- template-haskell - Haskell.org

Template Haskell - read function
by Thomas L. Bevan 02 Jul '03

02 Jul '03

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, I'm struggling to write the following function in Template Haskell. I have a phantom type, >data HTMLPage a = HTMLPage FilePath This type points to an HTML file, where 'a' refers to the return type of the posted form. So, mostly, I would expect types like HTMLPage (String, String, Date, Int) Where this represented a form the returned someone's first name, last name, dob and age. So, I want a function that will take the returned name/value pairs and cast them into the matching type. eg. >readForm :: Read a => [(String,String)] -> HTMLPage a -> a Can anyone give me any pointers or am I better off just abandoning this approach? Tom -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2 (GNU/Linux) iD8DBQE/AsiKYha8TWXIQwoRAvzuAKDPHD4oztAPLFShEWsLuclT7BaImwCfSB8e 9c2aTbTCSpVwabXVsht62DM= =orX3 -----END PGP SIGNATURE-----

1 1

Reification of top level variables - was: Re: Problem with Labelled Fields
by Eric Offermann 27 Jun '03

27 Jun '03

> What you will actually > get is the code for applying "c_succ" to lift c (it is lifted as it is > an argument to the function). However, lifting c needs to use this > instance again so loops. > > Ian Ian, thanks a lot for pointing that out. This leads me to the following modification: {- spliceC.hs -} module spliceC where import Language.Haskell.THSyntax data C a = C {c_succ :: a -> a ,c_pred :: a -> a } instance Lift (C a) where lift c = recCon "C" [fieldExp "c_succ" [| c_succ $(var "sampleC") |] ,fieldExp "c_pred" [| c_pred $(var "sampleC") |] ] c_succS :: C a -> ExpQ c_succS c = [| \ a -> c_succ c a |] sampleC :: C Int sampleC = C {c_succ = \ n -> n + 1 ,c_pred = \ n -> n - 1 } Thus, the possibility of using a "reifyVar c :: Q String" in the instance of lift c yielding the original name of c would be useful to keep the instance decleration generic. Any hint on implementing that? Thanks, Eric -- +++ GMX - Mail, Messaging & more http://www.gmx.net +++ Bitte lächeln! Fotogalerie online mit GMX ohne eigene Homepage!

2 1

Problem with Labelled Fields
by Eric Offermann 26 Jun '03

26 Jun '03

Hi! Perhaps there's anybody out there who can help me with the following problem with labelled fields (using ghc 6.0). The easy example modules are: {- spliceC.hs -} module spliceC where import Language.Haskell.THSyntax data C a = C {c_succ :: a -> a ,c_pred :: a -> a } instance Lift (C a) where lift c = recCon "C" [fieldExp "c_succ" [| c_succ c |] ,fieldExp "c_pred" [| c_pred c |] ] c_succS :: C a -> ExpQ c_succS c = [| \ a -> c_succ c a |] sampleC :: C Int sampleC = C {c_succ = \ n -> n + 1 ,c_pred = \ n -> n - 1 } {- main.hs -} module Main where import spliceC sampleC_succ = $(c_succS sampleC) I suppose that this should bring a function sampleC_succ into scope, instantiated for sampleC. But in fact I get *** Exception: stack overflow Who's wrong? Thanks, Eric

2 1

RE: Templates in ghc-6.0 & --make
by Simon Peyton-Jones 06 Jun '03

06 Jun '03

[redirecting to template Haskell. Are you on that list, Christoph?] Excellent point below. The dependency tracker assumes that if the type of an imported thing is unchanged, you don't need to recompile. But here it's a change in the *value* of an imported thing that should force recompilation. Sigh. Don't hold your breath for a fix! Simon | -----Original Message----- | From: glasgow-haskell-bugs-admin(a)haskell.org [mailto:glasgow-haskell-bugs-admin@haskell.org] On | Behalf Of Ch. A. Herrmann | Sent: 05 June 2003 12:21 | To: glasgow-haskell-bugs(a)haskell.org | Subject: Templates in ghc-6.0 & --make | | Hi, | | first, many thanks to the implementors | for releasing Template Haskell. | | A little flaw which can be observed with the | example Printf/Main from the web users guide. | | Changing in "Printf.hs" the parse function | to deliver a different format is not | recognized by the dependency tracker of | ghc --make | causing Main.hs to be skipped in the compilation, | but metaprogramming is supposed to change the | code of Main.hs. | | Many thanks in advance | -- | Dr. Christoph Herrmann | Teaching and Research Assistant | University of Passau, Germany | http://www.fmi.uni-passau.de/~herrmann | | _______________________________________________ | Glasgow-haskell-bugs mailing list | Glasgow-haskell-bugs(a)haskell.org | http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs

1 0

Template Haskell webpage
by Ian Lynagh 03 Jun '03

03 Jun '03

Hi all, I have created a page on Template Haskell (based in part on Simon Peyton Jones' existing page) at http://www.haskell.org/th/ If you're using Template Haskell then please drop me an e-mail with a paragraph or so saying what you are doing and giving a link to a webpage if you have a suitable one. Likewise if you have written a paper on Template Haskell then let me know where it is, along with a BiBTeX entry for it, and I'll add it to the page. Thanks! Ian, immensely proud of his logo :-)

1 0

RE: [Template-haskell] Release
by Simon Peyton-Jones 30 May '03

30 May '03

| I'm happy with the proposal :). Although now that 6.0 is out, is the | syntax still going to be changed? Oh yes, it's too early to start fixing TH just because it's out. In the sterling tradition of GHC (H98 subset aside) I think we are free to change the defn of TH! The user base is small enough and motivated enough to move with us. S

1 0

RE: [Template-haskell] Release
by Simon Peyton-Jones 30 May '03

30 May '03

| > How's that for a compromise? | | Does silence mean everyone is happy with this? | | Or have you all just given up caring? :-) I'd prefer not to suffix the Exp constructors (using something instead of Let to avoid the keyword conflict), but it seems I'm in a minority, so I'll beat a strategic retreat. The rest looks fine. Simon

1 0

RE: [Template-haskell] Release
by Simon Peyton-Jones 30 May '03

30 May '03

| | `Bind'/`bind' has the same number of characters as | `LetE'/`letE' and `Switch'/`switch' has even one more | character than `CaseE'/`caseE'. Indeed. It's the *other* constructors in Exp I'm worried about. | I still think the number one priority is to have no | exceptions in the naming scheme. In a large data type like Exp you are sure to have to refer to the constructor list to find out what they are all called. There's Var, App, Tup, Cond... and (LetE or Bind) and (CaseE or Switch). You have to look to see what they are. I'm agree 100% that the smart constructors should be derived systematically. So if we can't have 'let' for the smart constructor, then we can't use 'Let' for the dumb constructor either. What I'm missing is why the dumb constructors must all have a suffix. | Sorry for being stubborn on this, but I just don't see the | advantage of saving one character if that risks that people | have to refer to the documentation more often (or at least | get a compiler error before they remember again that they | are dealing with an exception). I agree with that. What I'm missing is why adding 'E' to the end of each Exp constructor will reduce the number of looks at documentation. It's a big data type. You have to look at it, don't you? S

6 6

RE: [Template-haskell] Release
by Simon Peyton-Jones 27 May '03

27 May '03

| The Maybe module defines one datatype. THSyntax defines | many. One problem is that constructors of some of these | data types clash. For examples both Exp and Pat have | variables and literals. Fair enough. Let's use 'P' for Pat, and nothing for Exp. Thus Var and VarP or PVar. I.e. leave Exp, the heaviest hitter of all, unannotated. Would that serve? Then the only oddities are 'Let' and 'Case' because the lower-case smart constructors (only) clash with the Haskell keywords. Is that a good enough reason to add 'E' to every constructor (concrete and smart)? My nose says 'no' but yours may differ. For these, I think Let/letE and Case/caseE are not so bad. An alternative would be to use Bind and Switch instead. this isn't a big deal in the end. I'm quite willing to be over-ruled. Simon

2 1

RE: [Template-haskell] Pattern matching against lists.
by Simon Peyton-Jones 26 May '03

26 May '03

Right. Then you need the magic names "GHC.Base::" etc. As Derek Ekins says, the thing to do is to write it quoted and then print; you'll then see the magic names. Tim and I are ruminating about this stuff. We're planning a way to let you name global things like (:) just by saying '(:) rather than with magic strings where you have to know about GHC.Base. But that won't arrive for a month or two Meanwhile this should get you unglued Simon | -----Original Message----- | From: template-haskell-admin(a)haskell.org [mailto:template-haskell-admin@haskell.org] On Behalf Of | Hampus Ram | Sent: 26 May 2003 14:52 | To: template-haskell(a)haskell.org | Subject: Re: [Template-haskell] Pattern matching against lists. | | On Mon, May 26 2003, Simon Peyton-Jones wrote: | > | > reading through THSyntax.hs I can find no way of matching against | > | > lists when creating functions. Is there something obvious that I've | > | > missed or is this something that is relly missing? | > | > [d| f (x:xs) = x |] | | Nothing, except that I needed to create pattern matching functions of | different lengths (so I need to construct them with the datatypes). I | tried conPat with just ":" and "[]" and that didn't work. I also tried | reifying declarations to see "behind the scenes", but reification does | seldom work for me... The solution was of course to use GHC.base:: and | GHC.base:[] and now everything works. One should perhaps not be too | quick to ask for help :) | | (btw. the new names are much better than the old) | | /Hampus | | -- | Homepage: http://www.dtek.chalmers.se/~d00ram | E-mail: d00ram(a)dtek.chalmers.se | | "Det är aldrig försent att ge upp" | _______________________________________________ | template-haskell mailing list | template-haskell(a)haskell.org | http://www.haskell.org/mailman/listinfo/template-haskell

2 1