$ cargo pretty-test
thread 'main' panicked at /home/josecelano/.cargo/registry/src/index.crates.io-6f17d22bba15001f/cargo-pretty-test-0.2.0/src/parsing.rs:430:5:
assertion `left == right` failed: the parsed amount of running tests [265, 0, 96, 31] should equal to the number in stats.total [265, 0, 96, 0]
  left: [265, 0, 96, 31]
 right: [265, 0, 96, 0]
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

The whole cargo test output I'm using:

cargo_test_ouput.txt

Accepts one of always,never,auto. And support environment variable CARGO_TERM_COLOR from cargo

e.g.

--color=never

--color never

CARGO_TERM_COLOR=never

I'm trying t use the command in a GitHub workflow but I get this error:

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

thread 'main' panicked at 'assertion failed: `(left == right)`
  left: `0`,
 right: `2`: the amount of test runners from stderr should equal to that from stdout', /home/runner/.cargo/registry/src/index.crates.io-6f17d22bba15001f/cargo-pretty-test-0.2.0/src/parsing.rs:41:5
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Error: Process completed with exit code 101.

The workflow: https://github.com/torrust/torrust-tracker/actions/runs/6299630094/job/17100743480

The cargo test output:

cargo_test_output.txt

I copied the code from https://www.rustexplorer.com/b/i058g3 in the "first commit". I just realized that the solution here was different:

fn main() {
    let output = Command::new("cargo").arg("test").output().unwrap();
    let text = String::from_utf8_lossy(&output.stdout);

    let re = Regex::new(r"(?m)^test \S+ \.\.\. \S+$").unwrap();
    println!(
        "{}",
        pretty_test(re.find_iter(&text).map(|m| m.as_str())).unwrap()
    );
}

Instead of capturing the input from the standard input, it runs the cargo test command. I think some people could use different commands, so I guess capturing the input is more generic.

The current solution is not good because the input cannot be empty.

• Install: cargo install cargo-pretty-test

• Use --color=always to force print in color: e.g. cargo pretty-test --color=always (demo)

• Use --color=never (to prevent escape chars) and $GITHUB_STEP_SUMMARY to print testing result in workflow summary (demo) Note: $GITHUB_STEP_SUMMARY will take the input as GFM (a markdown format), so you should put the result in text mode, like

  echo -n "<pre>" >> $GITHUB_STEP_SUMMARY
  echo -n "$(cargo pretty-test --color=never)" >> $GITHUB_STEP_SUMMARY
  echo -n "</pre>" >> $GITHUB_STEP_SUMMARY
  
  # or
  echo '```text' >> $GITHUB_STEP_SUMMARY
  echo "$(cargo pretty-test --color=never)" >> $GITHUB_STEP_SUMMARY
  echo '```' >> $GITHUB_STEP_SUMMARY
  

  

fix https://github.com/josecelano/cargo-pretty-test/issues/13 and suppress https://github.com/josecelano/cargo-pretty-test/pull/14

With this PR, you can use any of these commands to run tests:

• cargo test --features no-color
• cargo test -F no-color
• cargo t
• cargo test --all-features (which CI does)

$ cargo r -q
Generated by cargo-pretty-test
├── cargo_pretty_test
│   ├── tests/golden_master_test.rs
│   │   └─ ✅ golden_master_test
│   ├── tests/mocking_project.rs
│   │   ├─ ✅ snapshot_testing_for_parsed_output
│   │   └─ ✅ snapshot_testing_for_pretty_output
│   └── tests/parsing.rs
│       └─ ✅ parse_stderr_stdout
└── Doc
    └── cargo-pretty-test
        └─ ✅ src/doc.rs - doc (line 3)

$ cargo r -q -- --workspace --no-fail-fast
failures:

---- submod::panic::panicked stdout ----
thread 'submod::panic::panicked' panicked at tests/integration/src/lib.rs:9:13:
explicit panic
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

---- submod::panic::should_panic_but_didnt stdout ----
note: test did not panic as expected

failures:
    submod::panic::panicked
    submod::panic::should_panic_but_didnt

*************************************************************

Generated by cargo-pretty-test
├── cargo_pretty_test
│   ├── tests/golden_master_test.rs
│   │   └─ ✅ golden_master_test
│   ├── tests/mocking_project.rs
│   │   ├─ ✅ snapshot_testing_for_parsed_output
│   │   └─ ✅ snapshot_testing_for_pretty_output
│   └── tests/parsing.rs
│       └─ ✅ parse_stderr_stdout
├── integration
│   ├── src/lib.rs
│   │   ├── submod
│   │   │   ├─ 🔕 ignore
│   │   │   ├─ 🔕 ignore_without_reason
│   │   │   ├─ ✅ normal_test
│   │   │   └── panic
│   │   │       ├─ ❌ panicked
│   │   │       ├─ ✅ should_panic - should panic
│   │   │       ├─ ❌ should_panic_but_didnt - should panic
│   │   │       └─ ✅ should_panic_without_reanson - should panic
│   │   └─ ✅ works
│   ├── src/main.rs
│   │   └─ ✅ from_main_rs
│   └── tests/parsing.rs
│       └─ ✅ from_integration
└── Doc
    ├── cargo-pretty-test
    │   └─ ✅ src/doc.rs - doc (line 3)
    └── integration
        └─ ✅ tests/integration/src/lib.rs - doc (line 41)

Bumps regex-lite from 0.1.0 to 0.1.2.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Bumps indexmap from 2.0.1 to 2.0.2.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

close https://github.com/josecelano/cargo-pretty-test/issues/24

(unfortunately no color)

e.g. invalid argument

✗ cargo pretty-test --w
Error from cargo-pretty-test:
"" should contain `running (?P<amount>\d+) tests?` pattern

Error from cargo test:
error: unexpected argument '--w' found

  tip: a similar argument exists: '--workspace'

Usage: cargo test --workspace [TESTNAME] [-- [args]...]

For more information, try '--help'.

e.g. error in source code

✗ cargo pretty-test a
Error from cargo-pretty-test:
"" should contain `running (?P<amount>\d+) tests?` pattern

Error from cargo test:
   Compiling cargo-pretty-test v0.2.3 (/rust/tmp/urlo/pretty-test)
error[E0308]: mismatched types
  --> tests/mocking_project.rs:33:29
   |
33 |         Cache { raw_output, info }
   |                             ^^^^ expected `Vec<TestInfo<'_>>`, found `Result<Vec<TestInfo<'_>>, String>`
   |
   = note: expected struct `Vec<TestInfo<'static>>`
                found enum `Result<Vec<TestInfo<'_>>, std::string::String>`

For more information about this error, try `rustc --explain E0308`. error: could not compile `cargo-pretty-test` (test "mocking_project") due to previous error

When parsing wrong or invalid arguments, forward raw output instead panicking.

thread 'main' panicked at src/parsing.rs:413:5:
 should contain `running (?P<amount>\d+) tests?` pattern
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

Bumps regex-lite from 0.1.0 to 0.1.3.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Bumps insta from 1.33.0 to 1.34.0.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

   Doc-tests torrust-serde-bencode

running 5 tests
test src/ser.rs - ser::to_string (line 390) ... ok
test src/de.rs - de::from_str (line 347) ... ok
test src/de.rs - de::from_bytes (line 386) ... ok
test src/ser.rs - ser::to_bytes (line 353) ... ok
test src/lib.rs - (line 6) ... ok

test result: ok. 5 passed; 0 failed; 0 ignored; 0 measured; 0 filtered out; finished in 0.30s

I get this error:

Error from cargo-pretty-test:
the parsed amount of running tests [19, 53, 5] should equal to the number in stats.total [19, 53, 4]

It's not parsing one of the Doc-tests. I guess it's the last one.

cargo-nextest is faster. Maybe we can include a option to use it instead of cargo test. Besides cargo-nextest produces a XML report so parsing the report could be easier than parsing the output.

• https://nexte.st/book/junit.html

Ref: https://github.com/josecelano/cargo-pretty-test/issues/5

By default, -q is not passed, meaning raw output is forwarded as cargo test is running. If -q is passed in, don't display raw output, show test tree directly.

Note: this commit also makes --color default to always because not -q is the default, and we want forwarding looks pretty in color.

Press Enter to clear the screen to show the test tree:

https://github.com/josecelano/cargo-pretty-test/assets/25300418/a6f392e5-cd9e-4a17-a362-3a53e85f10c4

Some questions:

• Do we need this? / Do we need both the outputs from two CLIs?
  • Yes, because watching cargo test prints is fun and it's important to know the execution details in CI as well as unnecessary to run all test twice locally (cargo test + pretty-test)
  • No, why not run cargo test before this tool?
• Do we need to stop befor viewing the output from pretty-test?

• [x] -- --no-capture for test runners will be discarded because it messes up the output and hinders parsing
• [x] --help will prepend help info from pretty-test
  
  expand to see cargo pretty-test --help

  

• [x] #25