def hash_id(self):
content = tounicode(self.node)
hashed = md5()
try:
hashed.update(content.encode('utf-8', errors="replace"))
except Exception, e:
LOG.error("BOOM! " + str(e))
def generate_hash_id(node):
"""Generate a hash_id for the node in question.
:param node: lxml etree node
"""
content = tounicode(node)
hashed = md5()
try:
hashed.update(content.encode('utf-8', "replace"))
except Exception, e:
LOG.error("BOOM! " + str(e))
def generate_hash_id(node):
"""Generate a hash_id for the node in question.
:param node: lxml etree node
"""
content = tounicode(node)
hashed = md5()
try:
hashed.update(content.encode('utf-8', "replace"))
except Exception as e:
LOG.error("BOOM! " + str(e))
return hashed.hexdigest()[0:8]
def build_doc(page):
"""Requires that the `page` not be None"""
if page is None:
LOG.error("Page content is None, can't build_doc")
return ''
if isinstance(page, unicode):
page_unicode = page
else:
enc = get_encoding(page)
page_unicode = page.decode(enc, 'replace')
doc = document_fromstring(
page_unicode.encode('utf-8', 'replace'),
parser=utf8_parser)
return doc
def build_doc(page):
"""Requires that the `page` not be None"""
if page is None:
LOG.error("Page content is None, can't build_doc")
return ""
if isinstance(page, unicode):
page_unicode = page
else:
enc = get_encoding(page)
page_unicode = page.decode(enc, "replace")
try:
doc = document_fromstring(page_unicode.encode("utf-8", "replace"), parser=utf8_parser)
return doc
except XMLSyntaxError, exc:
LOG.error("Failed to parse: " + str(exc))
raise ValueError("Failed to parse document contents.")
def build_doc(page):
"""Requires that the `page` not be None"""
if page is None:
LOG.error("Page content is None, can't build_doc")
return ''
if isinstance(page, str):
page_unicode = page
else:
enc = get_encoding(page)
page_unicode = page.decode(enc, 'replace')
try:
doc = document_fromstring(
page_unicode.encode('utf-8', 'replace'),
parser=utf8_parser)
return doc
except XMLSyntaxError as exc:
LOG.error('Failed to parse: ' + str(exc))
raise ValueError('Failed to parse document contents.')
