Re[2]: [Nasm-devel] _NASM_FORMAT_ vs. __OUTPUT_FORMAT__

[Hotbox <un...@rb...>]

> Yes. Back when __OUTPUT_FORMAT__ support was added, I explicitly
> asked for its value to be unquoted, because the user can always
> surround it with a pair of single or double quotes (by using the
> preprocessor's support for unterminated string constants) while
> it would be impossible to remove the quotes from a string value.

Unquoted macro value is unsafe for it always can be expanded.
Querying anything as string is quite safely.

Also, I insist that using unterminated string for string creation
is a very ugly syntax. I assume, it's a bug. And, of course,
such syntax should not be recommended.

I've introduced a set of 'native' functions in my version of
Nasm preprocessor. These functions provide almost complete API
for string and list processing. I can move them into the current
Nasm preprocessor.

These functions are (the last revision):
 %strtok(<str>): <list>
 %detok(<list>): <str>
 %split(<str>, [<str>=" \t\r\n"]): <list>
 %join(<list>, [<str>=""]): <str>
 %strlen(<str>): <num>
 %listlen(<list>): <num>
 %substr(<str>, <expr>, [<expr>=1]): <str>
 %sublist(<list>, <expr>, [<expr>=1]): <list>
 %strrev(<str>): <str>
 %listrev(<str>): <str>
 %strpos(<str>, <str>, [<expr>=1]): <num>
 %listpos(<list>, <list>, [<expr>=1]): <num>
 %trim(<str>, [<str>=" \t\r\n"]): <str>
 %upper(<list>): <list>
 %lower(<list>): <list>
 %charcat(<list>): <str>

Tell me, if you have some questions or remarks about
function naming or functionality.

-- 
Best regards,
Nickolay Yurchenko           mailto:un...@rb...