Hello Don,
== Quote from BCS (n...@anon.com)'s article
OTOH, and IHMO it should be avoided where you can, kinda like goto.
(For one things, forget having usable line numbers in your error
messages.)
Whenever that happens, it's a compiler bug.
If you can produce a test case where that happens, please put it in
Bugzilla.
The issue isn't that the error has no line number but that the line number
is the line of the mixin or the line within the mixin (I forget which) to
be useful you need both AND the full text of the mixin.
Yes there are ways around this but my original point is that in my experience,
I can get just about as much done and for about the same effort with very
little code being mixed in. My approach to this kind of problem has typically
been to use CTFE to process input, template and or mixins to generate an
AST (or whatever the problem demands) represented as a type and normal template,
static ifs and tuple foreach to do the code generation with the occasional
line of code mixed in here and there. The result of this is that most of
the code is transparently visible as code rather than as string literals.
--
... <IXOYE><
