Re: [PATCH 1/6] reg.exe/tests: Parameterise error messages

[Stefan Dösinger]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Am 2014-10-20 22:24, schrieb Jonathan Vollebregt:
> -    ok(type == REG_DWORD, "got wrong type %d, expected %d\n", type, REG_DWORD);
> -    ok(size == sizeof(DWORD), "got wrong size %d, expected %d\n", size, (int)sizeof(DWORD));
> +    ok(type == REG_DWORD, "got wrong type %u, expected %u\n", type, REG_DWORD);
> +    ok(size == sizeof(DWORD), "got wrong size %u, expected %u\n", size, (DWORD)sizeof(DWORD));
> I don't know if this is the canonical way to handle this. My gut feeling is that a cast to unsigned int would be nicer than a cast of DWORD, but I hope others can give a better answer.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
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=3Os0
-----END PGP SIGNATURE-----

[PATCH 1/6] reg.exe/tests: Parameterise error messages

[Henri Verbeet]

On 21 October 2014 11:07, Stefan Dösinger <stefandoesinger(a)gmail.com> wrote:
> Am 2014-10-20 22:24, schrieb Jonathan Vollebregt:
> > -    ok(type == REG_DWORD, "got wrong type %d, expected %d\n", type, REG_DWORD);
> > -    ok(size == sizeof(DWORD), "got wrong size %d, expected %d\n", size, (int)sizeof(DWORD));
> > +    ok(type == REG_DWORD, "got wrong type %u, expected %u\n", type, REG_DWORD);
> > +    ok(size == sizeof(DWORD), "got wrong size %u, expected %u\n", size, (DWORD)sizeof(DWORD));
> > I don't know if this is the canonical way to handle this. My gut feeling is that a cast to unsigned int would be nicer than a cast of DWORD, but I hope others can give a better answer.

I'd just go with
+    ok(size == sizeof(DWORD), "Got unexpected size %u.\n", size);
The only tests that really need to print the expected value are ones
where it's not constant, and those should be fairly rare.