def filter_html_cb(s, jsrewrite):
cleaner = lxml.html.clean.Cleaner()
cleaner.scripts = False
cleaner.style = True
doc = lxml.html.fromstring(s)
clean = cleaner.clean_html(doc)
for el in clean.iter():
if el.tag == 'script':
el.text = jsrewrite(el.text)
for a in el.attrib:
del el.attrib[a]
if 'id' in el.attrib:
el.attrib['id'] = 'sandbox-' + el.attrib['id']
return lxml.html.tostring(clean)
def filter_html_cb(s, jsrewrite):
cleaner = lxml.html.clean.Cleaner()
cleaner.scripts = False
cleaner.style = True
doc = lxml.html.fromstring(s)
clean = cleaner.clean_html(doc)
for el in clean.iter():
if el.tag == 'script':
el.text = jsrewrite(el.text)
for a in el.attrib:
del el.attrib[a]
if 'id' in el.attrib:
el.attrib['id'] = 'sandbox-' + el.attrib['id']
return lxml.html.tostring(clean)
def _clean_html(element):
if element.tag == "style":
css = cssutils.parseString(element.text_content())
for rule in css:
try:
for sel in rule.selectorList:
# prefix all id and class selectors with google-
sel.selectorText = sel.selectorText.replace(
".", ".google-")
sel.selectorText = sel.selectorText.replace(
"#", "#google-")
# add #resultsFrame in front of every selector (beware of comma-separated parts)
selectors = sel.selectorText.split(",")
selectors = [
"#resultsFrame %s" % s.strip() for s in selectors
]
sel.selectorText = ", ".join(selectors)
except:
pass
return "<style>%s</style>" % css.cssText
elif element.tag == "script":
return lxml.html.tostring(element)
else:
# monkeypatch from http://stackoverflow.com/questions/15386605/lxml-cleaner-to-ignore-base64-image
# this prevents lxml from removing the data:image
new_pattern = '\s*(?:javascript:|jscript:|livescript:|vbscript:|data:[^(?:image/.+;base64)]+|about:|mocha:)'
lxml.html.clean._javascript_scheme_re = re.compile(new_pattern, re.I)
cleaner = lxml.html.clean.Cleaner()
cleaner.style = False
cleaner.safe_attrs_only = False
clean = cleaner.clean_html(element)
for e in clean.iter():
# prefix all ids and classes with google-
if 'id' in e.attrib:
e.attrib['id'] = 'google-' + e.attrib['id']
if 'class' in e.attrib:
# beware of elements with multiple space-separated classes
classes = e.attrib['class'].split(" ")
classes = ["google-%s" % c.strip() for c in classes]
e.attrib['class'] = " ".join(classes)
return lxml.html.tostring(clean)
def _clean_html(element):
if element.tag == "style":
css = cssutils.parseString(element.text_content())
for rule in css:
try:
for sel in rule.selectorList:
# prefix all id and class selectors with google-
sel.selectorText = sel.selectorText.replace(".", ".google-")
sel.selectorText = sel.selectorText.replace("#", "#google-")
# add #resultsFrame in front of every selector (beware of comma-separated parts)
selectors = sel.selectorText.split(",")
selectors = ["#resultsFrame %s" % s.strip() for s in selectors]
sel.selectorText = ", ".join(selectors)
except:
pass
return "<style>%s</style>" % css.cssText
elif element.tag == "script":
return lxml.html.tostring(element)
else:
# monkeypatch from http://stackoverflow.com/questions/15386605/lxml-cleaner-to-ignore-base64-image
# this prevents lxml from removing the data:image
new_pattern = '\s*(?:javascript:|jscript:|livescript:|vbscript:|data:[^(?:image/.+;base64)]+|about:|mocha:)'
lxml.html.clean._javascript_scheme_re = re.compile(new_pattern, re.I)
cleaner = lxml.html.clean.Cleaner()
cleaner.style = False
cleaner.safe_attrs_only = False
clean = cleaner.clean_html(element)
for e in clean.iter():
# prefix all ids and classes with google-
if 'id' in e.attrib:
e.attrib['id'] = 'google-' + e.attrib['id']
if 'class' in e.attrib:
# beware of elements with multiple space-separated classes
classes = e.attrib['class'].split(" ")
classes = ["google-%s" % c.strip() for c in classes]
e.attrib['class'] = " ".join(classes)
return lxml.html.tostring(clean)