| 1 |
This bug has been discovered in x86 too. |
| 2 |
|
| 3 |
On Friday 03 May 2002 20:28, you wrote: |
| 4 |
> So, gnome-extra/gal-0.19 fails to merge due to ./configure failing with : |
| 5 |
> |
| 6 |
> [ ... ] |
| 7 |
> checking for catalogs to be installed... az ca cs da de en_GB es fi fr gl |
| 8 |
> hu it ja ko lt nn no pl pt pt_BR ro ru sk sv ta tr uk zh_TW |
| 9 |
> sed: can't read ./intl/po2tbl.sed.in: No such file or directory |
| 10 |
> checking for libxml (aka gnome-xml) libraries 1.8.8 <= version < 2.0.0... |
| 11 |
> found |
| 12 |
> checking for iconv in -liconv... no |
| 13 |
> checking for iconv... yes |
| 14 |
> checking for gnu_get_libc_version... yes |
| 15 |
> checking if iconv() handles UTF-8... no |
| 16 |
> configure: error: You need to install a working iconv implementation, such |
| 17 |
> as http://clisp.cons.org/~haible/packages-libiconv.html |
| 18 |
> [ ... ] |
| 19 |
> |
| 20 |
> in config.log |
| 21 |
> |
| 22 |
> #include <string.h> |
| 23 |
> #ifdef HAVE_GNU_GET_LIBC_VERSION |
| 24 |
> #include <gnu/libc-version.h> |
| 25 |
> #endif |
| 26 |
> |
| 27 |
> int main (int argc, char **argv) |
| 28 |
> { |
| 29 |
> const char *jp = "\x1B\x24\x42\x46\x7C\x4B\x5C\x38\x6C"; |
| 30 |
> const char *utf8 = "\xE6\x97\xA5\xE6\x9C\xAC\xE8\xAA\x9E"; |
| 31 |
> char *transbuf = malloc (10), *trans = transbuf; |
| 32 |
> iconv_t cd; |
| 33 |
> size_t jp_len = strlen (jp), utf8_len = 10; |
| 34 |
> size_t utf8_real_len = strlen (utf8); |
| 35 |
> |
| 36 |
> #ifdef HAVE_GNU_GET_LIBC_VERSION |
| 37 |
> /* glibc 2.1.2's iconv is broken in hard to test ways. */ |
| 38 |
> if (!strcmp (gnu_get_libc_version (), "2.1.2")) |
| 39 |
> exit (1); |
| 40 |
> #endif |
| 41 |
> |
| 42 |
> cd = iconv_open ("UTF-8", "ISO-2022-JP"); |
| 43 |
> if (!cd) |
| 44 |
> exit (1); |
| 45 |
> if (iconv (cd, &jp, &jp_len, &trans, &utf8_len) == -1 || jp_len != 0) |
| 46 |
> exit (1); |
| 47 |
> if (memcmp (utf8, transbuf, utf8_real_len) != 0) |
| 48 |
> exit (1); |
| 49 |
> |
| 50 |
> exit (0); |
| 51 |
> } |
| 52 |
> |
| 53 |
> So at first I was looking into UTF-8 problems, but it appears, that since |
| 54 |
> % iconv -l |
| 55 |
> produces 'UTF-8', BUT NOT 'ISO-2022-JP' that the lack of iconv charset |
| 56 |
> support for JP specifically, but many others actually, is at issue. |
| 57 |
> |
| 58 |
> The tricky bit is that % iconv -l on my x86 box produces quite a lot of |
| 59 |
> output -- maybe 4 times. Of course, the glibc ebuilds are identical -- |
| 60 |
> with the exception that my x86 glibc was built during a system upgrade and |
| 61 |
> not build time. |
| 62 |
> |
| 63 |
> So to test a simple case, could someone who has rebuilt glibc via a portage |
| 64 |
> upgrade post their output from % iconv -l? |
| 65 |
> |
| 66 |
> thx. |
| 67 |
> Gontran |
| 68 |
> _______________________________________________ |
| 69 |
> gentooppc-dev mailing list |
| 70 |
> gentooppc-dev@g.o |
| 71 |
> http://lists.gentoo.org/mailman/listinfo/gentooppc-dev |
| 72 |
|
| 73 |
-- |
| 74 |
Pieter Van den Abeele |
| 75 |
pvdabeel@××××××.be |
Archives