On Mon, Jul 25, 2022 at 04:38:48PM +0200, Ævar Arnfjörð Bjarmason wrote:
>
> On Mon, Jul 25 2022, SZEDER Gábor wrote:
>
> > In 't0040-parse-options.sh' we thoroughly test the parsing of all
> > types and forms of options, but in all those tests parse_options() is
> > always invoked with a 0 flags parameter.
> >
> > Add a few tests to demonstrate how various 'enum parse_opt_flags'
> > values are supposed to influence option parsing.
> >
> > Signed-off-by: SZEDER Gábor <szeder.dev@xxxxxxxxx>
> > ---
> > t/helper/test-parse-options.c | 61 ++++++++++++++++++++++++++++++
> > t/helper/test-tool.c | 1 +
> > t/helper/test-tool.h | 1 +
> > t/t0040-parse-options.sh | 70 +++++++++++++++++++++++++++++++++++
> > 4 files changed, 133 insertions(+)
> >
> > diff --git a/t/helper/test-parse-options.c b/t/helper/test-parse-options.c
> > index 48d3cf6692..32b906bd6a 100644
> > --- a/t/helper/test-parse-options.c
> > +++ b/t/helper/test-parse-options.c
> > @@ -192,3 +192,64 @@ int cmd__parse_options(int argc, const char **argv)
> >
> > return ret;
> > }
> > +
> > +static int parse_options_flags__cmd(int argc, const char **argv,
> > + enum parse_opt_flags test_flags)
> > +{
> > + const char *usage[] = {
> > + "<...> cmd [options]",
> > + NULL
> > + };
> > + int opt = 0;
> > + const struct option options[] = {
> > + OPT_INTEGER('o', "opt", &opt, "an integer option"),
> > + OPT_END()
> > + };
> > +
> > + argc = parse_options(argc, argv, NULL, options, usage, test_flags);
> > +
> > + printf("opt: %d\n", opt);
> > + for (int i = 0; i < argc; i++)
> > + printf("arg %02d: %s\n", i, argv[i]);
> > +
> > + return 0;
> > +}
> > +
> > +static enum parse_opt_flags test_flags = 0;
> > +static const struct option test_flag_options[] = {
> > + OPT_GROUP("flag-options:"),
> > + OPT_BIT(0, "keep-dashdash", &test_flags,
> > + "pass PARSE_OPT_KEEP_DASHDASH to parse_options()",
> > + PARSE_OPT_KEEP_DASHDASH),
> > + OPT_BIT(0, "stop-at-non-option", &test_flags,
> > + "pass PARSE_OPT_STOP_AT_NON_OPTION to parse_options()",
> > + PARSE_OPT_STOP_AT_NON_OPTION),
> > + OPT_BIT(0, "keep-argv0", &test_flags,
> > + "pass PARSE_OPT_KEEP_ARGV0 to parse_options()",
> > + PARSE_OPT_KEEP_ARGV0),
> > + OPT_BIT(0, "keep-unknown", &test_flags,
> > + "pass PARSE_OPT_KEEP_UNKNOWN to parse_options()",
> > + PARSE_OPT_KEEP_UNKNOWN),
> > + OPT_BIT(0, "no-internal-help", &test_flags,
> > + "pass PARSE_OPT_NO_INTERNAL_HELP to parse_options()",
> > + PARSE_OPT_NO_INTERNAL_HELP),
> > + OPT_END()
> > +};
> > +
> > +int cmd__parse_options_flags(int argc, const char **argv)
> > +{
> > + const char *usage[] = {
> > + "test-tool parse-options-flags [flag-options] cmd [options]",
> > + NULL
> > + };
> > +
> > + argc = parse_options(argc, argv, NULL, test_flag_options, usage,
> > + PARSE_OPT_STOP_AT_NON_OPTION);
> > +
> > + if (argc == 0 || strcmp(argv[0], "cmd")) {
> > + error("'cmd' is mandatory");
> > + usage_with_options(usage, test_flag_options);
> > + }
> > +
> > + return parse_options_flags__cmd(argc, argv, test_flags);
> > +}
> > diff --git a/t/helper/test-tool.c b/t/helper/test-tool.c
> > index 318fdbab0c..6e62282b60 100644
> > --- a/t/helper/test-tool.c
> > +++ b/t/helper/test-tool.c
> > @@ -51,6 +51,7 @@ static struct test_cmd cmds[] = {
> > { "online-cpus", cmd__online_cpus },
> > { "pack-mtimes", cmd__pack_mtimes },
> > { "parse-options", cmd__parse_options },
> > + { "parse-options-flags", cmd__parse_options_flags },
> > { "parse-pathspec-file", cmd__parse_pathspec_file },
> > { "partial-clone", cmd__partial_clone },
> > { "path-utils", cmd__path_utils },
>
> I wanted to add some new parse_options() code to
> t/helper/test-parse-options.c in the past, but was stymied by its
> cmd_*() going through a singular parse_options().
>
> So just creating a new callback is a neat solution.
>
> But wouldn't it be better to just create a new
> t/helper/test-parse-options-something-.c & test file? It seems this
> doesn't really share anything with the current helper & tests...
Well, at least they share the concept, as they all test parse-options.
And 'parse-options-flags' and 'parse-subcommands' do share the options
array to specify the various parse_opt_flags.
A new test script for these flags and/or for subcommands would
definitely be worse, IMO. I've found it very convenient that whenever
I updated 'parse-options.{c,h}', I only needed to run a single
'./t0040-parse-options.sh' script to check my changes.
