Test cases for type inference - Haskell-Cafe - Haskell.org

newer
list -> sublists

Test cases for type inference

older
systemtimmetypeable: Workaround...

Peter Verswyvelen

20 Oct 2009 20 Oct '09

10:07 a.m.

For learning, I would like to develop my own implementation of type inference, based on the paper "Typing Haskell in Haskell". At first sight, the source code of THIH contains a small number of tests, but I was wandering if a large test set exist? Thanks, Peter

Reply

Sign in to reply online Use email software

Show replies by date

Martijn van Steenbergen

21 Oct 21 Oct

8:07 a.m.

Peter Verswyvelen wrote:

For learning, I would like to develop my own implementation of type inference, based on the paper "Typing Haskell in Haskell".

At first sight, the source code of THIH contains a small number of tests, but I was wandering if a large test set exist?

I'm pretty sure GHC maintains a large set of test cases. Maybe you can use some of those? Or maybe you can just pick a random project from hackage and run your type checker on that? Martijn.

Reply

Sign in to reply online Use email software

Job Vranish

10:47 a.m.

I was recently working on an type inference algorithm and to test it I did the following: Used the quickcheck Arbitrary typeclass to generate expressions Inferred types of the expressions using my algorithm, converted the expressions that passed inference to haskell and wrote them to a file (without type signatures) converted and wrote the failed expression to a separate file compiled the "passed" file with -Wall, and extracted all the type signatures that were spit out as warnings parsed the type signatures that -Wall spit out and compare with the signatures generated by my algorithm. compiled the "failed" file and make sure I get type errors for all my expressions My algorithm also infers infinite types (which haskell does not) so I had to test that functionality manually. Overall it was kinda messy, but it worked ok. I could possibly send you one of my lists. My test expressions are all very simple, no type classes, only 2 types (function and number), and the only expression components are let, lambda, apply, identifier, and number. If something like that would work, let me know. Hope that helps, - Job On Tue, Oct 20, 2009 at 10:07 AM, Peter Verswyvelen wrote:

For learning, I would like to develop my own implementation of type inference, based on the paper "Typing Haskell in Haskell".

At first sight, the source code of THIH contains a small number of tests, but I was wandering if a large test set exist?

Thanks, Peter _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Reply

Sign in to reply online Use email software

jur

2 p.m.

On Tue, Oct 20, 2009 at 10:07 AM, Peter Verswyvelen wrote: For learning, I would like to develop my own implementation of type inference, based on the paper "Typing Haskell in Haskell".

At first sight, the source code of THIH contains a small number of tests, but I was wandering if a large test set exist?

Thanks, Peter

Dear Peter, The sources of the Helium compiler contain quite a few test files for type inference and other issues like static checking. Go to http://www.cs.uu.nl/wiki/Helium go to Downloads and get the most recent version. Then unpack it. Then go to the directory heliumsystem/helium/test/typerrrors that contains a large number of type erroneous programs,and heliumsystem/helium/test/correct contains quite a few that should pass the test (but that does depend somewhat on how extensive the language you support is). hope this helps, Jurriaan

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Reply

Sign in to reply online Use email software

Peter Verswyvelen

2:31 p.m.

Thanks Job & Martijn for the feedback. I would like to write a type inferrer that is mostly Haskell 98 with Trex row polymorphism extensions (a bit like CAL I guess). That will certainly take a while because I'm also studying the theory in my spare time. So as soon as (or if) I have something working, I will need a larger set of tests. So Job, no need to do put any effort in preparing test sets, but I highly appreciate your offer. Cheers, Peter On Wed, Oct 21, 2009 at 4:47 PM, Job Vranish wrote:

I was recently working on an type inference algorithm and to test it I did the following: Used the quickcheck Arbitrary typeclass to generate expressions Inferred types of the expressions using my algorithm, converted the expressions that passed inference to haskell and wrote them to a file (without type signatures) converted and wrote the failed expression to a separate file compiled the "passed" file with -Wall, and extracted all the type signatures that were spit out as warnings parsed the type signatures that -Wall spit out and compare with the signatures generated by my algorithm. compiled the "failed" file and make sure I get type errors for all my expressions

My algorithm also infers infinite types (which haskell does not) so I had to test that functionality manually.

Overall it was kinda messy, but it worked ok.

I could possibly send you one of my lists. My test expressions are all very simple, no type classes, only 2 types (function and number), and the only expression components are let, lambda, apply, identifier, and number. If something like that would work, let me know.

Hope that helps,

- Job

On Tue, Oct 20, 2009 at 10:07 AM, Peter Verswyvelen wrote:

...
For learning, I would like to develop my own implementation of type inference, based on the paper "Typing Haskell in Haskell".

At first sight, the source code of THIH contains a small number of tests, but I was wandering if a large test set exist?

Thanks, Peter _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Reply

Sign in to reply online Use email software

5692

Age (days ago)

5693

Last active (days ago)

Download

4 comments

4 participants

tags

participants (4)

Job Vranish
jur
Martijn van Steenbergen
Peter Verswyvelen