- Haskell-Cafe - Haskell.org

Re: Unicode support
by Ketil Malde 09 Oct '01

09 Oct '01

[Posted to haskell-cafe, since it's getting quite off topic] "Kent Karlsson" <kentk(a)md.chalmers.se> writes: >>> for a long time. 16 bit unicode should be gotten rid of, being the worst >>> of both worlds, non backwards compatable with ascii, endianness issues >>> and no constant length encoding.... utf8 externally and utf32 when >>> worknig with individual characters is the way to go. >> I totally agree with you. > Now, what are your technical arguments for this position? > (B.t.w., UTF-16 isn't going to go away, it's very firmly established.) What's wrong with the ones already mentioned? You have endianness issues, and you need to explicitly type text files or insert BOMs. An UTF-8 stream limited to 7-bit ASCII simply is that ASCII stream. When not limited to ASCII, at least it avoids zero bytes and other potential problems. UTF-16 will among other things, be full of NULLs. I can understand UCS-2 looking attractive when it looked like a fixed-length encoding, but that no longer applies. > So it is not surprising that most people involved do not consider > UTF-16 a bad idea. The extra complexity is minimal, and further > surfaces rarely. But it needs to be there. It will introduce larger programs, more bugs, lower efficiency. > BMP characters are still (relatively) easy to process, and it saves > memory space and cache misses when large amounts of text data > is processed (e.g. databases). I couldn't find anything about the relative efficiencies of UTF-8 and UTF-16 on various languages. Do you have any pointers? From a Scandinavian POV, (using ASCII plus a handful of extra characters) UTF-8 should be a big win, but I'm sure there are counter examples. -kzm -- If I haven't seen further, it is by standing in the footprints of giants

2 2

Reasons behind the "one instance per type" limitation
by Diego Dainese 08 Oct '01

08 Oct '01

Hi, what are the reasons behind the rule stating that a type must not be declared as an instance of a particular class, more than once in the program? Thanks in advance, -- Diego To reply remove the 2 `x' from the address.

1 0

Re: UniCode
by Dylan Thurston 08 Oct '01

08 Oct '01

On Fri, Oct 05, 2001 at 11:23:50PM +1000, Andrew J Bromage wrote: > G'day all. > > On Fri, Oct 05, 2001 at 02:29:51AM -0700, Krasimir Angelov wrote: > > > Why Char is 32 bit. UniCode characters is 16 bit. > > It's not quite as simple as that. There is a set of one million > (more correctly, 1M) Unicode characters which are only accessible > using surrogate pairs (i.e. two UTF-16 codes). There are currently > none of these codes assigned, and when they are, they'll be extremely > rare. So rare, in fact, that the cost of strings taking up twice the > space that the currently do simply isn't worth the cost. This is no longer true, as of Unicode 3.1. Almost half of all characters currently assigned are outside of the BMP (i.e., require surrogate pairs in the UTF-16 encoding), including many Chinese characters. In current usage, these characters probably occur mainly in names, and are rare, but obviously important for the people involved. > However, you still need to be able to handle them. I don't know what > the "official" Haskell reasoning is (it may have more to do with word > size than Unicode semantics), but it makes sense to me to store single > characters in UTF-32 but strings in a more compressed format (UTF-8 or > UTF-16). Haskell already stores strings as lists of characters, so I see no advantage to anything other than UTF-32, since they'll take up a full machine word in any case. (Right?) There's even plenty of room for tags if any implementations want to use it. > See also: http://www.unicode.org/unicode/faq/utf_bom.html > > It just goes to show that strings are not merely arrays of characters > like some languages would have you believe. Right. In Unicode, the concept of a "character" is not really so useful; most functions that traditionally operate on characters (e.g., uppercase or display-width) fundamentally need to operate on strings. (This is due to properties of particular languages, not any design flaw of Unicode.) Err, this raises some questions as to just what the "Char" module from the standard library is supposed to do. Most of the functions are just not well-defined: isAscii, isLatin1 - OK isControl - I don't know about this. isPrint - Dubious. Is a non-spacing accent a printable character? isSpace - OK, by the comment in the report: "The isSpace function recognizes only white characters in the Latin-1 range". isUpper, isLower - Maybe OK. toUpper, toLower - Not OK. There are cases where upper casing a character yields two characters. etc. Any program using this library is bound to get confused on Unicode strings. Even before Unicode, there is much functionality missing; for instance, I don't see any way to compare strings using a localized order. Is anyone working on honest support for Unicode, in the form of a real Unicode library with an interface at the correct level? Best, Dylan Thurston

4 4

Re: Hashtable ADT
by John Hughes 06 Oct '01

06 Oct '01

> As I understand from the concepts of Functional Programming, it is not > possible to implement a Hashtable ADT in Haskell language, where one can > insert, and access values in O(1) complexity. It has to be implemented > with an external language. I don't know if it can be done in standard Haskell 98, but it can certainly be done using extensions provided by most or all implementations (IORef, IOArray). There is no need of using an external language, although it will not fit well the functional style. Better is to use the ST monad (which provides creation, reading, and writing of arrays and references in constant time). Side-effects in the ST monad can be encapsulated using runST :: (forall s. ST s a) -> a You give it a computation using updateable state as an argument, a "state thread", and it gives you the result as a pure value. The "forall s" is a typing trick which prevents an encapsulated state thread from referring to references belonging to a different state thread. Using this you can, for example, create a function share :: Hashable a => [a] -> [a] which takes a list of hashable values, and returns an equal list, in which equal values are always represented by the same pointer. Internally, there's a hash table which lets you map each element of the input to a unique representative. You could compose share with a lexical analyser, for example, to share all the copies of the same identifier. Encapsulated stateful operations like this fit very nicely with the functional style! Take a look at "State in Haskell", John Launchbury and Simon Peyton Jones, http://www.cse.ogi.edu/~jl/Papers/stateThreads.ps which explains all of this at length. John Hughes

1 0

Re: Hashtable ADT
by Marcin 'Qrczak' Kowalczyk 06 Oct '01

06 Oct '01

Sat, 6 Oct 2001 14:10:05 +0300, Cagdas Ozgenc <co19(a)cornell.edu> pisze: > As I understand from the concepts of Functional Programming, it is not > possible to implement a Hashtable ADT in Haskell language, where one can > insert, and access values in O(1) complexity. It has to be implemented > with an external language. I don't know if it can be done in standard Haskell 98, but it can certainly be done using extensions provided by most or all implementations (IORef, IOArray). There is no need of using an external language, although it will not fit well the functional style. -- __("< Marcin Kowalczyk * qrczak(a)knm.org.pl http://qrczak.ids.net.pl/ \__/ ^^ SYGNATURA ZASTÊPCZA QRCZAK

1 0

Hashtable ADT
by Cagdas Ozgenc 06 Oct '01

06 Oct '01

Hi, As I understand from the concepts of Functional Programming, it is not possible to implement a Hashtable ADT in Haskell language, where one can insert, and access values in O(1) complexity. It has to be implemented with an external language. Is my argument correct? Thanks for taking time.

1 0

Re: solutions to excercise
by S.J.Thompson 06 Oct '01

06 Oct '01

Sorry about this, but I have no online solutions. Simon

1 0

solutions to excercise
by riz er 06 Oct '01

06 Oct '01

hi iam a Computer Science student who is currently studying Haskell programming using the "Craft of Functional Programming second edition" by Simon Thompson. And i was wondering if you have the solutions to the excersise to the book. Becasue the book does not have them..a few students are having difficulty using the book because the that reason. If possible to send a url site where i can download them or a alternative way .. Thank you _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp

1 0

Re: newtype | data
by Marcin 'Qrczak' Kowalczyk 05 Oct '01

05 Oct '01

05 Oct 2001 11:53:05 -0700, Carl R. Witty <cwitty(a)newtonlabs.com> pisze: > Data> (T2 undefined) `seq` () > *** Exception: Prelude.undefined > Data> (T3 undefined) `seq` () > *** Exception: Prelude.undefined > > I can't think of a semantic difference between newtype and data > with a single unary strict constructor; It could indeed be represented in the same way, but they behave differently in pattern matching: case undefined of T _ -> () is () in the case of newtype and undefined in the case of strict data. -- __("< Marcin Kowalczyk * qrczak(a)knm.org.pl http://qrczak.ids.net.pl/ \__/ ^^ SYGNATURA ZASTÊPCZA QRCZAK

2 1

Re: newtype | data
by cwitty＠newtonlabs.com 05 Oct '01

05 Oct '01

Mark Carroll <mark(a)chaos.x-philes.com> writes: > Why does "newtype" exist, instead of letting people always use "data" and > still get maximum efficiency? After all, surely the implementation is an > implementation detail - a compiler could see the use of "data" with a > unary constructor and implement it as it does "newtype", instead of making > the programmer worry about how things are actually represented? > > I'm obviously missing something obvious here; I'm hoping to learn what. > (-: newtype is strict; data is not. Given > data T1 = T1 Int > newtype T2 = T2 Int > data T3 = T3 !Int you get the following results. Data> (T1 undefined) `seq` () () Data> (T2 undefined) `seq` () *** Exception: Prelude.undefined Data> (T3 undefined) `seq` () *** Exception: Prelude.undefined I can't think of a semantic difference between newtype and data with a single unary strict constructor; I suppose it would be possible to remove newtype from the language and make people declare things like T3 instead. Carl Witty

2 1