keithn/seatest

Seatest - Simple C based Unit Testing

seatest's Issues

strings.h file

Hi,

In the latest code, if WIN32 is not defined you include "strings.h" (line 12 of seatest.c). My compiler did not come with a "strings.h" file. It only has the standard library "string.h". Should "strings.h" be another include file in seatest repository?

I'm using the evaluation version of the Keil Arm compiler.

Marc

I would check everything Jia contributor did here in the eyes of the xz compromise

duplicate symbol _seatest_simple_test_result

in seatest.hpp 29 ，seatest_simple_test_result should defined more safer like this:

extern void (seatest_simple_test_result)(int passed, char reason, const char* function, unsigned int line);

because it's a global varible, and the real definition is in seatest.c 。 if use seatest in c++, every file who include the seatest.h, will define the seatest_simple_test_result again。and will result the compile error like this:

duplicate symbol _seatest_simple_test_result in:
/var/folders/wy/br32jzv15nqgy1xvn05r4gc80000gn/T/seatest-57864d.o
/var/folders/wy/br32jzv15nqgy1xvn05r4gc80000gn/T/test_main-339523.o

Multiple compilation warnings on assert_string_*

I'm using the svn version on Slackware 13.37 and get this warning:
envvar_tests.c:59:2: warning: passing argument 2 of 
‘seatest_assert_string_equal’ discards qualifiers from pointer target type
../seatest/seatest.h:33:6: note: expected ‘char *’ but argument is of type 
‘const char *’

I propose to change function prototypes from
void seatest_assert_string_equal(char* expected, char* actual, const char* 
function, unsigned int line);
to
void seatest_assert_string_equal(const char* expected, const char* actual, 
const char* function, unsigned int line);

Original issue reported on code.google.com by [email protected] on 17 Feb 2012 at 6:21

fatal error: conio.h: No such file or directory

When compiling my app with seatest, I get the following compiler error (using 
gcc (Ubuntu/Linaro 4.5.2-8ubuntu4) 4.5.2):

src/seatest/seatest.c:2:19: fatal error: conio.h: No such file or directory
compilation terminated.

I have attached a patch that corrects this problem, as well as a few other 
compiler warnings.

N.B. I have only tested that the patched seatest compiles on my dev machine, 
YMMV :-)

Original issue reported on code.google.com by [email protected] on 26 May 2011 at 11:19

Attachments:

• seatest.patch

Compile warning message

When I use seatest on Ubuntu Linux 9.04 with gcc version 4.4.1 (Ubuntu 
4.4.1-4ubuntu8).

GCC will display a warning message as follow
   expected ‘char *’ but argument is of type ‘const char *’

I think the each function prototype should be change form 
void seatest_simple_test_result(int passed, char* reason, const char* function, 
unsigned int line);
to 
void seatest_simple_test_result(int passed, char* reason, char* function, 
unsigned int line);

Add the "const" keyword for the __FUNCTION__ const 

thank you

Original issue reported on code.google.com by [email protected] on 1 Oct 2010 at 6:47

assert_string_equal() segfaults on null strings

If I include the following test in the suite I'm building for seatest itself:

assert_test_passes(assert_string_equal((char *)0, (char *)0));

... then I get a segfault, presumably from strcmp().

That behaviour is correct for strcmp, but should SeaTest really error out in 
this case?

I _think_ the right behaviour would be to consider null a value in its own 
right.  The logic could be:

 1. fail if one and only one string is null
 2. pass if both strings are null
 3. run a strcmp if neither string is null

Does that sound sensible?

Original issue reported on code.google.com by [email protected] on 29 Jun 2011 at 10:58

*NIX build support

Currently, Seatest has a Visual Studio project to demonstrate usage.

I'd like to put together a patch including:

 - autoconf setup & Makefiles

 - any changes necessary to compile demo project as ISO C w/o warnings

... basically everything necessary to build Seatest on *NIX as well as Windows. 


I can test it on Linux, Cygwin & possibly Mac OS as well (may  have to scrounge 
time on a Mac in exchange for beer).

Let me know if this is something you would like.

Original issue reported on code.google.com by [email protected] on 31 May 2011 at 12:02

String quotation

This patch adds quotes around expected and actual strings in assert_string_* 
functions. It's needed to make whitespace characters at the beginning and end 
of strings visible.

P.S. Thanks for seatest.

Original issue reported on code.google.com by [email protected] on 22 Jun 2011 at 7:27

Attachments:

• patch

assert_string_equal always gives true

Hi - 

I just pulled down v0.5 from the zip, and am looking at the example projects 
(which are very helpful, btw). 

It looks like the assert_string_equal assertion is always passing. Here's what 
I did:

I took the example/nix test_suite_one file, and changed the first test from:
void test_strings_equal()
{
    char *s = "hello";
    assert_string_equal("hello", s);
}

To:
{
    char *s = "hello";
    assert_string_equal("foo", s);
}

Or even:

{
    char *s = "hello";
    assert_string_equal("foo", "bar");
}

That test will always pass.

I ran make clean build, then ran ./example, and my output is below (notice the 
first suite shows 3 run with 0 failures):

---------------- test_suite_one.c ----------------
                 3 run  0 failed                  

---------------- test_suite_two.c ----------------
test_assert_fails    Line 12    Should of been true
                 2 run  1 failed                  



==================SEATEST v0.5====================

                      Failed
                 5 tests run
                    in 0 ms
==================================================

Original issue reported on code.google.com by [email protected] on 25 Feb 2014 at 10:15

Fix return value of run_tests

The way run_tests is currently implemented leads to problems with the exit 
status of the test run.

The line

return sea_tests_failed == 0;

returns 1 if the comparison succeeds. However, the exit status 1 stands for 
failure.

I would propose the following fix:

if (sea_tests_failed == 0) {
  return 0;
else {
  return -1;
}

Original issue reported on code.google.com by [email protected] on 17 Aug 2014 at 4:56

"should of" -> "should have"

Just a minor grammar glitch but it stands out every time I run my tests because 
I'm a grammar pedant :-).  

Patch attached.

Original issue reported on code.google.com by [email protected] on 30 May 2011 at 10:34

Attachments:

• seatest.patch

Return 0 in Seatest Hello World

Hi, there,
In *nix convention, return value 0 means success. But in the Seatest Hello World, it return 1 when success, which is a little surprise to caller. Is any better choice for Seatest Hello World other than return 1?

Corrected text

For:

void seatest_assert_true(int test, const char* function, unsigned int line)
{
    seatest_simple_test_result(test, "Should have been true", function, line);

}

void seatest_assert_false(int test, const char* function, unsigned int line)
{
    seatest_simple_test_result(!test, "Should have been false", function, line);
}

The text was corrected. A minor detail ;)

Thanks a lot for this testing library, it's very useful !

Original issue reported on code.google.com by [email protected] on 25 Nov 2013 at 7:16

Buffer overflow on string comparisons

strings when asserted, if they are over 100 characters will cause grief.

*rolls eyes* such a noob....

Original issue reported on code.google.com by [email protected] on 13 Jul 2010 at 4:40

fatal error

main.cpp:2:18: fatal error: conio.h: No such file or directory
compilation terminated.

Original issue reported on code.google.com by [email protected] on 11 Sep 2013 at 6:51