cmlenz@1: # -*- coding: utf-8 -*-
cmlenz@1: #
cmlenz@66: # Copyright (C) 2006 Edgewall Software
cmlenz@1: # All rights reserved.
cmlenz@1: #
cmlenz@1: # This software is licensed as described in the file COPYING, which
cmlenz@1: # you should have received as part of this distribution. The terms
cmlenz@66: # are also available at http://markup.edgewall.org/wiki/License.
cmlenz@1: #
cmlenz@1: # This software consists of voluntary contributions made by many
cmlenz@1: # individuals. For the exact contribution history, see the revision
cmlenz@66: # history and logs, available at http://markup.edgewall.org/log/.
cmlenz@1: 
cmlenz@1: """This module provides different kinds of serialization methods for XML event
cmlenz@1: streams.
cmlenz@1: """
cmlenz@1: 
cmlenz@123: from itertools import chain
cmlenz@1: try:
cmlenz@1:     frozenset
cmlenz@1: except NameError:
cmlenz@1:     from sets import ImmutableSet as frozenset
cmlenz@123: import re
cmlenz@1: 
cmlenz@73: from markup.core import escape, Markup, Namespace, QName
cmlenz@105: from markup.core import DOCTYPE, START, END, START_NS, END_NS, TEXT, COMMENT, PI
cmlenz@1: 
cmlenz@1: __all__ = ['Serializer', 'XMLSerializer', 'HTMLSerializer']
cmlenz@1: 
cmlenz@1: 
cmlenz@85: class DocType(object):
cmlenz@85:     """Defines a number of commonly used DOCTYPE declarations as constants."""
cmlenz@85: 
cmlenz@85:     HTML_STRICT = ('html', '-//W3C//DTD HTML 4.01//EN',
cmlenz@85:                    'http://www.w3.org/TR/html4/strict.dtd')
cmlenz@85:     HTML_TRANSITIONAL = ('html', '-//W3C//DTD HTML 4.01 Transitional//EN',
cmlenz@85:                          'http://www.w3.org/TR/html4/loose.dtd')
cmlenz@85:     HTML = HTML_STRICT
cmlenz@85: 
cmlenz@85:     XHTML_STRICT = ('html', '-//W3C//DTD XHTML 1.0 Strict//EN',
cmlenz@85:                     'http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd')
cmlenz@85:     XHTML_TRANSITIONAL = ('html', '-//W3C//DTD XHTML 1.0 Transitional//EN',
cmlenz@85:                           'http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd')
cmlenz@85:     XHTML = XHTML_STRICT
cmlenz@85: 
cmlenz@85: 
cmlenz@123: class XMLSerializer(object):
cmlenz@1:     """Produces XML text from an event stream.
cmlenz@1:     
cmlenz@1:     >>> from markup.builder import tag
cmlenz@20:     >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
cmlenz@123:     >>> print ''.join(XMLSerializer()(elem.generate()))
cmlenz@1:     <div><a href="foo"/><br/><hr noshade="True"/></div>
cmlenz@1:     """
cmlenz@123: 
cmlenz@123:     _PRESERVE_SPACE = frozenset()
cmlenz@123: 
cmlenz@123:     def __init__(self, doctype=None, strip_whitespace=True):
cmlenz@85:         """Initialize the XML serializer.
cmlenz@85:         
cmlenz@85:         @param doctype: a `(name, pubid, sysid)` tuple that represents the
cmlenz@85:             DOCTYPE declaration that should be included at the top of the
cmlenz@85:             generated output
cmlenz@123:         @param strip_whitespace: whether extraneous whitespace should be
cmlenz@123:             stripped from the output
cmlenz@85:         """
cmlenz@85:         self.preamble = []
cmlenz@85:         if doctype:
cmlenz@85:             self.preamble.append((DOCTYPE, doctype, (None, -1, -1)))
cmlenz@123:         self.filters = []
cmlenz@123:         if strip_whitespace:
cmlenz@123:             self.filters.append(WhitespaceFilter(self._PRESERVE_SPACE))
cmlenz@1: 
cmlenz@123:     def __call__(self, stream):
cmlenz@85:         have_doctype = False
cmlenz@1:         ns_attrib = []
cmlenz@1:         ns_mapping = {}
cmlenz@1: 
cmlenz@123:         stream = chain(self.preamble, stream)
cmlenz@123:         for filter_ in self.filters:
cmlenz@123:             stream = filter_(stream)
cmlenz@123:         stream = _PushbackIterator(stream)
cmlenz@1:         for kind, data, pos in stream:
cmlenz@1: 
cmlenz@109:             if kind is START:
cmlenz@1:                 tag, attrib = data
cmlenz@1: 
cmlenz@1:                 tagname = tag.localname
cmlenz@123:                 namespace = tag.namespace
cmlenz@123:                 if namespace:
cmlenz@123:                     if namespace in ns_mapping:
cmlenz@123:                         prefix = ns_mapping[namespace]
cmlenz@1:                         if prefix:
cmlenz@123:                             tagname = '%s:%s' % (prefix, tagname)
cmlenz@123:                     else:
cmlenz@123:                         ns_attrib.append((QName('xmlns'), namespace))
cmlenz@69:                 buf = ['<%s' % tagname]
cmlenz@1: 
cmlenz@123:                 for attr, value in attrib + ns_attrib:
cmlenz@1:                     attrname = attr.localname
cmlenz@1:                     if attr.namespace:
cmlenz@26:                         prefix = ns_mapping.get(attr.namespace)
cmlenz@1:                         if prefix:
cmlenz@69:                             attrname = '%s:%s' % (prefix, attrname)
cmlenz@73:                     buf.append(' %s="%s"' % (attrname, escape(value)))
cmlenz@123:                 ns_attrib = []
cmlenz@1: 
cmlenz@1:                 kind, data, pos = stream.next()
cmlenz@69:                 if kind is END:
cmlenz@1:                     buf.append('/>')
cmlenz@1:                 else:
cmlenz@1:                     buf.append('>')
cmlenz@1:                     stream.pushback((kind, data, pos))
cmlenz@1: 
cmlenz@1:                 yield Markup(''.join(buf))
cmlenz@1: 
cmlenz@69:             elif kind is END:
cmlenz@1:                 tag = data
cmlenz@1:                 tagname = tag.localname
cmlenz@1:                 if tag.namespace:
cmlenz@26:                     prefix = ns_mapping.get(tag.namespace)
cmlenz@26:                     if prefix:
cmlenz@69:                         tagname = '%s:%s' % (prefix, tag.localname)
cmlenz@1:                 yield Markup('</%s>' % tagname)
cmlenz@1: 
cmlenz@69:             elif kind is TEXT:
cmlenz@73:                 yield escape(data, quotes=False)
cmlenz@1: 
cmlenz@89:             elif kind is COMMENT:
cmlenz@89:                 yield Markup('<!--%s-->' % data)
cmlenz@89: 
cmlenz@109:             elif kind is DOCTYPE:
cmlenz@109:                 if not have_doctype:
cmlenz@109:                     name, pubid, sysid = data
cmlenz@109:                     buf = ['<!DOCTYPE %s']
cmlenz@109:                     if pubid:
cmlenz@109:                         buf.append(' PUBLIC "%s"')
cmlenz@109:                     elif sysid:
cmlenz@109:                         buf.append(' SYSTEM')
cmlenz@109:                     if sysid:
cmlenz@109:                         buf.append(' "%s"')
cmlenz@109:                     buf.append('>\n')
cmlenz@109:                     yield Markup(''.join(buf), *filter(None, data))
cmlenz@109:                     have_doctype = True
cmlenz@109: 
cmlenz@109:             elif kind is START_NS:
cmlenz@109:                 prefix, uri = data
cmlenz@109:                 if uri not in ns_mapping:
cmlenz@109:                     ns_mapping[uri] = prefix
cmlenz@109:                     if not prefix:
cmlenz@109:                         ns_attrib.append((QName('xmlns'), uri))
cmlenz@109:                     else:
cmlenz@109:                         ns_attrib.append((QName('xmlns:%s' % prefix), uri))
cmlenz@109: 
cmlenz@105:             elif kind is PI:
cmlenz@105:                 yield Markup('<?%s %s?>' % data)
cmlenz@105: 
cmlenz@1: 
cmlenz@96: class XHTMLSerializer(XMLSerializer):
cmlenz@96:     """Produces XHTML text from an event stream.
cmlenz@1:     
cmlenz@1:     >>> from markup.builder import tag
cmlenz@20:     >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
cmlenz@123:     >>> print ''.join(XHTMLSerializer()(elem.generate()))
cmlenz@96:     <div><a href="foo"></a><br /><hr noshade="noshade" /></div>
cmlenz@1:     """
cmlenz@1: 
cmlenz@18:     NAMESPACE = Namespace('http://www.w3.org/1999/xhtml')
cmlenz@1: 
cmlenz@1:     _EMPTY_ELEMS = frozenset(['area', 'base', 'basefont', 'br', 'col', 'frame',
cmlenz@1:                               'hr', 'img', 'input', 'isindex', 'link', 'meta',
cmlenz@1:                               'param'])
cmlenz@1:     _BOOLEAN_ATTRS = frozenset(['selected', 'checked', 'compact', 'declare',
cmlenz@1:                                 'defer', 'disabled', 'ismap', 'multiple',
cmlenz@1:                                 'nohref', 'noresize', 'noshade', 'nowrap'])
cmlenz@123:     _PRESERVE_SPACE = frozenset([QName('pre'), QName('textarea')])
cmlenz@1: 
cmlenz@123:     def __call__(self, stream):
cmlenz@85:         have_doctype = False
cmlenz@1:         ns_mapping = {}
cmlenz@1: 
cmlenz@123:         stream = chain(self.preamble, stream)
cmlenz@123:         for filter_ in self.filters:
cmlenz@123:             stream = filter_(stream)
cmlenz@123:         stream = _PushbackIterator(stream)
cmlenz@1:         for kind, data, pos in stream:
cmlenz@1: 
cmlenz@109:             if kind is START:
cmlenz@1:                 tag, attrib = data
cmlenz@18:                 if tag.namespace and tag not in self.NAMESPACE:
cmlenz@1:                     continue # not in the HTML namespace, so don't emit
cmlenz@1:                 buf = ['<', tag.localname]
cmlenz@96: 
cmlenz@96:                 for attr, value in attrib:
cmlenz@96:                     if attr.namespace and attr not in self.NAMESPACE:
cmlenz@96:                         continue # not in the HTML namespace, so don't emit
cmlenz@96:                     if attr.localname in self._BOOLEAN_ATTRS:
cmlenz@96:                         if value:
cmlenz@96:                             buf.append(' %s="%s"' % (attr.localname, attr.localname))
cmlenz@96:                     else:
cmlenz@96:                         buf.append(' %s="%s"' % (attr.localname, escape(value)))
cmlenz@96: 
cmlenz@96:                 if tag.localname in self._EMPTY_ELEMS:
cmlenz@96:                     kind, data, pos = stream.next()
cmlenz@96:                     if kind is END:
cmlenz@96:                         buf.append(' />')
cmlenz@96:                     else:
cmlenz@96:                         buf.append('>')
cmlenz@96:                         stream.pushback((kind, data, pos))
cmlenz@96:                 else:
cmlenz@96:                     buf.append('>')
cmlenz@96: 
cmlenz@96:                 yield Markup(''.join(buf))
cmlenz@96: 
cmlenz@96:             elif kind is END:
cmlenz@96:                 tag = data
cmlenz@96:                 if tag.namespace and tag not in self.NAMESPACE:
cmlenz@96:                     continue # not in the HTML namespace, so don't emit
cmlenz@96:                 yield Markup('</%s>' % tag.localname)
cmlenz@96: 
cmlenz@96:             elif kind is TEXT:
cmlenz@96:                 yield escape(data, quotes=False)
cmlenz@96: 
cmlenz@96:             elif kind is COMMENT:
cmlenz@96:                 yield Markup('<!--%s-->' % data)
cmlenz@96: 
cmlenz@109:             elif kind is DOCTYPE:
cmlenz@109:                 if not have_doctype:
cmlenz@109:                     name, pubid, sysid = data
cmlenz@109:                     buf = ['<!DOCTYPE %s']
cmlenz@109:                     if pubid:
cmlenz@109:                         buf.append(' PUBLIC "%s"')
cmlenz@109:                     elif sysid:
cmlenz@109:                         buf.append(' SYSTEM')
cmlenz@109:                     if sysid:
cmlenz@109:                         buf.append(' "%s"')
cmlenz@109:                     buf.append('>\n')
cmlenz@109:                     yield Markup(''.join(buf), *filter(None, data))
cmlenz@109:                     have_doctype = True
cmlenz@109: 
cmlenz@109:             elif kind is START_NS:
cmlenz@109:                 prefix, uri = data
cmlenz@109:                 if uri not in ns_mapping:
cmlenz@109:                     ns_mapping[uri] = prefix
cmlenz@109: 
cmlenz@105:             elif kind is PI:
cmlenz@105:                 yield Markup('<?%s %s?>' % data)
cmlenz@105: 
cmlenz@96: 
cmlenz@96: class HTMLSerializer(XHTMLSerializer):
cmlenz@96:     """Produces HTML text from an event stream.
cmlenz@96:     
cmlenz@96:     >>> from markup.builder import tag
cmlenz@96:     >>> elem = tag.div(tag.a(href='foo'), tag.br, tag.hr(noshade=True))
cmlenz@123:     >>> print ''.join(HTMLSerializer()(elem.generate()))
cmlenz@96:     <div><a href="foo"></a><br><hr noshade></div>
cmlenz@96:     """
cmlenz@96: 
cmlenz@123:     def __call__(self, stream):
cmlenz@96:         have_doctype = False
cmlenz@96:         ns_mapping = {}
cmlenz@96: 
cmlenz@123:         stream = chain(self.preamble, stream)
cmlenz@123:         for filter_ in self.filters:
cmlenz@123:             stream = filter_(stream)
cmlenz@123:         stream = _PushbackIterator(stream)
cmlenz@96:         for kind, data, pos in stream:
cmlenz@96: 
cmlenz@109:             if kind is START:
cmlenz@96:                 tag, attrib = data
cmlenz@96:                 if tag.namespace and tag not in self.NAMESPACE:
cmlenz@96:                     continue # not in the HTML namespace, so don't emit
cmlenz@96:                 buf = ['<', tag.localname]
cmlenz@96: 
cmlenz@1:                 for attr, value in attrib:
cmlenz@123:                     if attr.namespace and attr not in self.NAMESPACE \
cmlenz@123:                             or attr.localname.startswith('xml:'):
cmlenz@1:                         continue # not in the HTML namespace, so don't emit
cmlenz@1:                     if attr.localname in self._BOOLEAN_ATTRS:
cmlenz@1:                         if value:
cmlenz@1:                             buf.append(' %s' % attr.localname)
cmlenz@1:                     else:
cmlenz@73:                         buf.append(' %s="%s"' % (attr.localname, escape(value)))
cmlenz@1: 
cmlenz@1:                 if tag.localname in self._EMPTY_ELEMS:
cmlenz@1:                     kind, data, pos = stream.next()
cmlenz@69:                     if kind is not END:
cmlenz@1:                         stream.pushback((kind, data, pos))
cmlenz@1: 
cmlenz@1:                 yield Markup(''.join(buf + ['>']))
cmlenz@1: 
cmlenz@69:             elif kind is END:
cmlenz@1:                 tag = data
cmlenz@18:                 if tag.namespace and tag not in self.NAMESPACE:
cmlenz@1:                     continue # not in the HTML namespace, so don't emit
cmlenz@1:                 yield Markup('</%s>' % tag.localname)
cmlenz@1: 
cmlenz@69:             elif kind is TEXT:
cmlenz@73:                 yield escape(data, quotes=False)
cmlenz@1: 
cmlenz@89:             elif kind is COMMENT:
cmlenz@89:                 yield Markup('<!--%s-->' % data)
cmlenz@89: 
cmlenz@109:             elif kind is DOCTYPE:
cmlenz@109:                 if not have_doctype:
cmlenz@109:                     name, pubid, sysid = data
cmlenz@109:                     buf = ['<!DOCTYPE %s']
cmlenz@109:                     if pubid:
cmlenz@109:                         buf.append(' PUBLIC "%s"')
cmlenz@109:                     elif sysid:
cmlenz@109:                         buf.append(' SYSTEM')
cmlenz@109:                     if sysid:
cmlenz@109:                         buf.append(' "%s"')
cmlenz@109:                     buf.append('>\n')
cmlenz@109:                     yield Markup(''.join(buf), *filter(None, data))
cmlenz@109:                     have_doctype = True
cmlenz@109: 
cmlenz@109:             elif kind is START_NS:
cmlenz@109:                 prefix, uri = data
cmlenz@109:                 if uri not in ns_mapping:
cmlenz@109:                     ns_mapping[uri] = prefix
cmlenz@109: 
cmlenz@105:             elif kind is PI:
cmlenz@105:                 yield Markup('<?%s %s?>' % data)
cmlenz@105: 
cmlenz@1: 
cmlenz@123: class WhitespaceFilter(object):
cmlenz@123:     """A filter that removes extraneous ignorable white space from the
cmlenz@123:     stream."""
cmlenz@123: 
cmlenz@123:     _TRAILING_SPACE = re.compile('[ \t]+(?=\n)')
cmlenz@123:     _LINE_COLLAPSE = re.compile('\n{2,}')
cmlenz@123: 
cmlenz@123:     def __init__(self, preserve=None):
cmlenz@123:         """Initialize the filter.
cmlenz@123:         
cmlenz@123:         @param preserve: a sequence of tag names for which white-space should
cmlenz@123:             be ignored.
cmlenz@123:         """
cmlenz@123:         if preserve is None:
cmlenz@123:             preserve = []
cmlenz@123:         self.preserve = frozenset(preserve)
cmlenz@123: 
cmlenz@123:     def __call__(self, stream, ctxt=None):
cmlenz@123:         trim_trailing_space = self._TRAILING_SPACE.sub
cmlenz@123:         collapse_lines = self._LINE_COLLAPSE.sub
cmlenz@123:         mjoin = Markup('').join
cmlenz@123:         preserve = [False]
cmlenz@123: 
cmlenz@123:         textbuf = []
cmlenz@123:         for kind, data, pos in chain(stream, [(None, None, None)]):
cmlenz@123:             if kind is TEXT:
cmlenz@123:                 textbuf.append(data)
cmlenz@123:             else:
cmlenz@123:                 if kind is START:
cmlenz@123:                     preserve.append(data[0] in self.preserve or 
cmlenz@123:                                     data[1].get('xml:space') == 'preserve')
cmlenz@123:                 if textbuf:
cmlenz@123:                     if len(textbuf) > 1:
cmlenz@123:                         text = mjoin(textbuf, escape_quotes=False)
cmlenz@123:                         del textbuf[:]
cmlenz@123:                     else:
cmlenz@123:                         text = escape(textbuf.pop(), quotes=False)
cmlenz@123:                     if not preserve[-1]:
cmlenz@123:                         text = collapse_lines('\n', trim_trailing_space('', text))
cmlenz@123:                     yield TEXT, Markup(text), pos
cmlenz@123:                 if kind is END:
cmlenz@123:                     preserve.pop()
cmlenz@123:                 if kind is not None:
cmlenz@123:                     yield kind, data, pos
cmlenz@123: 
cmlenz@123: 
cmlenz@26: class _PushbackIterator(object):
cmlenz@1:     """A simple wrapper for iterators that allows pushing items back on the
cmlenz@1:     queue via the `pushback()` method.
cmlenz@1:     
cmlenz@1:     That can effectively be used to peek at the next item."""
cmlenz@1:     __slots__ = ['iterable', 'buf']
cmlenz@1: 
cmlenz@1:     def __init__(self, iterable):
cmlenz@1:         self.iterable = iter(iterable)
cmlenz@1:         self.buf = []
cmlenz@1: 
cmlenz@1:     def __iter__(self):
cmlenz@1:         return self
cmlenz@1: 
cmlenz@1:     def next(self):
cmlenz@1:         if self.buf:
cmlenz@1:             return self.buf.pop(0)
cmlenz@1:         return self.iterable.next()
cmlenz@1: 
cmlenz@1:     def pushback(self, item):
cmlenz@1:         self.buf.append(item)