There is a lot to be said about design philosophy, but let's first give a
simple example to give you the feel of how it works. (The example is in
-Perl, but there's also a functionally equivalent PHP version, and more
-languages should probably come soon.)
+Perl, but there are also functionally equivalent PHP and Python versions,
+and more languages should probably come soon.)
Template (simple.xml):
my $doc = XML::Template::process_file('../xml/clone.xml', {
'color' => 'red',
'#things' => [
- { 'li' => 'Raindrops on roses', 'li.class' => 'odd' },
- { 'li' => 'Whiskers on kittens', 'li.class' => 'even' },
- { 'li' => 'Bright copper kettles', 'li.class' => 'odd' },
- { 'li' => 'Warm, woolen mittens', 'li.class' => 'even' }
+ { 'li' => 'Raindrops on roses', 'li/class' => 'odd' },
+ { 'li' => 'Whiskers on kittens', 'li/class' => 'even' },
+ { 'li' => 'Bright copper kettles', 'li/class' => 'odd' },
+ { 'li' => 'Warm, woolen mittens', 'li/class' => 'even' }
]
});
print $doc->toString;
the entire DOM with wrappers for each language. (Thankfully, by relying on
the DOM support in each language, the code so far is under 200 lines per
implementation, so maintaining this hopefully shouldn't be much work.)
- As proof-of-concept, I've got Perl and PHP implementations that work and
- feel largely the same -- Python, Ruby and other implementations are welcome.
+ As proof-of-concept, I've got Perl, PHP and Python implementations that work
+ and feel largely the same -- Ruby and other implementations are welcome.
+ (This is backed up by a test suite, which ensures that all the different
+ implementations return structurally equivalent XML for a certain set of
+ test cases. Porting to a new language is not difficult, and once you've
+ got all the test cases to pass, your work is most likely done.)
As a side note to the second point, I've spent some time wondering exactly
_why_ you want to separate the back-end logic from your HTML, and why people