annotate markup/output.py @ 1:821114ec4f69

Initial import.
author cmlenz
date Sat, 03 Jun 2006 07:16:01 +0000
parents
children 4cbebb15a834
rev   line source
1
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
1 # -*- coding: utf-8 -*-
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
2 #
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
3 # Copyright (C) 2006 Christopher Lenz
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
4 # All rights reserved.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
5 #
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
6 # This software is licensed as described in the file COPYING, which
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
7 # you should have received as part of this distribution. The terms
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
8 # are also available at http://trac.edgewall.com/license.html.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
9 #
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
10 # This software consists of voluntary contributions made by many
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
11 # individuals. For the exact contribution history, see the revision
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
12 # history and logs, available at http://projects.edgewall.com/trac/.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
13
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
14 """This module provides different kinds of serialization methods for XML event
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
15 streams.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
16 """
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
17
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
18 try:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
19 frozenset
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
20 except NameError:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
21 from sets import ImmutableSet as frozenset
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
22
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
23 from markup.core import Markup, QName, Stream
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
24 from markup.filters import WhitespaceFilter
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
25
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
26 __all__ = ['Serializer', 'XMLSerializer', 'HTMLSerializer']
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
27
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
28
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
29 class Serializer(object):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
30 """Base class for serializers."""
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
31
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
32 def serialize(self, stream):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
33 raise NotImplementedError
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
34
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
35
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
36 class XMLSerializer(Serializer):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
37 """Produces XML text from an event stream.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
38
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
39 >>> from markup.builder import tag
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
40 >>> elem = tag.DIV(tag.A(href='foo'), tag.BR, tag.HR(noshade=True))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
41 >>> print ''.join(XMLSerializer().serialize(elem.generate()))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
42 <div><a href="foo"/><br/><hr noshade="True"/></div>
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
43 """
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
44
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
45 def serialize(self, stream):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
46 ns_attrib = []
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
47 ns_mapping = {}
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
48
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
49 stream = PushbackIterator(stream)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
50 for kind, data, pos in stream:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
51
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
52 if kind is Stream.DOCTYPE:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
53 # FIXME: what if there's no system or public ID in the input?
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
54 yield Markup('<!DOCTYPE %s "%s" "%s">\n' % data)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
55
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
56 elif kind is Stream.START_NS:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
57 prefix, uri = data
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
58 if uri not in ns_mapping:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
59 ns_mapping[uri] = prefix
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
60 if not prefix:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
61 ns_attrib.append((QName('xmlns'), uri))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
62 else:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
63 ns_attrib.append((QName('xmlns:%s' % prefix), uri))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
64
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
65 elif kind is Stream.START:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
66 tag, attrib = data
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
67
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
68 tagname = tag.localname
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
69 if tag.namespace:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
70 try:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
71 prefix = ns_mapping[tag.namespace]
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
72 if prefix:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
73 tagname = prefix + ':' + tag.localname
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
74 except KeyError:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
75 ns_attrib.append((QName('xmlns'), tag.namespace))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
76 buf = ['<', tagname]
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
77
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
78 if ns_attrib:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
79 attrib.extend(ns_attrib)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
80 ns_attrib = []
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
81 for attr, value in attrib:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
82 attrname = attr.localname
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
83 if attr.namespace:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
84 try:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
85 prefix = ns_mapping[attr.namespace]
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
86 except KeyError:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
87 # FIXME: synthesize a prefix for the attribute?
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
88 prefix = ''
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
89 if prefix:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
90 attrname = prefix + ':' + attrname
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
91 buf.append(' %s="%s"' % (attrname, Markup.escape(value)))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
92
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
93 kind, data, pos = stream.next()
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
94 if kind is Stream.END:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
95 buf.append('/>')
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
96 else:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
97 buf.append('>')
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
98 stream.pushback((kind, data, pos))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
99
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
100 yield Markup(''.join(buf))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
101
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
102 elif kind is Stream.END:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
103 tag = data
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
104 tagname = tag.localname
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
105 if tag.namespace:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
106 prefix = ns_mapping[tag.namespace]
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
107 if prefix:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
108 tagname = prefix + ':' + tag.localname
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
109 yield Markup('</%s>' % tagname)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
110
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
111 elif kind is Stream.TEXT:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
112 yield Markup.escape(data, quotes=False)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
113
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
114
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
115 class HTMLSerializer(Serializer):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
116 """Produces HTML text from an event stream.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
117
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
118 >>> from markup.builder import tag
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
119 >>> elem = tag.DIV(tag.A(href='foo'), tag.BR, tag.HR(noshade=True))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
120 >>> print ''.join(HTMLSerializer().serialize(elem.generate()))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
121 <div><a href="foo"></a><br><hr noshade></div>
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
122 """
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
123
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
124 NAMESPACE = 'http://www.w3.org/1999/xhtml'
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
125
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
126 _EMPTY_ELEMS = frozenset(['area', 'base', 'basefont', 'br', 'col', 'frame',
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
127 'hr', 'img', 'input', 'isindex', 'link', 'meta',
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
128 'param'])
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
129 _BOOLEAN_ATTRS = frozenset(['selected', 'checked', 'compact', 'declare',
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
130 'defer', 'disabled', 'ismap', 'multiple',
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
131 'nohref', 'noresize', 'noshade', 'nowrap'])
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
132
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
133 def serialize(self, stream):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
134 ns_mapping = {}
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
135
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
136 stream = PushbackIterator(stream)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
137 for kind, data, pos in stream:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
138
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
139 if kind is Stream.DOCTYPE:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
140 yield Markup('<!DOCTYPE %s "%s" "%s">\n' % data)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
141
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
142 elif kind is Stream.START_NS:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
143 prefix, uri = data
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
144 if uri not in ns_mapping:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
145 ns_mapping[uri] = prefix
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
146
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
147 elif kind is Stream.START:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
148 tag, attrib = data
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
149 if tag.namespace and tag.namespace != self.NAMESPACE:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
150 continue # not in the HTML namespace, so don't emit
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
151 buf = ['<', tag.localname]
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
152 for attr, value in attrib:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
153 if attr.namespace and attr.namespace != self.NAMESPACE:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
154 continue # not in the HTML namespace, so don't emit
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
155 if attr.localname in self._BOOLEAN_ATTRS:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
156 if value:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
157 buf.append(' %s' % attr.localname)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
158 else:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
159 buf.append(' %s="%s"' % (attr.localname,
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
160 Markup.escape(value)))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
161
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
162 if tag.localname in self._EMPTY_ELEMS:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
163 kind, data, pos = stream.next()
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
164 if kind is not Stream.END:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
165 stream.pushback((kind, data, pos))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
166
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
167 yield Markup(''.join(buf + ['>']))
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
168
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
169 elif kind is Stream.END:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
170 tag = data
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
171 if tag.namespace and tag.namespace != self.NAMESPACE:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
172 continue # not in the HTML namespace, so don't emit
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
173 yield Markup('</%s>' % tag.localname)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
174
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
175 elif kind is Stream.TEXT:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
176 yield Markup.escape(data, quotes=False)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
177
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
178
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
179 class PushbackIterator(object):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
180 """A simple wrapper for iterators that allows pushing items back on the
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
181 queue via the `pushback()` method.
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
182
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
183 That can effectively be used to peek at the next item."""
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
184 __slots__ = ['iterable', 'buf']
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
185
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
186 def __init__(self, iterable):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
187 self.iterable = iter(iterable)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
188 self.buf = []
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
189
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
190 def __iter__(self):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
191 return self
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
192
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
193 def next(self):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
194 if self.buf:
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
195 return self.buf.pop(0)
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
196 return self.iterable.next()
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
197
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
198 def pushback(self, item):
821114ec4f69 Initial import.
cmlenz
parents:
diff changeset
199 self.buf.append(item)
Copyright (C) 2012-2017 Edgewall Software