Richard E Maine wrote:
> Indeed. I understand the "done before" part, independent of language
> issues. But I have no idea why this would be "not particularly suited to
> Fortran". I personally have a horrid time dealing with even the most
> trivial of string processing tasks using C.
C is certainly no role model (for pretty much anything), as I already
said. But I personally find Fortran's character facilities rather
limiting and poorely designed (by virtue of history mostly). Not
"horrid", but certainly not enticing for a job like a parser. Maybe
finally with F2003's allocatable (variable-length) strings it is
starting to look descent. But I still dislike the unsymmetric treatment
of character arrays of length L and characters with kind length L...
> Part of this might be a function of me as much as the language
Of course, of course, if there is no product that one can use then one
should write in a language that is comfortable, even if it is not the
"best" one. I've written things in Fortran just because I can do it most
quickly in Fortran merely be knowing it best.
Speaking of comparing languages, this mention of PL/I made me look into
it (I had only heard of it but literally did not know anything). If
someone knows both Fortran 2003 and PL/I, a question: Is there some area
where PL/I still has far richer feature set than Fortran? From the quick
skim exception handling came to mind. I always like looking at other
languages to get design ideas...
Thanks,
Aleks
|