Malcolm Nooning wrote:
Hello,
I am trying to debug t/20-pp.t on WindowsXP, Perl 5.8.8,
PAR-Packer-0.970, in a directory that purposely has a space in the fully
qualified path name.
Test t/20-ppt is failing.
I went into t/20-pp.t and commented out "#do "automated_pp_test.pl";"
for brevity. I then added the print statement shown immediately below
the commented out line, as pasted below.
-------------------------------paste code
#do "automated_pp_test.pl";
print STDERR "\nSkipped the entire automated_pp_test.pl\n";
------------------------------end paste code
I then did "nmake test". The last number of lines of output are pasted
below.
------------------paste results
t/10-parl-generation....ok
t/20-pp.................Set up gcc environment - 3.4.2 (mingw-special)
Skipped the entire automated_pp_test.pl
FAILED before any test output arrived
t/30-current_exec.......# Please wait
Set up gcc environment - 3.4.2 (mingw-special)
t/30-current_exec.......ok
1/4 skipped: Can't get running executable that isn't in PATH on
MSWin32
Failed Test Stat Wstat Total Fail Failed List of Failed
-------------------------------------------------------------------------------
t/20-pp.t ?? ?? % ??
1 test and 1 subtest skipped.
Failed 1/4 test scripts, 75.00% okay. 0/35 subtests failed, 100.00% okay.
NMAKE : fatal error U1077: 'C:\perl\bin\perl.exe' : return code '0xff'
Stop.
C:\Documents and Settings\All
Users\Documents\webdownload\perl\par\PAR-Packer-0.970>
------------------end paste results
I do not understand all of the perl mechanisms used in t/20-ppt so I am
here going to just shell out and ask for help on this.
Can someone pick out what might be wrong?
Thanks
Two things.
For one, I only get the "fatal error" if I comment out the
afforementioned "do ... " line like I did. If I uncomment the "do" line
and make automated_pp_test.pl run one of it's tests, it goes away.
That still leaves the line "Can't get running executable that isn't in
PATH on MSWin32". However, the test is not actually failing. The print
out simply says that 1/4 of the test is being skipped, and letting me
know the reason why.
In the end, with the two prior reported bug fixes, all seems well on
Windows. I am simply going to drop this issue.
Thanks
