July 2010 - Haskell-prime

Propsal: NoDatatypeContexts
by Ian Lynagh 21 Oct '10

21 Oct '10

Hi all, H98 and H2010 allow a context to be given for datatypes, e.g. the "Eq a" in data Eq a => Foo a = Constr a I have made a proposal to remove support for that context (ticket #139). More details are on the proposal wiki page: http://hackage.haskell.org/trac/haskell-prime/wiki/NoDatatypeContexts Thanks Ian

11 13

prefix operators
by John Meacham 20 Jul '10

20 Jul '10

It occurred to me the other day that Haskell (w/ bang patterns) now has 3 prefix operators, all of which are defined independently and follow their own special rules for parsing. we have (-), (!) and (~). It would seem to me that we should somehow be able to unify the mechanism behind parsing these, as in practice, it seems that prefix operators are useful in haskell. We have some similarities, - and ! are both infix and prefix operators, ~ is not. ! and ~ can only be in patterns as prefix, (-) can be in both patterns and expressions. But it seems like we may be able to come up with a common way of parsing them all, prolog has had user defined infix, prefix, and postfix operators (sharing the same name even) and is still able to parse things properly so I don't think there will be a technical issue. My first impulse is to treat application as just another binary operator with a certain precedence and find appropriate precedences for !,~,- in the new framework. note: I am not proposing user defined prefix operators, just musing about whether we can unify the rules behind parsing the current three prefix operators, perhaps folding them into the fixity resolution algorithm. John -- John Meacham - ⑆repetae.net⑆john⑈ - http://notanumber.net/

6 8

Proposition: accessory functions which are local to type.
by Evgenij Merenkov 15 Jul '10

15 Jul '10

What I don't like about current accessory functions is that they pollute namespace. data Coordinate2D = Coordinate2D { x :: Double, y :: Double} will make x and y not available for other types: data Coordinate3D = Coordinate3D { x :: Double, y :: Double, z :: Double} What I am proposing is something like this: data Coordinate = Coordinate where { x :: Double, y :: Double} By using "where" (or other keyword) we mark those accesor functions as local to type. This means, that function can only be used in a context of a type, like this: coord <- getRandomCoordinate For modifying (same as current): let newCoord = coord {x = 3, y = 4} For accessing part of a type: let x = coord.x Current form "x coord" will generate error, as there is no standalone function "x". "x" can only be accessed in context of any function application that returns needed type. Meaning, it is valid to write something like this: let plus (Coordinate2D x1 y1) (Coordinate2D x2 y2) = Coordinate2D (x1+x2) (y1+y2) x = (plus coord coord).x Yes, that means that function composition will need to use another symbol, but I think it is a great trade-off, as it will reduce namespace pollution and will make use of notation familiar to mainstram programmers (Java, C#, C, C++, etc). _________________________________________________________________ Windows Live Hotmail: Your friends can get your Facebook updates, right from Hotmail®. http://www.microsoft.com/middleeast/windows/windowslive/see-it-in-action/so…

3 2

operators ~ and @
by Christian Maeder 15 Jul '10

15 Jul '10

Hi, the language description contains ~ and @ as reserved: reservedop → .. | : | :: | = | \ | | | <- | -> | @ | ~ | => These "operators" are only used within patterns. I see no reason to leave them user-definable and usable within expressions. The only change in the grammar would be, to disallow them as infix varop in funlhs: funlhs → var apat { apat } | pat varop⟨@|~⟩ pat ... They could be defined as (@) and (~). A similar restriction exists for the unary minus in right sections. Once the bang "!" moves into patterns it would be just a further symbol among ⟨@|~⟩. Is this worth creating a proper proposal? http://hackage.haskell.org/trac/haskell-prime/wiki/Process It is not that I recommend defining (@) or (~) as operators (although maybe useful in a few domains), it's just to make the language more orthogonal wrt "!" that is already used in types and will be possibly even more so via bang patterns. ("!" is currently not reserved and used in libraries as operator) Cheers Christian

1 0

fixity resolution
by Christian Maeder 15 Jul '10

15 Jul '10

http://www.haskell.org/~simonmar/haskell-2010-draft-report-2/haskellch10.ht… "Fixity resolution also applies to Haskell patterns, but patterns are a subset of expressions so in what follows we consider only expressions for simplicity." The string "1 * - 1" is legal as pattern, but rejected as expression! Furthermore fixity resolution does not distinguish between constructors and other operators as it should according to the grammar: pat → lpat qconop pat (infix constructor) | lpat funlhs → var apat { apat } | pat varop pat | ( funlhs ) apat { apat } "a : b * c : d = undefined" is currently rejected with: "cannot mix `:' [infixr 5] and `Main.*' [infixl 9] in the same infix expression" but should be fine by the given grammar (rule "pat varop pat"). Cheers Christian P.S. like in my proposal for infixexp I would change pat to: pat → pat qconop pat (infix constructor) | lpat for the sake of a better presentation only.

6 20

Re: ANNOUNCE: Haskell 2010 Report (final)
by Christian Maeder 12 Jul '10

12 Jul '10

Simon Marlow schrieb: > PS. don't forget there will be a new revision of Haskell this year, and > time is getting short to submit proposals. For instructions on getting > involved, see > > http://hackage.haskell.org/trac/haskell-prime In fact, I want to make a proposal "PrefixMinusResolution" as described in http://www.haskell.org/pipermail/haskell-prime/2010-July/003229.html related to http://hackage.haskell.org/trac/haskell-prime/wiki/FixityResolution http://hackage.haskell.org/trac/haskell-prime/ticket/30 but I could not create a wiki page or ticket. (I registered and verified as "maeder") Christian

1 0

Second draft of the Haskell 2010 report available
by Simon Marlow 08 Jul '10

08 Jul '10

The second draft of the Haskell 2010 report is now available in PDF and HTML formats (the PDF looks a lot nicer): http://www.haskell.org/~simonmar/haskell-2010-draft-report-2.pdf http://www.haskell.org/~simonmar/haskell-2010-draft-report-2/haskell.html relative to the first draft, which was only publicised on the haskell-prime mailing list, I have now updated the libraries too. Rather than update all the library documentation manually, I (perhaps rashly) decided to make a LaTeX backend for Haddock instead, and generate the report automatically from the library source code. Getting this to work turned out to be a lot more effort than I anticipated, but I think the results are quite attractive. Once the new Haddock backend is incorporated upstream, we'll finally have the ability to generate decent typeset API documentation. Furthermore, this should make it much easier to incorporate more libraries in future versions of the Haskell standard, should we decide to do so. Right now, the HTML version of the report is generated from the LaTeX sources, including the libraries. We could use the Haddock HTML output instead, but that would entail some difficulties with cross-references from the language part of the report to the libraries, which is why I've left it this way for now. This is why the libraries part of the report is bereft of hyperlinks in HTML; but at least it is well indexed in the PDF version. Summary of the library changes in Haskell 2010 relative to Haskell 98 and the FFI specification: * All libraries have been updated to their hierarchical names * The following library modules were dropped from the standard, due to being obsolete or superseded. Replacements are not part of the standard yet, but it is expected that they will be replaced in the future: Directory, System, Time, Locale, CPUTime, Random In the case of System, some functions have moved to the new modules System.Environment and System.Exit. * Foreign.Marshal.Error: functions on IOError moved to System.Error (this is where they've been in base for ever). * Data.List: added intercalate, subsequences, permutations, foldl', foldl1', stripPrefix (H2010 Data.List matches the current base version). * Data.Char: various additions of Unicode predicates (e.g. isLetter, isMark, isNumber), the GeneralCategory type and generalCategory. (matches the base version) * Control.Monad: added forM, forM_, (>=>), (<=<), forever, foldM_, replicateM, replicateM_ (matches the base version). * System.IO: added fixIO, hSetFileSize, hTell, hIsTerminalDevice, hSetEcho, hGetEcho, hShow. The base version has various additions: hGetBuf/hPutBuf, binary Handles, encodings, and newline support, but I erred on the side of being conservative here: these APIs need discussion, and in some cases are probably not suitable for the standard in their current state at all. * System.IO.Error: new module, providing functionality that was in Foreign.Marshal.Error in the FFI spec. * System.IO.Exit: new module, functionality moved from H98 System * System.Environment: new module, functionality moved from H98 System We expect to provide exactly these libraries in GHC 6.14, although the exact mechanism has yet to be decided; for discussion see http://www.haskell.org/pipermail/haskell-prime/2010-April/003158.html Comments on the draft report are welcome, before I finalise this and sign off on Haskell 2010. Cheers, Simon

7 16