I spent a few minutes seeing how difficult it would be to hack the parser to accept PYTHON blocks and pass the results to Inline::Python. Fairly easy, it turns out, provided you don't use template caching (the spacing upsets the python compiler), and you accept the caveat that I've not tested anything advanced at all. There's also no way to turn it on or off, like the EVAL_PERL config option. I wasn't feeling that ambitious.
My test template:
$ cat tmp/pytest.tt2
[% PYTHON %]
def test(whut):
print "Hello,", whut
test("world")
[% END -%]
Output:
$ make && perl -Mblib blib/script/tpage tmp/pytest.tt2
Hello, world
Not amazing, sure, but I had fun.
Attached is a diff against TT 2.22, freshly downloaded from CPAN.
To make it work, apply it to a freshly unpacked tarball, and recompile
the grammar:
$ patch -p1 < ../TT2+python.diff
patching file lib/Template/Directive.pm
patching file parser/Grammar.pm.skel
patching file parser/Parser.yp
$ (cd parser; sh yc)
/tmp/Template-Toolkit-2.22/parser
Compiling parser grammar (Parser.yp -> ../lib/Template/Grammar.pm)
4 shift/reduce conflicts and 1 reduce/reduce conflict
$ perl Makefile.PL
# ...
$ make
And then test it with -Mblib.
--
(darren)
TT2+python.diff
Description: Binary data
_______________________________________________ templates mailing list templates@template-toolkit.org http://mail.template-toolkit.org/mailman/listinfo/templates
