[FIX] doc: avoid conflict on extenstion and odoo folder

1 files changed, 691 insertions, 0 deletions

diff --git a/doc/_extensions/odoo_ext/translator.py b/doc/_extensions/odoo_ext/translator.py
new file mode 100644
index 00000000000..48c5e901f28
--- /dev/null
+++ b/doc/_extensions/odoo_ext/translator.py
@@ -0,0 +1,691 @@
+# -*- coding: utf-8 -*-
+import os.path
+import posixpath
+import re
+import urllib
+
+from docutils import nodes
+from sphinx import addnodes, util
+from sphinx.locale import admonitionlabels
+
+
+def _parents(node):
+    while node.parent:
+        node = node.parent
+        yield node
+
+class BootstrapTranslator(nodes.NodeVisitor, object):
+    head_prefix = 'head_prefix'
+    head = 'head'
+    stylesheet = 'stylesheet'
+    body_prefix = 'body_prefix'
+    body_pre_docinfo = 'body_pre_docinfo'
+    docinfo = 'docinfo'
+    body_suffix = 'body_suffix'
+    subtitle = 'subtitle'
+    header = 'header'
+    footer = 'footer'
+    html_prolog = 'html_prolog'
+    html_head = 'html_head'
+    html_title = 'html_title'
+    html_subtitle = 'html_subtitle'
+
+    # <meta> tags
+    meta = [
+        '<meta http-equiv="X-UA-Compatible" content="IE=edge">',
+        '<meta name="viewport" content="width=device-width, initial-scale=1">'
+    ]
+
+    def __init__(self, builder, document):
+        super(BootstrapTranslator, self).__init__(document)
+        self.builder = builder
+        self.body = []
+        self.fragment = self.body
+        self.html_body = self.body
+        # document title
+        self.title = []
+        self.start_document_title = 0
+        self.first_title = False
+
+        self.context = []
+        self.section_level = 0
+
+        self.highlightlang = self.highlightlang_base = self.builder.config.highlight_language
+        self.highlightopts = getattr(builder.config, 'highlight_options', {})
+
+        self.first_param = 1
+        self.optional_param_level = 0
+        self.required_params_left = 0
+        self.param_separator = ','
+
+    def encode(self, text):
+        return unicode(text).translate({
+            ord('&'): u'&amp;',
+            ord('<'): u'&lt;',
+            ord('"'): u'&quot;',
+            ord('>'): u'&gt;',
+            0xa0: u'&nbsp;'
+        })
+
+    def starttag(self, node, tagname, **attributes):
+        tagname = unicode(tagname).lower()
+
+        # extract generic attributes
+        attrs = {name.lower(): value for name, value in attributes.iteritems()}
+        attrs.update(
+            (name, value) for name, value in node.attributes.iteritems()
+            if name.startswith('data-')
+        )
+
+        prefix = []
+        postfix = []
+
+        # handle possibly multiple ids
+        assert 'id' not in attrs, "starttag can't be passed a single id attribute, use a list of ids"
+        ids = node.get('ids', []) + attrs.pop('ids', [])
+        if ids:
+            _ids = iter(ids)
+            attrs['id'] = next(_ids)
+            postfix.extend(u'<i id="{}"></i>'.format(_id) for _id in _ids)
+
+        # set CSS class
+        classes = set(node.get('classes', []) + attrs.pop('class', '').split())
+        if classes:
+            attrs['class'] = u' '.join(classes)
+
+        return u'{prefix}<{tag} {attrs}>{postfix}'.format(
+            prefix=u''.join(prefix),
+            tag=tagname,
+            attrs=u' '.join(u'{}="{}"'.format(name, self.attval(value))
+                            for name,  value in attrs.iteritems()),
+            postfix=u''.join(postfix),
+        )
+    # only "space characters" SPACE, CHARACTER TABULATION, LINE FEED,
+    # FORM FEED and CARRIAGE RETURN should be collapsed, not al White_Space
+    def attval(self, value, whitespace=re.compile(u'[ \t\n\f\r]')):
+        return self.encode(whitespace.sub(u' ', unicode(value)))
+
+    def astext(self):
+        return u''.join(self.body)
+
+    def unknown_visit(self, node):
+        print "unknown node", node.__class__.__name__
+        self.body.append(u'[UNKNOWN NODE {}]'.format(node.__class__.__name__))
+        raise nodes.SkipNode
+
+    def visit_highlightlang(self, node):
+        self.highlightlang = node['lang']
+    def depart_highlightlang(self, node):
+        pass
+
+    def visit_document(self, node):
+        self.first_title = True
+    def depart_document(self, node):
+        pass
+
+    def visit_section(self, node):
+        # close "parent" or preceding section, unless this is the opening of
+        # the first section
+        if self.section_level:
+            self.body.append(u'</section>')
+        self.section_level += 1
+
+        self.body.append(self.starttag(node, 'section'))
+    def depart_section(self, node):
+        self.section_level -= 1
+        # close last section of document
+        if not self.section_level:
+            self.body.append(u'</section>')
+
+    def is_compact_paragraph(self, node):
+        parent = node.parent
+        if isinstance(parent, (nodes.document, nodes.compound,
+                               addnodes.desc_content,
+                               addnodes.versionmodified)):
+            # Never compact paragraphs in document or compound.
+            return False
+
+        for key, value in node.attlist():
+            # we can ignore a few specific classes, all other non-default
+            # attributes require that a <p> node remains
+            if key != 'classes' or value not in ([], ['first'], ['last'], ['first', 'last']):
+                return False
+
+        first = isinstance(node.parent[0], nodes.label)
+        for child in parent.children[first:]:
+            # only first paragraph can be compact
+            if isinstance(child, nodes.Invisible):
+                continue
+            if child is node:
+                break
+            return False
+        parent_length = len([
+            1 for n in parent
+            if not isinstance(n, (nodes.Invisible, nodes.label))
+        ])
+        return parent_length == 1
+
+    def visit_paragraph(self, node):
+        if self.is_compact_paragraph(node):
+            self.context.append(u'')
+            return
+        self.body.append(self.starttag(node, 'p'))
+        self.context.append(u'</p>')
+    def depart_paragraph(self, node):
+        self.body.append(self.context.pop())
+    def visit_compact_paragraph(self, node):
+        pass
+    def depart_compact_paragraph(self, node):
+        pass
+
+    def visit_literal_block(self, node):
+        if node.rawsource != node.astext():
+            # most probably a parsed-literal block -- don't highlight
+            self.body.append(self.starttag(node, 'pre'))
+            return
+        lang = self.highlightlang
+        highlight_args = node.get('highlight_args', {})
+        if 'language' in node:
+            # code-block directives
+            lang = node['language']
+            highlight_args['force'] = True
+        linenos = node.get('linenos', False)
+        if lang is self.highlightlang_base:
+            # only pass highlighter options for original language
+            opts = self.highlightopts
+        else:
+            opts = {}
+
+        def warner(msg):
+            self.builder.warn(msg, (self.builder.current_docname, node.line))
+        highlighted = self.builder.highlighter.highlight_block(
+            node.rawsource, lang, opts=opts, warn=warner, linenos=linenos,
+            **highlight_args)
+        self.body.append(self.starttag(node, 'div', CLASS='highlight-%s' % lang))
+        self.body.append(highlighted)
+        self.body.append(u'</div>\n')
+        raise nodes.SkipNode
+    def depart_literal_block(self, node):
+        self.body.append(u'</pre>')
+
+    def visit_bullet_list(self, node):
+        self.body.append(self.starttag(node, 'ul'))
+    def depart_bullet_list(self, node):
+        self.body.append(u'</ul>')
+    def visit_enumerated_list(self, node):
+        self.body.append(self.starttag(node, 'ol'))
+    def depart_enumerated_list(self, node):
+        self.body.append(u'</ol>')
+    def visit_list_item(self, node):
+        self.body.append(self.starttag(node, 'li'))
+    def depart_list_item(self, node):
+        self.body.append(u'</li>')
+    def visit_definition_list(self, node):
+        self.body.append(self.starttag(node, 'dl'))
+    def depart_definition_list(self, node):
+        self.body.append(u'</dl>')
+    def visit_definition_list_item(self, node):
+        pass
+    def depart_definition_list_item(self, node):
+        pass
+    def visit_term(self, node):
+        self.body.append(self.starttag(node, 'dt'))
+    def depart_term(self, node):
+        self.body.append(u'</dt>')
+    def visit_termsep(self, node):
+        self.body.append(self.starttag(node, 'br'))
+        raise nodes.SkipNode
+    def visit_definition(self, node):
+        self.body.append(self.starttag(node, 'dd'))
+    def depart_definition(self, node):
+        self.body.append(u'</dd>')
+
+    def visit_admonition(self, node, type=None):
+        clss = {
+            # ???: 'alert-success',
+
+            'note': 'alert-info',
+            'hint': 'alert-info',
+            'tip': 'alert-info',
+            'seealso': 'alert-go_to',
+
+            'warning': 'alert-warning',
+            'attention': 'alert-warning',
+            'caution': 'alert-warning',
+            'important': 'alert-warning',
+
+            'danger': 'alert-danger',
+            'error': 'alert-danger',
+
+            'exercise': 'alert-exercise',
+        }
+        self.body.append(self.starttag(node, 'div', role='alert', CLASS='alert {}'.format(
+            clss.get(type, '')
+        )))
+        if 'alert-dismissible' in node.get('classes', []):
+            self.body.append(
+                u'<button type="button" class="close" data-dismiss="alert" aria-label="Close">'
+                u'<span aria-hidden="true">&times;</span>'
+                u'</button>')
+        if type:
+            node.insert(0, nodes.title(type, admonitionlabels[type]))
+    def depart_admonition(self, node):
+        self.body.append(u'</div>')
+    visit_note = lambda self, node: self.visit_admonition(node, 'note')
+    visit_warning = lambda self, node: self.visit_admonition(node, 'warning')
+    visit_attention = lambda self, node: self.visit_admonition(node, 'attention')
+    visit_caution = lambda self, node: self.visit_admonition(node, 'caution')
+    visit_danger = lambda self, node: self.visit_admonition(node, 'danger')
+    visit_error = lambda self, node: self.visit_admonition(node, 'error')
+    visit_hint = lambda self, node: self.visit_admonition(node, 'hint')
+    visit_important = lambda self, node: self.visit_admonition(node, 'important')
+    visit_tip = lambda self, node: self.visit_admonition(node, 'tip')
+    visit_exercise = lambda self, node: self.visit_admonition(node, 'exercise')
+    visit_seealso = lambda self, node: self.visit_admonition(node, 'seealso')
+    depart_note = depart_admonition
+    depart_warning = depart_admonition
+    depart_attention = depart_admonition
+    depart_caution = depart_admonition
+    depart_danger = depart_admonition
+    depart_error = depart_admonition
+    depart_hint = depart_admonition
+    depart_important = depart_admonition
+    depart_tip = depart_admonition
+    depart_exercise = depart_admonition
+    depart_seealso = depart_admonition
+    def visit_versionmodified(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS=node['type']))
+    def depart_versionmodified(self, node):
+        self.body.append(u'</div>')
+
+    def visit_title(self, node):
+        parent = node.parent
+        closing = u'</p>'
+        if isinstance(parent, nodes.Admonition):
+            self.body.append(self.starttag(node, 'p', CLASS='alert-title'))
+        elif isinstance(node.parent, nodes.document):
+            self.body.append(self.starttag(node, 'h1'))
+            closing = u'</h1>'
+            self.start_document_title = len(self.body)
+        else:
+            assert isinstance(parent, nodes.section), "expected a section node as parent to the title, found {}".format(parent)
+            if self.first_title:
+                self.first_title = False
+                raise nodes.SkipNode()
+            nodename = 'h{}'.format(self.section_level)
+            self.body.append(self.starttag(node, nodename))
+            closing = u'</{}>'.format(nodename)
+        self.context.append(closing)
+    def depart_title(self, node):
+        self.body.append(self.context.pop())
+        if self.start_document_title:
+            self.title = self.body[self.start_document_title:-1]
+            self.start_document_title = 0
+            del self.body[:]
+
+    # the rubric should be a smaller heading than the current section, up to
+    # h6... maybe "h7" should be a ``p`` instead?
+    def visit_rubric(self, node):
+        self.body.append(self.starttag(node, 'h{}'.format(min(self.section_level + 1, 6))))
+    def depart_rubric(self, node):
+        self.body.append(u'</h{}>'.format(min(self.section_level + 1, 6)))
+
+    def visit_block_quote(self, node):
+        self.body.append(self.starttag(node, 'blockquote'))
+    def depart_block_quote(self, node):
+        self.body.append(u'</blockquote>')
+    def visit_attribution(self, node):
+        self.body.append(self.starttag(node, 'footer'))
+    def depart_attribution(self, node):
+        self.body.append(u'</footer>')
+
+    def visit_container(self, node):
+        self.body.append(self.starttag(node, 'div'))
+    def depart_container(self, node):
+        self.body.append(u'</div>')
+    def visit_compound(self, node):
+        self.body.append(self.starttag(node, 'div'))
+    def depart_compound(self, node):
+        self.body.append(u'</div>')
+
+    def visit_image(self, node):
+        uri = node['uri']
+        if uri in self.builder.images:
+            uri = posixpath.join(self.builder.imgpath,
+                                 self.builder.images[uri])
+        attrs = {'src': uri, 'class': 'img-responsive'}
+        if 'alt' in node:
+            attrs['alt'] = node['alt']
+        if 'align' in node:
+            if node['align'] == 'center':
+                attrs['class'] += ' center-block'
+            else:
+                doc = None
+                if node.source:
+                    doc = node.source
+                    if node.line:
+                        doc += ':%d' % node.line
+                self.builder.app.warn(
+                    "Unsupported alignment value \"%s\"" % node['align'],
+                    location=doc
+                )
+        # todo: explicit width/height/scale?
+        self.body.append(self.starttag(node, 'img', **attrs))
+    def depart_image(self, node): pass
+    def visit_figure(self, node):
+        self.body.append(self.starttag(node, 'div'))
+    def depart_figure(self, node):
+        self.body.append(u'</div>')
+    def visit_caption(self, node):
+        # first paragraph of figure content
+        self.body.append(self.starttag(node, 'h4'))
+    def depart_caption(self, node):
+        self.body.append(u'</h4>')
+    def visit_legend(self, node): pass
+    def depart_legend(self, node): pass
+
+    def visit_line(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS='line'))
+        # ensure the line still takes the room it needs
+        if not len(node): self.body.append(u'<br />')
+    def depart_line(self, node):
+        self.body.append(u'</div>')
+
+    def visit_line_block(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS='line-block'))
+    def depart_line_block(self, node):
+        self.body.append(u'</div>')
+
+    def visit_table(self, node):
+        self.body.append(self.starttag(node, 'table', CLASS='table'))
+    def depart_table(self, node):
+        self.body.append(u'</table>')
+    def visit_tgroup(self, node): pass
+    def depart_tgroup(self, node): pass
+    def visit_colspec(self, node): raise nodes.SkipNode
+    def visit_thead(self, node):
+        self.body.append(self.starttag(node, 'thead'))
+    def depart_thead(self, node):
+        self.body.append(u'</thead>')
+    def visit_tbody(self, node):
+        self.body.append(self.starttag(node, 'tbody'))
+    def depart_tbody(self, node):
+        self.body.append(u'</tbody>')
+    def visit_row(self, node):
+        self.body.append(self.starttag(node, 'tr'))
+    def depart_row(self, node):
+        self.body.append(u'</tr>')
+    def visit_entry(self, node):
+        if isinstance(node.parent.parent, nodes.thead):
+            tagname = 'th'
+        else:
+            tagname = 'td'
+        self.body.append(self.starttag(node, tagname))
+        self.context.append(tagname)
+    def depart_entry(self, node):
+        self.body.append(u'</{}>'.format(self.context.pop()))
+
+    def visit_Text(self, node):
+        self.body.append(self.encode(node.astext()))
+    def depart_Text(self, node):
+        pass
+    def visit_literal(self, node):
+        self.body.append(self.starttag(node, 'code'))
+    def depart_literal(self, node):
+        self.body.append(u'</code>')
+    visit_literal_emphasis = visit_literal
+    depart_literal_emphasis = depart_literal
+    def visit_emphasis(self, node):
+        self.body.append(self.starttag(node, 'em'))
+    def depart_emphasis(self, node):
+        self.body.append(u'</em>')
+    def visit_strong(self, node):
+        self.body.append(self.starttag(node, 'strong'))
+    def depart_strong(self, node):
+        self.body.append(u'</strong>')
+    visit_literal_strong = visit_strong
+    depart_literal_strong = depart_strong
+    def visit_inline(self, node):
+        self.body.append(self.starttag(node, 'span'))
+    def depart_inline(self, node):
+        self.body.append(u'</span>')
+    def visit_abbreviation(self, node):
+        attrs = {}
+        if 'explanation' in node:
+            attrs['title'] = node['explanation']
+        self.body.append(self.starttag(node, 'abbr', **attrs))
+    def depart_abbreviation(self, node):
+        self.body.append(u'</abbr>')
+
+    def visit_reference(self, node):
+        attrs = {
+            'class': 'reference',
+            'href': node['refuri'] if 'refuri' in node else '#' + node['refid']
+        }
+        attrs['class'] += ' internal' if (node.get('internal') or 'refuri' not in node) else ' external'
+        if any(isinstance(ancestor, nodes.Admonition) for ancestor in _parents(node)):
+            attrs['class'] += ' alert-link'
+
+        if 'reftitle' in node:
+            attrs['title'] = node['reftitle']
+
+        self.body.append(self.starttag(node, 'a', **attrs))
+    def depart_reference(self, node):
+        self.body.append(u'</a>')
+    def visit_target(self, node): pass
+    def depart_target(self, node): pass
+    def visit_footnote(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS='footnote'))
+        self.footnote_backrefs(node)
+    def depart_footnote(self, node):
+        self.body.append(u'</div>')
+    def visit_footnote_reference(self, node):
+        self.body.append(self.starttag(
+            node, 'a', href='#' + node['refid'], CLASS="footnote-ref"))
+    def depart_footnote_reference(self, node):
+        self.body.append(u'</a>')
+    def visit_label(self, node):
+        self.body.append(self.starttag(node, 'span', CLASS='footnote-label'))
+        self.body.append(u'%s[' % self.context.pop())
+    def depart_label(self, node):
+        # Context added in footnote_backrefs.
+        self.body.append(u']%s</span> %s' % (self.context.pop(), self.context.pop()))
+    def footnote_backrefs(self, node):
+        # should store following data on context stack (in that order since
+        # they'll be popped so LIFO)
+        #
+        # * outside (after) label
+        # * after label text
+        # * before label text
+        backrefs = node['backrefs']
+        if not backrefs:
+            self.context.extend(['', '', ''])
+        elif len(backrefs) == 1:
+            self.context.extend([
+                '',
+                '</a>',
+                '<a class="footnote-backref" href="#%s">' % backrefs[0]
+            ])
+        else:
+            backlinks = (
+                '<a class="footnote-backref" href="#%s">%s</a>' % (backref, i)
+                for i, backref in enumerate(backrefs, start=1)
+            )
+            self.context.extend([
+                '<em class="footnote-backrefs">(%s)</em> ' % ', '.join(backlinks),
+                '',
+                ''
+            ])
+
+    def visit_desc(self, node):
+        self.body.append(self.starttag(node, 'section', CLASS='code-' + node['objtype']))
+    def depart_desc(self, node):
+        self.body.append(u'</section>')
+    def visit_desc_signature(self, node):
+        self.body.append(self.starttag(node, 'h6'))
+        self.body.append(u'<code>')
+    def depart_desc_signature(self, node):
+        self.body.append(u'</code>')
+        self.body.append(u'</h6>')
+    def visit_desc_addname(self, node): pass
+    def depart_desc_addname(self, node): pass
+    def visit_desc_type(self, node): pass
+    def depart_desc_type(self, node): pass
+    def visit_desc_returns(self, node):
+        self.body.append(u' → ')
+    def depart_desc_returns(self, node):
+        pass
+    def visit_desc_name(self, node): pass
+    def depart_desc_name(self, node): pass
+    def visit_desc_parameterlist(self, node):
+        self.body.append(u'(')
+        self.first_param = True
+        self.optional_param_level = 0
+        # How many required parameters are left.
+        self.required_params_left = sum(isinstance(c, addnodes.desc_parameter) for c in node.children)
+        self.param_separator = node.child_text_separator
+    def depart_desc_parameterlist(self, node):
+        self.body.append(u')')
+    # If required parameters are still to come, then put the comma after
+    # the parameter.  Otherwise, put the comma before.  This ensures that
+    # signatures like the following render correctly (see issue #1001):
+    #
+    #     foo([a, ]b, c[, d])
+    #
+    def visit_desc_parameter(self, node):
+        if self.first_param:
+            self.first_param = 0
+        elif not self.required_params_left:
+            self.body.append(self.param_separator)
+        if self.optional_param_level == 0:
+            self.required_params_left -= 1
+        if 'noemph' not in node: self.body.append(u'<em>')
+    def depart_desc_parameter(self, node):
+        if 'noemph' not in node: self.body.append(u'</em>')
+        if self.required_params_left:
+            self.body.append(self.param_separator)
+    def visit_desc_optional(self, node):
+        self.optional_param_level += 1
+        self.body.append(u'[')
+    def depart_desc_optional(self, node):
+        self.optional_param_level -= 1
+        self.body.append(u']')
+    def visit_desc_annotation(self, node):
+        self.body.append(self.starttag(node, 'em'))
+    def depart_desc_annotation(self, node):
+        self.body.append(u'</em>')
+    def visit_desc_content(self, node): pass
+    def depart_desc_content(self, node): pass
+    def visit_field_list(self, node):
+         self.body.append(self.starttag(node, 'div', CLASS='code-fields'))
+    def depart_field_list(self, node):
+        self.body.append(u'</div>')
+    def visit_field(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS='code-field'))
+    def depart_field(self, node):
+        self.body.append(u'</div>')
+    def visit_field_name(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS='code-field-name'))
+    def depart_field_name(self, node):
+        self.body.append(u'</div>')
+    def visit_field_body(self, node):
+        self.body.append(self.starttag(node, 'div', CLASS='code-field-body'))
+    def depart_field_body(self, node):
+        self.body.append(u'</div>')
+
+    def visit_glossary(self, node): pass
+    def depart_glossary(self, node): pass
+
+    def visit_comment(self, node): raise nodes.SkipNode
+    def visit_toctree(self, node):
+        # div class=row {{ section_type }}
+        #   h2 class=col-sm-12
+        #     {{ section title }}
+        #   div class=col-sm-6 col-md-3
+        #     figure class=card
+        #       a href=current_link style=background-image: document-image-attribute class=card-img
+        #         figcaption
+        #           {{ card title }}
+        env = self.builder.env
+        conf = self.builder.app.config
+        for title, ref in ((e[0], e[1]) for e in node['entries']):
+            # external URL, no toc, can't recurse into
+            if ref not in env.tocs:
+                continue
+            toc = env.tocs[ref].traverse(addnodes.toctree)
+
+            classes = env.metadata[ref].get('types', 'tutorials')
+            classes += ' toc-single-entry' if not toc else ' toc-section'
+            self.body.append(self.starttag(node, 'div', CLASS="row " + classes))
+            self.body.append(u'<h2 class="col-sm-12">')
+            self.body.append(title if title else util.nodes.clean_astext(env.titles[ref]))
+            self.body.append(u'</h2>')
+
+            entries = [(title, ref)] if not toc else ((e[0], e[1]) for e in toc[0]['entries'])
+            for subtitle, subref in entries:
+                baseuri = self.builder.get_target_uri(node['parent'])
+
+                if subref in env.metadata:
+                    cover = env.metadata[subref].get('banner', conf.odoo_cover_default)
+                elif subref in conf.odoo_cover_external:
+                    cover = conf.odoo_cover_external[subref]
+                else:
+                    cover = conf.odoo_cover_default_external
+
+                if cover:
+                    banner = '_static/' + cover
+                    base, ext = os.path.splitext(banner)
+                    small = "{}.small{}".format(base, ext)
+                    if os.path.isfile(urllib.url2pathname(small)):
+                        banner = small
+                    style = u"background-image: url('{}')".format(
+                        util.relative_uri(baseuri, banner) or '#')
+                else:
+                    style = u''
+
+                self.body.append(u"""
+                <div class="col-sm-6 col-md-3">
+                <figure class="card">
+                    <a href="{link}" class="card-img">
+                        <span style="{style}"></span>
+                        <figcaption>{title}</figcaption>
+                    </a>
+                </figure>
+                </div>
+                """.format(
+                    link=subref if util.url_re.match(subref) else util.relative_uri(
+                        baseuri, self.builder.get_target_uri(subref)),
+                    style=style,
+                    title=subtitle if subtitle else util.nodes.clean_astext(env.titles[subref]),
+                ))
+
+            self.body.append(u'</div>')
+        raise nodes.SkipNode
+
+    def visit_index(self, node): raise nodes.SkipNode
+
+    def visit_raw(self, node):
+        if 'html' in node.get('format', '').split():
+            t = 'span' if isinstance(node.parent, nodes.TextElement) else 'div'
+            if node['classes']:
+                self.body.append(self.starttag(node, t))
+            self.body.append(node.astext())
+            if node['classes']:
+                self.body.append('</%s>' % t)
+        # Keep non-HTML raw text out of output:
+        raise nodes.SkipNode
+
+    # internal node
+    def visit_substitution_definition(self, node): raise nodes.SkipNode
+
+    # without set_translator, add_node doesn't work correctly, so the
+    # serialization of html_domain nodes needs to be embedded here
+    def visit_div(self, node):
+        self.body.append(self.starttag(node, 'div'))
+    def depart_div(self, node):
+        self.body.append(u'</div>\n')
+    def visit_address(self, node):
+        self.body.append(self.starttag(node, 'address'))
+    def depart_address(self, node):
+        self.body.append(u'</address>')
+    # TODO: inline elements