dashes\/path_other\/foo_bar.txt"') >>> _test_js('"http:\/\/sub-site.example.com\/path-dashes\/path_other\/foo_bar.txt"')
'"/web/20131010/http:\\/\\/sub-site.example.com\\/path-dashes\\/path_other\\/foo_bar.txt"' '"/web/20131010/http:\\/\\/sub-site.example.com\\/path-dashes\\/path_other\\/foo_bar.txt"'
>>> _test_js('"a=b&amp;http:\/\/example.com/;c=d"')
'"a=b&amp;/web/20131010/http:\\/\\/example.com/;c=d"'
#================================================================= #=================================================================
# XML Rewriting # XML Rewriting