Mercurial > octave-nkf
view libinterp/octave-value/ov-flt-re-diag.cc @ 20299:bfe66db8addb
don't include hdf5.h or use HDF5 typedefs in public header files (bug #43180)
* oct-hdf5-types.h: Rename from oct-hdf5-id.h.
* oct-hdf5-types.cc: Rename from oct-hdf5-id.cc.
* libinterp/corefcn/module.mk: Update.
* oct-hdf5-types.h (octave_hdf5_err): New typedef.
* oct-hdf5-types.h, oct-hdf5-types.cc (check_hdf5_types): Rename from
check_hdf5_id_type. Also check size of herr_t.
* load-save.cc: Include oct-hdf5.h.
* ls-hdf5.cc: Include oct-hdf5.h instead of oct-hdf5-id.h.
Define hdf5_fstreambase functions that require HDF5 types here instead
of in ls-hdf5.h.
* ls-hdf5.h, ls-hdf5.cc: Use octave types in public interfaces.
* ls-hdf5.h, ov-base.h: Include oct-hdf5-types.h instead of oct-hdf5.h.
* oct-hdf5.h: Include oct-hdf5-types.h. Define H5T_NATIVE_IDX here.
* ls-hdf5.h: Not here.
author | John W. Eaton <jwe@octave.org> |
---|---|
date | Mon, 15 Jun 2015 17:42:44 -0400 |
parents | 4197fc428c7d |
children | f90c8372b7ba |
line wrap: on
line source
/* Copyright (C) 2008-2015 Jaroslav Hajek This file is part of Octave. Octave is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 3 of the License, or (at your option) any later version. Octave is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with Octave; see the file COPYING. If not, see <http://www.gnu.org/licenses/>. */ #ifdef HAVE_CONFIG_H #include <config.h> #endif #include "byte-swap.h" #include "ov-flt-re-diag.h" #include "ov-base-diag.cc" #include "ov-float.h" #include "ov-flt-re-mat.h" #include "ls-utils.h" template class octave_base_diag<FloatDiagMatrix, FloatMatrix>; DEFINE_OV_TYPEID_FUNCTIONS_AND_DATA (octave_float_diag_matrix, "float diagonal matrix", "single"); static octave_base_value * default_numeric_conversion_function (const octave_base_value& a) { CAST_CONV_ARG (const octave_float_diag_matrix&); return new octave_float_matrix (v.float_matrix_value ()); } octave_base_value::type_conv_info octave_float_diag_matrix::numeric_conversion_function (void) const { return octave_base_value::type_conv_info (default_numeric_conversion_function, octave_float_matrix::static_type_id ()); } octave_base_value * octave_float_diag_matrix::try_narrowing_conversion (void) { octave_base_value *retval = 0; if (matrix.nelem () == 1) retval = new octave_float_scalar (matrix (0, 0)); return retval; } DiagMatrix octave_float_diag_matrix::diag_matrix_value (bool) const { return DiagMatrix (matrix); } FloatDiagMatrix octave_float_diag_matrix::float_diag_matrix_value (bool) const { return matrix; } ComplexDiagMatrix octave_float_diag_matrix::complex_diag_matrix_value (bool) const { return ComplexDiagMatrix (matrix); } FloatComplexDiagMatrix octave_float_diag_matrix::float_complex_diag_matrix_value (bool) const { return FloatComplexDiagMatrix (matrix); } octave_value octave_float_diag_matrix::map (unary_mapper_t umap) const { switch (umap) { case umap_abs: return matrix.abs (); case umap_real: case umap_conj: return matrix; case umap_imag: return DiagMatrix (matrix.rows (), matrix.cols (), 0.0); case umap_sqrt: { FloatComplexColumnVector tmp = matrix.extract_diag ().map<FloatComplex> (rc_sqrt); FloatComplexDiagMatrix retval (tmp); retval.resize (matrix.rows (), matrix.columns ()); return retval; } default: return to_dense ().map (umap); } } bool octave_float_diag_matrix::save_binary (std::ostream& os, bool& /* save_as_floats*/) { int32_t r = matrix.rows (); int32_t c = matrix.cols (); os.write (reinterpret_cast<char *> (&r), 4); os.write (reinterpret_cast<char *> (&c), 4); FloatMatrix m = FloatMatrix (matrix.extract_diag ()); save_type st = LS_FLOAT; if (matrix.length () > 8192) // FIXME: make this configurable. { float max_val, min_val; if (m.all_integers (max_val, min_val)) st = get_save_type (max_val, min_val); } const float *mtmp = m.data (); write_floats (os, mtmp, st, m.numel ()); return true; } bool octave_float_diag_matrix::load_binary (std::istream& is, bool swap, oct_mach_info::float_format fmt) { int32_t r, c; char tmp; if (! (is.read (reinterpret_cast<char *> (&r), 4) && is.read (reinterpret_cast<char *> (&c), 4) && is.read (reinterpret_cast<char *> (&tmp), 1))) return false; if (swap) { swap_bytes<4> (&r); swap_bytes<4> (&c); } FloatDiagMatrix m (r, c); float *re = m.fortran_vec (); octave_idx_type len = m.length (); read_floats (is, re, static_cast<save_type> (tmp), len, swap, fmt); if (error_state || ! is) return false; matrix = m; return true; } bool octave_float_diag_matrix::chk_valid_scalar (const octave_value& val, float& x) const { bool retval = val.is_real_scalar (); if (retval) x = val.float_value (); return retval; }