Your program doesn't seem to have survived the transmission but try http://www.cse.ogi.edu/~mpj/pubs/springschool.html or http://www.cse.ogi.edu/~mpj/pubs/composing.html. For the theory, look at Triples, Toposes and Theories by Barr & Wells. I think it's chapter 7 but I haven't got it in front of me. Dominic Steinitz ----- Original Message ----- From: To: Sent: Saturday, April 12, 2003 5:01 PM Subject: Haskell-Cafe digest, Vol 1 #660 - 3 msgs

Send Haskell-Cafe mailing list submissions to haskell-cafe@haskell.org

To subscribe or unsubscribe via the World Wide Web, visit http://www.haskell.org/mailman/listinfo/haskell-cafe or, via email, send a message with subject or body 'help' to haskell-cafe-request@haskell.org

You can reach the person managing the list at haskell-cafe-admin@haskell.org

When replying, please edit your Subject line so it is more specific than "Re: Contents of Haskell-Cafe digest..."

Today's Topics:

1. Combining monads (Mark T.B. Carroll) 2. Re: Combining monads (Scott Turner) 3. Re: Combining monads (Glynn Clements)

--__--__--

Message: 1 Date: Sat, 12 Apr 2003 09:52:58 -0400 (EDT) From: "Mark T.B. Carroll" To: haskell-cafe@haskell.org Subject: Combining monads

This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info.

--481034957-302199926-1050155578=:555 Content-Type: TEXT/PLAIN; charset=US-ASCII

Hello! I am trying and failing to use one monad in another. A tiny example is enclosed, where I try to use IO to help along a State-ful computation. Is the real mistake in do_next? Is there a minimal rewrite that will get this working that I can use as a template for other things? Or, is there a good tutorial somewhere on this type of thing? Any pointers are much appreciated!

-- Mark

--481034957-302199926-1050155578=:555 Content-Type: TEXT/PLAIN; charset=US-ASCII; name="Test.hs" Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Example of incompetence Content-Disposition: attachment; filename="Test.hs"

aW1wb3J0IENvbnRyb2wuTW9uYWQuU3RhdGUNCmltcG9ydCBEYXRhLkxpc3QN Cg0KZ2V0X2xlbmd0aCA6OiBTdHJpbmcgLT4gSU8gSW50DQoNCmdldF9sZW5n dGggbmFtZSA9IA0KICAgIGRvIGNvbnRlbnRzIDwtIHJlYWRGaWxlIG5hbWUN CiAgICAgICByZXR1cm4gKGxlbmd0aCAobGluZXMgY29udGVudHMpKQ0KDQp1 cGRhdGVfYmVzdF93aXRoIDo6IChTdHJpbmcsIEludCkgLT4gU3RhdGVUIChT dHJpbmcsIEludCkgSU8gKCkNCg0KdXBkYXRlX2Jlc3Rfd2l0aCAoY3VycmVu dF9uYW1lLCBjdXJyZW50X2NvdW50KSA9DQogICAgZG8gKGJlc3RfbmFtZSwg YmVzdF9jb3VudCkgPC0gZ2V0DQogICAgICAgaWYgY3VycmVudF9jb3VudCA+ IGJlc3RfY291bnQgDQoJICB0aGVuIHB1dCAoY3VycmVudF9uYW1lLCBjdXJy ZW50X2NvdW50KQ0KCSAgZWxzZSByZXR1cm4gKCkNCg0KZG9fbmV4dCA6OiBT dHJpbmcgLT4gU3RhdGVUIChTdHJpbmcsIEludCkgSU8gKCkNCg0KZG9fbmV4 dCBuYW1lID0gDQogICAgZG8gY291bnQgPC0gZ2V0X2xlbmd0aCBuYW1lDQog ICAgICAgdXBkYXRlX2Jlc3Rfd2l0aCAobmFtZSwgY291bnQpDQoNCmxvbmdl c3RfZmlsZSA6OiBbU3RyaW5nXSAtPiBJTyAoU3RyaW5nLCBJbnQpDQoNCmxv bmdlc3RfZmlsZSBuYW1lcyA9DQogICAgZXhlY1N0YXRlVCAoc2VxdWVuY2Vf IChtYXAgZG9fbmV4dCBuYW1lcykpICgibm9uZSIsIDApDQoNCm1haW4gPSBk byAobmFtZSwgY291bnQpIDwtIGxvbmdlc3RfZmlsZSBbIi9ldGMvaG9zdG5h bWUiLCAiL2V0Yy9udHAuY29uZiIsICIvZXRjL3RpbWV6b25lIl0NCgkgIHB1 dFN0ckxuICgiTG9uZ2VzdCBmaWxlIGlzICIgKysgbmFtZSArKyAiIHdoaWNo IGlzICIgKysgc2hvdyBjb3VudCArKyAiIGxpbmVzIGxvbmciKQ0K --481034957-302199926-1050155578=:555--

--__--__--

Message: 2 From: Scott Turner To: "Mark T.B. Carroll" , haskell-cafe@haskell.org Subject: Re: Combining monads Date: Sat, 12 Apr 2003 09:14:45 -0500

On Saturday 12 April 2003 09:52, Mark T.B. Carroll wrote:

...

Hello! I am trying and failing to use one monad in another. Is there a minimal rewrite that will get this working ... ? Yes. Just add lift. do count <- lift $ get_length name

--__--__--

Message: 3 From: Glynn Clements Date: Sat, 12 Apr 2003 15:26:03 +0100 To: "Mark T.B. Carroll" Cc: haskell-cafe@haskell.org Subject: Re: Combining monads

Mark T.B. Carroll wrote:

...

Hello! I am trying and failing to use one monad in another. A tiny example is enclosed, where I try to use IO to help along a State-ful computation. Is the real mistake in do_next? Is there a minimal rewrite that will get this working that I can use as a template for other things? Or, is there a good tutorial somewhere on this type of thing? Any pointers are much appreciated!

Combining monads isn't straightforward. There is a paper on the topic, at:

http://cm.bell-labs.com/cm/cs/who/wadler/topics/monads.html#combining-monads

However, you can't really combine the IO monad in the way that the above paper describes.

The most direct way to fix your program is to use an IORef instead of a state monad, e.g.

...

import IORef

...

update_best_with :: IORef (String, Int) -> (String, Int) -> IO () update_best_with ref (current_name, current_count) = do (best_name, best_count) <- readIORef ref when (current_count > best_count) $ do writeIORef ref (current_name, current_count)

...

do_next :: IORef (String, Int) -> String -> IO () do_next ref name = do count <- get_length name update_best_with ref (name, count)

...

longest_file :: [String] -> IO (String, Int) longest_file names = do ref <- newIORef ("none", 0) mapM_ (do_next ref) names readIORef ref

A more elegant solution would be to generate a list of filename/length pairs (e.g. using mapM), then just use maximumBy to select the longest, e.g.

...

get_length' :: String -> IO (String, Int) get_length' name = do length <- get_length name return (name, length)

...

longest_file :: [String] -> IO (String, Int) longest_file names = do pairs <- mapM get_length' names return $ maximumBy cmp pairs where cmp (_, l1) (_, l2) = compare l1 l2

-- Glynn Clements

--__--__--

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

End of Haskell-Cafe Digest