flag to exit cmd2/stop the loop on command failure

[teto]

Follow up of #684.

This is some - late (sorry) - insight on #685. I've been trying to leverage cmd2 to help me with testing so here is my process: I have 2 scripts.
generate_transcripts et run_tests. run_tests fails as soon as a test fail, perfect !
Similarly, I would like generate_transcripts to exit when it fails to generate a transcript.

# Make it fail if any call fails
set -o errexit
function gen_transcript()
{
	suffix="$1"
	mptcpanalyzer "load tests/script_${suffix}.txt -t tests/trans_${suffix}.txt" "quit"
}

gen_transcript "tcp"
gen_transcript "mptcp"
...

E.g., I would like to have my program to stop with a non 0 exit code when "load script.txt -t text.txt" fails.
While the PR #685 targeted one command, I feel like there is the possibility to abstract the mechanism to apply it more generally. Like a flag --exit-on-error.
I don't know about the API, if it should be added per-command:

    @with_argparser(parser)
    @exit_on_error
    def do_mycommand

or as a global flag:
my_cmd2_program --exit-on-error "load test.txt -t test.txt"
What do you think ?

[tleonhardt]

@kmvanbrunt I think this request relates to some of the changes you were thinking of making.

[kmvanbrunt]

This should be fixed by #687. Neither _generate_transcript() or _test_transcript() stop when onecmd_plus_hooks() says to.

[teto]

@kmvanbrunt how does it translate as a cmd2 API user ? does generating a transcript always exit cmd2 ? or is it an option ? should I override onecmd_plus_hooks ?

[kmvanbrunt]

Generating a transcript will stop early if any of the commands it is running returns True for stop. So as long as the command that fails sets exit_code and returns True, then you should get what you need when the app exits.

This line of code you use to run tests should work with no changes:

mptcpanalyzer "load tests/script_${suffix}.txt -t tests/trans_${suffix}.txt" "quit"

[kmvanbrunt]

Closed by #696