def _parse(self, path):
try:
doc = etree.parse(path)
except etree.LxmlError as err:
# A workaround for cases when lxml (quite strangely)
# sets the filename to <string>.
if err.error_log[0].filename == "<string>":
filename = self._path
else:
filename = err.error_log[0].filename
loc = {
"file": os.path.basename(filename),
"line": err.error_log[0].line,
"col": err.error_log[0].column
}
exc = XmlProcessingError(err.error_log[0].message)
exc.set_loc(loc)
raise exc
except Exception as err:
loc = {
"file": os.path.basename(self._path),
"line": None,
"col": None
}
exc = XmlProcessingError(str(err))
exc.set_loc(loc)
raise exc
return doc
def _parse(self, path):
try:
if path.startswith('https'):
doc = etree.parse(urlopen(path))
else:
doc = etree.parse(path)
except etree.LxmlError as err:
# A workaround for cases when lxml (quite strangely)
# sets the filename to <string>.
if err.error_log[0].filename == "<string>":
filename = self._path
else:
filename = err.error_log[0].filename
loc = {"file": os.path.basename(filename),
"line": err.error_log[0].line,
"col": err.error_log[0].column}
exc = XmlProcessingError(err.error_log[0].message)
exc.set_loc(loc)
raise exc
except Exception as err:
loc = {"file": os.path.basename(self._path),
"line": None,
"col": None}
exc = XmlProcessingError(str(err))
exc.set_loc(loc)
raise exc
return doc
def parse(self):
doc = self._parse(self._path)
self._remove_comments(doc)
# Due to a weird implementation of XInclude in lxml, the
# XmlParser resolves included documents on it's own.
#
# To be able to tell later on where each tag was located
# in the XML document, we add a '__file' attribute to
# each element of the tree during the parsing.
#
# However, these special attributes are of course not
# valid according to our schemas. To solve this, a copy of
# the tree is made and the '__file' attributes are removed
# before validation.
#
# XXX This is a *EXTREMELY* dirty hack. Ideas/proposals
# for cleaner solutions are more than welcome!
root_tag = self._init_loc(doc.getroot(), self._path)
self._expand_xinclude(root_tag, os.path.dirname(self._path))
self._template_proc.process_aliases(root_tag)
try:
self._validate(doc)
except:
err = self._schema.error_log[0]
loc = {"file": os.path.basename(err.filename),
"line": err.line, "col": err.column}
exc = XmlProcessingError(err.message)
exc.set_loc(loc)
raise exc
return self._process(root_tag)
def parse(self):
doc = self._parse(self._path)
self._remove_comments(doc)
# Due to a weird implementation of XInclude in lxml, the
# XmlParser resolves included documents on it's own.
#
# To be able to tell later on where each tag was located
# in the XML document, we add a '__file' attribute to
# each element of the tree during the parsing.
#
# However, these special attributes are of course not
# valid according to our schemas. To solve this, a copy of
# the tree is made and the '__file' attributes are removed
# before validation.
#
# XXX This is a *EXTREMELY* dirty hack. Ideas/proposals
# for cleaner solutions are more than welcome!
root_tag = self._init_loc(doc.getroot(), self._path)
self._expand_xinclude(root_tag, os.path.dirname(self._path))
self._template_proc.process_aliases(root_tag)
try:
self._validate(doc)
except:
err = self._schema.error_log[0]
loc = {
"file": os.path.basename(err.filename),
"line": err.line,
"col": err.column
}
exc = XmlProcessingError(err.message)
exc.set_loc(loc)
raise exc
return self._process(root_tag)