refactor(bin/oli): use clap_derive to reduce boilerplate code

[koushiro]

Which issue does this PR close?

Closes #.

Rationale for this change

reduce boilerplate code

What changes are included in this PR?

use clap_derive (instead of clap_builder) to reduce boilerplate code

Are there any user-facing changes?

No

[koushiro]

oli's tests are not currently checked in CI. Do we need to add them to CI?

➜  oli git:(use-clap-derive) ✗ cargo test --release --all-features
   Compiling oli v0.41.12 (/Users/qinxuan/Code/rust-lang/opendal/bin/oli)
    Finished `release` profile [optimized] target(s) in 37.08s
     Running unittests src/lib.rs (target/release/deps/oli-c04c0e204779fc76)

running 7 tests
test config::tests::test_load_from_env ... ok
test config::tests::test_parse_s3_location3 ... ok
test config::tests::test_parse_fs_location ... ok
test config::tests::test_load_from_toml ... ok
test config::tests::test_load_config_from_file_and_env ... ok
test config::tests::test_parse_s3_location ... ok
test config::tests::test_parse_s3_location2 ... ok

test result: ok. 7 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running unittests src/bin/oli.rs (target/release/deps/oli-139e3cb8d6e37d4c)

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

     Running tests/cat.rs (target/release/deps/cat-7fa8be95eddce583)

running 2 tests
test test_cat_for_path_in_current_dir ... ok
test test_basic_cat ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.51s

     Running tests/cp.rs (target/release/deps/cp-f883dfcf27676c1a)

running 2 tests
test test_basic_cp ... ok
test test_cp_for_path_in_current_dir ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.03s

     Running tests/ls.rs (target/release/deps/ls-c1a38755bc9e6f86)

running 1 test
test test_basic_ls ... ok

test result: ok. 1 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/rm.rs (target/release/deps/rm-071bb9fe37b27c4c)

running 2 tests
test test_basic_rm ... ok
test test_rm_for_path_in_current_dir ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

     Running tests/stat.rs (target/release/deps/stat-ca4edaf8b3ac28e0)

running 2 tests
test test_stat_for_path_in_current_dir ... ok
test test_basic_stat ... ok

test result: ok. 2 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.01s

   Doc-tests oli

running 0 tests

test result: ok. 0 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.00s

[Xuanwo]

Hi, I implemented this to support the use of a proxy as follows:

ln oli ocp
ocp <src> <dst>

ln oli ols
ols <src>

Users can use ocp in exactly the same way as oli cp.

After this change, can we still operate in this manner?

[koushiro]

Hi, I implemented this to support the use of a proxy as follows:

ln oli ocp
ocp <src> <dst>

ln oli ols
ols <src>

Users can use ocp in exactly the same way as oli cp.

After this change, can we still operate in this manner?

No, but I can try to support this manner.
Do you think the other parts are ok?

[Xuanwo]

Do you think the other parts are ok?

Other parts looks good to me, thank you for working on this!

[koushiro]

Hi, I implemented this to support the use of a proxy as follows:

ln oli ocp
ocp <src> <dst>

ln oli ols
ols <src>

Users can use ocp in exactly the same way as oli cp.

I encountered two problems:

1. The previous (before this PR) implementation had the following problem:

➜  oli git:(main) ln target/release/oli ols
➜  oli git:(main) ✗ ./ols src
error: 'ols' requires a subcommand but one was not provided

Usage: ols [OPTIONS] <target>

For more information, try '--help'.

2. If we use clap_derive and want to support the proxy manner that you mentioned, then we cannot use the global default config arg, and we need to add a default config for each cmd, can you accept this implementation?

./target/release/oli -h
OpenDAL Command Line Interface

Usage: oli <COMMAND>

Commands:
  cat   Display object content
  cp    Copy object
  ls    List object
  rm    Remove object
  stat  Show object metadata
  help  Print this message or the help of the given subcommand(s)

Options:
  -h, --help     Print help
  -V, --version  Print version

➜  oli git:(use-clap-derive) ✗ ln target/release/oli ols
➜  oli git:(use-clap-derive) ✗ ./ols -h
List object

Usage: ols [OPTIONS] <TARGET>

Arguments:
  <TARGET>

Options:
      --config <CONFIG>  Path to the config file [default: "/Users/qinxuan/Library/Application Support/oli/config.toml"]
  -r, --recursive        List objects recursively
  -h, --help             Print help
➜  oli git:(use-clap-derive) ✗ ./ols src
src/

➜  oli git:(use-clap-derive) ✗ ln target/release/oli ocat
➜  oli git:(use-clap-derive) ✗ ./ocat -h
Display object content

Usage: ocat [OPTIONS] <TARGET>

Arguments:
  <TARGET>

Options:
      --config <CONFIG>  Path to the config file [default: "/Users/qinxuan/Library/Application Support/oli/config.toml"]
  -h, --help             Print help
➜  oli git:(use-clap-derive) ✗ ./ocat DEPENDENCIES.md
# Dependencies

Refer to [DEPENDENCIES.rust.tsv](DEPENDENCIES.rust.tsv) for full list.

[koushiro]

I add a new type ConfigParams in b7b0642 to reduce the same argument declaration, or you think it's not necessary?

[Xuanwo]

2. If we use clap_derive and want to support the proxy manner that you mentioned, then we cannot use the global default config arg, and we need to add a default config for each cmd, can you accept this implementation?

Hi, that's look to me.

[Xuanwo]

I add a new type ConfigArgs (or ConfigParams?) in 3808562 to reduce the same argument declaration, or you think it's not necessary?

Make sense to me.

[koushiro]

@Xuanwo All done, PTAL
BTW, do we need to add oli's tests to CI or just keep it.

[Xuanwo]

BTW, do we need to add oli's tests to CI or just keep it.

Including an oli test in the CI would be appreciated.

[Xuanwo]

Thank you @koushiro for this!