Example of Luhn parts 2 & 3, From Trait and Custom Trait #248

IanWhitney · 2017-01-16T15:48:26Z

This is an implementation of the idea I discussed here

After students do the original Luhn exercise, they can learn how to use
the From trait to allow validation of multiple types. Then they can implement it again with a custom type.

ijanos

I like this, because people will learn about implementing traits on primitives and learn it is not a good idea but can be useful sometimes.

petertseng

I really wish that it would just work if we add arbitrary slugs to config.json.

I applied the following diff to trackler

diff --git c/fixtures/tracks/fake/config.json i/fixtures/tracks/fake/config.json
index dea26e6..b942964 100644
--- c/fixtures/tracks/fake/config.json
+++ i/fixtures/tracks/fake/config.json
@@ -21,6 +21,11 @@
         ]
       },
       {
+        "slug": "my-new-awesome-problem" ,
+        "difficulty": 1,
+        "topics": []
+      },
+      {
         "difficulty": 3,
         "slug": "one",
         "topics": [
diff --git c/fixtures/tracks/fake/my-new-awesome-problem/HINTS.md i/fixtures/tracks/fake/my-new-awesome-problem/HINTS.md
new file mode 100644
index 0000000..b53cb5e
--- /dev/null
+++ i/fixtures/tracks/fake/my-new-awesome-problem/HINTS.md
@@ -0,0 +1 @@
+My new awesome problem.
diff --git c/fixtures/tracks/fake/my-new-awesome-problem/test.a i/fixtures/tracks/fake/my-new-awesome-problem/test.a
new file mode 100644
index 0000000..ce01362
--- /dev/null
+++ i/fixtures/tracks/fake/my-new-awesome-problem/test.a
@@ -0,0 +1 @@
+hello
diff --git c/test/trackler/implementation_test.rb i/test/trackler/implementation_test.rb
index 8b39d2b..0afe821 100644
--- c/test/trackler/implementation_test.rb
+++ i/test/trackler/implementation_test.rb
@@ -73,6 +73,18 @@ class ImplementationTest < Minitest::Test
     refute implementation.exists?, "Not expecting apple to be implemented for track TRACK_ID"
   end
 
+  def test_unknown_problem
+    slug = 'my-new-awesome-problem'
+    problem = Trackler::Problem.new(slug, PATH)
+    implementation = Trackler::Implementation.new(TRACK_ID, URL, problem, PATH)
+    begin
+      puts implementation.files
+    rescue => e
+      puts "Aw that's a shame, didn't find the files of #{slug} because #{e}"
+    end
+    assert implementation.exists?, "Expecting #{slug} to be implemented"
+  end
+
   def test_override_implementation_files
     implementation = Trackler::Implementation.new(TRACK_ID, URL, PROBLEM, PATH)
     files = { "filename" => "contents" }

then I ran the tests and got the following result:

$ ruby trackler/implementation_test.rb     (01-17 00:14)
Run options: --seed 23017

# Running:

.........Aw that's a shame, didn't find the files of my-new-awesome-problem because undefined method `[]' for nil:NilClass
F...

Fabulous run in 0.009562s, 1359.6007 runs/s, 2091.6933 assertions/s.

  1) Failure:
ImplementationTest#test_unknown_problem [trackler/implementation_test.rb:85]:
Expecting my-new-awesome-problem to be implemented

13 runs, 20 assertions, 1 failures, 0 errors, 0 skips

Let me know if you think I did the test wrong in some way. But I predict this means the new problems will not actually exist.

petertseng · 2017-01-17T00:15:39Z

exercises/luhn-from/HINTS.md

+
+`exercism fetch rust luhn`
+
+In the original Luhn exercise you only validated strings, but the Luhn algorithm can be applied integers as well.


"applied to"

petertseng · 2017-01-17T00:16:46Z

exercises/luhn-trait/HINTS.md

+- `exercism fetch rust luhn`
+- `exercism fetch rust luhn-from`
+
+In the original Luhn exercise you only validated strings, but the Luhn algorithm can be applied integers as well.


"applied to"

petertseng · 2017-01-17T00:17:05Z

exercises/luhn-trait/HINTS.md

+
+Instead of creating a Struct just to perform the validation, what if you you validated the primitives (i.e, String, u8, etc.) themselves?
+
+In this exercise you'll create and implement a custom [trait](https://doc.rust-lang.org/book/traits.html) that performs the validation


End sentence with a period, I think.

IanWhitney · 2017-01-17T01:19:07Z

I am unfamiliar with Trackler. Here's what I see after setting it up and pointing its version of the Rust track to this branch.

rust_track = Trackler.tracks["rust"]
implementations = rust_track.implementations
luhn_from = implementations.detect {|impl| impl.problem.slug == 'luhn-from'}
# => #<Trackler::Implementation:0x007fa6101799f0 @track_id="rust",...>
luhn_from.files
# => NoMethodError: undefined method `[]' for nil:NilClass

The error is thrown here, because the metadata method returns nil.

I think that's a fixable problem in Trackler. What it's trying to do here is pull together the language-independent files for an exercise. But if those files don't exist there should be fall-back behavior.

I think. I mean, I've only looked at that code for 5 minutes so I'm sure there's complexity I'm missing. I wanna dig into it more. Will have to wait until Wednesday at the earliest, though.

IanWhitney · 2017-02-21T01:06:49Z

Not sure when it'll be deployed, but track-specific exercises should be possible soon.

exercism/DEPRECATED.trackler#24

petertseng · 2017-02-21T02:28:24Z

config.json

@@ -132,6 +132,26 @@
      ]
    },
    {
+      "slug": "luhn-from",
+      "difficulty": 1,


now that I gave Luhn a 4, perhaps this should be as well

petertseng · 2017-02-21T02:28:30Z

config.json

+    },
+    {
+      "slug": "luhn-trait",
+      "difficulty": 1,


now that I gave Luhn a 4, perhaps this should be as well

same

petertseng · 2017-02-21T02:30:51Z

exercises/luhn-from/example.rs

+}
+
+impl From<&'static str> for Luhn {
+    fn from(s: &'static str) -> Self {


I understand it's only the example, but I'm not clear on why this would have to be &'static str specifically. What would happen if this were just &str? It seems unnecessarily restrictive to me to only accept &'static str.

petertseng · 2017-02-21T02:33:13Z

exercises/luhn-trait/HINTS.md

+
+In the original Luhn exercise you only validated strings, but the Luhn algorithm can be applied to integers as well.
+
+In Luhn From you implemented a From trait, which also required you to create a Luhn struct.


here, it says "Luhn From", whereas above it says "Luhn From Trait". In the header of the HINTS.md of the luhn-from, it says "Luhn with From Trait". Intentional? Seems OK I guess since they should all be distinguishable.

petertseng · 2017-02-21T02:33:55Z

exercises/luhn-trait/example.rs

+    }
+}
+
+impl Luhn for &'static str {


same - I'd ask if there is any requirement that these be &'static str rather than just &str.

IanWhitney · 2017-02-25T17:54:31Z

Rebased on master and made @petertseng's requested changes.

I'm not sure if the updated Trackler is deployed yet, so I don't know what will happen if this gets deployed.

IanWhitney · 2017-02-26T14:20:31Z

Nearly done. After rebasing down to a single commit I realized that I had the Luhn followups in the wrong part of the track.

"Luhn: Using the From Trait" now moves to just after Bracket Push (which is our example that uses From). In the Luhn implementation students will implement From for multiple types, so it makes sense to put it after Bracket Push where students only implement it for one type.

"Luhn: Using a Custom Trait" moves to after Space Age. Space Age solutions use custom traits and From, so pairing these exercises made sense.

IanWhitney · 2017-02-26T17:50:37Z

Checked with Katrina. Trackler is updated, so merging this is fine.

This implements the idea I discussed here exercism#247 (comment) After students do the original Luhn exercise, they can now do "Luhn: Using the From Trait" to allow validation of multiple types. And after that, they can do "Luhn: Using a Custom Trait" to develop their own trait. "Luhn: Using the From Trait" is placed just after Bracket Push (which also uses `From`). In the Luhn implementation students will implement `From` for multiple types, so it makes sense to put it after Bracket Push where students only implement it for one type. "Luhn: Using a Custom Trait" moves to after Space Age. Space Age solutions use custom traits and `From`, so pairing these exercises made sense.

kytrinyx · 2017-02-27T18:41:07Z

@IanWhitney trackler assumes that there's a description.md file either in x-common/exercises/:slug/description.md or in x:track_id/exercises/:slug/description.md. I don't see that here.

They were created for #248, but since they're not in the ignore_pattern they are being served. See for proof: http://x.exercism.io/v2/exercises/rust/luhn-from http://x.exercism.io/v2/exercises/rust/luhn-trait And note how description.md and metadata.yml appear inside the response.

Closes exercism#248 Approached a little different this time now that the core set of concepts start crystallizing. I've taken two reasonably different solutions and extracted applied language components. There's some overlap between some concepts (like `Ordering` and `conditionals`), but I think it's reasonable to touch such concepts in different concept-exercises.

IanWhitney changed the title ~~Example of Luhn part 2, From Trait~~ Example of Luhn parts 2 & 3, From Trait and Custom Trait Jan 16, 2017

ijanos approved these changes Jan 16, 2017

View reviewed changes

petertseng approved these changes Jan 17, 2017

View reviewed changes

IanWhitney mentioned this pull request Jan 17, 2017

Allow a Track to define an exercise that is does not have a language-independent definition exercism/DEPRECATED.trackler#22

Closed

petertseng mentioned this pull request Jan 17, 2017

Language-specific exercises exercism/problem-specifications#352

Closed

petertseng reviewed Feb 21, 2017

View reviewed changes

IanWhitney force-pushed the implement_luhn_from branch from bb5e7d1 to ed4d770 Compare February 25, 2017 17:53

petertseng approved these changes Feb 25, 2017

View reviewed changes

IanWhitney force-pushed the implement_luhn_from branch from 26ded73 to d2a257f Compare February 26, 2017 14:14

IanWhitney force-pushed the implement_luhn_from branch from d2a257f to 5c61101 Compare February 26, 2017 14:20

IanWhitney force-pushed the implement_luhn_from branch from 5c61101 to 68caa48 Compare February 26, 2017 17:51

IanWhitney merged commit e0a4c6e into exercism:master Feb 26, 2017

IanWhitney deleted the implement_luhn_from branch February 26, 2017 18:01

petertseng mentioned this pull request May 9, 2017

config: Ignore description.md and metadata.yml #284

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Example of Luhn parts 2 & 3, From Trait and Custom Trait #248

Example of Luhn parts 2 & 3, From Trait and Custom Trait #248

IanWhitney commented Jan 16, 2017 •

edited

Loading

ijanos left a comment

petertseng left a comment •

edited

Loading

petertseng Jan 17, 2017

petertseng Jan 17, 2017

petertseng Jan 17, 2017

IanWhitney commented Jan 17, 2017 •

edited

Loading

IanWhitney commented Feb 21, 2017

petertseng Feb 21, 2017

petertseng Feb 21, 2017

petertseng Feb 21, 2017

petertseng Feb 21, 2017

petertseng Feb 21, 2017

IanWhitney commented Feb 25, 2017

IanWhitney commented Feb 26, 2017

IanWhitney commented Feb 26, 2017

kytrinyx commented Feb 27, 2017


		`exercism fetch rust luhn`

		In the original Luhn exercise you only validated strings, but the Luhn algorithm can be applied integers as well.


		Instead of creating a Struct just to perform the validation, what if you you validated the primitives (i.e, String, u8, etc.) themselves?

		In this exercise you'll create and implement a custom [trait](https://doc.rust-lang.org/book/traits.html) that performs the validation


		In the original Luhn exercise you only validated strings, but the Luhn algorithm can be applied to integers as well.

		In Luhn From you implemented a From trait, which also required you to create a Luhn struct.

Example of Luhn parts 2 & 3, From Trait and Custom Trait #248

Example of Luhn parts 2 & 3, From Trait and Custom Trait #248

Conversation

IanWhitney commented Jan 16, 2017 • edited Loading

ijanos left a comment

Choose a reason for hiding this comment

petertseng left a comment • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

IanWhitney commented Jan 17, 2017 • edited Loading

IanWhitney commented Feb 21, 2017

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

IanWhitney commented Feb 25, 2017

IanWhitney commented Feb 26, 2017

IanWhitney commented Feb 26, 2017

kytrinyx commented Feb 27, 2017

IanWhitney commented Jan 16, 2017 •

edited

Loading

petertseng left a comment •

edited

Loading

IanWhitney commented Jan 17, 2017 •

edited

Loading