Pass raw params to less parser

[ilyapoz]

close #94

[ilyapoz]

I'm just starting out with ruby so if someone would help me with writing tests for this, I would appreciate it.

[ilyapoz]

This is not a failure of this diff, so I suppose this should not block it from merging, #117 created.

[ilyapoz]

@simi ? can we merge this?

[simi]

@ilyapoz can you provide some usecase and mention this change in readme and changelog?

[ilyapoz]

@simi did that

[simi]

Great! I have one idea. What about to write README section as universal way how to pass custom args to lessc and mention this one (relativeUrls) as example?

[simi]

@ilyapoz awesome! It will be really nice to add test. I can handle the ruby part. Can you just post me bare lessc command with for example --relativeUrls and some basic file and how it looks with and without that parameter? I'll handle the rest.

[ilyapoz]

nah, too late, I got it :)

[ilyapoz]

Cool that we added the tests, now it seems that default behavior is dependent on ruby versions (probably less.rb gem version, but I'm not sure). If you have any ideas, I would appreciate it. Meanwhile I must probably dig into less.rb history and take a closer look.

[ilyapoz]

Actually strictly speaking this tests less.rb behavior, not parameters passing, but for ppl using this gem as a blackbox this test should be useful.
But I think that now we have a bug, because less output depends on ruby version.

[ilyapoz]

I think this is suspicious, I must be not understanding something about the language or the test setup process as I managed to create an unstable test case.
Can I reliably test the default behavior here?

[simi]

You need to clean this parameter after test.

[simi]

@ilyapoz it was set here cowboyd/less.rb@4e72e0c#diff-0937a07e2f5bb236b1e021c5e94f047aR5 in less.rb 2.3.0.

Maybe it will make sense to test true and false relativeUrls explicitly instead of relying on less.rb defaults. Except this I think it is ready for merge. Can you confirm?

I think we should bump to 2.8 since it is adding new feature. But I can do it when I'll releasing.

[simi]

I digged deeper and relativeUrls was introduced in less/less.js@eb5c9fb#diff-41dc76a214b45826bb39925edcccf0f3R19 with false as default and it wasn't changed until today.

[ilyapoz]

In less.js it defaults to false, in less.rb to true (just to summarise).
Testing explicitly is double edged, I would think that people using less-rails and should the authors of less.rb for whatever reason change the default it will be at least good for you to know and bump the major version. If this is a stupid argument, just let me know, I'm not that experienced in releasing libraries via semver.

[ilyapoz]

So basically I think that it's a good idea to fix implicit guarantees (default behavior) in tests, whether it's a good or bad behavior. Because some users might rely on that, so why not state it in the tests.

[simi]

@ilyapoz sonuds reasonable.

[simi]

Merged!