On Mon, May 2, 2011 at 2:47 PM, Greg Stein <[email protected]> wrote: > On Mon, May 2, 2011 at 10:52, <[email protected]> wrote: >> Author: hwright >> Date: Mon May 2 14:52:26 2011 >> New Revision: 1098605 >> >> URL: http://svn.apache.org/viewvc?rev=1098605&view=rev >> Log: >> Followup to r1098594: apr_file_printf() doesn't return an error values, it >> returns the number of bytes printed, so just ignore its return value (and >> don't claim it's an error). >> >> * subversion/tests/libsvn_subr/translate-test.c >> (create_file): Ignore return for apr_file_printf(). >> >> Modified: >> subversion/trunk/subversion/tests/libsvn_subr/translate-test.c >> >> Modified: subversion/trunk/subversion/tests/libsvn_subr/translate-test.c >> URL: >> http://svn.apache.org/viewvc/subversion/trunk/subversion/tests/libsvn_subr/translate-test.c?rev=1098605&r1=1098604&r2=1098605&view=diff >> ============================================================================== >> --- subversion/trunk/subversion/tests/libsvn_subr/translate-test.c (original) >> +++ subversion/trunk/subversion/tests/libsvn_subr/translate-test.c Mon May >> 2 14:52:26 2011 >> @@ -237,9 +237,7 @@ create_file(const char *fname, const cha >> { >> const char *this_eol_str = eol_str ? eol_str : random_eol_marker(); >> >> - apr_err = apr_file_printf(f, "%s", lines[i]); >> - if (apr_err) >> - return svn_error_create(apr_err, NULL, fname); >> + apr_file_printf(f, "%s", lines[i]); > > If there is an ignored return value, then I like the form: > (void)apr_file_printf(...) > > It makes it very clear that you *intended* to ignore the return value, > rather than accidentally ignoring it.
In general I agree, but in all my years of C programming, I've very rarely seen that construct applied to printf() and its relatives. Since this is part of those family of functions, I think we can safely leave it as-is. -Hyrum
