Commit graph

277 commits

Author SHA1 Message Date
Ed Page
d21ee513fd test(parser): Verify subcommand required message 2022-11-15 10:12:05 -06:00
Ed Page
e9cbed34cd fix(parser): Don't panic on invalid UTF-8 values
Fixes #4473
2022-11-11 12:26:04 -06:00
Ed Page
45d26e0013 test(parser): Show UTF8 bug 2022-11-11 12:22:30 -06:00
Ed Page
ec03972023 test(assert): Verify empty positional assert exists
Wondered if we had this for #4467.  Figured we should actually test it.
2022-11-09 22:10:37 -06:00
Ed Page
9376a57d40 fix(help): Clarify that 'help' command accepts multiple
Making this plural can go either way as
- Clarify it is plural
- This is all really to simulate actually doing subcommands and you only
  do one at a time

For now, I lean towards clarifying it is plural

I also tweaked the message to be more consistent with how `--help` and
`-h` describe themselves.

Fixes #4342
2022-11-07 23:04:04 -06:00
Ed Page
dfe9e73880 fix(help): Update auto-next-line to use new padding
In clap v4, we changed the padding from 4 to 2 but we didn't update our
calculation for when to switch to next-line-help as it was a magic
number (we tried to catch all of these).

When updating the tests, we also missed that a test was being wrapped
too narrowly

This was found while discussing #3300
2022-11-07 10:36:51 -06:00
Ed Page
45dcf0ed22 fix(parser): Don't make Args exclusive with their ArgGroup
This is most obvious with the derive API as it creates `ArgGroup`s all
over the place now.

Fixes #4396
2022-10-17 19:02:39 -05:00
Ed Page
0f3c98a799 fix(error): Be consistent in puncutation 2022-10-13 13:56:58 -05:00
Ed Page
a4bcba6840 fix(error): Consistently indent suggestions 2022-10-13 12:10:13 -05:00
Ed Page
a964f45b1b fix(error): Properly quote, color subcommand suggestions 2022-10-13 10:51:17 -05:00
Ed Page
1039c61c53 fix(error): Be more consistent in error quoting 2022-10-13 10:45:38 -05:00
Ed Page
f8053fcedb fix(error): Don't suggest '--' when it doesn't help 2022-10-13 10:17:11 -05:00
Ed Page
035571fb5a test(error): Show good/bad trailing suggestions 2022-10-13 09:58:43 -05:00
Ed Page
d0dcaac2ab fix(parser): Only add ArgGroup to ArgMatches for command-line
This will fix `clap_derive`s behavior for optional-flattened groups as
it will properly detect when the group is present (#3566).

While I consider this a bug and not part of compatibility guarentees, I
still want to keep in mind user impact which could still prevent this.
Defaults will make the group always-present which has little value and
if anything is relying on this, it is probably an application bug.
2022-10-12 07:52:07 -05:00
Ed Page
5017f0d720 fxi(error): Don't report unrelated groups in requries errors 2022-09-30 11:18:47 -05:00
Ed Page
e1b02a11cc test: Demonstrate bad requires error 2022-09-30 11:02:44 -05:00
Ed Page
cb04c71105 fix(error): Don't report unrelated groups in conflict errors
Ideally, a conflict caused by a group itself would show the relevant
group but that is less important than showing a valid, non-garbage,
usage
2022-09-30 10:38:27 -05:00
Ed Page
76b7579341 test: Demonstrate bad group error 2022-09-30 10:00:24 -05:00
Ed Page
993b0a88d7 test: Expand group test 2022-09-30 09:53:09 -05:00
Ed Page
4e9f3cca2c fix(error): Specialize the self-conflicts error
Inspired by rust-lang/cargo#11159
2022-09-29 09:54:20 -05:00
Ed Page
2c097814e4 test(error): Self-conflict 2022-09-29 09:43:33 -05:00
Ed Page
5399f49572 fix(error): Quote literals consistently 2022-09-29 08:54:03 -05:00
Ed Page
3683e2c791 fix(parser): Allow one-off self-overrides
bat needed this.

See also #4261
2022-09-28 16:45:35 -05:00
Ed Page
2d7874948f fix(parser): SetFalse should also not allow self-override 2022-09-28 16:26:36 -05:00
Ed Page
cb1cd67009 fix(error): Include failed arg in usage in --flag=bad-value error 2022-09-27 09:25:24 -05:00
Ed Page
12d76d649a fix(error): Include 'Usage:' title in --flag=bad-value error 2022-09-27 09:20:42 -05:00
Ed Page
3a8d2a579b test(parser): Verify existing --flag=bad-value case 2022-09-27 09:19:27 -05:00
Ed Page
9bccded7ed fix(parser): Conflict-with-self is back on by default
See #4261 for more details
2022-09-26 13:29:48 -05:00
Ed Page
3c9bca5ead fix(help)!: Wrapping is behind wrap_help
If users don't want `wrap_help` feature, they can put newlines in where
needed.  Seems odd to support wrapping when the wrap size is fixed.  For
those hard coding the lines, this will save them a decent amount of
size.
2022-09-26 11:45:44 -05:00
Ed Page
4280fdfcbd fix(help): Replace help writers with renderers
The writer is less convenient and isn't offering any performance
benefits of avoidign the extra allocations, so let's render instead.

This supersedes #3874

Fixes #3873
2022-09-22 09:54:19 -05:00
Ed Page
652e71d616 fix(help)!: Provide styled usage to user
This will open us up to providing the user with access to the styled
version in the future.
2022-09-22 09:54:19 -05:00
Ed Page
4674e43493 fix(error): Remove RawFormatter
The likelihood of using this is a lot lower now with `KindFormatter` and
`error-context` feature flag.

People can implement it themselves.
2022-09-20 13:40:18 -05:00
Ed Page
90bcb7f75e fix(error): Use a non-generic Error alias
`clap::Error::raw` was producing ambiguity errors with a default generic
parameter on `clap::error::Error` (which `clap::Error` is a re-export
of).

I tried making `clap::Error` a type alias with a default generic
parameter but that ran into an ambiguity error with `map_err`.

So I'm going ahead and hard coding `clap::Error`.  We don't expect
people to change this all that often.
2022-09-20 13:37:02 -05:00
Ed Page
bfa365a2cc feat(help): 'usage' feature flag for auto-genned usage 2022-09-19 13:15:47 -05:00
Ed Page
7a5dad89ff feat(help): Break out help feature flag
This removes auto-generated help, saving about 50 KiB.
2022-09-19 11:54:10 -05:00
Ed Page
2d83a7b12e feat(error): Break out error-context feature flag
This is a cheap pass at creating this to allow cutting out the cost of
rich error information / programmatic error information.

This cuts about 20 KiB off of the binary.

There is more we could cut out, like collecting of used arguments for
the usage, but I want to keep the conditionals simple.
2022-09-19 10:23:55 -05:00
Ed Page
b4788d51f1 fix(error): Provide a no-context alternative to RichFormatter 2022-09-19 09:48:54 -05:00
Ed Page
bbbaca2ffe perf: Hint to the compiler when once_cell isn't needed 2022-09-16 16:14:32 -05:00
Ed Page
acb0fb7809 fix: Generalize mut_arg, like mut_subcommand
This makes us accept `str` and not do any allocations at the cost of
panicing if unsupported which I think fits our overall story in trying
to catch development-time errors.
2022-09-16 15:31:49 -05:00
Ed Page
0184cf008a fix(parser): Quote the suggested help
We do it elsewhere but here it is only distinguished coloring.

Inspired by #4218
2022-09-15 16:24:59 -05:00
Ed Page
b7d13dfb88 fix(parser): Prefer invalid subcommands over invalid args
Just having `--help` or `--version` can make us get invalid args instead
of invalid subcommands.   It doesn't make sense to do this unless
positionals are used.  Even then it might not make sense but this is at
least a step in the right direction.

Unsure how I feel about this being backported to clap 3.  It most likely
would be fine?

This was noticed while looking into #4218
2022-09-15 10:30:03 -05:00
Ed Page
f68500d9fd fix: Replace Arg::env with Arg::env_os
With `Into<OsStr>`, the separate function isn't needed.
2022-09-13 14:55:12 -05:00
Ed Page
c9eef44213 fix: Make arg!(--flag <value>) optional by default
This was ported over from the usage parser which modeled after docopt.
We just never got around to implementing the rest of the syntax.

However, when considering this as a standalone feature, an
`arg!(--flag <value>)`, outside of other context, should be optional.
This is how the help would display it.

Fixes #4206
2022-09-12 17:10:01 -05:00
Ed Page
c90a4eabae fix(help): Make output more dense
In looking at other help output, I noticed that they use two spaces, in
place of clap's 4, and it doesn't suffer from legibility.  If it
doesn't make the output worse, let's go ahead and make it as dense so we
fit more content on the screen.

This is a part of #4132
2022-09-07 17:13:55 -05:00
Ed Page
56fe50eab8 fix(help): I think this fixes a bug?
The existing behavior is hard to explain, so this is the best I've
figured is going on.
2022-09-07 16:38:33 -05:00
Ed Page
bbb6c38bad fix(help):Be dense on short next line help
If short help is too long for the terminal, clap will automatically
switch to next line help.  As part of next line help for longs, we add a
blank line between args.  This helps make the args clearer when dealing
with multiple paragraphs.  However, its not as much needed for short and
subcommands (always short), so now short matches subcommands.

This was inspired by #3300 and a part of #4132
2022-09-07 14:05:17 -05:00
Ed Page
9a645d2d19 fix(help): Collapse usage to one line
After looking at more examples, I've become more attached to this
briefer format.

Part of #4132
2022-09-07 11:03:57 -05:00
Ed Page
1258f3e5f6 fix: Deprecate Command::allow_negative_numbers
Better to set on individual args
2022-09-07 07:24:42 -05:00
Ed Page
8ad29ef337 feat: Allow specifying negative nums on Arg, like hyphens 2022-09-07 07:18:19 -05:00
Ed Page
d45e4be14b fix(derive): Deprecate Command::allow_hyphen_values
Fixes #3450
2022-09-06 20:50:09 -05:00