grep: New exercise

[pgaspar]

Closes #946.

Implementation

The implementation was heavily influenced by python's. It can certainly be improved, but we may want to leave that for the mentor notes themselves, since the main goal here is to get CI to pass.

Test Generation

The tests are generated using the canonical data, but the file contents are hardcoded in a custom test template. I ended up doing this because I couldn't easily pass custom variables from the test case file to the ERB test template. Without that I can't parse the canonical data comments and make the file contents dynamic, like some tracks do.

I couldn't find other exercises that did anything close to this and looking at the generator code I didn't find a clear way to pass a file_data hash to the template file.

Note: I really want to merge #929... Having the generator write <<-EXPECTED.gsub(/^ */, '').strip instead of <<~EXPECTED on the test file adds a layer of extra complexity we could easily avoid. We've merged this and I updated the code 👍.

@exercism/ruby

TODO

• Add to config.json
• Add issue to create mentor notes

[kotp]

The rest of it looks 'generally' good to me.

[kotp]

I noticed the conflglet failure as well as reported by Travis CI. I may have some time to look at the file name problem indicated as well as the configlet issue, if someone doesn't beat me to it, or already have some ideas.

[pgaspar]

I noticed the conflglet failure as well as reported by Travis CI.

Yes, I noted above that before merging we need to add this exercise to config.json. I didn't do that because I wasn't sure where @F3PiX wanted it.

I may have some time to look at the file name problem indicated as well as the configlet issue, if someone doesn't beat me to it, or already have some ideas.

Not sure what file name problem you're referring to 🤔

[kotp]

I may have some time to look at the file name problem indicated as well as the configlet issue, if someone doesn't beat me to it, or already have some ideas.

Not sure what file name problem you're referring to 🤔

I think it was corrected... so that part is good to go.

[emcoding]

we need to add this exercise to config.json. I didn't do that because I wasn't sure where @F3PiX wanted it.

It should be somewhere at the end of the track. The file reading is a feature above 'level 3'. And moreover, it's also not the simplest form of reading a file. And it's also totally new.

I'd say: let's make it "core": false, "unlocked_by": "12-days"` and difficulty: 8. (Difficulty and topics are not yet part of the Track Anatomy set up, so I don't have opinions on those. For the difficulty I tend to use 2 difficulties for each level but that's not official.)

So, it starts as a side, so the mentors can play with it first, and we can come up with mentor notes.
When we know more know how difficult it really is for students, we can add it to the cores.

Looking sooooooooo forward to trying this :-))))

[emcoding]

Thinking TDD, would it make sense to put all the no-flags tests at the beginning?

[pgaspar]

The order is defined by the canonical data. Do we want to go against that? (It may be worth considering opening a PR on the problem-specifications repo if you feel strongly about changing the order everywhere).

[emcoding]

Aha, no, we don't want to go against that. And as the TDD sequence may differ from language to language, I don't think it's worth to change it in the prob-specs.

[emcoding]

It looks marvellous :-)
I'll leave the technical details to Victor to review.
Thanks so much @pgaspar, for yet another exercise 🎉

[kotp]

This is good enough. We can always make improvements and changes later.

Let's get it in! ;)

[pgaspar]

Thank you both!