pythonpoint working again

git-svn-id: http://svn.code.sf.net/p/docutils/code/trunk@1380 929543f6-e4f2-0310-98a6-ba3bd3dd1d04

1 files changed, 177 insertions, 219 deletions

diff --git a/sandbox/richard/pythonpoint/pythonpoint.py b/sandbox/richard/pythonpoint/pythonpoint.py
index 3313f0513..bcea529d6 100644
--- a/sandbox/richard/pythonpoint/pythonpoint.py
+++ b/sandbox/richard/pythonpoint/pythonpoint.py
@@ -1,153 +1,133 @@
 import cStringIO, cgi, sys, urllib
-import docutils.utils
-from docutils.parsers import get_parser_class
+import docutils.core, docutils.io
+from docutils import writers, nodes, languages
 
-class DumbPythonPointFormatter:
-    def __init__(self):
-        self.out = cStringIO.StringIO()
-        self.w = self.out.write
+class Writer(writers.Writer):
+
+    settings_spec = ()
+    settings_default_overrides = {}
+    output = None
+
+    def translate(self):
+        visitor = DumbPythonPointFormatter(self.document)
+        self.document.walkabout(visitor)
+        self.output = visitor.astext()
+        #self.head_prefix = visitor.head_prefix
+        #self.head = visitor.head
+        #self.body_prefix = visitor.body_prefix
+        #self.body = visitor.body
+        #self.body_suffix = visitor.body_suffix
+
+class DumbPythonPointFormatter(nodes.NodeVisitor):
+
+    def __init__(self, document):
+        nodes.NodeVisitor.__init__(self, document)
         self.section = 0
         self.closers = []
-        self.slidenum = 1
+        self.slidenum = 0
+        self.body = []
+        self.w = self.body.append
+        self.suppress_para = 0
 
-    def format(self, node):
-        '''Format a node
-        '''
-        for entry in node:
-            self.formatOneTag(entry)
+    def astext(self):
+        return ''.join(self.body)
 
-    def formatOneTag(self, tag):
-        if tag.tagname == '#text':
-            meth = self.format__text
-        else:
-            if not hasattr(self, 'format_'+tag.tagname):
-                print >>sys.stderr, '**skipping %s'%tag.tagname
-                return
-            else:
-                meth = getattr(self, 'format_'+tag.tagname)
-        meth(tag)
+    def visit_reference(self, node):
+        pass
+    def depart_reference(self, node):
+        pass
+
+    def visit_document(self, node):
+        self.w('<presentation>\n'
+        '<stylesheet module="modern" function="getParagraphStyles"/>\n'
+        '<section name="Main">\n')
+    def depart_document(self, node):
+        self.w('</section>\n'
+        '</presentation>\n')
 
-    def open_slide(self, node):
+    def visit_section(self, node):
         if node.attributes.has_key('dupname'):
             name = node.attributes['dupname']
         else:
             name = node.attributes['name']
-        self.w('<slide id="Slide%03d" title="%s">\n'%(self.slidenum, name))
         self.slidenum += 1
+        self.w('<slide id="Slide%03d" title="%s">\n'
+            '<frame x="160" y="72" width="600" height="432" leftmargin="36"'
+            'rightmargin="0">\n'%(self.slidenum, name))
+    def depart_section(self, node):
+        self.w('</frame>\n</slide>\n')
 
-    def open_frame(self):
-        self.w('<frame x="160" y="72" width="600" height="432" leftmargin="36" rightmargin="0">\n')
-
-    #
-    # Root Element
-    #
-    # ((title, subtitle?)?, docinfo?, %structure.model;)
-    #
-    def format_document(self, document):
-        ''' ((title, subtitle?)?, docinfo?, %structure.model;)
-
-            
-        '''
-        self.document = document
-
-        # make sure the structure is what we're expecting
-        self.w('<presentation>\n')
-
-        # TODO: get this into the stx
-        self.w('<stylesheet module="modern" function="getParagraphStyles"/>\n')
-
-        self.w('<section name="Main">\n')
-
-        # TODO: get this into the stx
-        self.w('<fixedimage filename="logo.gif" x="0" y="0" width="134" height="70"/>\n')
-
-        # now for the body
-        for entry in document:
-            assert entry.tagname == 'section'
-            self.formatOneTag(entry)
-
-        self.w('</section>\n')
-        self.w('</presentation>\n')
-
-        return self.out.getvalue()
-
-    def format_title(self, node):
+    def visit_title(self, node):
         self.w('<para style="Heading2">')
-        if node.children: self.format(node)
+        self.suppress_para = 1
+    def depart_title(self, node):
+        self.suppress_para = 0
         self.w('</para>\n')
 
-    def format_section(self, node):
-        self.open_slide(node)
-        self.open_frame()
-        if node.children: self.format(node)
-        self.w('</frame>\n')
-        self.w('</slide>\n')
-
-    def format_paragraph(self, node):
-        ''' %text.model;
-        '''
-        # TODO: there are situations where the <p> </p> are unnecessary
-        self.w('<para>')
-        if node.children: self.format(node)
-        self.w('</para>\n')
+    def visit_paragraph(self, node):
+        if not self.suppress_para: self.w('<para>')
+    def depart_paragraph(self, node):
+        if not self.suppress_para: self.w('</para>\n')
 
     # Simple lists
-    def format_bullet_list(self, node):
-        if node.children: self.format(node)
-
-    def format_enumerated_list(self, node):
-        if node.children: self.format(node)
+    def visit_bullet_list(self, node):
+        pass
+    def depart_bullet_list(self, node):
+        pass
+    def visit_enumerated_list(self, node):
+        pass
+    def depart_enumerated_list(self, node):
+        pass
 
-    def format_list_item(self, node):
+    def visit_list_item(self, node):
         self.w('<para style="Bullet">')
-        if node.children: self.format(node[0])
+        self.suppress_para = 1
+    def depart_list_item(self, node):
+        self.suppress_para = 0
         self.w('</para>\n')
 
     # Definition List
-    def format_definition_list(self, node):
-        if node.children: self.format(node)
-
-    def format_definition_list_item(self, node):
-        '''  (term, classifier?, definition)
-        '''
-        if node.children: self.format(node)
+    def visit_definition_list(self, node):
+        pass
+    def depart_definition_list(self, node):
+        pass
+    def visit_definition_list_item(self, node):
+        pass
+    def depart_definition_list_item(self, node):
+        pass
 
-    def format_term(self, node):
-        ''' %text.model;
-        '''
+    def visit_term(self, node):
         self.w('<para><b>')
-        if node.children:self.format(node[0])
+    def depart_term(self, node):
         self.w('</b>')
 
-    def format_classifier(self, node):
-        ''' %text.model;
-        '''
-        # TODO: handle the classifier better
+    def visit_classifier(self, node):
         self.w('<i>')
-        if node.children: self.format(node[0])
+    def depart_classifier(self, node):
         self.w('</i>')
 
-    def format_definition(self, node):
-        ''' (%body.elements;)+
-        '''
+    def visit_definition(self, node):
         self.w('</para>\n<para style="Definition">')
-        for child in node.children:
-            self.w(child[0].data)
+        self.suppress_para = 1
+    def depart_definition(self, node):
+        self.suppress_para = 0
         self.w('</para>\n')
 
     # Literal Block
-    def format_literal_block(self, node):
+    def visit_literal_block(self, node):
         self.w('<pre>')
-        if node.children: self.format(node)
+        self.suppress_para = 1
+    def depart_literal_block(self, node):
+        self.suppress_para = 0
         self.w('</pre>\n')
 
     # Block Quote
-    def format_block_quote(self, node):
+    def visit_block_quote(self, node):
         self.w('<para style="Indent">')
-        if node.children: self.format(node)
+    def depart_block_quote(self, node):
         self.w('</para>\n')
 
-    def format_image(self, node):
+    def visit_image(self, node):
         ''' EMPTY
             uri       CDATA     #REQUIRED
             alt       CDATA     #IMPLIED
@@ -158,13 +138,15 @@ class DumbPythonPointFormatter:
         attrs = node.attributes
         l = ['src="%(uri)s"'%attrs]
         # TODO: scale
-        self.w('<image filename="%s">'%node.attributes['uri'])
+        self.w('<image filename="%s"/>'%node.attributes['uri'])
+    def depart_image(self, node):
+        pass
 
     #
     # Tables:
     #  NOT IN DOM YET
     #
-    def format_table(self, node):
+    def visit_table(self, node):
         '''
             +------------------------+------------+----------+----------+
             | Header row, column 1   | Header 2   | Header 3 | Header 4 |
@@ -180,100 +162,73 @@ class DumbPythonPointFormatter:
             +------------------------+------------+---------------------+
         '''
         self.w('<table border=1>\n')
-        if node.children: self.format(node)
+    def depart_table(self, node):
         self.w('</table>\n')
 
-    def format_tgroup(self, node):
-        # we get the number of columns, if that's important
-        if node.children: self.format(node)
+    def visit_tgroup(self, node):
+        pass
+    def depart_tgroup(self, node):
+        pass
 
-    def format_colspec(self, node):
-        # we get colwidth, but don't need it
+    def visit_colspec(self, node):
+        pass
+    def depart_colspec(self, node):
         pass
 
-    def format_thead(self, node):
-        for row in node.children:
-            self.w('<tr>')
-            for cell in row.children:
-                s = ''
-                attrs = cell.attributes
-                if attrs.has_key('morecols'):
-                    s = s + ' colspan=%d'%(attrs['morecols']+1)
-                if attrs.has_key('morerows'):
-                    s = s + ' rowspan=%d'%(attrs['morerows']+1)
-                self.w('<th valign="top" align="left"%s>'%s)
-                if cell.children: self.format(cell)
-                self.w('</th>\n')
-            self.w('</tr>\n')
-
-    def format_tbody(self, node):
-        for row in node.children:
-            self.w('<tr>')
-            for cell in row.children:
-                s = ''
-                attrs = cell.attributes
-                if attrs.has_key('morecols'):
-                    s = s + ' colspan=%d'%(attrs['morecols']+1)
-                if attrs.has_key('morerows'):
-                    s = s + ' rowspan=%d'%(attrs['morerows']+1)
-                self.w('<td valign="top" align="left"%s>'%s)
-                if cell.children: self.format(cell)
-                self.w('</td>\n')
-            self.w('</tr>\n')
+    def visit_row(self, node):
+        self.body.append(self.starttag(node, 'tr', ''))
 
-    #
-    # Inline Elements
-    #
-    # Inline elements occur within the text contents of body elements. Some
-    # nesting of inline elements is allowed by these definitions, with the
-    # following caveats:
-    # - An inline element may not contain a nested element of the same type
-    #   (e.g. <strong> may not contain another <strong>).
-    # - Nested inline elements may or may not be supported by individual
-    #   applications using this DTD.
-    # - The inline elements <footnote_reference>, <literal>, and <image> do
-    #   not support nesting.
-    #
-    #  What that means is that all of these take (%text.model;) except:
-    #   literal (#PCDATA)
-    #   footnote_reference (#PCDATA)
-    #
-    # text.model:
-    # (#PCDATA | %inline.elements;)*
-    #
-    def format_emphasis(self, node):
-        ''' (%text.model;)
-        '''
+    def depart_row(self, node):
+        self.body.append('</tr>\n')
+
+    def visit_thead(self, node):
+        self.thead = 1
+    def depart_thead(self, node):
+        self.thead = 0
+    def visit_tbody(self, node):
+        self.thead = 1
+    def depart_tbody(self, node):
+        self.thead = 0
+
+    def visit_entry(self, node):
+        if self.thead:
+            s = 'th '
+        else:
+            s = 'td '
+        attrs = node.attributes
+        if attrs.has_key('morecols'):
+            s = s + 'colspan=%d '%(attrs['morecols']+1)
+        if attrs.has_key('morerows'):
+            s = s + 'rowspan=%d '%(attrs['morerows']+1)
+        self.w('<%svalign="top" align="left">'%s)
+
+    def depart_entry(self, node):
+        if self.thead:
+            self.w('</th>\n')
+        else:
+            self.w('</td>\n')
+
+    def visit_emphasis(self, node):
         self.w('<i>')
-        if node.children: self.format(node)
+    def depart_emphasis(self, node):
         self.w('</i>')
 
-    def format_strong(self, node):
-        ''' (%text.model;)
-        '''
+    def visit_strong(self, node):
         self.w('<b>')
-        if node.children: self.format(node)
+    def depart_strong(self, node):
         self.w('</b>')
 
-    def format_interpreted(self, node):
-        ''' (%text.model;)
-            type      CDATA     #IMPLIED
-        '''
+    def visit_interpreted(self, node):
+        pass #raise NotImplementedError, node
+    def depart_interpreted(self, node):
         pass #raise NotImplementedError, node
 
-    def format_literal(self, node):
-        ''' (#PCDATA)
-        '''
+    def visit_literal(self, node):
         self.w('<tt>')
-        for literal in node.children:
-            self.w(cgi.escape(literal.data))
+    def depart_literal(self, node):
         self.w('</tt>')
 
-    def format_reference(self, node):
-        ''' (%text.model;)
-            %reference.atts;
-            %anonymous.att;
-        '''
+    def visit_reference(self, node):
         attrs = node.attributes
         doc = self.document
         ok = 1
@@ -292,53 +247,56 @@ class DumbPythonPointFormatter:
             name = attrs['refname']
             self.w('<a href="#%s">'%urllib.quote(name))
         else:
-            ok = 0
             self.w('<span class="formatter_error">target "%s" '
                 'undefined</span>'%attrs['refname'])
-        if node.children: self.format(node)
-        if ok:
-            self.w('</a>')
-
-    def format_footnote_reference(self, node):
-        ''' (#PCDATA)
-            %reference.atts;
-            %auto.att;
-        '''
+
+    def depart_reference(self, node):
+        self.w('</a>')
+
+    def visit_footnote_reference(self, node):
         raise NotImplementedError, node
 
-    def format_substitution_reference(self, node):
-        ''' (%text.model;)
-            %refname.att;
-        '''
-        #raise NotImplementedError, node
+    def visit_substitution_reference(self, node):
+        pass
+    def depart_substitution_reference(self, node):
         pass
 
-    def format_problematic(self, node):
-        ''' (%text.model;)
-        '''
+    def visit_problematic(self, node):
+        raise NotImplementedError, node
+    def depart_problematic(self, node):
         raise NotImplementedError, node
 
-    def format_system_message(self, node):
-        ''' just print it to stderr
-        '''
+    def visit_system_message(self, node):
         print >>sys.stderr, '%s: %s'%(node.attributes['type'], node[0][0].data)
+    def depart_system_message(self, node):
+        pass
 
-    #
-    # Finally, #text
-    #
-    def format__text(self, node):
+    def visit_comment(self, node):
+        pass
+    def depart_comment(self, node):
+        pass
+
+    def visit_Text(self, node):
         self.w(cgi.escape(node.data))
+    def depart_Text(self, node):
+        pass
 
 def main(filename, debug=0):
-    parser = get_parser_class('rest')()
-    input = open(filename).read()
-    document = docutils.utils.new_document(parser)
-    parser.parse(input, document)
+    pub = docutils.core.Publisher()
+    pub.set_reader('standalone', None, 'restructuredtext')
+    pub.writer = Writer()
+    pub.get_settings()
+    pub.settings._destination = ''
+    pub.source = docutils.io.StringInput(source=open(filename).read(),
+        encoding='latin-1')
+    pub.destination = docutils.io.StringOutput(encoding='latin-1')
+    document = pub.reader.read(pub.source, pub.parser, pub.settings)
+    pub.apply_transforms(document)
+
     if debug == 1:
         print document.pformat()
     else:
-        formatter = DumbPythonPointFormatter()
-        print formatter.format_document(document)
+        print pub.writer.write(document, pub.destination)
 
 if __name__ == '__main__':
     if len(sys.argv) > 2: