Unify cli input parsing #549
Conversation
|
Addresses issue #521 |
There was a problem hiding this comment.
It's obvious but I wanted to point out it's a breaking change for the frontend, so it has to be coordinated with them.
The last thing I think is awkward with our CLIs is that subcommands take a -- (like migration-engine --can-connect-to-database instead of migration-engine can-connect-to-database 'dev.db'), which is kind of unusual.
| use structopt::StructOpt; | ||
|
|
||
| #[derive(Debug, StructOpt, Clone)] | ||
| #[structopt(version = env!("GIT_HASH"))] |
There was a problem hiding this comment.
oh I didn't know that option, it's great!
| stdin: bool, | ||
| /// Whether to ignore warnings from the migration engine regarding data loss. Default: false. | ||
| #[structopt(long = "force")] | ||
| #[structopt(long)] |
There was a problem hiding this comment.
didn't know that either, great!
| fn main() { | ||
| match FmtOpts::from_args() { | ||
| FmtOpts::Lint(opts) => lint::run(opts), | ||
| FmtOpts::Format(opts) => format::run(opts), | ||
| } | ||
| } |
Yeah. But this way we have to provide only one of them. If this would be a struct, you could provide both but then the code would decide the order, like only execute |
|
Yep I meant not making non-subcommands, but removing the |
|
That was a mistake, and I'm adding them back now.
Sent from [ProtonMail](https://protonmail.ch), encrypted email based in Switzerland.
‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
…On Tuesday, March 3, 2020 5:31 PM, Tom Houlé ***@***.***> wrote:
Yep I meant not making non-subcommands, but removing the -- (ideally).
—
You are receiving this because you authored the thread.
Reply to this email directly, [view it on GitHub](#549?email_source=notifications&email_token=AAAIRF7IW74JU5HHUO46KALRFU5FLA5CNFSM4LAO6RG2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOENUNEIY#issuecomment-594072099), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AAAIRFZOWBPOYIWAADU4VZ3RFU5FLANCNFSM4LAO6RGQ).
|
pimeys
force-pushed
the
unified-cli-param-handling
branch
from
c454130
to
d815f89
Compare
|
wait, I meant like the subcommands better without |
pimeys
force-pushed
the
unified-cli-param-handling
branch
from
d815f89
to
ad88c54
Compare
pimeys
force-pushed
the
unified-cli-param-handling
branch
from
ad88c54
to
56404d1
Compare
Yes. Actually. You're right. Let's go this way. |
pimeys
force-pushed
the
unified-cli-param-handling
branch
from
56404d1
to
4ec1451
Compare
- Version command displays the binary name and commit id everywhere - All long cli params are now kebab-case instead of snake_case, the standard - The prisma-fmt has now subcommands `format` and `lint` instead of one boolean flag `lint` that when set lints and when not set formats.
pimeys
force-pushed
the
unified-cli-param-handling
branch
from
4ec1451
to
9ebc450
Compare
formatandlintinstead of oneboolean flag
lintthat when set lints and when not set formats.This requires us to change the vscode plugin and prisma-client and prisma2 cli to support the new options.