class Mpfr < Formula desc "C library for multiple-precision floating-point computations" homepage "https://www.mpfr.org/" url "https://ftp.gnu.org/gnu/mpfr/mpfr-4.0.2.tar.xz" mirror "https://ftpmirror.gnu.org/mpfr/mpfr-4.0.2.tar.xz" sha256 "1d3be708604eae0e42d578ba93b390c2a145f17743a744d8f3f8c2ad5855a38a" bottle do cellar :any sha256 "cfce7ab866e98360c9364cd924da620ae7936d3a224d686aacc209c2107d19aa" => :mojave sha256 "bf5d21e7e8e549f7e8d07791a90f44a487f4c7151897b7c64d02928b5bd73520" => :high_sierra sha256 "4fb1860a481e24c70eefc8e5782030612840f1bb8f50586ca76a4c6c53629deb" => :sierra end depends_on "gmp" def install system "./configure", "--disable-dependency-tracking", "--prefix=#{prefix}", "--disable-silent-rules" system "make" system "make", "check" system "make", "install" end test do (testpath/"test.c").write <<~EOS #include #include #include int main() { mpfr_t x, y; mpfr_inits2 (256, x, y, NULL); mpfr_set_ui (x, 2, MPFR_RNDN); mpfr_root (y, x, 2, MPFR_RNDN); mpfr_pow_si (x, y, 4, MPFR_RNDN); mpfr_add_si (y, x, -4, MPFR_RNDN); mpfr_abs (y, y, MPFR_RNDN); if (fabs(mpfr_get_d (y, MPFR_RNDN)) > 1.e-30) abort(); return 0; } EOS system ENV.cc, "test.c", "-L#{lib}", "-L#{Formula["gmp"].opt_lib}", "-lgmp", "-lmpfr", "-o", "test" system "./test" end end