def whereIsYadis(resp):
"""Given a HTTPResponse, return the location of the Yadis document.
May be the URL just retrieved, another URL, or None, if I can't
find any.
[non-blocking]
@returns: str or None
"""
# Attempt to find out where to go to discover the document
# or if we already have it
content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if (content_type
and content_type.split(';', 1)[0].lower() == YADIS_CONTENT_TYPE):
return resp.final_url
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
try:
yadis_loc = findHTMLMeta(StringIO(resp.body))
except MetaNotFound:
pass
return yadis_loc
def whereIsYadis(resp):
"""Given a HTTPResponse, return the location of the Yadis document.
May be the URL just retrieved, another URL, or None, if I can't
find any.
[non-blocking]
@returns: str or None
"""
# Attempt to find out where to go to discover the document
# or if we already have it
content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if (content_type and
content_type.split(';', 1)[0].lower() == YADIS_CONTENT_TYPE):
return resp.final_url
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
try:
yadis_loc = findHTMLMeta(StringIO(resp.body))
except MetaNotFound:
pass
return yadis_loc
def whereIsYadis(resp):
"""Given a HTTPResponse, return the location of the Yadis document.
May be the URL just retrieved, another URL, or None, if I can't
find any.
[non-blocking]
@returns: str or None
"""
# Attempt to find out where to go to discover the document
# or if we already have it
content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if content_type and content_type.split(';',
1)[0].lower() == YADIS_CONTENT_TYPE:
return resp.final_url
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
# Decode body by encoding of file
content_type = content_type or ''
encoding = content_type.rsplit(';', 1)
if len(encoding) == 2 and encoding[1].strip().startswith(
'charset='):
encoding = encoding[1].split('=', 1)[1].strip()
else:
encoding = 'UTF-8'
try:
content = resp.body.decode(encoding)
except UnicodeError:
# Keep encoded version in case yadis location can be found before encoding shut this up.
# Possible errors will be caught lower.
content = resp.body
try:
yadis_loc = findHTMLMeta(StringIO(content))
except (MetaNotFound, UnicodeError):
# UnicodeError: Response body could not be encoded and xrds location
# could not be found before troubles occurs.
pass
return yadis_loc
def whereIsYadis(resp):
"""Given a HTTPResponse, return the location of the Yadis document.
May be the URL just retrieved, another URL, or None, if I can't
find any.
[non-blocking]
@returns: str or None
"""
# Attempt to find out where to go to discover the document
# or if we already have it
content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if (content_type and
content_type.split(';', 1)[0].lower() == YADIS_CONTENT_TYPE):
return resp.final_url
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
# Decode body by encoding of file
content_type = content_type or ''
encoding = content_type.rsplit(';', 1)
if len(encoding) == 2 and encoding[1].strip().startswith('charset='):
encoding = encoding[1].split('=', 1)[1].strip()
else:
encoding = 'UTF-8'
try:
content = resp.body.decode(encoding)
except UnicodeError:
# Keep encoded version in case yadis location can be found before encoding shut this up.
# Possible errors will be caught lower.
content = resp.body
try:
yadis_loc = findHTMLMeta(StringIO(content))
except (MetaNotFound, UnicodeError):
# UnicodeError: Response body could not be encoded and xrds location
# could not be found before troubles occurs.
pass
return yadis_loc
def test_multiple_headers(self):
buff = StringIO(
'<html><head>'
'<meta http-equiv="X-XRDS-Location" content="found">'
'<meta http-equiv="X-XRDS-Location" content="not-found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_javascript_in_head(self):
buff = StringIO(
'<html><head><script type="text/javascript">document.write("<body>");</script>'
'<META http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_missing_html_tag(self):
buff = StringIO(
'<head><meta http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_top_level_bogus(self):
buff = StringIO(
'</porky><html><head><meta http-equiv="X-XRDS-Location" content="found">'
)
self.assertEqual(findHTMLMeta(buff), 'found')
def test_case_insensitive_header_name(self):
buff = StringIO(
'<html><head><meta http-equiv="x-xrds-location" content="found"></head></html>'
)
self.assertEqual(findHTMLMeta(buff), 'found')
def test_missing_html_tag(self):
buff = StringIO('<head><meta http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_empty_string(self):
buff = StringIO('<head><meta http-equiv="X-XRDS-Location" content="">')
self.assertEqual(findHTMLMeta(buff), '')
def test_case_insensitive_header_name(self):
buff = StringIO('<html><head><meta http-equiv="x-xrds-location" content="found"></head></html>')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_decimal_entity(self):
buff = StringIO('<head><meta http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_empty_string(self):
buff = StringIO('<head><meta http-equiv="X-XRDS-Location" content="">')
self.assertEqual(findHTMLMeta(buff), '')
def test_standard_entity(self):
buff = StringIO('<head><meta http-equiv="X-XRDS-Location" content="&">')
self.assertEqual(findHTMLMeta(buff), '&')
def test_multiple_headers(self):
buff = StringIO('<html><head>'
'<meta http-equiv="X-XRDS-Location" content="found">'
'<meta http-equiv="X-XRDS-Location" content="not-found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_javascript_in_head(self):
buff = StringIO('<html><head><script type="text/javascript">document.write("<body>");</script>'
'<META http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_standard_entity(self):
buff = StringIO(
'<head><meta http-equiv="X-XRDS-Location" content="&">')
self.assertEqual(findHTMLMeta(buff), '&')
def test_xhtml(self):
buff = StringIO('<html><head><meta http-equiv="X-XRDS-Location" content="found" /></head></html>')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_decimal_entity(self):
buff = StringIO(
'<head><meta http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def test_xhtml(self):
buff = StringIO(
'<html><head><meta http-equiv="X-XRDS-Location" content="found" /></head></html>'
)
self.assertEqual(findHTMLMeta(buff), 'found')
def whereIsYadis(resp):
"""Given a HTTPResponse, return the location of the Yadis document.
May be the URL just retrieved, another URL, or None if no suitable URL can
be found.
[non-blocking]
@returns: str or None
"""
# Attempt to find out where to go to discover the document
# or if we already have it
content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if (content_type and
content_type.split(';', 1)[0].lower() == YADIS_CONTENT_TYPE):
return resp.final_url
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
# Decode body by encoding of file
content_type = content_type or ''
encoding = content_type.rsplit(';', 1)
if (len(encoding) == 2 and
encoding[1].strip().startswith('charset=')):
encoding = encoding[1].split('=', 1)[1].strip()
else:
encoding = 'utf-8'
if isinstance(resp.body, bytes):
try:
content = resp.body.decode(encoding)
except UnicodeError:
# All right, the detected encoding has failed. Try with
# UTF-8 (even if there was no detected encoding and we've
# defaulted to UTF-8, it's not that expensive an operation)
try:
content = resp.body.decode('utf-8')
except UnicodeError:
# At this point the content cannot be decoded to a str
# using the detected encoding or falling back to utf-8,
# so we have to resort to replacing undecodable chars.
# This *will* result in broken content but there isn't
# anything else that can be done.
content = resp.body.decode(encoding, 'replace')
else:
content = resp.body
try:
yadis_loc = findHTMLMeta(StringIO(content))
except (MetaNotFound, UnicodeError):
# UnicodeError: Response body could not be encoded and xrds
# location could not be found before troubles occur.
pass
return yadis_loc
def test_top_level_bogus(self):
buff = StringIO('</porky><html><head><meta http-equiv="X-XRDS-Location" content="found">')
self.assertEqual(findHTMLMeta(buff), 'found')
def whereIsYadis(resp):
"""Given a HTTPResponse, return the location of the Yadis document.
May be the URL just retrieved, another URL, or None if no suitable URL can
be found.
[non-blocking]
@returns: str or None
"""
# Attempt to find out where to go to discover the document
# or if we already have it
content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if (content_type
and content_type.split(';', 1)[0].lower() == YADIS_CONTENT_TYPE):
return resp.final_url
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
# Decode body by encoding of file
content_type = content_type or ''
encoding = content_type.rsplit(';', 1)
if (len(encoding) == 2
and encoding[1].strip().startswith('charset=')):
encoding = encoding[1].split('=', 1)[1].strip()
else:
encoding = 'utf-8'
if isinstance(resp.body, bytes):
try:
content = resp.body.decode(encoding)
except UnicodeError:
# All right, the detected encoding has failed. Try with
# UTF-8 (even if there was no detected encoding and we've
# defaulted to UTF-8, it's not that expensive an operation)
try:
content = resp.body.decode('utf-8')
except UnicodeError:
# At this point the content cannot be decoded to a str
# using the detected encoding or falling back to utf-8,
# so we have to resort to replacing undecodable chars.
# This *will* result in broken content but there isn't
# anything else that can be done.
content = resp.body.decode(encoding, 'replace')
else:
content = resp.body
try:
yadis_loc = findHTMLMeta(StringIO(content))
except (MetaNotFound, UnicodeError):
# UnicodeError: Response body could not be encoded and xrds
# location could not be found before troubles occur.
pass
return yadis_loc
def discover(uri):
"""Discover services for a given URI.
@param uri: The identity URI as a well-formed http or https
URI. The well-formedness and the protocol are not checked, but
the results of this function are undefined if those properties
do not hold.
@return: DiscoveryResult object
@raises Exception: Any exception that can be raised by fetching a URL with
the given fetcher.
"""
result = DiscoveryResult(uri)
resp = fetchers.fetch(uri, headers={'Accept': YADIS_ACCEPT_HEADER})
if resp.status != 200:
raise DiscoveryFailure(
'HTTP Response status from identity URL host is not 200. '
'Got status %r' % (resp.status,), resp)
# Note the URL after following redirects
result.normalized_uri = resp.final_url
# Attempt to find out where to go to discover the document
# or if we already have it
result.content_type = resp.headers.get('content-type')
# According to the spec, the content-type header must be an exact
# match, or else we have to look for an indirection.
if (result.content_type and
result.content_type.split(';', 1)[0].lower() == YADIS_CONTENT_TYPE):
result.xrds_uri = result.normalized_uri
else:
# Try the header
yadis_loc = resp.headers.get(YADIS_HEADER_NAME.lower())
if not yadis_loc:
# Parse as HTML if the header is missing.
#
# XXX: do we want to do something with content-type, like
# have a whitelist or a blacklist (for detecting that it's
# HTML)?
try:
yadis_loc = findHTMLMeta(StringIO(resp.body))
except MetaNotFound:
pass
# At this point, we have not found a YADIS Location URL. We
# will return the content that we scanned so that the caller
# can try to treat it as an XRDS if it wishes.
if yadis_loc:
result.xrds_uri = yadis_loc
resp = fetchers.fetch(yadis_loc)
if resp.status != 200:
exc = DiscoveryFailure(
'HTTP Response status from Yadis host is not 200. '
'Got status %r' % (resp.status,), resp)
exc.identity_url = result.normalized_uri
raise exc
result.content_type = resp.headers.get('content-type')
result.response_text = resp.body
return result
