Welcome to mirror list, hosted at ThFree Co, Russian Federation.

git.blender.org/blender.git - Unnamed repository; edit this file 'description' to name the repository.
summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat (limited to 'source/blender/blenlib/intern/vectorops.c')
-rw-r--r--source/blender/blenlib/intern/vectorops.c165
1 files changed, 165 insertions, 0 deletions
diff --git a/source/blender/blenlib/intern/vectorops.c b/source/blender/blenlib/intern/vectorops.c
new file mode 100644
index 00000000000..9314e7fca3e
--- /dev/null
+++ b/source/blender/blenlib/intern/vectorops.c
@@ -0,0 +1,165 @@
+/*
+ *
+ * Some vector operations.
+ *
+ * Always use
+ * - vector with x components : float x[3], int x[3], etc
+ *
+ * $Id$
+ *
+ * ***** BEGIN GPL/BL DUAL LICENSE BLOCK *****
+ *
+ * This program 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 2
+ * of the License, or (at your option) any later version. The Blender
+ * Foundation also sells licenses for use in proprietary software under
+ * the Blender License. See http://www.blender.org/BL/ for information
+ * about this.
+ *
+ * This program 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 this program; if not, write to the Free Software Foundation,
+ * Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+ *
+ * The Original Code is Copyright (C) 2001-2002 by NaN Holding BV.
+ * All rights reserved.
+ *
+ * The Original Code is: all of this file.
+ *
+ * Contributor(s): none yet.
+ *
+ * ***** END GPL/BL DUAL LICENSE BLOCK *****
+ */
+
+/* ------------------------------------------------------------------------- */
+/* General format: op(a, b, c): a = b op c */
+/* Copying is done cp <from, to> */
+/* ------------------------------------------------------------------------- */
+
+#include "MTC_vectorops.h"
+#include <math.h>
+
+void MTC_diff3Int(int v1[3], int v2[3], int v3[3])
+{
+ v1[0] = v2[0] - v3[0];
+ v1[1] = v2[1] - v3[1];
+ v1[2] = v2[2] - v3[2];
+}
+
+/* ------------------------------------------------------------------------- */
+void MTC_diff3Float(float v1[3], float v2[3], float v3[3])
+{
+ v1[0] = v2[0] - v3[0];
+ v1[1] = v2[1] - v3[1];
+ v1[2] = v2[2] - v3[2];
+}
+
+/* ------------------------------------------------------------------------- */
+
+void MTC_cross3Int(int v1[3], int v2[3], int v3[3])
+{
+ v1[0] = v2[1]*v3[2] - v2[2]*v3[1];
+ v1[1] = v2[2]*v3[0] - v2[0]*v3[2];
+ v1[2] = v2[0]*v3[1] - v2[1]*v3[0];
+}
+
+/* ------------------------------------------------------------------------- */
+
+void MTC_cross3Float(float v1[3], float v2[3], float v3[3])
+{
+ v1[0] = v2[1]*v3[2] - v2[2]*v3[1];
+ v1[1] = v2[2]*v3[0] - v2[0]*v3[2];
+ v1[2] = v2[0]*v3[1] - v2[1]*v3[0];
+}
+/* ------------------------------------------------------------------------- */
+
+void MTC_cross3Double(double v1[3], double v2[3], double v3[3])
+{
+ v1[0] = v2[1]*v3[2] - v2[2]*v3[1];
+ v1[1] = v2[2]*v3[0] - v2[0]*v3[2];
+ v1[2] = v2[0]*v3[1] - v2[1]*v3[0];
+}
+
+/* ------------------------------------------------------------------------- */
+
+int MTC_dot3Int(int v1[3], int v2[3])
+{
+ return (v1[0]*v2[0] + v1[1]*v2[1] + v1[2]*v2[2]);
+}
+
+/* ------------------------------------------------------------------------- */
+
+float MTC_dot3Float(float v1[3], float v2[3])
+{
+ return (v1[0]*v2[0] + v1[1]*v2[1] + v1[2]*v2[2]);
+}
+
+/* ------------------------------------------------------------------------- */
+
+void MTC_cp3Float(float v1[3], float v2[3])
+{
+ v2[0] = v1[0];
+ v2[1] = v1[1];
+ v2[2] = v1[2];
+}
+
+/* ------------------------------------------------------------------------- */
+
+void MTC_cp3FloatInv(float v1[3], float v2[3])
+{
+ v2[0] = -v1[0];
+ v2[1] = -v1[1];
+ v2[2] = -v1[2];
+}
+
+/* ------------------------------------------------------------------------- */
+
+void MTC_swapInt(int *i1, int *i2)
+{
+ int swap;
+ swap = *i1;
+ *i1 = *i2;
+ *i2 = swap;
+}
+
+/* ------------------------------------------------------------------------- */
+
+void MTC_diff3DFF(double v1[3], float v2[3], float v3[3])
+{
+ v1[0] = v2[0] - v3[0];
+ v1[1] = v2[1] - v3[1];
+ v1[2] = v2[2] - v3[2];
+}
+
+/* ------------------------------------------------------------------------- */
+float MTC_normalise3DF(float n[3])
+{
+ float d;
+
+ d= n[0]*n[0]+n[1]*n[1]+n[2]*n[2];
+ /* FLT_EPSILON is too large! A larger value causes normalise errors in */
+ /* a scaled down utah teapot */
+ if(d>0.0000000000001) {
+
+ /* d= sqrt(d); This _should_ be sqrt, but internally it's a double*/
+ /* anyway. This is safe. */
+ d = sqrt(d);
+
+ n[0]/=d;
+ n[1]/=d;
+ n[2]/=d;
+ } else {
+ n[0]=n[1]=n[2]= 0.0;
+ d= 0.0;
+ }
+ return d;
+}
+
+/* ------------------------------------------------------------------------- */
+
+/* eof */