Implement exercise Poker#347
Conversation
Update master to exercism/rust:master
Description and tests copied from https://github.com/exercism/problem-specifications/tree/master/exercises/poker A `lib.rs` stub is included so that users don't have to puzzle out the type signature.
* Ignore all but the first test * Add entry to `config.json` and ensure `configlet lint` has no output Tried running `configlet generate . --only poker`, but that eliminated all problem-specific text from the README; reverted that change before commit.
|
Travis passes, so I think this should be all set for someone with merge authority. Note that the command |
| /// | ||
| /// Note the type signature: this function should return _the same_ reference to | ||
| /// the winning hand(s) as were passed in, not reconstructed strings which happen to be equal. | ||
| pub fn winning_hands<'a>(_: &[&'a str]) -> Option<Vec<&'a str>> { |
There was a problem hiding this comment.
I would prefer using a variable name instead of _
There was a problem hiding this comment.
- Any variable name we use will need to begin with an underscore anyway, to prevent compiler warnings about unused variables.
- Using
_encourages the implementer choose their own variable name. - The only point of including the function stub is so that users don't have to guess and check the expected signature by inspecting the tests. While the name used is technically part of the public signature of the function, I figured it best to minimize the degree to which we constrain the individual implementers.
If you feel very strongly about it, I can change it, but I think there are good reasons to leave it as is.
|
Here is what I'm thinking about: I think the rust track would really benefit from an exercise that teaches implementing
I think the second option would be better, what do you think? |
|
I'd agree that traits are an important concept to learn, and that this exercise is a pretty natural fit for using that trait implementation. After all, that's the path I went in my own solution. That said, I tend to prefer exercises which allow a broad range of implementations over those which railroad the implementer into a particular design choice, and the required function signature already requires that people know about lifetimes in function signatures. In short, I agree that the second path is better. I'm going to adjust the readme now to suggest the use of |
Possibly the simplest way to solve this problem is to create a PokerHand type which implements PartialOrd, and then sort the input and take the highest values. However, this may not be obvious to novice coders. The new hints in the readme should point people in the right direction.
|
This is awesome, thank you! I will wait a few days with the merge to give a chance to other maintainers to chime in, if you don't mind. |
|
@ijanos It's been two weeks with no further comments; perhaps you could merge this now? |
| "uuid": "0a33f3ac-cedd-4a40-a132-9d044b0e9977", | ||
| "slug": "poker", | ||
| "core": false, | ||
| "unlocked_by": null, |
There was a problem hiding this comment.
Note that this will be our first exercise of difficulty 6.
Historically, this track has only used difficulties 1, 4, 7, and 10, corresponding to the four sections (Introduction = 1, Getting Rusty = 4, Rust Gets Strange = 7, Putting it all Together = 10) in https://github.com/exercism/rust/blob/aea470a76a9c6f0005adebb2a00d969512da3b6e/problems.md. That file does not exist anymore as it has been replaced with https://github.com/exercism/rust/blob/master/problem_ordering.md . So we are not restricted to only doing 1 4 7 10. But you would be breaking new ground.
I am assuming all parties involved are OK with this. I'm also inferring that everyone agreed that this is harder than all exercises of difficulty 4 and easier than all exercises of difficulty 7.
This comment therefore does not require a response or anything to be changed, but I was obligated to write this comment for the sake of anyone who looks back on this PR and wonders about the origin of the first difficulty 6 exercise.
|
|
||
| ## Hints | ||
|
|
||
| - Ranking a list of poker hands can be considered a sorting problem. |
There was a problem hiding this comment.
these hints are Rust-specific, therefore for the README generator to not remove them, they need to be placed in .meta/hints.md . I did so in a commit.
|
Question: README generator writes the description wrapped to some number of columns, because those are how the files appear in https://github.com/exercism/problem-specifications/blob/master/exercises/poker/description.md / https://raw.githubusercontent.com/exercism/problem-specifications/master/exercises/poker/description.md and https://github.com/exercism/rust/blob/master/config/exercise-readme-insert.md / https://raw.githubusercontent.com/exercism/rust/master/config/exercise-readme-insert.md . So, regenerating the README results in this diff: I'm wondering whether the hints should get wrapped too, to match. |
|
|
||
| ## Source | ||
|
|
||
| J Dalbey's Programming Practice problems <http://users.csc.calpoly.edu/~jdalbey/103/Projects/ProgrammingPractice.html> |
There was a problem hiding this comment.
this is not the source listed in https://github.com/exercism/problem-specifications/blob/master/exercises/poker/metadata.yml.
|
Well, my constraints are that I don't want to maintain any README file that some generator does not generate (even if it isn't configlet's generator), so I would prefer to use the README as it was appears in f4b2ae5 , hoping nobody has objections to that (the fact that some lines are wrapped and some aren't). Assuming nobody has complaints, then there is nothing more to be said |
|
Thanks for implementing this exercise! |
Per the comment [here](#347 (comment)), there's continued interest in a Rust-specific exercise focusing on teaching trait implementation. My idea is to have users implement an arbitrary-precision `Decimal` type[1], and ensure that standard math is possible for them. I.e. they must implement at least `Add`, `Sub`, `Mul`, `PartialEq`, and `PartialOrd`; tests will ensure that the `==`, `+`, `-`, `*`, `>`, and `<` operators all work as expected. --- [1] We'd have to ask people not to simply import [bigdecimal](https://crates.io/crates/bigdecimal), which does exactly what we require.
3 participants
Staking a claim on implementing the poker exercise.