On Tue, Apr 16, 2002 at 02:10:15PM -0700, Bill Moseley wrote:
> > [% stag = "[\%"
> > etag = "%\]"
> > %]
>
> What's the logic for the backslash? I would have expected
The parser works in 2 passes. First, the split_text() method splits
the template into directives and text. The backslash confounds it and
stops it from thinking it's a real start/end of directive.
Then, the tokenise_directive() method kicks in, parses the double quoted
string and unescapes the escaped characters so that you end up with what
you want.
> [% stag = "\[%"
> etag = "\%]"
> %]
This would get parsed as:
[% stag = "\[%" etag = "\%]" %]
^^^^^^^^^^^^^^^^^^^^^^^^^^^
[% DIRECTIVE %]
> Or more perlish
>
> [% stag = '[%'
> etag = '%]'
> %]
Similarly,
[% stag = '[%' etag = '%]' %]
^^^^^^^^^^^^^^^^^^^^^^^^^
[% DIRECTIVE %]
> > [% TAGS star %]
> > [% INCLUDE foo %] # not a directive, just plain text, passed through
> > [* INCLUDE foo *] # is a directive
>
> More work for the existing template. It would be cool if TAGS was somehow
> block scoped, although I can't see how you would END the block -- with new
> or old tag?
In v3 the parser configuration (start/end tag, interpolate flags, etc) can
indeed be block scoped. Not sure of the syntax, but probably something like
this:
[% PARSE
start_tag = '[*'
end_tag = '*]'
interpolate = 1
%]
...
[% END %]
A
