Added more unit tests around Shape #18
Conversation
| specify "== on different shapes" $ (1:* 2:* 3 :* Nil) == (1:* 2:* 2 :* Nil) `shouldBe` False | ||
|
|
||
| specify "fmap on Nil" $ fmap (+1) Nil `shouldBe` Nil | ||
| specify "fmap on size n shape" $ fmap (+1) (1 :* 3 :* Nil) `shouldBe` (2:* 4 :* Nil) |
There was a problem hiding this comment.
i hadn't tested fmap? good point
| specify "foldr on shape" $ S.foldr (+) 0 (1:* 2:* 3 :* Nil ) `shouldBe` P.foldr (+) 0 [1,2,3] | ||
| specify "foldl1 on shape" $ S.foldl1 (+) (1:* 2:* 3 :* Nil ) `shouldBe` P.foldl1 (+) [1,2,3] | ||
| specify "foldr1 on shape" $ S.foldr1 (+) (1:* 2:* 3 :* Nil ) `shouldBe` P.foldr1 (+) [1,2,3] | ||
|
|
||
| specify "== on Nil" $ Nil == Nil `shouldBe` True | ||
| specify "== on identical shapes" $ (1:* 2:* 3 :* Nil) == (1:* 2:* 3 :* Nil) `shouldBe` True | ||
| specify "== on different shapes" $ (1:* 2:* 3 :* Nil) == (1:* 2:* 2 :* Nil) `shouldBe` False |
There was a problem hiding this comment.
i nearly didn't see the 1 2 3 vs the 1 2 2
from a code coverage perspective, even a singleton list / shape should suffice i think
There was a problem hiding this comment.
also i'd say "shapes with different element", :)
| specify "map2 on shape" $ S.map2 (+) (1:* 2:* 3 :* Nil) (2:* 3:* 4 :* Nil) `shouldBe` (3:* 5:* 7 :* Nil) | ||
|
|
||
| specify "shapeToList on shape" $ shapeToList (1:* 2:* 3 :* Nil) `shouldBe` [1, 2, 3] | ||
| specify "shapeSize on shape" $ shapeSize (1:* 2:* 3 :* Nil) `shouldBe` length (shapeToList (1:* 2:* 3 :* Nil)) |
There was a problem hiding this comment.
maybe i should define shapeSize in terms of the Foldable.length function? hrmmm
| specify "backwards traverse on shape" $ backwards traverse (:[]) (1:* 2 :* Nil) `shouldBe` [(1:* 2 :* Nil)] | ||
|
|
||
| specify "unShapeVector on shape" $ | ||
| do a <- return (uvFromList [1:* 2:* Nil,2:* 3:* Nil,3:* 4:* Nil :: Shape (S (S Z)) Int]) |
There was a problem hiding this comment.
heere I think we're testing the Unboxed representation of a sized list (aka shape) right?
| @@ -42,6 +88,20 @@ unitTestShape = describe "unit tests for Shape" $ do | |||
| do a <- return (uvFromList [3:* 4:* Nil,1:*2:*Nil :: Shape (S (S Z)) Int]) ; | |||
| UV.toList a `shouldBe` [3:* 4:* Nil,1:*2:*Nil] | |||
|
|
|||
| specify "weaklyDominates equal to" $ (1:* 2:* 3 :* Nil ) `weaklyDominates` (1:* 2:* 3 :* Nil ) `shouldBe` True | |||
There was a problem hiding this comment.
yeah, these are good to test too i suppose
|
some phrasing was confusing to me, possibly because i'm rusty. that said: i do think you'd benefit from learning how to run the haskell code coverage tools, they let you see what branches of a function have been executed or not (which is of course in general isn't a good state space coverage metric, but is probably enough in this module, and i'm semi confident that at least you'd learn about whats actually adding coverage in a meaninful way vs more complex than needed :) ) |
|
Thanks for taking a look! I had a similar feeling about some of the tests I was writing. I kept going back and forth about whether some of my cases were too trivial or even covering what the function was supposed to accomplish. I'll definitely look in the code coverage tools. |
|
Cabal has a flag called hpc, though it may not work with cabal 2.0
newbuild. I think the fix is in 2.4 or 2.2? But try test plus vanilla
build
…On Sat, Mar 3, 2018 at 7:54 PM George Granberry ***@***.***> wrote:
Thanks for taking a look! I had a similar feeling about some of the tests
I was writing. I kept going back and forth about whether some of my cases
were too trivial or even covering what the function was supposed to
accomplish. I'll definitely look in the code coverage tools.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#18 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AAAQwpjQDNQPAzkLNQ2ottJXKn6ZvbbOks5tazsxgaJpZM4SCYX8>
.
|
|
Also I’m not sure how much of the testing is needed. That’s module is
super stable and there’s no complexity in the meanings. Point being: not
sure if we even need to add all the nill ones. .. and the language in the
unboxed vector one could be written more clearly
On Sun, Mar 4, 2018 at 12:56 PM Carter Schonwald <[email protected]>
wrote:
… Cabal has a flag called hpc, though it may not work with cabal 2.0
newbuild. I think the fix is in 2.4 or 2.2? But try test plus vanilla
build
On Sat, Mar 3, 2018 at 7:54 PM George Granberry ***@***.***>
wrote:
> Thanks for taking a look! I had a similar feeling about some of the tests
> I was writing. I kept going back and forth about whether some of my cases
> were too trivial or even covering what the function was supposed to
> accomplish. I'll definitely look in the code coverage tools.
>
> —
> You are receiving this because you commented.
>
>
> Reply to this email directly, view it on GitHub
> <#18 (comment)>,
> or mute the thread
> <https://github.com/notifications/unsubscribe-auth/AAAQwpjQDNQPAzkLNQ2ottJXKn6ZvbbOks5tazsxgaJpZM4SCYX8>
> .
>
|
cartazio
force-pushed
the
master
branch
3 times, most recently
from
0f731a9 to
650ddea
Compare
|
i'll look at these again, past few years were nuts, pardon the slow folloow up |
No description provided.