On Sat, 18 Oct 2014, David Bremner <david@tethera.net> wrote: > Most of the existing tests for pre/post-new hook don't seem to apply. > --- > test/T400-hooks.sh | 18 ++++++++++++++++++ > 1 file changed, 18 insertions(+) > > diff --git a/test/T400-hooks.sh b/test/T400-hooks.sh > index 77e8569..e741211 100755 > --- a/test/T400-hooks.sh > +++ b/test/T400-hooks.sh > @@ -30,6 +30,8 @@ rm_hooks () { > > # add a message to generate mail dir and database > add_message > +# create maildir structure for notmuch-insert > +mkdir -p "$MAIL_DIR"/{cur,new,tmp} > > test_begin_subtest "pre-new is run" > rm_hooks > @@ -45,6 +47,16 @@ create_echo_hook "post-new" expected output > notmuch new > /dev/null > test_expect_equal_file expected output > > +test_begin_subtest "post-insert hook is run" > +rm_hooks > +generate_message > +create_echo_hook "post-insert" expected output > +echo $gen_msg_filename > +cat output I presume the two lines above... > +notmuch insert < "$gen_msg_filename" > +echo $? ...and this line are leftover debug messages? Otherwise LGTM. I guess this could be expanded with a subtest checking that the hook does not get run on errors, and does get run with some errors that are ignored with --keep. BR, Jani. > +test_expect_equal_file expected output > + > test_begin_subtest "pre-new is run before post-new" > rm_hooks > generate_message > @@ -82,6 +94,12 @@ test_expect_equal_file expected output > # depends on the previous subtest leaving broken hook behind > test_expect_code 1 "post-new non-zero exit status (notmuch status)" "notmuch new" > > +rm_hooks > +generate_message > +create_failing_hook "post-insert" > +test_expect_success "post-insert hook does not affect insert status" \ > + "notmuch insert < \"$gen_msg_filename\" > /dev/null" > + > # test_begin_subtest "hook without executable permissions" > rm_hooks > mkdir -p ${HOOK_DIR} > -- > 2.1.1 > > _______________________________________________ > notmuch mailing list > notmuch@notmuchmail.org > http://notmuchmail.org/mailman/listinfo/notmuch