aboutsummaryrefslogtreecommitdiffstats
path: root/sleekxmpp/xmlstream/tostring
diff options
context:
space:
mode:
authorThibaut Horel <thibaut.horel@gmail.com>2010-11-08 00:59:14 +0100
committerThibaut Horel <thibaut.horel@gmail.com>2010-11-08 00:59:14 +0100
commitb0a2a305028bf284fc5dcf7e1a696d85787f128f (patch)
treee6463e36e381b4342b7c864200a3482cca182618 /sleekxmpp/xmlstream/tostring
parentb8499306ce329ca3881b1d1dfc3362a3a5c115d0 (diff)
downloadalias-b0a2a305028bf284fc5dcf7e1a696d85787f128f.tar.gz
Add the sleekxmpp library (will be added as a submodule later)
Diffstat (limited to 'sleekxmpp/xmlstream/tostring')
-rw-r--r--sleekxmpp/xmlstream/tostring/__init__.py19
-rw-r--r--sleekxmpp/xmlstream/tostring/tostring.py95
-rw-r--r--sleekxmpp/xmlstream/tostring/tostring26.py101
3 files changed, 215 insertions, 0 deletions
diff --git a/sleekxmpp/xmlstream/tostring/__init__.py b/sleekxmpp/xmlstream/tostring/__init__.py
new file mode 100644
index 0000000..5852cba
--- /dev/null
+++ b/sleekxmpp/xmlstream/tostring/__init__.py
@@ -0,0 +1,19 @@
+"""
+ SleekXMPP: The Sleek XMPP Library
+ Copyright (C) 2010 Nathanael C. Fritz
+ This file is part of SleekXMPP.
+
+ See the file LICENSE for copying permission.
+"""
+
+import sys
+
+# Import the correct tostring and xml_escape functions based on the Python
+# version in order to properly handle Unicode.
+
+if sys.version_info < (3, 0):
+ from sleekxmpp.xmlstream.tostring.tostring26 import tostring, xml_escape
+else:
+ from sleekxmpp.xmlstream.tostring.tostring import tostring, xml_escape
+
+__all__ = ['tostring', 'xml_escape']
diff --git a/sleekxmpp/xmlstream/tostring/tostring.py b/sleekxmpp/xmlstream/tostring/tostring.py
new file mode 100644
index 0000000..d8f5c5b
--- /dev/null
+++ b/sleekxmpp/xmlstream/tostring/tostring.py
@@ -0,0 +1,95 @@
+"""
+ SleekXMPP: The Sleek XMPP Library
+ Copyright (C) 2010 Nathanael C. Fritz
+ This file is part of SleekXMPP.
+
+ See the file LICENSE for copying permission.
+"""
+
+
+def tostring(xml=None, xmlns='', stanza_ns='', stream=None, outbuffer=''):
+ """
+ Serialize an XML object to a Unicode string.
+
+ If namespaces are provided using xmlns or stanza_ns, then elements
+ that use those namespaces will not include the xmlns attribute in
+ the output.
+
+ Arguments:
+ xml -- The XML object to serialize. If the value is None,
+ then the XML object contained in this stanza
+ object will be used.
+ xmlns -- Optional namespace of an element wrapping the XML
+ object.
+ stanza_ns -- The namespace of the stanza object that contains
+ the XML object.
+ stream -- The XML stream that generated the XML object.
+ outbuffer -- Optional buffer for storing serializations during
+ recursive calls.
+ """
+ # Add previous results to the start of the output.
+ output = [outbuffer]
+
+ # Extract the element's tag name.
+ tag_name = xml.tag.split('}', 1)[-1]
+
+ # Extract the element's namespace if it is defined.
+ if '}' in xml.tag:
+ tag_xmlns = xml.tag.split('}', 1)[0][1:]
+ else:
+ tag_xmlns = ''
+
+ # Output the tag name and derived namespace of the element.
+ namespace = ''
+ if tag_xmlns not in ['', xmlns, stanza_ns]:
+ namespace = ' xmlns="%s"' % tag_xmlns
+ if stream and tag_xmlns in stream.namespace_map:
+ mapped_namespace = stream.namespace_map[tag_xmlns]
+ if mapped_namespace:
+ tag_name = "%s:%s" % (mapped_namespace, tag_name)
+ output.append("<%s" % tag_name)
+ output.append(namespace)
+
+ # Output escaped attribute values.
+ for attrib, value in xml.attrib.items():
+ if '{' not in attrib:
+ value = xml_escape(value)
+ output.append(' %s="%s"' % (attrib, value))
+
+ if len(xml) or xml.text:
+ # If there are additional child elements to serialize.
+ output.append(">")
+ if xml.text:
+ output.append(xml_escape(xml.text))
+ if len(xml):
+ for child in xml.getchildren():
+ output.append(tostring(child, tag_xmlns, stanza_ns, stream))
+ output.append("</%s>" % tag_name)
+ elif xml.text:
+ # If we only have text content.
+ output.append(">%s</%s>" % (xml_escape(xml.text), tag_name))
+ else:
+ # Empty element.
+ output.append(" />")
+ if xml.tail:
+ # If there is additional text after the element.
+ output.append(xml_escape(xml.tail))
+ return ''.join(output)
+
+
+def xml_escape(text):
+ """
+ Convert special characters in XML to escape sequences.
+
+ Arguments:
+ text -- The XML text to convert.
+ """
+ text = list(text)
+ escapes = {'&': '&amp;',
+ '<': '&lt;',
+ '>': '&gt;',
+ "'": '&apos;',
+ '"': '&quot;'}
+ for i, c in enumerate(text):
+ text[i] = escapes.get(c, c)
+ return ''.join(text)
diff --git a/sleekxmpp/xmlstream/tostring/tostring26.py b/sleekxmpp/xmlstream/tostring/tostring26.py
new file mode 100644
index 0000000..0ee432c
--- /dev/null
+++ b/sleekxmpp/xmlstream/tostring/tostring26.py
@@ -0,0 +1,101 @@
+"""
+ SleekXMPP: The Sleek XMPP Library
+ Copyright (C) 2010 Nathanael C. Fritz
+ This file is part of SleekXMPP.
+
+ See the file LICENSE for copying permission.
+"""
+
+from __future__ import unicode_literals
+import types
+
+
+def tostring(xml=None, xmlns='', stanza_ns='', stream=None, outbuffer=''):
+ """
+ Serialize an XML object to a Unicode string.
+
+ If namespaces are provided using xmlns or stanza_ns, then elements
+ that use those namespaces will not include the xmlns attribute in
+ the output.
+
+ Arguments:
+ xml -- The XML object to serialize. If the value is None,
+ then the XML object contained in this stanza
+ object will be used.
+ xmlns -- Optional namespace of an element wrapping the XML
+ object.
+ stanza_ns -- The namespace of the stanza object that contains
+ the XML object.
+ stream -- The XML stream that generated the XML object.
+ outbuffer -- Optional buffer for storing serializations during
+ recursive calls.
+ """
+ # Add previous results to the start of the output.
+ output = [outbuffer]
+
+ # Extract the element's tag name.
+ tag_name = xml.tag.split('}', 1)[-1]
+
+ # Extract the element's namespace if it is defined.
+ if '}' in xml.tag:
+ tag_xmlns = xml.tag.split('}', 1)[0][1:]
+ else:
+ tag_xmlns = u''
+
+ # Output the tag name and derived namespace of the element.
+ namespace = u''
+ if tag_xmlns not in ['', xmlns, stanza_ns]:
+ namespace = u' xmlns="%s"' % tag_xmlns
+ if stream and tag_xmlns in stream.namespace_map:
+ mapped_namespace = stream.namespace_map[tag_xmlns]
+ if mapped_namespace:
+ tag_name = u"%s:%s" % (mapped_namespace, tag_name)
+ output.append(u"<%s" % tag_name)
+ output.append(namespace)
+
+ # Output escaped attribute values.
+ for attrib, value in xml.attrib.items():
+ if '{' not in attrib:
+ value = xml_escape(value)
+ output.append(u' %s="%s"' % (attrib, value))
+
+ if len(xml) or xml.text:
+ # If there are additional child elements to serialize.
+ output.append(u">")
+ if xml.text:
+ output.append(xml_escape(xml.text))
+ if len(xml):
+ for child in xml.getchildren():
+ output.append(tostring(child, tag_xmlns, stanza_ns, stream))
+ output.append(u"</%s>" % tag_name)
+ elif xml.text:
+ # If we only have text content.
+ output.append(u">%s</%s>" % (xml_escape(xml.text), tag_name))
+ else:
+ # Empty element.
+ output.append(u" />")
+ if xml.tail:
+ # If there is additional text after the element.
+ output.append(xml_escape(xml.tail))
+ return u''.join(output)
+
+
+def xml_escape(text):
+ """
+ Convert special characters in XML to escape sequences.
+
+ Arguments:
+ text -- The XML text to convert.
+ """
+ if type(text) != types.UnicodeType:
+ text = list(unicode(text, 'utf-8', 'ignore'))
+ else:
+ text = list(text)
+ escapes = {u'&': u'&amp;',
+ u'<': u'&lt;',
+ u'>': u'&gt;',
+ u"'": u'&apos;',
+ u'"': u'&quot;'}
+ for i, c in enumerate(text):
+ text[i] = escapes.get(c, c)
+ return u''.join(text)