Mercurial > genshi > mirror

annotate markup/output.py @ 142:349b3ff5367d trunk

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Minor cleanup in XInclude filter.
author cmlenz
date Fri, 11 Aug 2006 08:31:58 +0000
parents 520a5b7dd6d2
children 3d4c214c979a
rev   line source
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
1 # -*- coding: utf-8 -*-
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
2 #
66
59eb24184e9c Switch copyright to Edgewall and URLs to markup.edgewall.org.
cmlenz
parents: 27
diff changeset
3 # Copyright (C) 2006 Edgewall Software
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
4 # All rights reserved.
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
5 #
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
6 # This software is licensed as described in the file COPYING, which
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
7 # you should have received as part of this distribution. The terms
66
59eb24184e9c Switch copyright to Edgewall and URLs to markup.edgewall.org.
cmlenz
parents: 27
diff changeset
8 # are also available at http://markup.edgewall.org/wiki/License.
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
9 #
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
10 # This software consists of voluntary contributions made by many
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
11 # individuals. For the exact contribution history, see the revision
66
59eb24184e9c Switch copyright to Edgewall and URLs to markup.edgewall.org.
cmlenz
parents: 27
diff changeset
12 # history and logs, available at http://markup.edgewall.org/log/.
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
13
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
14 """This module provides different kinds of serialization methods for XML event
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
15 streams.
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
16 """
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
17
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
18 from itertools import chain
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
19 try:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
20 frozenset
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
21 except NameError:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
22 from sets import ImmutableSet as frozenset
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
23 import re
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
24
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
25 from markup.core import escape, Markup, Namespace, QName, XML_NAMESPACE
105
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
26 from markup.core import DOCTYPE, START, END, START_NS, END_NS, TEXT, COMMENT, PI
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
27
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
28 __all__ = ['Serializer', 'XMLSerializer', 'HTMLSerializer']
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
29
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
30
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
31 class DocType(object):
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
32 """Defines a number of commonly used DOCTYPE declarations as constants."""
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
33
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
34 HTML_STRICT = ('html', '-//W3C//DTD HTML 4.01//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
35 'http://www.w3.org/TR/html4/strict.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
36 HTML_TRANSITIONAL = ('html', '-//W3C//DTD HTML 4.01 Transitional//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
37 'http://www.w3.org/TR/html4/loose.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
38 HTML = HTML_STRICT
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
39
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
40 XHTML_STRICT = ('html', '-//W3C//DTD XHTML 1.0 Strict//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
41 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
42 XHTML_TRANSITIONAL = ('html', '-//W3C//DTD XHTML 1.0 Transitional//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
43 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
44 XHTML = XHTML_STRICT
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
45
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
46
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
47 class XMLSerializer(object):
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
48 """Produces XML text from an event stream.
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
49
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
50 >>> from markup.builder import tag
20
cc92d74ce9e5 Fix tests broken in [20].
cmlenz
parents: 19
diff changeset
51 >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
52 >>> print ''.join(XMLSerializer()(elem.generate()))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
53 <div><a href="foo"/><br/><hr noshade="True"/></div>
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
54 """
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
55
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
56 _PRESERVE_SPACE = frozenset()
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
57
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
58 def __init__(self, doctype=None, strip_whitespace=True):
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
59 """Initialize the XML serializer.
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
60
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
61 @param doctype: a `(name, pubid, sysid)` tuple that represents the
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
62 DOCTYPE declaration that should be included at the top of the
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
63 generated output
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
64 @param strip_whitespace: whether extraneous whitespace should be
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
65 stripped from the output
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
66 """
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
67 self.preamble = []
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
68 if doctype:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
69 self.preamble.append((DOCTYPE, doctype, (None, -1, -1)))
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
70 self.filters = []
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
71 if strip_whitespace:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
72 self.filters.append(WhitespaceFilter(self._PRESERVE_SPACE))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
73
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
74 def __call__(self, stream):
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
75 have_doctype = False
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
76 ns_attrib = []
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
77 ns_mapping = {XML_NAMESPACE.uri: 'xml'}
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
78
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
79 stream = chain(self.preamble, stream)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
80 for filter_ in self.filters:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
81 stream = filter_(stream)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
82 stream = _PushbackIterator(stream)
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
83 pushback = stream.pushback
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
84 for kind, data, pos in stream:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
85
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
86 if kind is START:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
87 tag, attrib = data
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
88
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
89 tagname = tag.localname
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
90 namespace = tag.namespace
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
91 if namespace:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
92 if namespace in ns_mapping:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
93 prefix = ns_mapping[namespace]
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
94 if prefix:
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
95 tagname = '%s:%s' % (prefix, tagname)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
96 else:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
97 ns_attrib.append((QName('xmlns'), namespace))
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
98 buf = ['<', tagname]
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
99
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
100 for attr, value in attrib + ns_attrib:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
101 attrname = attr.localname
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
102 if attr.namespace:
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
103 prefix = ns_mapping.get(attr.namespace)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
104 if prefix:
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
105 attrname = '%s:%s' % (prefix, attrname)
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
106 buf += [' ', attrname, '="', escape(value), '"']
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
107 ns_attrib = []
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
108
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
109 kind, data, pos = stream.next()
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
110 if kind is END:
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
111 buf += ['/>']
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
112 else:
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
113 buf += ['>']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
114 pushback((kind, data, pos))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
115
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
116 yield Markup(''.join(buf))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
117
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
118 elif kind is END:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
119 tag = data
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
120 tagname = tag.localname
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
121 if tag.namespace:
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
122 prefix = ns_mapping.get(tag.namespace)
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
123 if prefix:
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
124 tagname = '%s:%s' % (prefix, tag.localname)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
125 yield Markup('</%s>' % tagname)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
126
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
127 elif kind is TEXT:
73
1da51d718391 Some more performance tweaks.
cmlenz
parents: 69
diff changeset
128 yield escape(data, quotes=False)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
129
89
80386d62814f Support comments in templates that are not included in the output, in the same way Kid does: if the comment text starts with a `!` character, it is stripped from the output.
cmlenz
parents: 85
diff changeset
130 elif kind is COMMENT:
80386d62814f Support comments in templates that are not included in the output, in the same way Kid does: if the comment text starts with a `!` character, it is stripped from the output.
cmlenz
parents: 85
diff changeset
131 yield Markup('<!--%s-->' % data)
80386d62814f Support comments in templates that are not included in the output, in the same way Kid does: if the comment text starts with a `!` character, it is stripped from the output.
cmlenz
parents: 85
diff changeset
132
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
133 elif kind is DOCTYPE and not have_doctype:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
134 name, pubid, sysid = data
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
135 buf = ['<!DOCTYPE %s']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
136 if pubid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
137 buf += [' PUBLIC "%s"']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
138 elif sysid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
139 buf += [' SYSTEM']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
140 if sysid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
141 buf += [' "%s"']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
142 buf += ['>\n']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
143 yield Markup(''.join(buf), *filter(None, data))
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
144 have_doctype = True
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
145
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
146 elif kind is START_NS:
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
147 prefix, uri = data
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
148 if uri not in ns_mapping:
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
149 ns_mapping[uri] = prefix
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
150 if not prefix:
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
151 ns_attrib.append((QName('xmlns'), uri))
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
152 else:
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
153 ns_attrib.append((QName('xmlns:%s' % prefix), uri))
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
154
105
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
155 elif kind is PI:
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
156 yield Markup('<?%s %s?>' % data)
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
157
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
158
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
159 class XHTMLSerializer(XMLSerializer):
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
160 """Produces XHTML text from an event stream.
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
161
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
162 >>> from markup.builder import tag
20
cc92d74ce9e5 Fix tests broken in [20].
cmlenz
parents: 19
diff changeset
163 >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
164 >>> print ''.join(XHTMLSerializer()(elem.generate()))
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
165 <div><a href="foo"></a><br /><hr noshade="noshade" /></div>
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
166 """
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
167
18
5420cfe42d36 Actually make use of the `markup.core.Namespace` class, and add a couple of doctests.
cmlenz
parents: 1
diff changeset
168 NAMESPACE = Namespace('http://www.w3.org/1999/xhtml')
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
169
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
170 _EMPTY_ELEMS = frozenset(['area', 'base', 'basefont', 'br', 'col', 'frame',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
171 'hr', 'img', 'input', 'isindex', 'link', 'meta',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
172 'param'])
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
173 _BOOLEAN_ATTRS = frozenset(['selected', 'checked', 'compact', 'declare',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
174 'defer', 'disabled', 'ismap', 'multiple',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
175 'nohref', 'noresize', 'noshade', 'nowrap'])
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
176 _PRESERVE_SPACE = frozenset([QName('pre'), QName('textarea')])
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
177
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
178 def __call__(self, stream):
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
179 namespace = self.NAMESPACE
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
180 ns_attrib = []
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
181 ns_mapping = {XML_NAMESPACE.uri: 'xml'}
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
182 boolean_attrs = self._BOOLEAN_ATTRS
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
183 empty_elems = self._EMPTY_ELEMS
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
184 have_doctype = False
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
185
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
186 stream = chain(self.preamble, stream)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
187 for filter_ in self.filters:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
188 stream = filter_(stream)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
189 stream = _PushbackIterator(stream)
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
190 pushback = stream.pushback
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
191 for kind, data, pos in stream:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
192
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
193 if kind is START:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
194 tag, attrib = data
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
195
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
196 tagname = tag.localname
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
197 namespace = tag.namespace
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
198 if namespace:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
199 if namespace in ns_mapping:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
200 prefix = ns_mapping[namespace]
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
201 if prefix:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
202 tagname = '%s:%s' % (prefix, tagname)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
203 else:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
204 ns_attrib.append((QName('xmlns'), namespace))
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
205 buf = ['<', tagname]
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
206
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
207 for attr, value in attrib + ns_attrib:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
208 attrname = attr.localname
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
209 if attr.namespace:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
210 prefix = ns_mapping.get(attr.namespace)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
211 if prefix:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
212 attrname = '%s:%s' % (prefix, attrname)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
213 if attrname in boolean_attrs:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
214 if value:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
215 buf += [' ', attrname, '="', attrname, '"']
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
216 else:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
217 buf += [' ', attrname, '="', escape(value), '"']
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
218 ns_attrib = []
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
219
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
220 if (not tag.namespace or tag in namespace) and \
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
221 tagname in empty_elems:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
222 kind, data, pos = stream.next()
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
223 if kind is END:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
224 buf += [' />']
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
225 else:
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
226 buf += ['>']
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
227 pushback((kind, data, pos))
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
228 else:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
229 buf += ['>']
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
230
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
231 yield Markup(''.join(buf))
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
232
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
233 elif kind is END:
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
234 tag = data
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
235 tagname = tag.localname
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
236 if tag.namespace:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
237 prefix = ns_mapping.get(tag.namespace)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
238 if prefix:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
239 tagname = '%s:%s' % (prefix, tag.localname)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
240 yield Markup('</%s>' % tagname)
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
241
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
242 elif kind is TEXT:
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
243 yield escape(data, quotes=False)
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
244
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
245 elif kind is COMMENT:
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
246 yield Markup('<!--%s-->' % data)
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
247
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
248 elif kind is DOCTYPE and not have_doctype:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
249 name, pubid, sysid = data
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
250 buf = ['<!DOCTYPE %s']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
251 if pubid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
252 buf += [' PUBLIC "%s"']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
253 elif sysid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
254 buf += [' SYSTEM']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
255 if sysid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
256 buf += [' "%s"']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
257 buf += ['>\n']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
258 yield Markup(''.join(buf), *filter(None, data))
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
259 have_doctype = True
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
260
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
261 elif kind is START_NS:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
262 prefix, uri = data
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
263 if uri not in ns_mapping:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
264 ns_mapping[uri] = prefix
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
265 if not prefix:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
266 ns_attrib.append((QName('xmlns'), uri))
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
267 else:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
268 ns_attrib.append((QName('xmlns:%s' % prefix), uri))
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
269
105
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
270 elif kind is PI:
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
271 yield Markup('<?%s %s?>' % data)
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
272
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
273
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
274 class HTMLSerializer(XHTMLSerializer):
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
275 """Produces HTML text from an event stream.
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
276
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
277 >>> from markup.builder import tag
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
278 >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
279 >>> print ''.join(HTMLSerializer()(elem.generate()))
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
280 <div><a href="foo"></a><br><hr noshade></div>
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
281 """
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
282
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
283 _NOESCAPE_ELEMS = frozenset([QName('script'), QName('style')])
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
284
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
285 def __init__(self, doctype=None, strip_whitespace=True):
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
286 """Initialize the HTML serializer.
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
287
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
288 @param doctype: a `(name, pubid, sysid)` tuple that represents the
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
289 DOCTYPE declaration that should be included at the top of the
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
290 generated output
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
291 @param strip_whitespace: whether extraneous whitespace should be
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
292 stripped from the output
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
293 """
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
294 super(HTMLSerializer, self).__init__(doctype, False)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
295 if strip_whitespace:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
296 self.filters.append(WhitespaceFilter(self._PRESERVE_SPACE,
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
297 self._NOESCAPE_ELEMS))
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
298
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
299 def __call__(self, stream):
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
300 namespace = self.NAMESPACE
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
301 ns_mapping = {}
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
302 boolean_attrs = self._BOOLEAN_ATTRS
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
303 empty_elems = self._EMPTY_ELEMS
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
304 noescape_elems = self._NOESCAPE_ELEMS
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
305 have_doctype = False
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
306 noescape = False
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
307
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
308 stream = chain(self.preamble, stream)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
309 for filter_ in self.filters:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
310 stream = filter_(stream)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
311 stream = _PushbackIterator(stream)
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
312 pushback = stream.pushback
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
313 for kind, data, pos in stream:
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
314
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
315 if kind is START:
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
316 tag, attrib = data
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
317 if not tag.namespace or tag in namespace:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
318 tagname = tag.localname
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
319 buf = ['<', tagname]
96
fa08aef181a2 Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
cmlenz
parents: 89
diff changeset
320
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
321 for attr, value in attrib:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
322 attrname = attr.localname
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
323 if not attr.namespace or attr in namespace:
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
324 if attrname in boolean_attrs:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
325 if value:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
326 buf += [' ', attrname]
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
327 else:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
328 buf += [' ', attrname, '="', escape(value), '"']
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
329
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
330 if tagname in empty_elems:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
331 kind, data, pos = stream.next()
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
332 if kind is not END:
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
333 pushback((kind, data, pos))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
334
140
c1f4390d50f8 Fix bug in HTML serializer, plus some other minor tweaks.
cmlenz
parents: 136
diff changeset
335 buf += ['>']
c1f4390d50f8 Fix bug in HTML serializer, plus some other minor tweaks.
cmlenz
parents: 136
diff changeset
336 yield Markup(''.join(buf))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
337
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
338 if tagname in noescape_elems:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
339 noescape = True
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
340
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
341 elif kind is END:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
342 tag = data
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
343 if not tag.namespace or tag in namespace:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
344 yield Markup('</%s>' % tag.localname)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
345
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
346 noescape = False
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
347
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
348 elif kind is TEXT:
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
349 if noescape:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
350 yield data
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
351 else:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
352 yield escape(data, quotes=False)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
353
89
80386d62814f Support comments in templates that are not included in the output, in the same way Kid does: if the comment text starts with a `!` character, it is stripped from the output.
cmlenz
parents: 85
diff changeset
354 elif kind is COMMENT:
80386d62814f Support comments in templates that are not included in the output, in the same way Kid does: if the comment text starts with a `!` character, it is stripped from the output.
cmlenz
parents: 85
diff changeset
355 yield Markup('<!--%s-->' % data)
80386d62814f Support comments in templates that are not included in the output, in the same way Kid does: if the comment text starts with a `!` character, it is stripped from the output.
cmlenz
parents: 85
diff changeset
356
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
357 elif kind is DOCTYPE and not have_doctype:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
358 name, pubid, sysid = data
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
359 buf = ['<!DOCTYPE %s']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
360 if pubid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
361 buf += [' PUBLIC "%s"']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
362 elif sysid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
363 buf += [' SYSTEM']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
364 if sysid:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
365 buf += [' "%s"']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
366 buf += ['>\n']
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
367 yield Markup(''.join(buf), *filter(None, data))
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
368 have_doctype = True
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
369
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
370 elif kind is START_NS and data[1] not in ns_mapping:
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
371 ns_mapping[data[1]] = data[0]
109
230ee6a2c6b2 Reorder the conditional branches in the serializers so that the more common event kinds are on top.
cmlenz
parents: 105
diff changeset
372
105
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
373 elif kind is PI:
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
374 yield Markup('<?%s %s?>' % data)
71f3db26eecb Include processing instructions in serialized streams.
cmlenz
parents: 96
diff changeset
375
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
376
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
377 class WhitespaceFilter(object):
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
378 """A filter that removes extraneous ignorable white space from the
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
379 stream."""
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
380
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
381 _TRAILING_SPACE = re.compile('[ \t]+(?=\n)')
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
382 _LINE_COLLAPSE = re.compile('\n{2,}')
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
383 _XML_SPACE = XML_NAMESPACE['space']
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
384
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
385 def __init__(self, preserve=None, noescape=None):
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
386 """Initialize the filter.
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
387
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
388 @param preserve: a set or sequence of tag names for which white-space
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
389 should be ignored.
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
390 @param noescape: a set or sequence of tag names for which text content
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
391 should not be escaped
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
392
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
393 Both the `preserve` and `noescape` sets are expected to refer to
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
394 elements that cannot contain further child elements.
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
395 """
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
396 if preserve is None:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
397 preserve = []
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
398 self.preserve = frozenset(preserve)
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
399 if noescape is None:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
400 noescape = []
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
401 self.noescape = frozenset(noescape)
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
402
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
403 def __call__(self, stream, ctxt=None):
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
404 trim_trailing_space = self._TRAILING_SPACE.sub
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
405 collapse_lines = self._LINE_COLLAPSE.sub
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
406 xml_space = self._XML_SPACE
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
407 mjoin = Markup('').join
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
408 preserve_elems = self.preserve
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
409 preserve = False
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
410 noescape_elems = self.noescape
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
411 noescape = False
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
412
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
413 textbuf = []
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
414 push_text = textbuf.append
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
415 pop_text = textbuf.pop
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
416 for kind, data, pos in chain(stream, [(None, None, None)]):
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
417 if kind is TEXT:
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
418 if noescape:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
419 data = Markup(data)
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
420 push_text(data)
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
421 else:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
422 if textbuf:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
423 if len(textbuf) > 1:
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
424 text = mjoin(textbuf, escape_quotes=False)
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
425 del textbuf[:]
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
426 else:
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
427 text = escape(pop_text(), quotes=False)
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
428 if not preserve:
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
429 text = collapse_lines('\n', trim_trailing_space('', text))
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
430 yield TEXT, Markup(text), pos
141
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
431
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
432 if kind is START:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
433 tag, attrib = data
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
434 if tag.localname in preserve_elems or \
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
435 data[1].get(xml_space) == 'preserve':
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
436 preserve = True
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
437
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
438 if tag.localname in noescape_elems:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
439 noescape = True
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
440
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
441 elif kind is END:
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
442 preserve = noescape = False
520a5b7dd6d2 * No escaping of `<script>` or `<style>` tags in HTML output (see #24)
cmlenz
parents: 140
diff changeset
443
136
b86f496f6035 Minor performance improvements in serialization.
cmlenz
parents: 123
diff changeset
444 if kind:
123
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
445 yield kind, data, pos
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
446
10279d2eeec9 Fix for #18: whitespace in space-sensitive elements such as `<pre>` and `<textarea>` is now preserved.
cmlenz
parents: 109
diff changeset
447
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
448 class _PushbackIterator(object):
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
449 """A simple wrapper for iterators that allows pushing items back on the
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
450 queue via the `pushback()` method.
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
451
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
452 That can effectively be used to peek at the next item."""
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
453 __slots__ = ['iterable', 'buf']
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
454
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
455 def __init__(self, iterable):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
456 self.iterable = iter(iterable)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
457 self.buf = []
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
458
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
459 def __iter__(self):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
460 return self
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
461
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
462 def next(self):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
463 if self.buf:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
464 return self.buf.pop(0)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
465 return self.iterable.next()
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
466
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
467 def pushback(self, item):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
468 self.buf.append(item)
Copyright (C) 2012-2017 Edgewall Software