1878ac9335
It's mainly refactoring and mass-rebuild simplifications without any real impact (besides better readability).
53 lines
1.7 KiB
Nix
53 lines
1.7 KiB
Nix
{ fetchurl, stdenv, lib }:
|
|
|
|
assert (!stdenv.isLinux);
|
|
|
|
stdenv.mkDerivation rec {
|
|
name = "libiconv-1.14";
|
|
|
|
src = fetchurl {
|
|
url = "mirror://gnu/libiconv/${name}.tar.gz";
|
|
sha256 = "04q6lgl3kglmmhw59igq1n7v3rp1rpkypl366cy1k1yn2znlvckj";
|
|
};
|
|
|
|
patches = lib.optionals stdenv.isCygwin [
|
|
./libiconv-1.14-reloc.patch
|
|
./libiconv-1.14-wchar.patch
|
|
];
|
|
|
|
configureFlags =
|
|
# On Cygwin, Libtool produces a `.dll.a', which is not a "real" DLL
|
|
# (Windows' linker would need to be used somehow to produce an actual
|
|
# DLL.) Thus, build the static library too, and this is what Gettext
|
|
# will actually use.
|
|
lib.optional stdenv.isCygwin "--enable-static"
|
|
++ lib.optional stdenv.isFreeBSD "--with-pic";
|
|
|
|
crossAttrs = {
|
|
# Disable stripping to avoid "libiconv.a: Archive has no index" (MinGW).
|
|
dontStrip = true;
|
|
dontCrossStrip = true;
|
|
};
|
|
|
|
meta = {
|
|
description = "An iconv(3) implementation";
|
|
|
|
longDescription = ''
|
|
Some programs, like mailers and web browsers, must be able to convert
|
|
between a given text encoding and the user's encoding. Other programs
|
|
internally store strings in Unicode, to facilitate internal processing,
|
|
and need to convert between internal string representation (Unicode)
|
|
and external string representation (a traditional encoding) when they
|
|
are doing I/O. GNU libiconv is a conversion library for both kinds of
|
|
applications.
|
|
'';
|
|
|
|
homepage = http://www.gnu.org/software/libiconv/;
|
|
license = lib.licenses.lgpl2Plus;
|
|
|
|
maintainers = [ ];
|
|
|
|
# This library is not needed on GNU platforms.
|
|
hydraPlatforms = with lib.platforms; cygwin ++ darwin ++ freebsd;
|
|
};
|
|
}
|