mirrors/gstreamer
mirrors/gstreamer
1
1
Fork 0
mirror of https://gitlab.freedesktop.org/gstreamer/gstreamer.git synced 2024-11-26 11:41:09 +00:00
Code Issues 1 Projects Releases Wiki Activity

testsuite/caps/fraction-convert.c: Make the test less pedantic about float roundoff errors.

Browse source 
Original commit message from CVS:
* testsuite/caps/fraction-convert.c: (check_from_double_convert),
(check_from_fraction_convert), (transform_test), (main):
Make the test less pedantic about float roundoff errors.
This commit is contained in:
Wim Taymans 2004-07-29 15:40:00 +00:00
parent 3d67acc6cd
commit 8615072540
3 changed files with 54 additions and 24 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
ChangeLog
View file

@ -1,3 +1,9 @@
2004-07-29 Wim Taymans <wim@fluendo.com>
* testsuite/caps/fraction-convert.c: (check_from_double_convert),
(check_from_fraction_convert), (transform_test), (main):
Make the test less pedantic about float roundoff errors.
2004-07-29 Benjamin Otte <otte@gnome.org> 2004-07-29 Benjamin Otte <otte@gnome.org>
* gst/elements/gstfilesrc.c: (gst_filesrc_open_file), * gst/elements/gstfilesrc.c: (gst_filesrc_open_file),

36
tests/old/testsuite/caps/fraction-convert.c
View file

@ -25,11 +25,12 @@
#include <glib.h> #include <glib.h>
static void static void
check_from_double_convert (gdouble value, gint nom, gint denom, gdouble prec) check_from_double_convert (gdouble value, gint num, gint denom, gdouble prec)
{ {
GValue value1 = { 0 }; GValue value1 = { 0 };
GValue value2 = { 0 }; GValue value2 = { 0 };
gdouble check; gdouble check;
gint res_num, res_denom;
g_value_init (&value1, G_TYPE_DOUBLE); g_value_init (&value1, G_TYPE_DOUBLE);
g_value_init (&value2, GST_TYPE_FRACTION); g_value_init (&value2, GST_TYPE_FRACTION);
@ -37,9 +38,20 @@ check_from_double_convert (gdouble value, gint nom, gint denom, gdouble prec)
g_value_set_double (&value1, value); g_value_set_double (&value1, value);
g_value_transform (&value1, &value2); g_value_transform (&value1, &value2);
g_print ("%s = %s ? (expected: %d/%d )\n", g_print ("%s = %s ? (expected: %d/%d )\n",
gst_value_serialize (&value1), gst_value_serialize (&value2), nom, denom); gst_value_serialize (&value1), gst_value_serialize (&value2), num, denom);
g_assert (gst_value_get_fraction_numerator (&value2) == nom);
g_assert (gst_value_get_fraction_denominator (&value2) == denom); res_num = gst_value_get_fraction_numerator (&value2);
res_denom = gst_value_get_fraction_denominator (&value2);
if (res_num == num && res_denom == denom) {
g_print ("best conversion.\n");
} else {
if (fabs (value - res_num / (gdouble) res_denom) <= prec) {
g_print ("acceptable suboptimal conversion.\n");
} else {
g_print ("unacceptable suboptimal conversion.\n");
g_assert_not_reached ();
}
}
g_value_transform (&value2, &value1); g_value_transform (&value2, &value1);
g_print ("%s = %s\n", g_print ("%s = %s\n",
gst_value_serialize (&value2), gst_value_serialize (&value1)); gst_value_serialize (&value2), gst_value_serialize (&value1));
@ -80,15 +92,15 @@ check_from_fraction_convert (gint nom, gint denom, gdouble prec)
static void static void
transform_test (void) transform_test (void)
{ {
check_from_fraction_convert (30000, 1001, 1.0e-12); check_from_fraction_convert (30000, 1001, 1.0e-9);
check_from_fraction_convert (1, G_MAXINT, 1.0e-12); check_from_fraction_convert (1, G_MAXINT, 1.0e-9);
check_from_fraction_convert (G_MAXINT, 1, 1.0e-12); check_from_fraction_convert (G_MAXINT, 1, 1.0e-9);
check_from_double_convert (0.0, 0, 1, 1.0e-12); check_from_double_convert (0.0, 0, 1, 1.0e-9);
check_from_double_convert (1.0, 1, 1, 1.0e-12); check_from_double_convert (1.0, 1, 1, 1.0e-9);
check_from_double_convert (-1.0, -1, 1, 1.0e-12); check_from_double_convert (-1.0, -1, 1, 1.0e-9);
check_from_double_convert (M_PI, 1881244168, 598818617, 1.0e-12); check_from_double_convert (M_PI, 1881244168, 598818617, 1.0e-9);
check_from_double_convert (-M_PI, -1881244168, 598818617, 1.0e-12); check_from_double_convert (-M_PI, -1881244168, 598818617, 1.0e-9);
check_from_double_convert (G_MAXDOUBLE, G_MAXINT, 1, G_MAXDOUBLE); check_from_double_convert (G_MAXDOUBLE, G_MAXINT, 1, G_MAXDOUBLE);
check_from_double_convert (G_MINDOUBLE, 0, 1, G_MAXDOUBLE); check_from_double_convert (G_MINDOUBLE, 0, 1, G_MAXDOUBLE);

36
testsuite/caps/fraction-convert.c
View file

@ -25,11 +25,12 @@
#include <glib.h> #include <glib.h>
static void static void
check_from_double_convert (gdouble value, gint nom, gint denom, gdouble prec) check_from_double_convert (gdouble value, gint num, gint denom, gdouble prec)
{ {
GValue value1 = { 0 }; GValue value1 = { 0 };
GValue value2 = { 0 }; GValue value2 = { 0 };
gdouble check; gdouble check;
gint res_num, res_denom;
g_value_init (&value1, G_TYPE_DOUBLE); g_value_init (&value1, G_TYPE_DOUBLE);
g_value_init (&value2, GST_TYPE_FRACTION); g_value_init (&value2, GST_TYPE_FRACTION);
@ -37,9 +38,20 @@ check_from_double_convert (gdouble value, gint nom, gint denom, gdouble prec)
g_value_set_double (&value1, value); g_value_set_double (&value1, value);
g_value_transform (&value1, &value2); g_value_transform (&value1, &value2);
g_print ("%s = %s ? (expected: %d/%d )\n", g_print ("%s = %s ? (expected: %d/%d )\n",
gst_value_serialize (&value1), gst_value_serialize (&value2), nom, denom); gst_value_serialize (&value1), gst_value_serialize (&value2), num, denom);
g_assert (gst_value_get_fraction_numerator (&value2) == nom);
g_assert (gst_value_get_fraction_denominator (&value2) == denom); res_num = gst_value_get_fraction_numerator (&value2);
res_denom = gst_value_get_fraction_denominator (&value2);
if (res_num == num && res_denom == denom) {
g_print ("best conversion.\n");
} else {
if (fabs (value - res_num / (gdouble) res_denom) <= prec) {
g_print ("acceptable suboptimal conversion.\n");
} else {
g_print ("unacceptable suboptimal conversion.\n");
g_assert_not_reached ();
}
}
g_value_transform (&value2, &value1); g_value_transform (&value2, &value1);
g_print ("%s = %s\n", g_print ("%s = %s\n",
gst_value_serialize (&value2), gst_value_serialize (&value1)); gst_value_serialize (&value2), gst_value_serialize (&value1));
@ -80,15 +92,15 @@ check_from_fraction_convert (gint nom, gint denom, gdouble prec)
static void static void
transform_test (void) transform_test (void)
{ {
check_from_fraction_convert (30000, 1001, 1.0e-12); check_from_fraction_convert (30000, 1001, 1.0e-9);
check_from_fraction_convert (1, G_MAXINT, 1.0e-12); check_from_fraction_convert (1, G_MAXINT, 1.0e-9);
check_from_fraction_convert (G_MAXINT, 1, 1.0e-12); check_from_fraction_convert (G_MAXINT, 1, 1.0e-9);
check_from_double_convert (0.0, 0, 1, 1.0e-12); check_from_double_convert (0.0, 0, 1, 1.0e-9);
check_from_double_convert (1.0, 1, 1, 1.0e-12); check_from_double_convert (1.0, 1, 1, 1.0e-9);
check_from_double_convert (-1.0, -1, 1, 1.0e-12); check_from_double_convert (-1.0, -1, 1, 1.0e-9);
check_from_double_convert (M_PI, 1881244168, 598818617, 1.0e-12); check_from_double_convert (M_PI, 1881244168, 598818617, 1.0e-9);
check_from_double_convert (-M_PI, -1881244168, 598818617, 1.0e-12); check_from_double_convert (-M_PI, -1881244168, 598818617, 1.0e-9);
check_from_double_convert (G_MAXDOUBLE, G_MAXINT, 1, G_MAXDOUBLE); check_from_double_convert (G_MAXDOUBLE, G_MAXINT, 1, G_MAXDOUBLE);
check_from_double_convert (G_MINDOUBLE, 0, 1, G_MAXDOUBLE); check_from_double_convert (G_MINDOUBLE, 0, 1, G_MAXDOUBLE);