Using the XmlReader

The XML use case

Any grammar read with the default Tokenizer and LalrProcessor implementations can be read as XML by using the TokenXmlReader class. This can be processed with an XSLT style sheet to generate a different representation of data not available as XML, basically the combination of the parser and an XSLT stylesheet enable parsing almost anything to a nice XML representation.

Possible use cases:

• Convert textual formulas to MathML
• Convert drawing instructions to SVG
• Convert some Wiki markup or markdown to (X)HTML or even XSL-FO
• Implement syntax highlighting or pretty printing

What it does

It exposes the grammar tree previously parsed into a Token as XML tree, using elements with the same name as their symbol (as far as XML can handle those) in the grammar and using namespaces (urn:nonterminal for reductions and urn:terminal for text tokens).

Things you should know:

• Only terminals have line information attached to them
• All insignificant data, such as comments and whitespace, is not included in the XML output
• Symbol names are converted to XML compatible names by removing all invalid characters from the name. If the resulting name is empty, an _ (underscore) character is used as name. Note that XML names may therefore be ambiguous compared to the original names in the grammar, the symbols <A B> and <AB> for instance would get the same name (AB), as well as '==' and '.' etc. would (_). Therefore you may want to name all terminals (in the GOLD grammar) which you want to transform later on:

EQUAL = '=='
DOT = '.'
! etc.