You can invoke an arbitrary shell script when your program exits or otherwise terminates by using the -run-at-exit option. For example, you can use a script to move or remove log files or view files, to note test failure, or to interact with PureCoverage data.
In addition to the %V, %v, and %p conversion characters, you can use these conversion characters:
|
Character |
Converts to |
|
%z |
The string value "true" or "false" indicating whether any call chains were printed, for example, in error or leak reports |
|
%x |
The program's exit status (0 if the program did not call exit) |
|
%e |
The number of distinct access errors printed |
|
%E |
The total number of error occurrences noted |
|
%l |
The number of bytes of memory leaked |
|
%L |
The number of bytes of memory potentially leaked |
Use the %z sequence to have your exit script act conditionally when Purify finds something of interest to you. If a call chain is printed in an access error, a memory leak, or as a result of calling purify_printf_with_call_chain(), the string is true; otherwise, it is false.
For example, if you set the option:
setenv PURIFYOPTIONS '-run-at-exit="if %z ; then \
echo \"%v: %e errors, %l+%L bytes leaked.\"
; fi"'
When your program exits, you might see on stdout:
testprog: 2 errors, 1+10 bytes leaked.
If your program is running in the Viewer, Purify sends any command output to the X Window where you started the Viewer.