December 2011 - Haskell-Cafe

Chuan-Kai Lin's Unimo framework
by Jean-Luc Delatre 13 Dec '11

13 Dec '11

Does anybody has an hands on experience of using Chuan-Kai Lin's Unimo framework? https://sites.google.com/site/chklin/research It looks interesting but it seems it amounts to add an extra layer of interpretation (the monads encoding as a data type) and the paper lacks examples of actual uses of the framework. jld

2 2

Linear types using kinds?
by Gregory Crosswhite 13 Dec '11

13 Dec '11

Hey everyone! So, I was on LtU reading about F* http://lambda-the-ultimate.org/node/4318 and an idea just occurred to me which *must* have occurred to someone smarter than me, so I would love your feedback on the following question: Could Haskell be extended to support linear/affine types (in order to have, for example, more efficient resource management) by extending the kind system to include such types? Or, even if the answer were yes, does the use of monads and higher-order types (like those used in "runST" function for the ST monad) make linear/affine types completely redundant in Haskell. Thanks a lot! Greg

1 0

indentation blues
by Ben Kolera 13 Dec '11

13 Dec '11

I am fairly new to haskell, but I really like the emacs haskell mode. It is a bit strict but it generally does what I want it to. Unfortunately I can't really compare to the haskell vim mode since I only did Scala and Perl back when I was a heavy vim user. The one useful thing that I can add is that there are some really good packages out there for modal vi keybindings in emacs. If you truly like the vim keybindings better then you can still use them in emacs. I hope that helps a little bit. On Tue, Dec 13, 2011 at 10:50 AM, Martin DeMello <martindemello(a)gmail.com> wrote: > The vim autoindent for haskell is really bad :( Is there a better > indent.hs file floating around somewhere? Alternatively, is the emacs > haskell mode better enough that it's worth my time learning my way > around emacs and evil? > > martin > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe

1 0

Module name space question
by Christoph Breitkopf 12 Dec '11

12 Dec '11

Hi, I recently asked about what interfaces to implement for a new data type. Following the rule that the last 10% of work take the second 90% of time, some other questions have come up. If anyone wants to look at the code in question: http://www.chr-breitkopf.de/comp/IntervalMap Some time ago, I was looking for a data structure to search in sets of possibly overlapping intervals, and found only Data.SegmentTree, which did not fit my needs (nice term for "I did not understand the type signatures"). So I started to roll my own IntervalMap implementation. But yesterday I used Hayoo instead of Hoogle for the first time, and found Data.IntervalMap.FingerTree. This has an Interface that I _do_ understand, and, while slightly different from my aims (depite the name, it seems more like a multi-map and does not offer lookup of specific keys), it raises some tasks/questions: Tasks: Benchmark against my implementation, and maybe scrap own if it performs worse. And questions: 1. Why doesn't Hoogle find this? 2. Module name space pollution I would have been so bold to name my module Data.IntervalMap. Now, since even Ross Patterson uses a submodule name instead, what is the accepted practice? If I decide to release my class after all, what should be the name? Data.IntervalMap.RedBlackTree ? Or even Data.IntervalMap.MyDomaindNameOrSomething. An I even have a module Data.IntervalMap.Interval, so if Ross had decided to put his Interval type into a separate module, even FingerTree vs RedBlackTree would not have prevented a name clash. 3. Are there more wothwile operations on this data structure? Data.IntervalMap.FingerTree has 'dominators', which I missed. A function to extract the interval with the largest endpoint might be useful, too. Any more ideas? TIA, Chris

6 5

Cabal: Disjunctive Build-Depends
by Matthew Farkas-Dyck 12 Dec '11

12 Dec '11

Hello all. I have written some code that can be compiled with either of two libraries, with no modification. How can I tell Cabal? I tried "||" but it failed to parse. I could find which package is available in the build script, and then call defaultMainNoRead with the appropriate GenericPackageDescription, but I'm not sure how to find this system-agnostically. Thanks for any help. Cheers, Matthew Farkas-Dyck

2 2

HaXml 1.13 -> 1.22 upgrade
by Michael Orlitzky 12 Dec '11

12 Dec '11

I'm trying to migrate one my programs from the old HaXml API to the new. Please, someone save me. I'm currently stuck with this, which works in 1.13. All of the filters work on Content, so I make some from the root element with the "root_elem = CElem root" line. -- |Takes an XML String as an argument, and returns the -- status that was parsed from it. Should only be used -- on XML string where a <status> is a top-level element. parse_status :: String -> [Status] parse_status xml_data = catMaybes maybe_status where (Document _ _ root _) = xmlParse xml_file_name xml_data root_elem = CElem root status_element = (single_status root_elem) maybe_status = map status_from_content status_element In the new API, xmlParse returns the root element with type (Element i) rather than just Element. And the Content constructor I have to use is (CElem (Element i) i), but I have no way to pass the correct 'i' to it. I just want to parse a few elements from an XML file.

3 4

Re: [Haskell-cafe] Module name space question
by John Lato 12 Dec '11

12 Dec '11

> From: Christoph Breitkopf <chbreitkopf(a)googlemail.com> > > Hi, > > I recently asked about what interfaces to implement for a new data type. > Following the rule that the last 10% of work take the second 90% of time, > some other questions have come up. > > If anyone wants to look at the code in question: > http://www.chr-breitkopf.de/comp/IntervalMap > > Some time ago, I was looking for a data structure to search in sets of > possibly > overlapping intervals, and found only Data.SegmentTree, which did not fit > my needs > (nice term for "I did not understand the type signatures"). I can't answer any of your specific questions, although I've published a related structure in the splaytree package (http://hackage.haskell.org/package/splaytree) The interface is minimal, although it suits my needs and I'd be happy to extend it if someone else found it worthwhile. The difference between my data structure (which I've called a RangeSet) and a standard IntervalSet is that overlapping ranges are combined, and if part of a range is deleted, the affected range is modified and possibly split e.g. > let set1 = singleton $ range 0 5 > set2 = insert (range 2 5) set1 after this, set2 has one node representing the range 0-7. Staying on topic, my package puts everything under `Data.SplayTree.xxx`, which is different from the fingertree package organization. AFAICT the selection of one convention over the other is fairly arbitrary, although more packages seem to use the "one second-level hierarchy" structure (e.g. bytestring, text, containers). John L.

1 0

No instance for (Read POSIXTime)
by dokondr 12 Dec '11

12 Dec '11

Hi, I need to read / write epoch seconds from / to plain text files. When I try to read POSIXTime that I use in my own data type: data TimedClassRecT = TCR {timeStamp :: POSIXTime, classCosMap :: Map.Map String Float} deriving (Eq, Read, Show) I get the following error: No instance for (Read POSIXTime) arising from the 'deriving' clause of a data type declaration Possible fix: add an instance declaration for (Read POSIXTime) or use a standalone 'deriving instance' declaration, so you can specify the instance context yourself When deriving the instance for (Read TimedClassRecT) What should I do to provide Read instance for POSIXTime? I would rather not implement it myself. Thanks!

2 3

Tracing Prelude.read exceptions
by dokondr 12 Dec '11

12 Dec '11

Hi, I got quite used to a sequence providing simple data persistence : 1) Store my data to a file: writeFile fileName (show someData) 2) Some time later read this data back: line <- readFile fileName let someData = read line :: SomeDataType Having this done hundreds of times I now got stuck with step 2) trying to read moderately complex structure back. I get read exception in run-time: fromList *** Exception: Prelude.read: no parse I have checked and rechecked my types, data files, etc. - and still no idea. So my question: Is there any way to trace Prelude.read exceptions to see exactly on what data element read fails in run-time? Thanks!

3 2

Recommended class instances for container type
by Christoph Breitkopf 12 Dec '11

12 Dec '11

Hello, I'm in the process of implementing a container data type, and wonder what class instances are generally considered necessary. E.g. is it ok to start out with a Show that's adequate for debugging, or is it a 'must' to include instances of everything possible (Eq, Ord if possible, Read, Show, Functor, ...). And what about the more experimental things? Say, DeepSeq, Typeable, Data? I'd like to keep this simple at start, and I've admittedly not followed recent developments in Haskell-land (recent meaning the last 10 years or so. I _do_ know about hierachical modules ;-) ). OTOH, if not having such instances makes it impossible to do things the modern way, I'd probably take the time to implement (and maybe understand) them. Thanks, Chris

6 8