gstreamer/tests/old/testsuite/test_gst_init.c

#include <gst/gst.h>

/* This tests that gst_init() doesn't segfault when passed two NULLs as
 * parameters, and that it doesn't fail if gst_init() happens to get called
 * a second time. */
int
main (int argc, char *argv[])
{
  gst_init(NULL, NULL);
  gst_init(&argc, &argv);

  return 0;
}
Added a testsuite. This currently only has one test, which replicates a problem just submitted to the list. Original commit message from CVS: Added a testsuite. This currently only has one test, which replicates a problem just submitted to the list. I hope noone disagrees that we need a testsuite. Also, hopefully we all think that a tiny testsuite which is easy to add tests to is much better than a huge project to create a complete testsuite, which never gets started because its too big and scary. I suggest that tests get created whenever someone has a spare moment, or whenever a bug is reported (to replicate the problem, and thus be sure that when fixed it stays fixed), and before making a change to a piece of code to ensure that it remains working. If we all add little tests occasionally, then we'll have a good useful testsuite before we know it. All you need to do is make a 2001-03-09 12:08:38 +00:00			`#include <gst/gst.h>`

Updated test to have a comment explaining its purpose (very important this, especially for regression tests), and to ... Original commit message from CVS: Updated test to have a comment explaining its purpose (very important this, especially for regression tests), and to actually test what I meant it to. 2001-03-09 12:29:51 +00:00			`/* This tests that gst_init() doesn't segfault when passed two NULLs as`
			`* parameters, and that it doesn't fail if gst_init() happens to get called`
			`* a second time. */`
Added a testsuite. This currently only has one test, which replicates a problem just submitted to the list. Original commit message from CVS: Added a testsuite. This currently only has one test, which replicates a problem just submitted to the list. I hope noone disagrees that we need a testsuite. Also, hopefully we all think that a tiny testsuite which is easy to add tests to is much better than a huge project to create a complete testsuite, which never gets started because its too big and scary. I suggest that tests get created whenever someone has a spare moment, or whenever a bug is reported (to replicate the problem, and thus be sure that when fixed it stays fixed), and before making a change to a piece of code to ensure that it remains working. If we all add little tests occasionally, then we'll have a good useful testsuite before we know it. All you need to do is make a 2001-03-09 12:08:38 +00:00			`int`
			`main (int argc, char *argv[])`
			`{`
Updated test to have a comment explaining its purpose (very important this, especially for regression tests), and to ... Original commit message from CVS: Updated test to have a comment explaining its purpose (very important this, especially for regression tests), and to actually test what I meant it to. 2001-03-09 12:29:51 +00:00			`gst_init(NULL, NULL);`
			`gst_init(&argc, &argv);`

			`return 0;`
Added a testsuite. This currently only has one test, which replicates a problem just submitted to the list. Original commit message from CVS: Added a testsuite. This currently only has one test, which replicates a problem just submitted to the list. I hope noone disagrees that we need a testsuite. Also, hopefully we all think that a tiny testsuite which is easy to add tests to is much better than a huge project to create a complete testsuite, which never gets started because its too big and scary. I suggest that tests get created whenever someone has a spare moment, or whenever a bug is reported (to replicate the problem, and thus be sure that when fixed it stays fixed), and before making a change to a piece of code to ensure that it remains working. If we all add little tests occasionally, then we'll have a good useful testsuite before we know it. All you need to do is make a 2001-03-09 12:08:38 +00:00			`}`