def serialize(self, stream, encoding=None, fragment=False, format=None):
try:
text = ''.join(self.generate(stream, encoding, fragment, format))
except TypeError: # workaround for bug 905389 in Python < 2.5
text = ''.join(
tuple(self.generate(stream, encoding, fragment, format)))
if not fragment:
text = Format.strip(text)
return text
def serialize(self, stream, encoding=None,
fragment=False, format=None):
try:
text = ''.join(self.generate(stream, encoding, fragment, format))
except TypeError: # workaround for bug 905389 in Python < 2.5
text = ''.join(tuple(
self.generate(stream, encoding, fragment, format)))
if not fragment:
text = Format.strip(text)
return text
def generate(self, stream, encoding=None, fragment=False, format=None):
"""Serializes an event stream to bytes of the specified encoding.
This function yields an encoded string over and over until the
stream is exhausted.
"""
decl = self.decl
doctype = self.doctype
encoding = encoding or self.encoding or 'utf-8'
entity_map = self.entity_map
format = self._get_format(format)
if format:
if format.decl is not None:
decl = format.decl
if format.doctype is not None:
doctype = format.doctype
if format.entity_map is not None:
entity_map = format.entity_map
if entity_map == True:
# if True, use default HTML entity map
entity_map = default_entity_map
elif entity_map == False:
entity_map = None
if isinstance(doctype, basestring):
# allow doctype strings
doctype = doctypes[self.doctype]
escape_cdata = XMLSerializer.escape_cdata
escape_attrib = XMLSerializer.escape_attrib
lastev = None
stream = iter(stream)
names = NamespaceStack(self.namespaces)
if not fragment:
if decl:
yield '<?xml version="1.0" encoding="%s"?>\n' % encoding
if doctype is not None:
yield serialize_doctype(doctype) + '\n'
text = None
for ev, item in self.apply_filters(stream, format):
if ev in (START, END) and item.tag == Fragment:
continue
elif ev == TEXT:
if text is not None:
text = u''.join([text, item])
else:
text = item
continue
if lastev == START:
if ev == END and (not text or not (Format.strip(text)
or self.is_formatted(item.tag))) \
and self.can_be_empty_element(item.tag):
yield ' />'
lastev = END
text = None
names.pop()
continue
yield ">"
if text:
yield escape_cdata(text, encoding, entity_map)
text = None
if ev == START:
if item.tag == Comment:
yield "<!--%s-->" % item.text.encode(encoding)
lastev = COMMENT
continue
elif item.tag == ProcessingInstruction:
yield "<?%s?>" % item.text.encode(encoding)
lastev = PI
continue
else:
current_names = names.current
names.push(namespaces(item, remove=True))
qname = names.qname(item.tag, default=True)
yield "<" + qname.encode(encoding)
for k, v in item.attrib.items():
k = names.qname(k, default=False).encode(encoding)
v = escape_attrib(v, encoding)
yield ' %s="%s"' % (k, v)
for prefix, uri in names.current.items():
if prefix not in current_names \
or current_names[prefix] != uri:
v = escape_attrib(uri, encoding)
if prefix:
k = 'xmlns:' + prefix.encode(encoding)
else:
k = 'xmlns'
yield ' %s="%s"' % (k, v)
elif ev == END and item.tag not in (Comment,
ProcessingInstruction):
qname = names.qname(item.tag, default=True)
yield "</%s>" % qname.encode(encoding)
names.pop()
lastev = ev
if fragment and text:
yield escape_cdata(text, encoding, entity_map)
return
def generate(self, stream, encoding=None,
fragment=False, format=None):
"""Serializes an event stream to bytes of the specified encoding.
This function yields an encoded string over and over until the
stream is exhausted.
"""
decl = self.decl
doctype = self.doctype
encoding = encoding or self.encoding or 'utf-8'
entity_map = self.entity_map
format = self._get_format(format)
if format:
if format.decl is not None:
decl = format.decl
if format.doctype is not None:
doctype = format.doctype
if format.entity_map is not None:
entity_map = format.entity_map
if entity_map == True:
# if True, use default HTML entity map
entity_map = default_entity_map
elif entity_map == False:
entity_map = None
if isinstance(doctype, basestring):
# allow doctype strings
doctype = doctypes[self.doctype]
escape_cdata = XMLSerializer.escape_cdata
escape_attrib = XMLSerializer.escape_attrib
lastev = None
stream = iter(stream)
names = NamespaceStack(self.namespaces)
if not fragment:
if decl:
yield '<?xml version="1.0" encoding="%s"?>\n' % encoding
if doctype is not None:
yield serialize_doctype(doctype) + '\n'
text = None
for ev, item in self.apply_filters(stream, format):
if ev in (START, END) and item.tag == Fragment:
continue
elif ev == TEXT:
if text is not None:
text = u''.join([text, item])
else:
text = item
continue
if lastev == START:
if ev == END and (not text or not (Format.strip(text)
or self.is_formatted(item.tag))) \
and self.can_be_empty_element(item.tag):
yield ' />'
lastev = END
text = None
names.pop()
continue
yield ">"
if text:
yield escape_cdata(text, encoding, entity_map)
text = None
if ev == START:
if item.tag == Comment:
yield "<!--%s-->" % item.text.encode(encoding)
lastev = COMMENT
continue
elif item.tag == ProcessingInstruction:
yield "<?%s?>" % item.text.encode(encoding)
lastev = PI
continue
else:
current_names = names.current
names.push(namespaces(item, remove=True))
qname = names.qname(item.tag, default=True)
yield "<" + qname.encode(encoding)
for k, v in item.attrib.items():
k = names.qname(k, default=False).encode(encoding)
v = escape_attrib(v, encoding)
yield ' %s="%s"' % (k, v)
for prefix, uri in names.current.items():
if prefix not in current_names \
or current_names[prefix] != uri:
v = escape_attrib(uri, encoding)
if prefix:
k = 'xmlns:' + prefix.encode(encoding)
else:
k = 'xmlns'
yield ' %s="%s"' % (k, v)
elif ev == END and item.tag not in (
Comment, ProcessingInstruction):
qname = names.qname(item.tag, default=True)
yield "</%s>" % qname.encode(encoding)
names.pop()
lastev = ev
if fragment and text:
yield escape_cdata(text, encoding, entity_map)
return
