For some reason, this was done in trunk but never got to
branch.
xs << html::StartTag("div", "class='bibtexentry'");
// FIXME XHTML
// The same name/id problem we have elsewhere.
- string const attr = "id='LyXCite-" + to_utf8(entry.key()) + "'";
+ string const attr = "id='LyXCite-" +
+ to_utf8(html::cleanAttr(entry.key())) + "'";
xs << html::CompTag("a", attr);
docstring citekey;
if (numbers)
return content;
// we have to do the escaping here, because we will ultimately
// write this as a raw string, so as not to escape the tags.
- return "<a href='#LyXCite-" + key + "'>" +
+ return "<a href='#LyXCite-" + html::cleanAttr(key) + "'>" +
html::htmlize(content, XHTMLStream::ESCAPE_ALL) + "</a>";
}
- Fix problem with HTML output of simple formulas like "y[i]" (bug 8609).
+- Clean ids for citations before outputting them. Characters like ":" are
+ not allowed in HTML identifiers, apparently (bug 8606).
+
* USER INTERFACE