Mercurial > genshi > mirror

annotate markup/output.py @ 96:fa08aef181a2 trunk

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Add an XHTML serialization method. Now really need to get rid of some code duplication in the `markup.output` module.
author cmlenz
date Fri, 21 Jul 2006 11:39:32 +0000
parents 80386d62814f
children 71f3db26eecb
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
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
18 try:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
19 frozenset
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
20 except NameError:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
21 from sets import ImmutableSet as frozenset
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
22 from itertools import chain
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
23
73
1da51d718391 Some more performance tweaks.
cmlenz
parents: 69
diff changeset
24 from markup.core import escape, Markup, Namespace, QName
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
25 from markup.core import DOCTYPE, START, END, START_NS, END_NS, TEXT, COMMENT
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
26
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
27 __all__ = ['Serializer', 'XMLSerializer', 'HTMLSerializer']
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
28
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
29
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
30 class Serializer(object):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
31 """Base class for serializers."""
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
32
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
33 def serialize(self, stream):
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
34 """Must be implemented by concrete subclasses to serialize the given
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
35 stream.
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
36
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
37 This method must be implemented as a generator, producing the
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
38 serialized output incrementally as unicode strings.
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
39 """
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
40 raise NotImplementedError
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
41
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
42
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
43 class DocType(object):
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
44 """Defines a number of commonly used DOCTYPE declarations as constants."""
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
45
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
46 HTML_STRICT = ('html', '-//W3C//DTD HTML 4.01//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
47 'http://www.w3.org/TR/html4/strict.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
48 HTML_TRANSITIONAL = ('html', '-//W3C//DTD HTML 4.01 Transitional//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
49 'http://www.w3.org/TR/html4/loose.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
50 HTML = HTML_STRICT
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
51
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
52 XHTML_STRICT = ('html', '-//W3C//DTD XHTML 1.0 Strict//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
53 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
54 XHTML_TRANSITIONAL = ('html', '-//W3C//DTD XHTML 1.0 Transitional//EN',
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
55 'http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
56 XHTML = XHTML_STRICT
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
57
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
58
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
59 class XMLSerializer(Serializer):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
60 """Produces XML text from an event stream.
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
61
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
62 >>> from markup.builder import tag
20
cc92d74ce9e5 Fix tests broken in [20].
cmlenz
parents: 19
diff changeset
63 >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
64 >>> print ''.join(XMLSerializer().serialize(elem.generate()))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
65 <div><a href="foo"/><br/><hr noshade="True"/></div>
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
66 """
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
67 def __init__(self, doctype=None):
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
68 """Initialize the XML serializer.
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
69
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
70 @param doctype: a `(name, pubid, sysid)` tuple that represents the
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
71 DOCTYPE declaration that should be included at the top of the
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
72 generated output
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
73 """
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
74 self.preamble = []
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
75 if doctype:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
76 self.preamble.append((DOCTYPE, doctype, (None, -1, -1)))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
77
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
78 def serialize(self, stream):
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
79 have_doctype = False
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
80 ns_attrib = []
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
81 ns_mapping = {}
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
82
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
83 stream = _PushbackIterator(chain(self.preamble, stream))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
84 for kind, data, pos in stream:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
85
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
86 if kind is DOCTYPE:
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
87 if not have_doctype:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
88 name, pubid, sysid = data
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
89 buf = ['<!DOCTYPE %s']
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
90 if pubid:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
91 buf.append(' PUBLIC "%s"')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
92 elif sysid:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
93 buf.append(' SYSTEM')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
94 if sysid:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
95 buf.append(' "%s"')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
96 buf.append('>\n')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
97 yield Markup(''.join(buf), *filter(None, data))
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
98 have_doctype = True
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
99
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
100 elif kind is START_NS:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
101 prefix, uri = data
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
102 if uri not in ns_mapping:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
103 ns_mapping[uri] = prefix
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
104 if not prefix:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
105 ns_attrib.append((QName('xmlns'), uri))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
106 else:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
107 ns_attrib.append((QName('xmlns:%s' % prefix), uri))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
108
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
109 elif kind is START:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
110 tag, attrib = data
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
111
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
112 tagname = tag.localname
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
113 if tag.namespace:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
114 try:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
115 prefix = ns_mapping[tag.namespace]
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
116 if prefix:
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
117 tagname = '%s:%s' % (prefix, tag.localname)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
118 except KeyError:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
119 ns_attrib.append((QName('xmlns'), tag.namespace))
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
120 buf = ['<%s' % tagname]
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
121
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
122 if ns_attrib:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
123 attrib.extend(ns_attrib)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
124 ns_attrib = []
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
125 for attr, value in attrib:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
126 attrname = attr.localname
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
127 if attr.namespace:
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
128 prefix = ns_mapping.get(attr.namespace)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
129 if prefix:
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
130 attrname = '%s:%s' % (prefix, attrname)
73
1da51d718391 Some more performance tweaks.
cmlenz
parents: 69
diff changeset
131 buf.append(' %s="%s"' % (attrname, escape(value)))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
132
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
133 kind, data, pos = stream.next()
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
134 if kind is END:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
135 buf.append('/>')
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
136 else:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
137 buf.append('>')
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
138 stream.pushback((kind, data, pos))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
139
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
140 yield Markup(''.join(buf))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
141
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
142 elif kind is END:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
143 tag = data
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
144 tagname = tag.localname
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
145 if tag.namespace:
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
146 prefix = ns_mapping.get(tag.namespace)
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
147 if prefix:
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
148 tagname = '%s:%s' % (prefix, tag.localname)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
149 yield Markup('</%s>' % tagname)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
150
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
151 elif kind is TEXT:
73
1da51d718391 Some more performance tweaks.
cmlenz
parents: 69
diff changeset
152 yield escape(data, quotes=False)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
153
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
154 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
155 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
156
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
157
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
158 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
159 """Produces XHTML text from an event stream.
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
160
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
161 >>> from markup.builder import tag
20
cc92d74ce9e5 Fix tests broken in [20].
cmlenz
parents: 19
diff changeset
162 >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
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
163 >>> print ''.join(XHTMLSerializer().serialize(elem.generate()))
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
164 <div><a href="foo"></a><br /><hr noshade="noshade" /></div>
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
165 """
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
166
18
5420cfe42d36 Actually make use of the `markup.core.Namespace` class, and add a couple of doctests.
cmlenz
parents: 1
diff changeset
167 NAMESPACE = Namespace('http://www.w3.org/1999/xhtml')
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
168
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
169 _EMPTY_ELEMS = frozenset(['area', 'base', 'basefont', 'br', 'col', 'frame',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
170 'hr', 'img', 'input', 'isindex', 'link', 'meta',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
171 'param'])
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
172 _BOOLEAN_ATTRS = frozenset(['selected', 'checked', 'compact', 'declare',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
173 'defer', 'disabled', 'ismap', 'multiple',
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
174 'nohref', 'noresize', 'noshade', 'nowrap'])
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
175
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
176 def serialize(self, stream):
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
177 have_doctype = False
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
178 ns_mapping = {}
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
179
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
180 stream = _PushbackIterator(chain(self.preamble, stream))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
181 for kind, data, pos in stream:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
182
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
183 if kind is DOCTYPE:
85
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
184 if not have_doctype:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
185 name, pubid, sysid = data
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
186 buf = ['<!DOCTYPE %s']
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
187 if pubid:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
188 buf.append(' PUBLIC "%s"')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
189 elif sysid:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
190 buf.append(' SYSTEM')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
191 if sysid:
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
192 buf.append(' "%s"')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
193 buf.append('>\n')
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
194 yield Markup(''.join(buf), *filter(None, data))
4938c310d904 Improve handling of DOCTYPE declarations.
cmlenz
parents: 73
diff changeset
195 have_doctype = True
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
196
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
197 elif kind is START_NS:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
198 prefix, uri = data
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
199 if uri not in ns_mapping:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
200 ns_mapping[uri] = prefix
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
201
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
202 elif kind is START:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
203 tag, attrib = data
18
5420cfe42d36 Actually make use of the `markup.core.Namespace` class, and add a couple of doctests.
cmlenz
parents: 1
diff changeset
204 if tag.namespace and tag not in self.NAMESPACE:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
205 continue # not in the HTML namespace, so don't emit
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
206 buf = ['<', tag.localname]
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
207
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
208 for attr, value in attrib:
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
209 if attr.namespace and attr not in self.NAMESPACE:
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
210 continue # not in the HTML namespace, so don't emit
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
211 if attr.localname in self._BOOLEAN_ATTRS:
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
212 if value:
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
213 buf.append(' %s="%s"' % (attr.localname, attr.localname))
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
214 else:
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
215 buf.append(' %s="%s"' % (attr.localname, escape(value)))
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
216
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
217 if tag.localname in self._EMPTY_ELEMS:
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
218 kind, data, pos = stream.next()
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
219 if 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
220 buf.append(' />')
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
221 else:
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
222 buf.append('>')
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
223 stream.pushback((kind, data, pos))
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
224 else:
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 buf.append('>')
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
226
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
227 yield Markup(''.join(buf))
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
228
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
229 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
230 tag = 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
231 if tag.namespace and tag not in self.NAMESPACE:
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 continue # not in the HTML namespace, so don't emit
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 yield Markup('</%s>' % tag.localname)
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
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
235 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
236 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
237
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
238 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
239 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
240
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 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
243 """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
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 >>> 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
246 >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
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 >>> print ''.join(HTMLSerializer().serialize(elem.generate()))
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
248 <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
249 """
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
250
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
251 def serialize(self, 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
252 have_doctype = 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
253 ns_mapping = {}
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
254
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
255 stream = _PushbackIterator(chain(self.preamble, 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
256 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
257
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
258 if kind is DOCTYPE:
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
259 if not have_doctype:
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
260 name, pubid, sysid = 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
261 buf = ['<!DOCTYPE %s']
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
262 if pubid:
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
263 buf.append(' PUBLIC "%s"')
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
264 elif sysid:
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
265 buf.append(' SYSTEM')
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
266 if sysid:
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
267 buf.append(' "%s"')
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
268 buf.append('>\n')
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
269 yield Markup(''.join(buf), *filter(None, 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
270 have_doctype = True
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
271
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
272 elif kind is START_NS:
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 prefix, uri = 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
274 if uri not in ns_mapping:
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 ns_mapping[uri] = prefix
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 elif kind is START:
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 tag, attrib = 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
279 if tag.namespace and tag not in self.NAMESPACE:
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 continue # not in the HTML namespace, so don't emit
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 buf = ['<', tag.localname]
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
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
283 for attr, value in attrib:
18
5420cfe42d36 Actually make use of the `markup.core.Namespace` class, and add a couple of doctests.
cmlenz
parents: 1
diff changeset
284 if attr.namespace and attr not in self.NAMESPACE:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
285 continue # not in the HTML namespace, so don't emit
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
286 if attr.localname in self._BOOLEAN_ATTRS:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
287 if value:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
288 buf.append(' %s' % attr.localname)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
289 else:
73
1da51d718391 Some more performance tweaks.
cmlenz
parents: 69
diff changeset
290 buf.append(' %s="%s"' % (attr.localname, escape(value)))
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
291
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
292 if tag.localname in self._EMPTY_ELEMS:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
293 kind, data, pos = stream.next()
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
294 if kind is not END:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
295 stream.pushback((kind, data, pos))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
296
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
297 yield Markup(''.join(buf + ['>']))
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
298
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
299 elif kind is END:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
300 tag = data
18
5420cfe42d36 Actually make use of the `markup.core.Namespace` class, and add a couple of doctests.
cmlenz
parents: 1
diff changeset
301 if tag.namespace and tag not in self.NAMESPACE:
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
302 continue # not in the HTML namespace, so don't emit
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
303 yield Markup('</%s>' % tag.localname)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
304
69
c40a5dcd2b55 A couple of minor performance improvements.
cmlenz
parents: 66
diff changeset
305 elif kind is TEXT:
73
1da51d718391 Some more performance tweaks.
cmlenz
parents: 69
diff changeset
306 yield escape(data, quotes=False)
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
307
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
308 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
309 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
310
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
311
26
3c1a022be04c * Split out the XPath tests into a separate `unittest`-based file.
cmlenz
parents: 20
diff changeset
312 class _PushbackIterator(object):
1
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
313 """A simple wrapper for iterators that allows pushing items back on the
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
314 queue via the `pushback()` method.
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
315
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
316 That can effectively be used to peek at the next item."""
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
317 __slots__ = ['iterable', 'buf']
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
318
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
319 def __init__(self, iterable):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
320 self.iterable = iter(iterable)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
321 self.buf = []
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
322
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
323 def __iter__(self):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
324 return self
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
325
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
326 def next(self):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
327 if self.buf:
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
328 return self.buf.pop(0)
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
329 return self.iterable.next()
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
330
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
331 def pushback(self, item):
5479aae32f5a Initial import.
cmlenz
parents:
diff changeset
332 self.buf.append(item)
Copyright (C) 2012-2017 Edgewall Software