def parse_xml(xml_str_or_file, parser=None, **parser_kwargs):
""" simple lxml parser to execute the correct parsing method and use ETree class as the default
so we can use xpathbuilder directly without str casting
Args:
xml_str_or_file (str or IOBase):
parser:
**parser_kwargs:
Returns:
"""
if not parser:
parser_lookup = ET.ElementDefaultClassLookup(
element=ElementBaseXpathStr)
parser = ET.XMLParser(**parser_kwargs)
parser.set_element_class_lookup(parser_lookup)
if isinstance(xml_str_or_file, str):
return ET.fromstring(xml_str_or_file, parser=parser)
elif isinstance(xml_str_or_file, bytes):
return ET.parse(io.BytesIO(xml_str_or_file), parser=parser).getroot()
elif hasattr(xml_str_or_file, 'read'):
return ET.parse(xml_str_or_file, parser=parser).getroot()
else:
raise NotImplementedError(
'We only know how to parse string, bytes or file objects. Use straight lxml methods'
)
def fromfile(self,
source,
tags_factory=Tag,
fragment=False,
no_leading_text=False,
encoding='utf-8',
**kw):
"""Parse a XML file
In:
- ``source`` -- can be a filename or a file object
- ``fragment`` -- if ``True``, can parse a XML fragment i.e a XML without
a unique root
- ``no_leading_text`` -- if ``fragment`` is ``True``, ``no_leading_text``
is ``False`` and the XML to parsed begins by a text, this text is keeped
- ``kw`` -- keywords parameters are passed to the XML parser
Return:
- the root element of the parsed XML, if ``fragment`` is ``False``
- a list of XML elements, if ``fragment`` is ``True``
"""
if isinstance(source, (str, type(u''))):
if source.startswith(('http://', 'https://', 'ftp://')):
source = urlopen(source)
else:
source = fileopen(source, encoding=encoding)
# Create a dedicated parser with the ``kw`` parameter
parser = self._parser.__class__(encoding=encoding, **kw)
# This parser will generate nodes of type ``Tag``
parser.set_element_class_lookup(
etree.ElementDefaultClassLookup(element=tags_factory))
if not fragment:
# Parse a tree (only one root)
# ----------------------------
root = etree.parse(source, parser).getroot()
source.close()
# Attach the renderer to the root
root._renderer = self
return root
# Parse a fragment (multiple roots)
# ---------------------------------
# Create a dummy root
xml = BufferIO(b'<html><body>%s</body></html>' % source.read())
source.close()
root = etree.parse(xml, parser).getroot()[0]
for e in root:
if isinstance(e, tags_factory):
# Attach the renderer to each roots
e._renderer = self
# Return the children of the dummy root
return ((root.text.encode(encoding), )
if root.text and not no_leading_text else ()) + tuple(root[:])
def parse_html(self,
source,
fragment=False,
no_leading_text=False,
xhtml=False,
**kw):
"""Parse a (X)HTML file
In:
- ``source`` -- can be a filename or a file object
- ``fragment`` -- if ``True``, can parse a HTML fragment i.e a HTML without
a unique root
- ``no_leading_text`` -- if ``fragment`` is ``True``, ``no_leading_text``
is ``False`` and the HTML to parsed begins by a text, this text is keeped
- ``xhtml`` -- is the HTML to parse a valid XHTML ?
- ``kw`` -- keywords parameters are passed to the HTML parser
Return:
- the root element of the parsed HTML, if ``fragment`` is ``False``
- a list of HTML elements, if ``fragment`` is ``True``
"""
parser = ET.XMLParser(**kw) if xhtml else ET.HTMLParser(**kw)
parser.setElementClassLookup(
ET.ElementDefaultClassLookup(element=_HTMLTag))
return self._parse_html(parser, source, fragment, no_leading_text,
**kw)
class Renderer(xml.XmlRenderer):
doctype = '<!DOCTYPE document SYSTEM "rml.dtd">'
content_type = 'application/pdf'
namespace = 'http://namespaces.zope.org/rml'
_parser = etree.XMLParser()
_parser.set_element_class_lookup(etree.ElementDefaultClassLookup(element=Tag))
def __init__(self, parent=None, *args, **kw):
super(Renderer, self).__init__(parent, *args, **kw)
self.namespaces = {None: self.namespace}
@classmethod
def get_tags(cls, tags, tag, signature):
if tag not in tags:
tags[tag] = set(schema.getFields(signature))
for child in signature.queryTaggedValue('directives', ()):
cls.get_tags(tags, child.tag, child.signature)
return tags
@classmethod
def create_RML_tags(cls):
for tag, signature in cls.get_tags({}, 'document', document.IDocument).items():
setattr(cls, tag, xml.TagProp(tag, signature))
@classmethod
def create_para_extra_tags(cls):
tags = [method[8:] for method in Para.__dict__ if method.startswith('compile_')]
for tag in tags:
setattr(cls, tag, xml.TagProp(tag))
async def __aiter__(self) -> _AsyncGenerator[NSElement, None]:
if not Api.agent:
raise RuntimeError("The API's user agent is not yet set.")
url = self.value
# pylint: disable=E1101
tag = self.name.upper().rstrip("S")
parser = etree.XMLPullParser(["end"],
base_url=url,
remove_blank_text=True,
tag=tag)
parser.set_element_class_lookup(
etree.ElementDefaultClassLookup(element=NSElement))
events = parser.read_events()
dobj = zlib.decompressobj(16 + zlib.MAX_WBITS)
async with Api.session.request("GET",
url,
headers={"User-Agent":
Api.agent}) as response:
async for data, _ in response.content.iter_chunks():
parser.feed(dobj.decompress(data))
for _, element in events:
yield element
element.clear()
while element.getparent(
) is not None and element.getprevious() is not None:
del element.getparent()[0]
def createDefaultParser(self):
parser = _etree.XMLParser(**self.parser_config)
element_class = self.element_class
if self.element_class is not None:
lookup = _etree.ElementDefaultClassLookup(element=element_class)
parser.set_element_class_lookup(lookup)
return parser
def __getattr__(cls, name: str) -> str:
"""
Magic method to call a HTML element.
Args:
name: The HTML element name
Returns:
HTML element
Raises:
AttributeError: If an HTML element name is invalid
"""
if name == '__elements__' or name not in cls.__elements__:
raise AttributeError('Invalid element')
parser = etree.HTMLParser()
parser.set_element_class_lookup(
etree.ElementDefaultClassLookup(element=HtmlElement))
def _wrapper(*args, **kwargs):
return \
getattr(builder.ElementMaker(makeelement=parser.makeelement),
name)(*args, **kwargs)
return _wrapper
def createFromExisting(cls, source):
"""
Create a TouchOSCLayout instance from an existing TouchOSC Layout.
@type source: filename or fileobject
@param source: Path to an existing .touchosc file, or
TouchOSC index.xml file (from unzipping .touchosc file)
@rtype: Layout
@return: An instance containing the layout
"""
fallback = etree.ElementDefaultClassLookup()
lookupTabpages = etree.ElementNamespaceClassLookup(fallback)
namespace = lookupTabpages.get_namespace(None)
namespace['tabpage'] = tabpage.Tabpage
lookupControls = etree.AttributeBasedElementClassLookup(
'type', controls.type_class_mapping, lookupTabpages)
layoutParser = etree.XMLParser(remove_blank_text=True)
layoutParser.setElementClassLookup(lookupControls)
if type(source) is str:
(path, fname) = os.path.split(source)
(name, extension) = os.path.splitext(fname)
if extension == ".touchosc":
f = ZipFile(source, "r")
layoutTree = etree.parse(StringIO(f.read("index.xml")),
layoutParser)
f.close()
elif extension == ".xml":
name = None
layoutTree = etree.parse(source, layoutParser)
return Layout(layoutTree, name)
def parseXML(filename):
""" Parse an XML document, thus also suitable for XHTML """
# XML doesn't require jumping through the same hoops as HTML since there
# are no existing custom element classes.
parser_lookup = _etree.ElementDefaultClassLookup(element=HtHtmlElement)
parser = _etree.XMLParser()
parser.set_element_class_lookup(parser_lookup)
return _etree.parse(filename, parser=parser)
def Element(name, attrib={}, **extra):
attrib = attrib.copy()
attrib.update(extra)
parser_lookup = etree.ElementDefaultClassLookup(element=MyWriter)
parser = etree.XMLParser()
parser.set_element_class_lookup(parser_lookup)
tag = parser.makeelement(name, attrib=attrib)
return tag
def class_init(cls, special_tags):
"""Class initialisation
In:
-- ``special_tags`` -- tags that have a special factory
"""
# Create a XML parser that generate ``_Tag`` nodes
cls._xml_parser = ET.XMLParser()
cls._xml_parser.setElementClassLookup(
ET.ElementDefaultClassLookup(element=_Tag))
def parse_xml(self, source, fragment=False, no_leading_text=False, **kw):
"""Parse a XML file
In:
- ``source`` -- can be a filename or a file object
- ``fragment`` -- if ``True``, can parse a XML fragment i.e a XML without
a unique root
- ``no_leading_text`` -- if ``fragment`` is ``True``, ``no_leading_text``
is ``False`` and the XML to parsed begins by a text, this text is keeped
- ``kw`` -- keywords parameters are passed to the XML parser
Return:
- the root element of the parsed XML, if ``fragment`` is ``False``
- a list of XML elements, if ``fragment`` is ``True``
"""
if isinstance(source, basestring):
if source.startswith(('http://', 'https://', 'ftp://')):
source = urllib.urlopen(source)
else:
source = open(source)
# Create a dedicated XML parser with the ``kw`` parameter
parser = ET.XMLParser(**kw)
# This parser will generate nodes of type ``_Tag``
parser.setElementClassLookup(
ET.ElementDefaultClassLookup(element=_Tag))
if not fragment:
# Parse a XML file
# ----------------
root = ET.parse(source, parser).getroot()
source.close()
# Attach the renderer to the root
root._renderer = self
return root
# Parse a XML fragment
# --------------------
# Create a dummy root
xml = cStringIO.StringIO('<dummy>%s</dummy>' % source.read())
source.close()
root = ET.parse(xml, parser).getroot()
for e in root[:]:
# Attach the renderer to each roots
e._renderer = self
# Return the children of the dummy root
return ([root.text]
if root.text and not no_leading_text else []) + root[:]
def create_tag(name, value, attrib={}, **extra):
# print ("tworze tag %s o wartości %s"%(name,value))
attrib = attrib.copy()
attrib.update(extra)
parser_lookup = etree.ElementDefaultClassLookup(element=MyWriter)
parser = etree.XMLParser()
parser.set_element_class_lookup(parser_lookup)
tag = parser.makeelement(name, attrib=attrib)
if value:
# tag.text = value.decode("utf-8")
tag.text = value
return tag
def __init__(self, file_location):
"""
Parser/iterator for the OAIRecord class. Iterates over record elements in any namespace (repox or oai-pmh).
:param file_location:
"""
oai_parser_registration = etree.ElementDefaultClassLookup(
element=OAIRecord)
oai_parser = etree.XMLParser()
oai_parser.set_element_class_lookup(oai_parser_registration)
super(OAIReader, self).__init__(file_location,
'{*}record',
parser=oai_parser)
def __init__(self, file_location):
"""
Parser/iterator for the MODSRecord class. Iterates on mods:mods elements.
:param file_location:
"""
mods_parser_registration = etree.ElementDefaultClassLookup(
element=MODSRecord)
mods_parser = etree.XMLParser()
mods_parser.set_element_class_lookup(mods_parser_registration)
super(MODSReader, self).__init__(file_location,
'{0}mods'.format(NAMESPACES['mods']),
parser=mods_parser)
def metadata(self):
"""
Exposes the metadata content of an OAIRecord.
:return: A reparsed root element either in the MODSRecord or DCRecord class, as appropriate.
"""
record_data = self.find('./{*}metadata')
if record_data is not None:
try:
if 'mods' in record_data[0].tag:
mods_parser_registration = etree.ElementDefaultClassLookup(
element=MODSRecord)
mods_parser = etree.XMLParser()
mods_parser.set_element_class_lookup(
mods_parser_registration)
return etree.XML(etree.tostring(
record_data[0], encoding='UTF-8').decode('utf-8'),
parser=mods_parser)
elif 'qualified' in record_data[0].tag:
qdc_parser_registration = etree.ElementDefaultClassLookup(
element=DCRecord)
qdc_parser = etree.XMLParser()
qdc_parser.set_element_class_lookup(
qdc_parser_registration)
return etree.XML(etree.tostring(
record_data[0], encoding='UTF-8').decode('utf-8'),
parser=qdc_parser)
elif 'dc' in record_data[0].tag:
dc_parser_registration = etree.ElementDefaultClassLookup(
element=DCRecord)
dc_parser = etree.XMLParser()
dc_parser.set_element_class_lookup(dc_parser_registration)
return etree.XML(etree.tostring(
record_data[0], encoding='UTF-8').decode('utf-8'),
parser=dc_parser)
except IndexError:
pass
def _build(self):
""" create layout and define widget attribute by tkouter html """
if not self.layout:
return
env = Environment(loader=self.loader)
if '.html' in self.layout or 'xml' in self.layout:
template = env.get_template(self.layout)
self._html = template.render(self.context)
else:
self._html = Template(self.layout).render(self.context)
# lxml parser
parser_lookup = etree.ElementDefaultClassLookup(element=TkOutElement)
self._parser = etree.XMLParser()
self._parser.set_element_class_lookup(parser_lookup)
self._tree = etree.parse(StringIO(self._html), self._parser)
# we should cache the elements for storing data to it
self._proxy_cache = list(self._tree.getroot().iter())
# css
css = None
for e in self._tree.getroot().iter():
if e.is_css and e.get('href'):
self._css = env.get_template(e.get('href')).render()
self._css_parser = tinycss.make_parser()
self._stylesheet = self._css_parser.parse_stylesheet(self._css)
for rule in self._stylesheet.rules:
for e in self._select(rule.selector.as_css()):
for d in rule.declarations:
if e.get(d.name) is None:
e.set(d.name, d.value.as_css())
# post init etree elements and display their widgets
for e in self._tree.getroot().iter():
try:
e.init(self)
e.display()
except TagError as err:
print('Error when parsing tag: ')
print(
etree.tostring(e,
pretty_print=True,
encoding=str,
method='html'))
raise err
def set_parser_to_relaxed(cls):
"""
Creates a XML parser which attempts to recover syntactically-flawed XML.
Returns:
None
"""
# Creates an `etree.XMLParser` object, equivalent to the default parser used
# by the parser `lxml.fromstring()` (see `lxml.GlobalParserTLS.createDefaultParser()`),
# except enabling the `recover=True` attribute.
relaxed_xml_parser = etree.XMLParser(recover=True,
resolve_entities=False)
lookup = etree.ElementDefaultClassLookup(element=RestrictedElement)
relaxed_xml_parser.set_element_class_lookup(lookup)
# Inject parser
cls._parse_etree = partial(super()._parse_etree,
parser=relaxed_xml_parser)
def class_init(cls, specialTags):
"""Class initialisation
In:
-- ``special_tags`` -- tags that have a special factory
"""
class CustomLookup(ET.CustomElementClassLookup):
def __init__(self, specialTags, defaultLookup):
super(CustomLookup, self).__init__(defaultLookup)
self._specialTags = specialTags
def lookup(self, node_type, document, namespace, name):
return self._specialTags.get(name)
cls._specialTags.update(specialTags)
cls._custom_lookup = CustomLookup(cls._specialTags, ET.ElementDefaultClassLookup(element=xhtml_base._HTMLTag))
cls._html_parser = ET.HTMLParser()
cls._html_parser.setElementClassLookup(cls._custom_lookup)
async def __aiter__(self):
url = self.value
parser = etree.XMLPullParser(["end"],
base_url=url,
remove_blank_text=True)
parser.set_element_class_lookup(
etree.ElementDefaultClassLookup(element=_NSElement))
events = parser.read_events()
dobj = zlib.decompressobj(16 + zlib.MAX_WBITS)
async with Api.session.request("GET",
url,
headers={"User-Agent":
Api.agent}) as response:
yield parser.makeelement("HEADERS", attrib=response.headers)
async for data, _ in response.content.iter_chunks():
parser.feed(dobj.decompress(data))
for _, element in events:
yield element
element.clear()
async def __aiter__(
self,
*,
no_clear: bool = False) -> _AsyncGenerator[NSElement, None]:
if not self.agent:
raise RuntimeError("The API's user agent is not yet set.")
if not self:
# Preempt the request to conserve ratelimit
raise ValueError("Bad request")
if "a" in self and self["a"].lower() == "sendtg":
raise RuntimeError(
"This API wrapper does not support API telegrams.")
url = str(self)
parser = etree.XMLPullParser(["end"],
base_url=url,
remove_blank_text=True)
parser.set_element_class_lookup(
etree.ElementDefaultClassLookup(element=NSElement))
events = parser.read_events()
async with self.session.request("GET",
url,
headers={"User-Agent":
self.agent}) as response:
encoding = response.headers["Content-Type"].split(
"charset=")[1].split(",")[0]
async for data, _ in response.content.iter_chunks():
parser.feed(data.decode(encoding))
for _, element in events:
if not no_clear and (element.getparent() is None
or element.getparent().getparent()
is not None):
continue
yield element
if no_clear:
continue
element.clear()
while element.getprevious() is not None:
del element.getparent()[0]
async def __aiter__(self, *, clear: bool = True):
if not self:
raise ValueError("Bad request")
url = str(self)
parser = etree.XMLPullParser(["end"],
base_url=url,
remove_blank_text=True)
parser.set_element_class_lookup(
etree.ElementDefaultClassLookup(element=_NSElement))
events = parser.read_events()
async with type(self).session.request(
"GET", url, headers={"User-Agent":
type(self).agent}) as response:
yield parser.makeelement("HEADERS", attrib=response.headers)
encoding = response.headers["Content-Type"].split(
"charset=")[1].split(",")[0]
async for data, _ in response.content.iter_chunks():
parser.feed(data.decode(encoding))
for _, element in events:
yield element
if clear:
element.clear()
def __init__(self):
self.parser = etree.XMLParser()
fallback = etree.ElementDefaultClassLookup(PDFXML)
lookup = etree.ElementNamespaceClassLookup(fallback)
namespace = lookup.get_namespace(None)
#leafs
namespace['name'] = PDFName
namespace['string'] = PDFString
namespace['number'] = PDFNumber
namespace['null'] = PDFNull
namespace['bool'] = PDFBool
namespace['R'] = PDFR
namespace['header'] = PDFHeader
namespace['startxref'] = PDFStartxref
namespace['data'] = PDFData
#trees
namespace['entry'] = PDFEntry
namespace['dictionary'] = PDFDictionary
namespace['stream'] = PDFStream
namespace['pdf'] = PDFPdf
namespace['pdf_update'] = PDFUpdate
namespace['indirect_object'] = PDFIndirect
namespace['array'] = PDFArray
self.parser.set_element_class_lookup(lookup)
def main(benchmark_class):
import_lxml = True
callgrind_zero = False
if len(sys.argv) > 1:
try:
sys.argv.remove('-i')
# run benchmark 'inplace'
sys.path.insert(0, 'src')
except ValueError:
pass
try:
sys.argv.remove('-nolxml')
# run without lxml
import_lxml = False
except ValueError:
pass
try:
sys.argv.remove('-z')
# reset callgrind after tree setup
callgrind_zero = True
except ValueError:
pass
initArgs(sys.argv)
_etrees = []
if import_lxml:
from lxml import etree
_etrees.append(etree)
try:
sys.argv.remove('-fel')
except ValueError:
pass
else:
# use fast element creation in lxml.etree
etree.set_element_class_lookup(etree.ElementDefaultClassLookup())
if len(sys.argv) > 1:
if '-a' in sys.argv or '-c' in sys.argv:
# 'all' or 'C-implementations' ?
try:
sys.argv.remove('-c')
except ValueError:
pass
try:
import cElementTree as cET
_etrees.append(cET)
except ImportError:
try:
import xml.etree.cElementTree as cET
_etrees.append(cET)
except ImportError:
pass
try:
# 'all' ?
sys.argv.remove('-a')
except ValueError:
pass
else:
try:
from elementtree import ElementTree as ET
_etrees.append(ET)
except ImportError:
try:
from xml.etree import ElementTree as ET
_etrees.append(ET)
except ImportError:
pass
if not _etrees:
print("No library to test. Exiting.")
sys.exit(1)
print("Preparing test suites and trees ...")
selected = set(sys.argv[1:])
benchmark_suites, benchmarks = \
buildSuites(benchmark_class, _etrees, selected)
print("Running benchmark on",
', '.join(b.lib_name for b in benchmark_suites))
print('')
printSetupTimes(benchmark_suites)
if callgrind_zero:
cmd = open("callgrind.cmd", 'w')
cmd.write('+Instrumentation\n')
cmd.write('Zero\n')
cmd.close()
runBenchmarks(benchmark_suites, benchmarks)
class RDFXMLReader:
CORE_SYNTAX_TERMS = {
RDF.RDF, RDF.ID, RDF.about, RDF.parseType, RDF.resource, RDF.nodeID,
RDF.datatype
}
SYNTAX_TERMS = CORE_SYNTAX_TERMS | {RDF.Description, RDF.li}
OLD_TERMS = {RDF.aboutEach, RDF.aboutEachPrefix, RDF.bagID}
XML_TERMS = {XML.base, XML.lang}
ILLEGAL_NODE_TAGS = CORE_SYNTAX_TERMS | {RDF.li} | OLD_TERMS
ILLEGAL_PROPERTY_TAGS = CORE_SYNTAX_TERMS | {RDF.Description} | OLD_TERMS
ILLEGAL_PROPERTY_ATTRS = SYNTAX_TERMS | OLD_TERMS
_PARSER_LOOKUP = etree.ElementDefaultClassLookup(element=Element)
def __init__(self, parser=None):
if parser is None:
parser = etree.XMLParser(remove_comments=True, remove_pis=True)
parser.set_element_class_lookup(self._PARSER_LOOKUP)
self.parser = parser
def read(self, lines, base_uri=None):
root = etree.parse(lines, self.parser, base_url=base_uri).getroot()
ids = set()
# rdf:RDF is not necessarily the root element.
for element in root if root.uri == RDF.RDF else [root]:
self._validate(element)
for triple in self._node_element(element, ids):
yield triple
def _validate(self, element):
for attr, value in element.items():
attr = QName(attr)
# Ignore unknown and reserved XML attributes.
if attr.namespace is None or (attr.namespace == XML
and attr not in _XML_ATTRS):
del element.attrib[attr]
# Validate but ignore old syntax terms.
elif attr == QName(RDF, 'bagID'):
if not _NCNAME.match(value):
raise ParseError(
"rdf:bagID does not match NCName: {!r}".format(value))
del element.attrib[attr]
elif attr in _OLD_ATTRS:
raise ParseError
def _node_element(self, element, ids):
# 7.2.11 Production nodeElement
self._validate(element)
if element.uri in self.ILLEGAL_NODE_TAGS:
raise ParseError("Illegal node element: {!s}".format(element.tag))
element.subject = self._subject(element, ids)
# 2.13 Typed Node Elements
if element.uri != RDF.Description:
yield (element.subject, RDF.type, element.uri)
for triple in self._property_attrs(element):
yield triple
for triple in self._property_elements(element, ids):
yield triple
def _subject(self, element, ids):
id_ = self._id(element, ids)
node_id = element.get(QName(RDF, 'nodeID'))
about = element.get(QName(RDF, 'about'))
if id_ is not None:
if node_id is None:
if about is None:
return id_
raise ParseError
raise ParseError
elif node_id is not None:
if about is None:
if _NCNAME.match(node_id):
return BlankNode(node_id)
raise ParseError
raise ParseError
elif about is not None:
return self._uri(about, element.base_uri)
return BlankNode()
def _uri(self, uri, base_uri=None):
if base_uri and not uri:
base_uri = base_uri.rsplit('#', 1)[0]
return URI(urllib.parse.urljoin(base_uri or '', uri))
def _id(self, element, ids):
name = element.get(QName(RDF, 'ID'))
if name is not None:
if _NCNAME.match(name):
uri = self._uri('#' + name, element.base_uri)
if uri not in ids:
ids.add(uri)
return uri
else:
raise ParseError("rdf:ID is not unique: {!r}".format(uri))
else:
raise ParseError(
"rdf:ID does not match NCName: {!r}".format(name))
def _property_attrs(self, element):
# 2.5 Property Attributes
for attr, value in element.items():
if attr not in _XML_ATTRS:
predicate = URI(QName(attr))
if predicate not in self.ILLEGAL_PROPERTY_ATTRS:
if predicate != RDF.type:
object_ = PlainLiteral(value, element.language)
else:
object_ = URI(value)
yield (element.subject, predicate, object_)
elif predicate == RDF.li:
raise ParseError("rdf:li is not allowed as attribute")
def _property_elements(self, parent, ids):
# 7.2.13 Production propertyEltList
li_counter = 1
for element in parent:
# 7.2.14 Production propertyElt
self._validate(element)
if element.uri in self.ILLEGAL_PROPERTY_TAGS:
raise ParseError("Illegal property element: {!s}".format(
element.tag))
elif element.uri == RDF.li:
# Container Membership Property Elements: rdf:li and rdf:_n
element.uri = RDF['_' + str(li_counter)]
li_counter += 1
parse_type = element.attrib.get(QName(RDF, 'parseType'))
legal_attrs = _XML_ATTRS | {QName(RDF, 'ID')}
if parse_type is not None:
legal_attrs.add(QName(RDF, 'parseType'))
if any(attr not in legal_attrs for attr in element.keys()):
raise ParseError
elif parse_type == 'Resource':
triples = self._parse_type_resource_property(
element, parent, ids)
elif parse_type == 'Collection':
triples = self._parse_type_collection_property(
element, parent, ids)
else:
triples = self._parse_type_literal_property(
element, parent, ids)
elif len(element) == 1:
if all(attr not in legal_attrs for attr in element.keys()):
triples = self._resource_property(element, parent, ids)
else:
raise ParseError
elif len(element) == 0:
if element.text:
legal_attrs.add(QName(RDF, 'datatype'))
if all(attr in legal_attrs for attr in element.keys()):
triples = self._literal_property(element, parent, ids)
else:
raise ParseError
else:
triples = self._empty_property(element, parent, ids)
for triple in triples:
yield triple
def _reify(self, uri, triple):
yield (uri, RDF.type, RDF.Statement)
yield (uri, RDF.subject, triple[0])
yield (uri, RDF.predicate, triple[1])
yield (uri, RDF.object, triple[2])
def _resource_property(self, element, parent, ids):
# 7.2.15 Production resourcePropertyElt
node_element = element[0]
for triple in self._node_element(node_element, ids):
yield triple
triple = (parent.subject, element.uri, node_element.subject)
yield triple
id_ = self._id(element, ids)
if id_ is not None:
# 7.3 Reification Rules
for triple in self._reify(id_, triple):
yield triple
def _literal_property(self, element, parent, ids):
# 7.2.16 Production literalPropertyElt
datatype = element.get(QName(RDF, 'datatype'))
if datatype is not None:
object_ = TypedLiteral(element.text, URI(datatype))
else:
object_ = PlainLiteral(element.text, element.language)
triple = (parent.subject, element.uri, object_)
yield triple
id_ = self._id(element, ids)
if id_ is not None:
# 7.3 Reification Rules
for triple in self._reify(id_, triple):
yield triple
def _parse_type_resource_property(self, element, parent, ids):
# 7.2.18 Production parseTypeResourcePropertyElt
node_element = element.makeelement(QName(RDF, 'Description'))
node_element[:] = element
for triple in self._node_element(node_element, ids):
yield triple
triple = (parent.subject, element.uri, node_element.subject)
yield triple
id_ = self._id(element, ids)
if id_ is not None:
# 7.3 Reification Rules
for triple in self._reify(id_, triple):
yield triple
def _parse_type_collection_property(self, element, parent, ids):
# 7.2.19 Production parseTypeCollectionPropertyElt
node_ids = []
for node_element in element:
for triple in self._node_element(node_element, ids):
yield triple
node_ids.append((node_element, BlankNode()))
for node_element, object_ in node_ids:
break
else:
object_ = RDF.nil
triple = (parent.subject, element.uri, object_)
yield triple
id_ = self._id(element, ids)
if id_ is not None:
# 7.3 Reification Rules
for triple in self._reify(id_, triple):
yield triple
for i, (node_element, object_) in enumerate(node_ids):
yield (object_, RDF.first, node_element.subject)
try:
next_pair = node_ids[i + 1]
except IndexError:
next_object = RDF.nil
else:
next_element, next_object = next_pair
yield (object_, RDF.rest, next_object)
def _parse_type_literal_property(self, element, parent, ids):
literal = element.text or ""
if len(element):
tree = etree.ElementTree(element[0])
bytes_io = BytesIO()
tree.write_c14n(bytes_io, exclusive=True, with_comments=True)
literal += bytes_io.getvalue().decode('utf-8')
literal += element[0].tail or ""
object_ = TypedLiteral(literal, RDF.XMLLiteral)
triple = (parent.subject, element.uri, object_)
yield triple
id_ = self._id(element, ids)
if id_ is not None:
# 7.3 Reification Rules
for triple in self._reify(id_, triple):
yield triple
def _empty_property(self, element, parent, ids):
# 7.2.21 Production emptyPropertyElt
id_ = self._id(element, ids)
literal_attrs = _XML_ATTRS | {QName(RDF, 'ID')}
if all(attr in literal_attrs for attr in element.keys()):
object_ = PlainLiteral("", element.language)
triple = (parent.subject, element.uri, object_)
yield triple
if id_ is not None:
for triple in self._reify(id_, triple):
yield triple
else:
resource = element.attrib.get(QName(RDF, 'resource'))
node_id = element.attrib.get(QName(RDF, 'nodeID'))
if resource is not None:
if node_id is None:
object_ = self._uri(resource, element.base_uri)
else:
raise ParseError
elif node_id is not None:
if _NCNAME.match(node_id):
object_ = BlankNode(node_id)
else:
raise ParseError(
"rdf:nodeID does not match NCName: {!r}".format(
node_id))
else:
object_ = BlankNode()
triple = (parent.subject, element.uri, object_)
yield triple
if id_ is not None:
for triple in self._reify(id_, triple):
yield triple
subject = object_
property_attrs = set(element.keys())
property_attrs -= literal_attrs | {
QName(RDF, 'resource'),
QName(RDF, 'nodeID')
}
for attr in property_attrs:
predicate = URI(QName(attr))
if predicate in self.XML_TERMS:
continue
elif predicate in self.ILLEGAL_PROPERTY_ATTRS:
raise ParseError
value = element.get(attr)
if predicate != RDF.type:
object_ = PlainLiteral(value, element.language)
else:
object_ = self._uri(value, element.base_uri)
yield (subject, predicate, object_)
def HTMLParser(*args, **kwargs):
lookup = etree.ElementDefaultClassLookup(element=HTMLElement)
parser = etree.HTMLParser(*args, **kwargs)
parser.set_element_class_lookup(lookup)
return parser
def HTMLParser(*args, **kwargs):
kwargs.setdefault('encoding', 'utf-8')
lookup = etree.ElementDefaultClassLookup(element=HTMLElement)
parser = etree.HTMLParser(*args, **kwargs)
parser.set_element_class_lookup(lookup)
return parser
return True
class ExactlyOneError(ValueError):
pass
def one(mylist):
"""
assert that there's only one thing, and get it.
"""
if len(mylist) != 1:
raise ExactlyOneError(
'Expected exactly one item. Got %i: %r' % (
len(mylist),
[
item.tostring()
if isinstance(item, etree.ElementBase)
else item
for item in mylist
]
)
)
return mylist[0]
node_lookup = etree.ElementDefaultClassLookup(element=RefactorLibNodeBase)
__all__ = ('RefactorLibNodeBase',)
@staticmethod
def to_string(element):
if callable(element):
element = element()
return etree.tostring(
element,
encoding='utf-8',
xml_declaration=True,
pretty_print=True,
)
XMLParserLookup = etree.ElementNamespaceClassLookup(
fallback=etree.ElementDefaultClassLookup(element=ElementBase))
XMLParser = etree.XMLParser(encoding='utf-8', no_network=False)
XMLParser.set_element_class_lookup(XMLParserLookup)
E = ElementMaker(
nsmap=SOAP_NSMAP,
makeelement=XMLParser.makeelement,
)
S = ElementMaker(
namespace=SOAP_ENV_URI,
nsmap=SOAP_NSMAP,
makeelement=XMLParser.makeelement,
)
# create etree parser using custom Element class
class LayoutElement(etree.ElementBase):
@property
def layout(self):
if not hasattr(self, '_layout'):
self._layout = None
return self._layout
@layout.setter
def layout(self, value):
self._layout = value
parser_lookup = etree.ElementDefaultClassLookup(element=LayoutElement)
parser = etree.XMLParser()
parser.set_element_class_lookup(parser_lookup)
# main class
class PDFQuery(object):
def __init__(
self,
file,
merge_tags=('LTChar', 'LTAnno'),
round_floats=True,
round_digits=3,
input_text_formatter=None,
normalize_spaces=True,
resort=True,