[DOXIA-480] XhtmlBaseParser ignores XHTML default entities

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.4
• Fix Version/s: 1.4
• Component/s: Core, Module - Xhtml
• Flags: Patch
• Assignee: Olivier Lamy
• Reporter: Andrius Velykis
• Created: 27/Oct/12 16:40
• Updated: 05/Apr/15 14:44
• Resolved: 21/Jan/13 16:00

Description

XHTML defines a number of default entities that can appear in valid XHTML files (http://www.w3.org/TR/xhtml1/#h-A2), such as left/right quotes: “, ’, and many others.

XhtmlBaseParser, however, ignores XHTML default entities appearing in the source code. This is because it delegates the parsing to AbstractXmlParser, which uses vanilla MXParser to parse. MXParser only recognises default XML entities.

Because the HTML entities are not resolved by the XML parser, and thus by the XHTML parser, they are not rendered by the XHTML module. I have attached a sample project for Maven site that uses XHTML module. The source file has double/single quotes, however the output file does not.

This also affects other parsers that extend XhtmlParser, e.g. MarkdownParser (see DOXIA-473 for a reported bug). This is because Pegdown library, used to parse Markdown, generates “ for quotes and other entities.

I have attached a patch that fixes this problem. It exposes the XmlPullParser (MXParser) for configuration before parsing, so that extending classes could define default entities. Then XhtmlBaseParser adds default XHTML entities to the parser. This patch will also fix DOXIA-473, because MarkdownParser extends XhtmlParser.

Issue Links

• supercedes: DOXIA-473 Quotes and double quotes are stripped from the HTML output.