def _escape_cdata(text, encoding=None):
# escape character data
try:
if encoding:
try:
text = text.encode(encoding)
except UnicodeError:
return _encode_entity(text)
text = nonentity_re.sub("&", text)
text = text.replace("<", "<")
return text
except (TypeError, AttributeError):
_raise_serialization_error(text)
def _escape_cdata(text, encoding=None):
# escape character data
try:
if encoding:
try:
text = text.encode(encoding)
except UnicodeError:
return _encode_entity(text)
text = nonentity_re.sub('&', text)
text = text.replace("<", "<")
return text
except (TypeError, AttributeError):
_raise_serialization_error(text)
def _escape_attrib(text, encoding=None):
# escape attribute value
try:
if encoding:
try:
text = text.encode(encoding)
except UnicodeError:
return _encode_entity(text)
# don't requote properly-quoted entities
text = nonentity_re.sub("&", text)
text = text.replace("<", "<")
text = text.replace('"', """)
return text
except (TypeError, AttributeError):
_raise_serialization_error(text)
def _escape_attrib(text, encoding=None):
# escape attribute value
try:
if encoding:
try:
text = text.encode(encoding)
except UnicodeError:
return _encode_entity(text)
# don't requote properly-quoted entities
text = nonentity_re.sub('&', text)
text = text.replace("<", "<")
text = text.replace('"', """)
return text
except (TypeError, AttributeError):
_raise_serialization_error(text)
def _write_html(write, node, encoding, namespaces, depth=-1, maxdepth=None):
" Write HTML to file " ""
if encoding is None:
encoding = "utf-8"
tag = node.tag
tail = node.tail
text = node.text
tail = node.tail
to_write = ""
if tag is Replace:
if not node.structure:
if cdata_needs_escaping(text):
text = _escape_cdata(text)
write(text.encode(encoding))
elif tag is Comment:
if cdata_needs_escaping(text):
text = _escape_cdata(text)
write("<!-- " + text + " -->".encode(encoding))
elif tag is ProcessingInstruction:
if cdata_needs_escaping(text):
text = _escape_cdata(text)
write("<!-- " + text + " -->".encode(encoding))
else:
xmlns_items = [] # new namespaces in this scope
try:
if tag[:1] == "{":
if tag[:_XHTML_PREFIX_LEN] == _XHTML_PREFIX:
tag = tag[_XHTML_PREFIX_LEN:]
else:
tag, xmlns = fixtag(tag, namespaces)
if xmlns:
xmlns_items.append(xmlns)
except TypeError:
_raise_serialization_error(tag)
to_write += "<%s" % tag.encode(encoding)
attrib = node.attrib
if attrib is not None:
if len(attrib) > 1:
attrib_keys = attrib.keys()
attrib_keys.sort()
else:
attrib_keys = attrib
for k in attrib_keys:
try:
if k[:1] == "{":
continue
except TypeError:
_raise_serialization_error(k)
if k in _HTMLATTRS_BOOLEAN:
to_write += " " + k.encode(encoding)
else:
v = attrib[k]
to_write += ' %s="%s"' % (k, v)
for k, v in xmlns_items:
to_write += ' %s="%s"' % (k, v)
to_write += ">"
if text is not None and text:
if tag in _HTMLTAGS_NOESCAPE:
to_write += text.encode(encoding)
elif cdata_needs_escaping(text):
to_write += _escape_cdata(text)
else:
to_write += text.encode(encoding)
write(to_write)
for child in node._children:
if maxdepth is not None:
depth = depth + 1
if depth < maxdepth:
_write_html(write, child, encoding, namespaces, depth, maxdepth)
elif depth == maxdepth and text:
write(" [...]\n")
else:
_write_html(write, child, encoding, namespaces, depth, maxdepth)
if text or node._children or tag not in _HTMLTAGS_UNBALANCED:
write("</" + tag.encode(encoding) + ">")
if tail:
if cdata_needs_escaping(tail):
write(_escape_cdata(tail))
else:
write(tail.encode(encoding))
def _write_xml(write, node, encoding, namespaces, pipeline, xhtml=False):
""" Write XML to a file """
if encoding is None:
encoding = "utf-8"
tag = node.tag
if tag is Comment:
write("<!-- %s -->" % _escape_cdata(node.text, encoding))
elif tag is ProcessingInstruction:
write("<?%s?>" % _escape_cdata(node.text, encoding))
elif tag is Replace:
if node.structure:
# this may produce invalid xml
write(node.text.encode(encoding))
else:
write(_escape_cdata(node.text, encoding))
else:
if xhtml:
if tag[:_XHTML_PREFIX_LEN] == _XHTML_PREFIX:
tag = tag[_XHTML_PREFIX_LEN:]
if node.attrib:
items = node.attrib.items()
else:
items = [] # must always be sortable.
xmlns_items = [] # new namespaces in this scope
try:
if tag[:1] == "{":
tag, xmlns = fixtag(tag, namespaces)
if xmlns:
xmlns_items.append(xmlns)
except TypeError:
_raise_serialization_error(tag)
write("<" + tag.encode(encoding))
if items or xmlns_items:
items.sort() # lexical order
for k, v in items:
try:
if k[:1] == "{":
if not pipeline:
if k == _MELD_ID:
continue
k, xmlns = fixtag(k, namespaces)
if xmlns:
xmlns_items.append(xmlns)
if not pipeline:
# special-case for HTML input
if k == "xmlns:meld":
continue
except TypeError:
_raise_serialization_error(k)
write(' %s="%s"' % (k.encode(encoding), _escape_attrib(v, encoding)))
for k, v in xmlns_items:
write(' %s="%s"' % (k.encode(encoding), _escape_attrib(v, encoding)))
if node.text or node._children:
write(">")
if node.text:
write(_escape_cdata(node.text, encoding))
for n in node._children:
_write_xml(write, n, encoding, namespaces, pipeline, xhtml)
write("</" + tag.encode(encoding) + ">")
else:
write(" />")
for k, v in xmlns_items:
del namespaces[v]
if node.tail:
write(_escape_cdata(node.tail, encoding))
def _write_html_no_encoding(write, node, namespaces):
""" Append HTML to string without any particular unicode encoding.
We have a separate function for this due to the fact that encoding
while recursing is very expensive if this will get serialized out to
utf8 anyway (the encoding can happen afterwards). We append to a string
because it's faster than calling any 'write' or 'append' function."""
tag = node.tag
tail = node.tail
text = node.text
tail = node.tail
to_write = ""
if tag is Replace:
if not node.structure:
if cdata_needs_escaping(text):
text = _escape_cdata_noencoding(text)
write(text)
elif tag is Comment:
if cdata_needs_escaping(text):
text = _escape_cdata_noencoding(text)
write("<!-- " + text + " -->")
elif tag is ProcessingInstruction:
if cdata_needs_escaping(text):
text = _escape_cdata_noencoding(text)
write("<!-- " + text + " -->")
else:
xmlns_items = [] # new namespaces in this scope
try:
if tag[:1] == "{":
if tag[:_XHTML_PREFIX_LEN] == _XHTML_PREFIX:
tag = tag[_XHTML_PREFIX_LEN:]
else:
tag, xmlns = fixtag(tag, namespaces)
if xmlns:
xmlns_items.append(xmlns)
except TypeError:
_raise_serialization_error(tag)
to_write += "<" + tag
attrib = node.attrib
if attrib is not None:
if len(attrib) > 1:
attrib_keys = attrib.keys()
attrib_keys.sort()
else:
attrib_keys = attrib
for k in attrib_keys:
try:
if k[:1] == "{":
continue
except TypeError:
_raise_serialization_error(k)
if k in _HTMLATTRS_BOOLEAN:
to_write += " " + k
else:
v = attrib[k]
to_write += ' %s="%s"' % (k, v)
for k, v in xmlns_items:
to_write += ' %s="%s"' % (k, v)
to_write += ">"
if text is not None and text:
if tag in _HTMLTAGS_NOESCAPE:
to_write += text
elif cdata_needs_escaping(text):
to_write += _escape_cdata_noencoding(text)
else:
to_write += text
write(to_write)
for child in node._children:
_write_html_no_encoding(write, child, namespaces)
if text or node._children or tag not in _HTMLTAGS_UNBALANCED:
write("</" + tag + ">")
if tail:
if cdata_needs_escaping(tail):
write(_escape_cdata_noencoding(tail))
else:
write(tail)
def _write_xml(write, node, encoding, namespaces, pipeline, xhtml=False):
""" Write XML to a file """
if encoding is None:
encoding = 'utf-8'
tag = node.tag
if tag is Comment:
write("<!-- %s -->" % _escape_cdata(node.text, encoding))
elif tag is ProcessingInstruction:
write("<?%s?>" % _escape_cdata(node.text, encoding))
elif tag is Replace:
if node.structure:
# this may produce invalid xml
write(node.text.encode(encoding))
else:
write(_escape_cdata(node.text, encoding))
else:
if xhtml:
if tag[:_XHTML_PREFIX_LEN] == _XHTML_PREFIX:
tag = tag[_XHTML_PREFIX_LEN:]
if node.attrib:
items = node.attrib.items()
else:
items = [] # must always be sortable.
xmlns_items = [] # new namespaces in this scope
try:
if tag[:1] == "{":
tag, xmlns = fixtag(tag, namespaces)
if xmlns:
xmlns_items.append(xmlns)
except TypeError:
_raise_serialization_error(tag)
write("<" + tag.encode(encoding))
if items or xmlns_items:
items.sort() # lexical order
for k, v in items:
try:
if k[:1] == "{":
if not pipeline:
if k == _MELD_ID:
continue
k, xmlns = fixtag(k, namespaces)
if xmlns: xmlns_items.append(xmlns)
if not pipeline:
# special-case for HTML input
if k == 'xmlns:meld':
continue
except TypeError:
_raise_serialization_error(k)
write(" %s=\"%s\"" %
(k.encode(encoding), _escape_attrib(v, encoding)))
for k, v in xmlns_items:
write(" %s=\"%s\"" %
(k.encode(encoding), _escape_attrib(v, encoding)))
if node.text or node._children:
write(">")
if node.text:
write(_escape_cdata(node.text, encoding))
for n in node._children:
_write_xml(write, n, encoding, namespaces, pipeline, xhtml)
write("</" + tag.encode(encoding) + ">")
else:
write(" />")
for k, v in xmlns_items:
del namespaces[v]
if node.tail:
write(_escape_cdata(node.tail, encoding))
def _write_html_no_encoding(write, node, namespaces):
""" Append HTML to string without any particular unicode encoding.
We have a separate function for this due to the fact that encoding
while recursing is very expensive if this will get serialized out to
utf8 anyway (the encoding can happen afterwards). We append to a string
because it's faster than calling any 'write' or 'append' function."""
tag = node.tag
tail = node.tail
text = node.text
tail = node.tail
to_write = ""
if tag is Replace:
if not node.structure:
if cdata_needs_escaping(text):
text = _escape_cdata_noencoding(text)
write(text)
elif tag is Comment:
if cdata_needs_escaping(text):
text = _escape_cdata_noencoding(text)
write('<!-- ' + text + ' -->')
elif tag is ProcessingInstruction:
if cdata_needs_escaping(text):
text = _escape_cdata_noencoding(text)
write('<!-- ' + text + ' -->')
else:
xmlns_items = [] # new namespaces in this scope
try:
if tag[:1] == "{":
if tag[:_XHTML_PREFIX_LEN] == _XHTML_PREFIX:
tag = tag[_XHTML_PREFIX_LEN:]
else:
tag, xmlns = fixtag(tag, namespaces)
if xmlns:
xmlns_items.append(xmlns)
except TypeError:
_raise_serialization_error(tag)
to_write += "<" + tag
attrib = node.attrib
if attrib is not None:
if len(attrib) > 1:
attrib_keys = attrib.keys()
attrib_keys.sort()
else:
attrib_keys = attrib
for k in attrib_keys:
try:
if k[:1] == "{":
continue
except TypeError:
_raise_serialization_error(k)
if k in _HTMLATTRS_BOOLEAN:
to_write += ' ' + k
else:
v = attrib[k]
to_write += " %s=\"%s\"" % (k, v)
for k, v in xmlns_items:
to_write += " %s=\"%s\"" % (k, v)
to_write += ">"
if text is not None and text:
if tag in _HTMLTAGS_NOESCAPE:
to_write += text
elif cdata_needs_escaping(text):
to_write += _escape_cdata_noencoding(text)
else:
to_write += text
write(to_write)
for child in node._children:
_write_html_no_encoding(write, child, namespaces)
if text or node._children or tag not in _HTMLTAGS_UNBALANCED:
write("</" + tag + ">")
if tail:
if cdata_needs_escaping(tail):
write(_escape_cdata_noencoding(tail))
else:
write(tail)
def _write_html(write, node, encoding, namespaces, depth=-1, maxdepth=None):
" Write HTML to file " ""
if encoding is None:
encoding = 'utf-8'
tag = node.tag
tail = node.tail
text = node.text
tail = node.tail
to_write = ""
if tag is Replace:
if not node.structure:
if cdata_needs_escaping(text):
text = _escape_cdata(text)
write(text.encode(encoding))
elif tag is Comment:
if cdata_needs_escaping(text):
text = _escape_cdata(text)
write('<!-- ' + text + ' -->'.encode(encoding))
elif tag is ProcessingInstruction:
if cdata_needs_escaping(text):
text = _escape_cdata(text)
write('<!-- ' + text + ' -->'.encode(encoding))
else:
xmlns_items = [] # new namespaces in this scope
try:
if tag[:1] == "{":
if tag[:_XHTML_PREFIX_LEN] == _XHTML_PREFIX:
tag = tag[_XHTML_PREFIX_LEN:]
else:
tag, xmlns = fixtag(tag, namespaces)
if xmlns:
xmlns_items.append(xmlns)
except TypeError:
_raise_serialization_error(tag)
to_write += "<%s" % tag.encode(encoding)
attrib = node.attrib
if attrib is not None:
if len(attrib) > 1:
attrib_keys = attrib.keys()
attrib_keys.sort()
else:
attrib_keys = attrib
for k in attrib_keys:
try:
if k[:1] == "{":
continue
except TypeError:
_raise_serialization_error(k)
if k in _HTMLATTRS_BOOLEAN:
to_write += ' ' + k.encode(encoding)
else:
v = attrib[k]
to_write += " %s=\"%s\"" % (k, v)
for k, v in xmlns_items:
to_write += " %s=\"%s\"" % (k, v)
to_write += ">"
if text is not None and text:
if tag in _HTMLTAGS_NOESCAPE:
to_write += text.encode(encoding)
elif cdata_needs_escaping(text):
to_write += _escape_cdata(text)
else:
to_write += text.encode(encoding)
write(to_write)
for child in node._children:
if maxdepth is not None:
depth = depth + 1
if depth < maxdepth:
_write_html(write, child, encoding, namespaces, depth,
maxdepth)
elif depth == maxdepth and text:
write(' [...]\n')
else:
_write_html(write, child, encoding, namespaces, depth,
maxdepth)
if text or node._children or tag not in _HTMLTAGS_UNBALANCED:
write("</" + tag.encode(encoding) + ">")
if tail:
if cdata_needs_escaping(tail):
write(_escape_cdata(tail))
else:
write(tail.encode(encoding))