
On Thu, Jun 25, 2015 at 12:23:43 -0400, John Ferlan wrote:
On 06/23/2015 01:15 PM, Peter Krempa wrote:
The helper makes sure that strings passed to APIs are non-NULL and non-empty. This allows to drop some inlined checks where it does not make sense. --- src/internal.h | 11 +++++++++++ src/libvirt-domain.c | 4 ++-- src/qemu/qemu_driver.c | 11 ----------- src/util/virerror.h | 11 +++++++++++ 4 files changed, 24 insertions(+), 13 deletions(-)
...
diff --git a/src/util/virerror.h b/src/util/virerror.h index ad3a946..c1a445e 100644 --- a/src/util/virerror.h +++ b/src/util/virerror.h @@ -95,6 +95,17 @@ void virReportSystemErrorFull(int domcode, 0, 0, \ _("%s in %s must not be NULL"), \ #argname, __FUNCTION__) +# define virReportInvalidEmptyStringArg(argname) \ + virRaiseErrorFull(__FILE__, __FUNCTION__, __LINE__, \ + VIR_FROM_THIS, \ + VIR_ERR_INVALID_ARG, \ + VIR_ERR_ERROR, \ + __FUNCTION__, \ + #argname, \ + NULL, \ + 0, 0, \ + _("string %s in %s must not be non empty"), \
This reads strangely... The double negative (not and non)
"string %s in %s must not be empty" or "string %s in %s must be non empty"
I went with the first option. I goofed when I tried to touch up the message after stealing it from an existing macro :)
Or more technically
"string %s in %s cannot start with the NUL character" (yes, correct spelling for NUL - that's the '\0')
Yep, but this option does not sound user-friendly to me.
+ #argname, __FUNCTION__) # define virReportInvalidPositiveArg(argname) \ virRaiseErrorFull(__FILE__, __FUNCTION__, __LINE__, \ VIR_FROM_THIS, \
ACK with some sort of change
Thanks. Peter