|Perl: the Markov chain saw|
Reusable template components with HTML::Templateby dws (Chancellor)
|on Aug 04, 2004 at 01:18 UTC||Need Help??|
When you're using HTML::Template and want to put a table into a page, you have two options: You can embed loop logic in the template, or you can inject a table constructed elsewhere into the template. If you favor reuse, you're going to lean towards injection. But that doesn't mean you can't use templates for both parts. Here's a variation that does both, in a reusable way.
A simple embedded table looks like
and the code to expand it looks like
You can isolate part of the above into a separate template, and pull it in to the main template via
This expands the included template in-line, with no visible difference to the calling program, though you do get the benefit of having a reusable table template. But you can do better, as we'll see below.
When injecting a table, the template looks like
You're then on the hook for providing a chunk of HTML to supply as the template parameter table. You can construct the HTML from code, from a heredoc, by using CGI.pm tag support, or by some other means. Whichever way you go, you'll end up doing something like
A possibly non-obvious "other means" is to produce the table HTML by using a separate HTML::Template instance that uses its own template for the table part. This looks like
You now have the basis for a reusable component that you can pull out any time you need to embed tables in a template. Variations include passing a query into the component, which then extracts data from some data source. This technique also extends nicely to other types of components.