[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [bug-gnulib] Re: iconv made easy
From: |
Paul Eggert |
Subject: |
Re: [bug-gnulib] Re: iconv made easy |
Date: |
Mon, 13 Dec 2004 16:05:19 -0800 |
User-agent: |
Gnus/5.1006 (Gnus v5.10.6) Emacs/21.3 (gnu/linux) |
Simon Josefsson <address@hidden> writes:
> I like this. But it would be quite wasteful on long strings;
> MB_LEN_MAX is 16 on glibc hosts.
I suspect it wouldn't be that wasteful of memory if you realloc the
result. There may be some memory fragmentation and with some
applications this might be decisively bad; but for typical usage and a
good allocator I'd guess it wouldn't be a problem.
> This would be fast on short strings (one malloc, one iconv, one
> realloc), reasonable fast on long strings (two calls to iconv), and it
> would even handle internal bugs when iconv exceed the computed upper
> bound.
Hmmm, I don't see why it would handle internal bugs any better; if
iconv is buggy, all bets are off.
I was hoping to avoid the "two calls to iconv", by trading space for
time. If the string is sufficiently large, the memory allocator
should "do a good job" in reallocating it (i.e., realloc shouldn't
need to copy the data). Also, it's simpler to stick with one
algorithm rather than to have an if-then-else.
(But it's a minor issue, of course. Also, one could always
measure....)