This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Universal Character Names, v2
Martin v. L?wis wrote:-
> Would there be anything wrong with converting the entire *file*?
Some things to keep in mind:
1) How do you keep track of line / col? My pet hates about C
preprocessing, trigraphs and escaped newlines, make it awkward.
This problem also exists when doing it a line at a time like I
suggested.
2) You can easily get diagnostics for line 1000 before line 10 in
a file. e.g. our escaped newline at end of file, missing
newline at EOF, -Wtrigraph, space between '\' and newline
warnings. It's nice to have warnings appear in order.
3) Doing a file at a time can consume more memory and trash more
CPU cache.
But then, it does have the advantage of simplicity.
Neil.