spkg-configure.m4 for flint

[dimpase]

Flint on Sage is linked with NTL, and this can be checked by testing for presence of
fmpz_poly_get_ZZX there. So, this:

SAGE_SPKG_CONFIGURE([flint], [
    AC_CHECK_HEADER(flint/flint.h, [], [sage_spkg_install_flint=yes])
dnl fmpz_mat_is_hadamard appears in Flint 2.5.0
    AC_SEARCH_LIBS([fmpz_mat_is_hadamard], [flint], [], [sage_spkg_install_flint=yes])
    AC_SEARCH_LIBS([fmpz_poly_get_ZZX], [flint], [break], [sage_spkg_install_flint=yes])
])

CC: @embray

Component: build

Author: Dima Pasechnik

Branch/Commit: u/dimpase/packages/flintconfig @ 8fc4fb6

Issue created by migration from https://trac.sagemath.org/ticket/27264

[embray]

comment:1

Do you want to go ahead and add a branch? Obviously, it seems fine to me at first glance.

[dimpase]

comment:2

Yes, but need to do #27265 (NTL) first

[dimpase]

Dependencies: #27265

[dimpase]

Commit: 82772d7

[dimpase]

Branch: u/dimpase/packages/flintconfig

[dimpase]

Changed dependencies from #27265 to #27265, #27212

[dimpase]

Last 10 new commits:

af9ab90	deal with --with-ntl, proper version check
5e1a30b	spkg-configure for mpfr, adjustments for its dependees
18fbf7b	add SAGE_CONFIGURE_MPFR, fix sage-env-config.in
50aebab	implement --with-mpfr=system/install
0fc433f	update for Sage 8.7.rc0, a typo fix
38427e1	spkg-configure for mpc, adjustments for its dependees
ec32471	Merge branch 'u/dimpase/packages/mpc-config' of trac.sagemath.org:sage into mpf
91f423f	rebase, deal with --with-mpc
33edfa8	Merge branch 'mpf' into flintconf - no spkg-configure yet
82772d7	spkg-configure for flint

[embray]

comment:5

Should we still look at NTL first, or maybe try testing with both this ticket and #27265 merged together?

[dimpase]

comment:6

Naturally, I am testing them all together...
(this branch has them all).

Actually, there is a lot of common boilerplate in spkg-configure.m4's for mpfr, mpc, ntl, and flint. But OK, perhaps this simplification can wait---depending upon how many more spkgs need --with-...= support.

[embray]

comment:7

I agree, I think that for now it will be simpler and easier to understand if we are somewhat repetitive. If, once we've made more progress, a clear pattern emerges then I can certainly collapse the repetitive bits into a macro.