HTML and CSS Preprocessors

I was a little surprised to see so much dedicated to HAML in an HTML/CSS tutorial. The author's background is as an instructor at the starter league which is heavy on Rails. HAML is widely used in that community so his perspective might be slightly bias.

But it is still useful to know and see there are other ways to handle to write HTML.

And that is because it is more than "markup should be beautiful". I cringe when that is stated as the only reason. Markup that is simplified can lead to easier readability and fewer mistakes. It removes the worry if you forgot to close a tag and removes them all together.

I was hesitant at first after being so used to working with HTML but I am happy as ever after making the switch. I find my time is more productive and updating the html is much quicker.

Oh, come on. You got hung up on an adjective used in a hero unit.

http://haml.info/docs/yardoc/file.REFERENCE.html

Filters, passing Ruby objects into it to generate classes, hell, even markup comments that don't show up in html.

I've used HAML for 4 years now, my take on it is that HAML gets the computer to do work it can dependably do (writing out closing tags) and reducing substantially how much I have to type. Less typing means less code to maintain but also easier to see the purpose of the code.

I know a lot of ruby on rails developers who use HAML for their template framework instead of ERB. Some people prefer HAML's use of indentation instead of using tags to enclose everything.

I like it in my projects mostly for how clean and easy it is to read. It takes a little bit to get used to initially but it is worth it.

You might be interested in shpaml-jinja[0]. It extends the shpaml[1] syntax to provide jinja commands.

[0]: https://github.com/p/shpaml-jinja [1]: http://shpaml.webfactional.com/examples

I'm finding that it's a lot cleaner to use with Javascript. For example, say you have an array of items and you want to create a fieldset for each item, and add an attribute of "data-id", and "data-category" to each fieldset so you can grab them with jQuery later on.

In HAML all you have to do is this...

- @items.each do |item|

  %fieldset{:data => {:id => item.id, :category => item.category}}

===

In ERB the equivalent would be...

<% @items.each do |item| %>

  <fieldset data-id="<%= item.id %>" data-category="<%= item.category %>">

  </fieldset>

<% end %>

===

It's a lot cleaner in HAML, especially extrapolated out over larger views.

My old company used HAML exclusively for all of our RoR view code. In most general cases it makes the coding faster. THere are of course some wonky edge cases in HAML vs code style best practices.

For instance: try making an html element with a long list of attributes and see how it looks. Or worse, run a translate on a block of text while maintaining reasonable line lengths.

Good? Yes. Silver bullet html replacement? Lol.

Anything that is self-described as an "abstraction markup language" and must be compiled into the language it abstracts is really a pre-processor.

Regardless of intended use cases, i cant think of a better definition.