Robert Feldt <feldt / ce.chalmers.se> writes: > Since ri now supports abbreviations via prefix matching we should be able > to give abbreviations of length 1. Patch is below. Thanks > Another thought on ri: I proposed some time ago that we should base a > simple documentation format on the formats used by Dave&Andy in the book > and now (in another form) in ri. I still think this would be a good > thing. Anyone else got ideas on this? Would it suffice if I/we do a simple > parser outputting stuff the way ri documentation is formatted? (Or is > too much info lost there to get back to XML?) Would also > be good with a short list of advice on how to write docs the "pickaxe > way". Dave? If people want to write using rhe original LaTeX, I'd be happy to release our style files. That's pretty clean. For example, from Array: \begin{method}{reverse}{\self.reverse \returns{\obj{anArray}}}{A} Returns a new array using \self's elements in reverse order. \begin{ruby}[comment] [ "a", "b", "c" ].reverse [ 1 ].reverse \end{ruby} \end{method} To generate stuff for ri, I have a Ruby program and some XSL that converts the above to: <method name="reverse"> <callseq> <i>arr</i>.reverse -> <i>anArray</i> </callseq> <desc> <p/> Returns a new array using <i>arr</i>'s elements in reverse order. <p/> <verbatim> <MULTI><m1><code>[ "a", "b", "c" ].reverse</code></m1><m2><code>["c", "b", "a"]</code></m2></MULTI> <MULTI><m1><code>[ 1 ].reverse</code></m1><m2><code>[1]</code></m2></MULTI> </verbatim> </desc> </method> However, it you wanted to use this as a general format, I'd want to step back a bit, and instead use something like <method name="reverse"> <callseq> <obj>arr</obj>.reverse <returns><obj>anArray</obj></returns> </callseq> <desc> Returns a new array using <obj>arr</obj>'s elements in reverse order. <code show_output_as="comments"> [ "a", "b", "c" ].reverse [ 1 ].reverse </code> </desc> </method> This would be usable to generate decent HTML and ri stuff. Was this the kind of thing you were looking for? Dave