benansell / lobo Goto Github PK

Hi there!

This is a cool library, with some cool features. However, the defacto test runner will be shipped under elm-test. It would be awesome for the entire Elm community to add any features you thought were missing to elm-test, rather than go it alone where less people get to see some of the things you have. Come chat to us on #testing on the Elm Slack channel, and we can discuss!

Cheers,
eeue56

From your README:

filter -> instead use skip

Elm-test 4.0.0 removed the filter function and replaced it withskip. You will want to update the README accordingly.

First, fantastic tool! I love it.

One sticky point for me - I often code on the bus, where I have no wifi. I noticed today that lobo does some kind of package check before running any tests, which results in an error, meaning, I can't run my tests if I'm offline. I know elm-test node runner works offline, so I'm hoping you can make lobo work offline too. Maybe give a warning, but still run the tests.

Here is the specific output:

-----------------------------------[ BUILD ]------------------------------------
Error: The following HTTP request failed.
<http://package.elm-lang.org/all-packages?elm-package-version=0.18&since=2017-03-03%2001%3A17%3A11.4254%20UTC>

FailedConnectionException2 "package.elm-lang.org" 80 False getAddrInfo: does not exist (nodename nor servname provided, or not known)

----------------------------------[ WAITING ]-----------------------------------

Thanks!!

I've been using elm-test and the node runner while we took care of a few lobo issues. I do prefer lobo for 2 main reasons:

1. it will watch even if the build fails on the initial run (elm-test is working on this)
2. it will manage test deps while running/watching while elm-test requires a separate command

One thing I have come to depend on elm-test for though, is the fact that it will find all functions with the Test type. It will warn if a top level module has a function with that type and doesn't expose it. It will also automatically run them nested in a describe named after the module's name. It's been nice to break up my tests without having to roll them up into the main Test module. Is this sort of behavior something you would consider adding to lobo?

p.s. Thank you for your work on lobo.

I have a test that is testing that my json decoders work correctly, but am having some issues with case statements in my tests. For example:

test "json parser" <|
    \() ->
        case parseAllTheThings of
            Ok _ ->
                Expect.pass

            Err e ->
                Expect.fail e

I then went and broke the json, but I didn’t get any failures. So I tried this:

test "basic case statement" <|
    \() ->
        case 1 of
            1 ->
                Expect.fail "should fail"

            _ ->
                Expect.fail "would also fail"

Neither of those tests fail with elm-test or the lobo test extra frameworks. Once I switched to the elm-test cli runner the tests fail as expected.

Let me know if you need any other info and thank you for your hard work.

I migrated my app to Elm 0.19
When launching my test with Lobo, I have this message:

Unable to find elm-package.json in the current directory

It's not relevant with ELM 0.19.

Hello and thanks for creating lobo.
When adding lobo to an existing project with tests, the CLI prompt doesn't accept the default argument

The source-directories of the test elm-package.json needs to be updated to contain:
../node_modules/lobo/runner
../node_modules/lobo/plugin/elm-test-extra

May I add them to elm-package.json for you? (Y/n)

After this prompt, I expected that if I press the Enter key, the default value of Y should be chosen, rather than having me manually input the Y.
However, that is not currently implemented.

When tests fail, it would be great to see the output for the test printed out in the summary. Perhaps an option like --showTodo and --showSkip (which are why we moved to lobo).