def validate(self, schema):
"""
Validate the XML in *file_* against a given schema.
| *Parameters:*
| *schema*........path to the schema file (*must* be RelaxNG)
| *Return:*
| True if schema validates successfully, False otherwise
"""
try:
from lxml import etree
from lxml.etree import RelaxNG
except ImportError:
raise Exception(
'[XMLObject.validate] Need lxml to validate xml!')
try:
relaxng_doc = etree.parse(schema)
relaxng = RelaxNG(relaxng_doc)
xml_doc = etree.parse(self._file)
except etree.XMLSyntaxError as e:
raise Exception(
'[XMLObject.validate] Cannot validate xml: ' + str(e))
return relaxng.validate(xml_doc)
def validate(self, schema):
"""
Validate the XML in *file_* against a given schema.
| *Parameters:*
| *schema*........path to the schema file (*must* be RelaxNG)
| *Return:*
| True if schema validates successfully, False otherwise
"""
try:
from lxml import etree
from lxml.etree import RelaxNG
except ImportError:
raise Exception('[XMLObject.validate] Need lxml to validate xml!')
try:
relaxng_doc = etree.parse(schema)
relaxng = RelaxNG(relaxng_doc)
xml_doc = etree.parse(self._file)
except etree.XMLSyntaxError as e:
raise Exception('[XMLObject.validate] Cannot validate xml: ' +
str(e))
return relaxng.validate(xml_doc)
def validate(self, xml_tree):
if self._start is None:
rng_tree = self._tree
else:
rng_tree = self._relocate_xslt(self._tree, newref="'%s'" % self._start)
# ugly hack to get around namespace (?) issues
rng_tree = parse(StringIO(str(rng_tree)))
rng = RelaxNG(rng_tree)
self.validate = rng.validate # replace the object method by the real thing
return rng.validate(xml_tree)
def validate(self, xml_tree):
if self._start is None:
rng_tree = self._tree
else:
rng_tree = self._relocate_xslt(self._tree,
newref="'%s'" % self._start)
# ugly hack to get around namespace (?) issues
rng_tree = parse(StringIO(str(rng_tree)))
rng = RelaxNG(rng_tree)
self.validate = rng.validate # replace the object method by the real thing
return rng.validate(xml_tree)
def validate_tei(xmldoc): # , filename=""
'''Check if an XML document is conform to the guidelines of the Text Encoding Initiative'''
global TEI_RELAXNG
if TEI_RELAXNG is None:
# load validator
with lzma.open(TEI_SCHEMA, 'rb') as schemafile:
schema_data = load_pickle(schemafile)
TEI_RELAXNG = RelaxNG(fromstring(schema_data))
result = TEI_RELAXNG.validate(xmldoc)
if result is False:
LOGGER.warning('not a valid TEI document: %s', TEI_RELAXNG.error_log.last_error)
return result
