- template-haskell - Haskell.org

What happened to currentLoc ?
by Alfonso Acosta 25 Oct '07

25 Oct '07

Hi, I've been trying to improve error reporting in my TH-based embedded compiler. I sometimes build data structures using splices. Those structures can later cause an error at runtime which, to be more informative, should give the line and module in which the splice was executed (i.e. tell to what structure the error refering by telling where it was exactly created) I have just read about currentLoc in the "Notes on Template Haskell 2" document. currentLoc :: Q (FilePath, Int) -- Returns the location of the top-level splice being executed. Was it ever implemented? Are there any plans to do it? The path to the module is not really important (it should be enough with currentModule) but the line information can be really helpful. Thanks, Fons

2 3

Defining a function and using it in a splice within the same module
by Alfonso Acosta 25 Oct '07

25 Oct '07

Hi, The user guide from latest GHC release makes clear that doing what I wrote in the subject of this message is not possible: >From 7.6: "You can only run a function at compile time if it is imported from another module. That is, you can't define a function in a module, and call it from within a splice in the same module." Unfortunately I have encountered this problem more times than I'd like while using TH to implement a DSL. My usual solution is to generate a function to process the extra arguments which cannot be passed to the splice (i.e. push the arguments out of the splice so that I can use functions defined in the same module). Nevertheless, I'd like to know if there are any chances of seeing this problem solved anytime soon. Thanks, Alfonso

2 5

Any chances of having TH included in Haskell II?
by Alfonso Acosta 04 Oct '07

04 Oct '07

Hi, I recently developed an embedded DSL compiler making use of TH (and my plan is to continue with its development). However, my employers are naturally concerned about the long-term support of Template Haskell. I've been trying to gather some information about it, but I don't seem to find any opened tickets in haskell-prime. Can anyone confirm if there are any plans of including Template Haskell in the next standard and/or any information regarding long-term support? Best Regards, Alfonso Acosta

3 3

type quotation brackets work!
by Wolfgang Jeltsch 14 Sep '07

14 Sep '07

Hello, wow, I didn’t know that the [t|…|] syntax has been implemented already. The GHC User’s Guide says, it isn’t. Could someone please update it? Best wishes, Wolfgang

2 1

Generating classes
by Barney Hilken 11 Sep '07

11 Sep '07

Here's a fun exercise in TH. I've noticed in several bits of Haskell code how people define a concrete datatype to represent some data, but also a typeclass with corresponding member functions in case someone wants to use a different representation. Mathematically the relationship between the type and the class is like the relationship between a free algebra and the class of all algebras. Logically, it's like the relationship between the abstract syntax and the class of models. I've been meaning to try out Template Haskell, so why not write some TH code to do it all automatically? So here it the exercise: 1. Write a TH function which takes (the name of) a recursive datatype, and declares the corresponding typeclass. 2. Extend it to declare the type an instance of the class. These are pretty straightforward, once you've got the hang of TH. The key mathematical property of term algebras is that they are initial: you can map them (uniquely) into any algebra. This extends to any recursive definition, provided it is functorial in the recursion (so T = X -> T is ok, but T = T -> X is not). 3. Automatically generate the polymorphic function from the datatype to an arbitrary type in the class. This is as far as I've got. But we could easily go further: 4. Add support for record labels and the terminal coalgebra property. 5. Generalise to turn n mutually recursive datatypes into an n- parameter typeclass. If the TH 'reify' function wasn't limited to '98 datatypes, we could try: 6. How about existentials? 7. GADTs? I've appended my solution to parts 1-3. If anyone's interested, I'll have a go at 4 & 5. Is this kind of code of any use to anyone? Barney. Attachments: UnivAlg.hs defines the function which does all the work, Test.hs gives some recursive datatypes for testing, and FunctorN.hs generalises the functor class to multiple arguments.

1 0

reifyDecl and reifyType not working
by Wolfgang Jeltsch 11 Sep '07

11 Sep '07

Hello, reifyDecl and reifyType seem to be recognized as identifiers and not as part of TH syntax by GHC 6.7.20070903. Take, for example, the following code: {-# LANGUAGE TemplateHaskell #-} x = reifyDecl () If I compile this (with or without -fth), I get the error message “Not in scope: `reifyDecl'”. What’s wrong here? Best wishes, Wolfgang

4 7

Newbie question - class dependencies
by Hugo Pacheco 08 Jul '07

08 Jul '07

Hi, I'm wondering if I can (ab)use from TH in order to generalize a context-dependant function into a generic function without that context, as long as I now that the arguments I'm applying the function to belong to the class. I guess this would work similarly to an automatic instance generation example. My example would be: module Main where myprint :: a -> String myprint = show main = do putStr $ myprint 1 putStr $ myprint '2' My ideia would be to stage the computation of *myprint* until it is called, where the *Show* instance can be satisfied. hugo

2 2

Couple of Template Haskell Questions
by Adrian Hey 04 Jun '07

04 Jun '07

Hello, I've been considering using this and have read the 2 papers papers about the basic design, but these don't seem to cover a couple of issues that I anticipate as possible "show stoppers" for me. The first is, how to add generated top level declarations to the export list (and how to augment the import list too possibly). Paragraph 7 of the second paper seems to acknowledge the problem and suggests a solution, but also says it's not implemented (as of Nov. 2003). Has this been implemented? The second is, how to get Haddock to properly document the resulting module? I gather it is possible to get TH to emit another module (as source text), so that may be 1 possible solution to both these problems. I'm not sure how this will interact with --make option and Cabal though. Also, the old TH web page says something about not being able to splice types. Does that mean no type sigs for generated top level declarations? Thanks -- Adrian Hey

3 2

Call for Papers: Generative Programming and Component Engineering (GPCE'07)
by Emir Pasalic 20 Mar '07

20 Mar '07

- Apologies for multiple posts - Call for Papers Sixth International Conference on Generative Programming and Component Engineering (GPCE'07) October 1-3, 2007 Salzburg, Austria (co-located with ESWEEK'07) http://www.gpce.org/07 Important Dates: * Submission of abstracts: April 17, 2007 * Submission: April 20, 2007 * Notification: June 10, 2007 * Tutorial and workshop proposals: March 16, 2007 * Tutorial and workshop notification: April 9, 2007 Scope Generative and component approaches are revolutionizing software development similar to how automation and components revolutionized manufacturing. Generative Programming (developing programs that synthesize other programs), Component Engineering (raising the level of modularization and analysis in application design), and Domain-Specific Languages (elevating program specifications to compact domain-specific notations that are easier to write, maintain, and analyze) are key technologies for automating program development. GPCE provides a venue for researchers and practitioners interested in foundational techniques for enhancing the productivity, quality, and time-to-market in software development that stems from deploying standard componentry and automating program generation. In addition to exploring cutting-edge techniques for developing generative and component-based software, our goal is to foster further cross-fertilization between the software engineering research community and the programming languages community. As GPCE is co-located with ESWEEK this year, we also especially encourage papers from the embedded systems community. Submissions 10 pages in SIGPLAN proceedings style (sigplanconf.cls) reporting research results and/or experience related to the topics above (PC chair can advise on appropriateness). We particularly encourage original high-quality reports on applying GPCE technologies to real-world problems, relating ideas and concepts from several topics, or bridging the gap between theory and practice. Please note that in contrast to last year, GPCE 2007 is not using a double-blind reviewing process. Topics GPCE seeks contributions in software engineering and in programming languages related (but not limited) to: * Generative programming o Reuse, meta-programming, partial evaluation, multi-stage and multi-level languages, and step-wise refinement o Semantics, type systems, symbolic computation, linking and explicit substitution, in-lining and macros, templates, and program transformation o Runtime code generation, compilation, active libraries, synthesis from specifications, development methods, generation of non-code artifacts, formal methods, and reflection * Generative techniques for o Product-line architectures o Distributed, real-time and embedded systems o Model-driven development and architecture o Resource bounded/safety critical systems. * Component-based software engineering o Reuse, distributed platforms and middleware, distributed systems, evolution, patterns, development methods, deployment and configuration techniques, and formal methods * Integration of generative and component-based approaches * Domain engineering and domain analysis o Domain-specific languages (DSLs) including visual and UML-based DSLs * Separation of concerns o Aspect-oriented and feature-oriented programming, o Intentional programming and multi-dimensional separation of concerns * Industrial applications * Applications in embedded systems Reports on applications of these techniques to real-world problems are especially encouraged, as are submissions that relate ideas and concepts from several of these topics, or bridge the gap between theory and practice. The program chair is happy to advise on the appropriateness of a particular subject. Submissions must adhere to SIGPLAN's republication policy. Please contact the program chair if you have any questions about how this policy applies to your paper (gpce07 at diku.dk). General Chair * Charles Consel (LABRI/INRIA, Bordeaux) Program Committee Program Chair: * Julia Lawall (DIKU, University of Copenhagen) Program Committee Members: * Edwin Brady (University of St Andrews, UK) * Johan Brichau (UniversitÈ Catholique de Louvain, Belgium) * Rastislav Bodik (UC Berkeley, USA) * Jacques Carette (McMaster University, Canada) * Albert Cheng (University of Houston, USA) * Remi Douence (Ecole des Mines de Nantes-Inria, Lina, France) * Lidia Fuentes (University of M·laga, Spain) * Ian Gorton (Pacific Northwest National Lab) * Jean-Marc Jezequel (IRISA (INRIA & Univ. Rennes 1), France) * Kyo Kang (Pohang University of Science and Technology, Korea) * Siau Cheng Khoo (National University of Singapore, Singapore) * Paul Kelly (Imperial College London, UK) * Anne-Francoise Le Meur (University of Lille 1, France) * Christian Lengauer (University of Passau, Germany) * Sandeep Neema (Vanderbilt University, USA) * Scott Owens (University of Cambridge, UK) * Jens Palsberg (UCLA, USA) * Renaud Pawlak (Rensselaer Polytechnic Institute, USA) * Zoltan Porkolab (Eotvos Lorand University, Hungary) * Robby (Kansas State University, USA) * Peter Sestoft (IT University of Copenhagen, Denmark) * Jeremy Siek (University of Colorado at Boulder, USA) * Tony Sloane (Macquarie University, Australia) * Kevin J. Sullivan (University of Virginia, USA) * Peri Tarr (IBM Thomas J. Watson Research Center, USA)

1 0

Saving the AST generated by Template Haskell
by Alfonso Acosta 22 Feb '07

22 Feb '07

Hi all, I'm using Template Haskell to design a small subset of a Hardware Description DSEL (Domain Specific Embedded Language). My language supports higher order as the user can supply small functions as arguments. I chose to parse them with TH because it allows me to use plain Haskell for the function syntax (instead of reinventing the wheel) but mostly because gives parsing for free. The AST of the functions must be kept by the embedded compiler so that it can be later translated to a target language by one of the potential backends of the embedded compiler. The problem is ... how to store that AST? Let me show an example ---- import Language.Haskell.TH.Syntax -- sample function from the DSLE library hdMapSY :: (HDPrimType a, HDPrimType b) => HDFun (a -> b) -> HDSignal a -> HDSignal b -- We keep the function's AST (to make program transformations in the backends) newtype HDPrimFun = HDPrimFun [Dec] deriving Show -- Type safety layer, we keep the function to make sure TH checks the type (and possible further simulations) data HDFun a = HDFun [Dec] a deriving Show -- Helper constructor function, which suffers from the Saving-the-AST problem -- mkMetaAST currently returns a phony value mkHDFun :: Q [Dec] -> Q Exp mkHDFun qd = do dx <- qd metaASTnm <- newName "metaAST" let funnm = getFunName dx metaAST = mkMetaAST dx metaASTdec = ValD (VarP metaASTnm) (NormalB metaAST) [] return $ LetE (metaASTdec:dx) (AppE (AppE (ConE $ mkName "HDFun") (AppE (ConE $ mkName "HDPrimFun") (VarE metaASTnm))) (VarE funnm)) where getFunName :: [Dec] -> Name getFunName [FunD nm _] = nm getFunName _ = error "mkHDFun: toy example, just and exactly one dec!" -- This function should create an AST expression from the AST -- but it would be a big pain to code mkMetaAST :: [Dec] -> Exp mkMetaAST _ = AppE (ConE (mkName "LitE")) (LitE (StringL "big pain to code!")) --- An example program coded in the DSLE could could be something like .. myCircuit :: HDSignal Int -> HDSignal Int myCircuit = hdMapSY ($mkHDFun [d| f input = input+1 |]) So the question is .. How can mkHDfun save the AST of "f input = input+1" (for which it needs to create and return a metaAST) without the effort of having to create boiler plate code for the whole TH library types? Did anyone do something similar before? I workaround would be saving the String of the AST with show, but Dec nor Exp belong to the Read class, :S Thanks in advance, Alfonso Acosta

2 3