For a long, long time, I have thought, "It will be way cool to have better text-to-html in Rubric!" I really wanted a decent markup system for doing this, and I didn't want to make any decisions. Decisions are hard. Instead, I punted: "I'll figure it out later!"
I just put in a default filter to keep my paragraph breaks, and I was happy. The only problem was that code snippets came out ugly, but that wasn't a big deal. I didn't filter HTML, so I just wrapped them in CODE elements and coped. This made it easy-ish for them to get passed on to use.perl.org, too. Still, it pretty much sucked.
Later on, I added the @private tag for "entries that only I can see." I recently got a bug that it still leaks a data (like tag counts), but that's not the point. The point was that I'd finally picked a way (the @-prefix) to designate tags reserved for the system.
Someone decided that it would be a good idea to run a public (open registration) Rubric installation, and quickly noticed that he was open to cross-site scripting. Since I wrote Rubric for use just by me, I wasn't scrubbing HTML (remember? I just stuck in <code> blocks, above!). I told him how to change the TT2 filter that was being used, but it got me thinking, again, about making this easier.
Even later, I implemented a feature that a number of people had wanted: tag values. Honestly, I didn't really see much use. If I wanted "language:perl" I would just use "language perl" and avoid the weird mixing of tags with attributes. Lately I've had a few ideas about when this might be interesting to use, but at the time I had just one reason to implement it: @markup.
The idea was that I'd create a system tag called @markup, and it would have a value. That value would indicate the way in which the entry's body should be rendered form. This would not be part of the Rubric::Renderer, but would be a set of pairs, mapping @markup values to Rubric::Entry::Formatter::Whatever classes. Rubric::Entry would provide an
as_whatever method (possibly soon to be
as('whatever')) that would return the formatted entry body; the renderer would use that.
I implemented this a while ago, providing two bundled formatters: Nil and HTMLEscape. Nil is the default: it does nothing but maintain HTML line breaks. HTMLEscape does what you think: it escapes anything that needs to be escaped for HTML, so you see just what the text contained.
I should probably rename these.
Anyway, nobody much noticed when I released this, but I started getting more questions about making the text-to-HTML better. Now I had a simple answer: write a plugin and use it! Dan said he was writing one, and I was really excited to see someone do it, but I don't think it's been released yet. I wrote a feature that I thought would be cool, but nothing cool was being done with it!
Well, last week I brought some music from work that got me wanting to write an entry quoting some lyrics. I didn't want to have to actually type out all those blockquotes! Blockquote is way, way too long for an element name, you know. It was time to write a useful formatter!
I was almost sad when it only took about two minutes. It certainly took much longer updating my post-from-vim-and-also-to-use.perl script than writing the formatter.
The new formatter, which is formatting this entry (I hope!) is Rubric::Entry::Formatter::Markdown. It uses Text::Markdown, which couldn't be much simpler to use. Ignoring blank lines and POD, it's just twelve lines of code. I'll be uploading it tonight.
I am pretty happy about this. I may not need to write another formatter for a while, now... unless I wrote VimColorize!