haskelldb-users — For questions, suggestions, bug reports and patches

Re: [Haskelldb-users] ghc 6.6.1 compilation

[Peter S. <si...@cr...>]

Bjorn Bringert writes:

 > Does anyone know how to do something like darcs' motd? Or some
 > other way to mark a CVS repo as obsolete?

Maybe it's possible to have Sourceforge disable the CVS
repository?

Anyway, I'd like to say that HaskellDB is great. I was sceptical
at first because I thought that feeding plain text SQL queries
into HSQL would be good enough, but type-safe access really is a
different thing. Thanks a lot to everyone involved for providing
this fine library as free software,

Best regards,
Peter

Re: [Haskelldb-users] ghc 6.6.1 compilation

[Bjorn B. <bri...@cs...>]

On Sep 10, 2007, at 21:07 , Peter Simons wrote:

>> darcs get http://darcs.haskell.org/haskelldb/
>
> Thanks to everyone who responded; the latest darcs version does
> work fine. It might be a good idea to empty the CVS repository at
> sourceforge (or otherwise mark it as obsolete). I didn't realize
> that it is no longer used.
>
> Best regards,
> Peter

That sounds like a good idea. I'm reluctant to delete it, but marking =20=

it as obsolete with a reference to the darcs version would be nice. =20
Does anyone know how to do something like darcs' motd? Or some other =20
way to mark a CVS repo as obsolete?

/Bj=F6rn

Re: [Haskelldb-users] ghc 6.6.1 compilation

[Peter S. <si...@cr...>]

 > darcs get http://darcs.haskell.org/haskelldb/

Thanks to everyone who responded; the latest darcs version does
work fine. It might be a good idea to empty the CVS repository at
sourceforge (or otherwise mark it as obsolete). I didn't realize
that it is no longer used. 

Best regards,
Peter

Re: [Haskelldb-users] ghc 6.6.1 compilation

[Bjorn B. <bri...@cs...>]

On Sep 9, 2007, at 2:00 , Peter Simons wrote:

> Hi,
>
> I tried to build the current CVS head with GHC 6.6.1. To make
> Cabal work, I had to change the extensions field as follows:
>
> --- a/haskelldb.cabal
> +++ b/haskelldb.cabal
> @@ -7,7 +7,7 @@ License: BSD3
>  build-depends: haskell98, base
>  Extensions: ExistentialQuantification,
>              OverlappingInstances,
> -            AllowUndecidableInstances,
> +            UndecidableInstances,
>              MultiParamTypeClasses
>  Synopsis: SQL unwrapper for Haskell.
>
> During the build, however, I get this error:
>
> src/Database/HaskellDB/Database.hs:116:16:
>     The scoped type variables `ef' and `f'
>       are bound to the same type (variable)
>       Distinct scoped type variables must be distinct
>     In the pattern: _ :: Rel (RecCons ef (Expr ea) er)
>     In the definition of `getRec':
> 	getRec vfs (_ :: Rel (RecCons ef (Expr ea) er)) (f : fs) stmt
> 		 =3D do x <- getValue vfs stmt f
> 		      r <- getRec vfs (undefined :: Rel er) fs stmt
> 		      return ((RecCons x) . r)
>     In the definition for method `getRec'
>
> I'm not sure how to fix that. Can anyone help me, please?

You said "CVS head", if that's correct, use the darcs repo instead:

darcs get http://darcs.haskell.org/haskelldb/

The currect darcs version builds with GHC 6.6, I haven't tested 6.6.1 =20=

myself.

/Bj=F6rn

[Haskelldb-users] ghc 6.6.1 compilation

[Peter S. <si...@cr...>]

Hi,

I tried to build the current CVS head with GHC 6.6.1. To make
Cabal work, I had to change the extensions field as follows:

--- a/haskelldb.cabal
+++ b/haskelldb.cabal
@@ -7,7 +7,7 @@ License: BSD3
 build-depends: haskell98, base
 Extensions: ExistentialQuantification,
             OverlappingInstances,
-            AllowUndecidableInstances,
+            UndecidableInstances,
             MultiParamTypeClasses
 Synopsis: SQL unwrapper for Haskell.

During the build, however, I get this error:

src/Database/HaskellDB/Database.hs:116:16:
    The scoped type variables `ef' and `f'
      are bound to the same type (variable)
      Distinct scoped type variables must be distinct
    In the pattern: _ :: Rel (RecCons ef (Expr ea) er)
    In the definition of `getRec':
	getRec vfs (_ :: Rel (RecCons ef (Expr ea) er)) (f : fs) stmt
		 = do x <- getValue vfs stmt f
		      r <- getRec vfs (undefined :: Rel er) fs stmt
		      return ((RecCons x) . r)
    In the definition for method `getRec'

I'm not sure how to fix that. Can anyone help me, please?

Peter

Re: [Haskelldb-users] Status of HaskellDB

[<bri...@cs...>]

Luke Plant wrote:
> Hi,

Hi Luke,

> I am trying to use HaskellDB for a real project, and due to the amount 
> of broken stuff, and the lack of any recent updates to web pages and so 
> on, is anyone in a position to tell me the status of the project -- 
> like is it dead or not?

I'd say that the code is generally not dead, but the homepage and 
documentation are. I use HaskellDB in Hope (http://hope.bringert.net/), 
and I at least keep the parts needed for that alive. That does not 
include DBDirect or the test suite.

I don't have a lot of time, so it would be very welcome if anyone would 
like to fix the homepage and/or documentation.

> Broken stuff includes:
> 
> - test suite doesn't even compile (patch attached)
>   This suggests that no-one is running tests.

True. Thanks for the patch, I've pushed it to the darcs repo.

> - I can't get test suite or DBDirect to run at all.
>   I get errors like this:
> 
> Connecting to database...
> DBDirect: user error (Couldn't load 
> Database.HaskellDB.HDBC.PostgreSQL.driver from package 
> haskelldb-hdbc-postgresql-0.10)
> 
> From the archives, it looks like this might be a problem with 
> hs-plugins, but I don't have the expertise to know how to diagnose or 
> fix this (I'm no Haskell expert, I just want a database library!).

Yes, this is a hs-plugins problem. I'm not sure what the problem is, but 
I believe it is some change of functionality in current hs-plugins 
versions. Again, help is welcome.

> Out of date information includes:
> 
> * http://haskelldb.sourceforge.net/getting-started.html -- pretty out of 
> date compared to version in darcs
> 
> * Good sources of information are supposedly:
>  http://www.scannedinavian.org/AvianWiki/HaskellDbTutorial
>  http://www.haskell.org/hawiki/HaskellDbTutorial
> 
> but both of these 404.  The old 'hawiki' wiki disappeared a month or so 
> ago, I don't know why.

I think those were pretty out of date as well by the time they disappeared.

/Björn

Re: [Haskelldb-users] Status of HaskellDB

[Shae M. E. <sh...@Sc...>]

Luke Plant <L.P...@ca...> writes:

> * http://haskelldb.sourceforge.net/getting-started.html -- pretty out of=
=20
> date compared to version in darcs
>
> * Good sources of information are supposedly:
>  http://www.scannedinavian.org/AvianWiki/HaskellDbTutorial

I looked on my old drive, but I couldn't find a copy of this.
Lemmih had updated it most recently, he might have something.

>  http://www.haskell.org/hawiki/HaskellDbTutorial

This was just a version of the previous url.

> but both of these 404.  The old 'hawiki' wiki disappeared a month or so=20
> ago, I don't know why.

The old hawiki had some misleading information, but had been made immutable=
,=20
so it was best to get rid of it.
=2D-=20
Programming is the Magic Executable Fridge Poetry, | www.ScannedInAvian.com
It is machines made of thought, fueled by ideas.   | -- Shae Matijs Erisson

[Haskelldb-users] Status of HaskellDB

[Luke P. <L.P...@ca...>]

Hi,

I am trying to use HaskellDB for a real project, and due to the amount=20
of broken stuff, and the lack of any recent updates to web pages and so=20
on, is anyone in a position to tell me the status of the project --=20
like is it dead or not?

Broken stuff includes:

=2D test suite doesn't even compile (patch attached)
  This suggests that no-one is running tests.

=2D I can't get test suite or DBDirect to run at all.
  I get errors like this:

Connecting to database...
DBDirect: user error (Couldn't load=20
Database.HaskellDB.HDBC.PostgreSQL.driver from package=20
haskelldb-hdbc-postgresql-0.10)

=46rom the archives, it looks like this might be a problem with=20
hs-plugins, but I don't have the expertise to know how to diagnose or=20
fix this (I'm no Haskell expert, I just want a database library!).

Out of date information includes:

* http://haskelldb.sourceforge.net/getting-started.html -- pretty out of=20
date compared to version in darcs

* Good sources of information are supposedly:
 http://www.scannedinavian.org/AvianWiki/HaskellDbTutorial
 http://www.haskell.org/hawiki/HaskellDbTutorial

but both of these 404.  The old 'hawiki' wiki disappeared a month or so=20
ago, I don't know why.

Thanks for your time,

Luke

=2D-=20
"If we could just get everyone to close their eyes and visualise world
 peace for an hour, imagine how serene and quiet it would be until the
 looting started" -- Anon

Luke Plant || http://lukeplant.me.uk/

[Haskelldb-users] This caused problems with different characters sets that also used these codes.

[Lepinsy M. <Lep...@an...>]

I don't think its the drivers, because I can get a full list of table in VB.

Re: [Haskelldb-users] Parse SQL queries into SqlSelect

[<bri...@cs...>]

Henning Thielemann wrote:
> Is there some routine with signature
>  String -> SqlSelect
>   for converting an SQL query into the SqlSelect structure?

No. But you are very welcome to write one! It sounds like a fun project.

/Björn

[Haskelldb-users] Parse SQL queries into SqlSelect

[Henning T. <le...@he...>]

Is there some routine with signature
 String -> SqlSelect
  for converting an SQL query into the SqlSelect structure?

Re: [Haskelldb-users] Using Savepoints

[Bjorn B. <bri...@cs...>]

On Jan 24, 2007, at 22:50 , Marco T=FAlio Gontijo e Silva wrote:

> Hello,
>
> I'm using haskelldb with postgresql and I'd like to use savepoints. I
> couldn't find a way to use it or to hack sql to do it by hand.
>
> Is it possible?
>
> I'm sending this question here because I want to be sure before
> migrating to HDBC.

Hi Marco,

There is no support for savepoints in the current HaskellDB. It would =20=

be interesting to add it, with some nice API. It should probably =20
generate the savepoint name automatically, and return some handle =20
that can be passed to rollbackTo. Patches are welcome!

/Bj=F6rn=

[Haskelldb-users] Using Savepoints

[Marco G. e S. <mal...@ri...>]

Hello,

I'm using haskelldb with postgresql and I'd like to use savepoints. I
couldn't find a way to use it or to hack sql to do it by hand.

Is it possible?

I'm sending this question here because I want to be sure before
migrating to HDBC.

Thanks.

--=20
malebria
Marco T=C3=BAlio Gontijo e Silva
Correio (MSN): mal...@ri...
Jabber (GTalk): mal...@ja...
Ekiga: mal...@ek...
IRC: mal...@ir...
     mal...@ir...
Skype: marcotmarcot
Telefone: 33346720
Celular: 98116720
Endere=C3=A7o:
    Rua Paula C=C3=A2ndido, 257/201
    Gutierrez 30430-260
    Belo Horizonte/MG Brasil

Re: [Haskelldb-users] Intel Mac Build Problems

[<bri...@cs...>]

Kelly Norton wrote:
> Hi folks,
>=20
> I haven't had very much luck building haskelldb on an intel mac. The=20
> error that finally put an end to forward progress is copied below. I am=
=20
> running ghc-6.6 and ./configure seemed to like my setup (although I did=
=20
> have to run ./configure --build i386-apple-macosx8.8.0, since=20
> config.guess couldn't make out my mac). The summary given by ./configur=
e=20
> was:
>=20
>=20
> HaskellDB will support:
> -----------------------
>=20
> GHC:               yes
>   HSQL:            yes
>     ODBC:          no
>     MySql:         yes
>     SQLite:        no
>     PostgreSQL:    no
>   WxHaskell:       no
> Hugs:              no
>   HSQL:            no
>     ODBC:          no
>     MySql:         no
>     SQLite:        no
>     PostgreSQL:    no
>   WxHaskell:       no
>=20
> The actual error emitted during make.
>=20
> /usr/local/bin/ghc -fglasgow-exts -fallow-overlapping-instances -packag=
e=20
> hsql  -i../build/ghc -package-name haskelldb -c=20
> ../build/ghc/Database/HaskellDB/HDBRec.hs
>=20
> ../build/ghc/Database/HaskellDB/HDBRec.hs:90:0:
>     Illegal instance declaration for `RecCat (RecCons f a r1)
>                                              r2
>                                              (RecCons f a r3)'
>         (the Coverage Condition fails for one of the functional=20
> dependencies)
>     In the instance declaration for `RecCat (RecCons f a r1) r2 (RecCon=
s=20
> f a r3)'
>=20
> ../build/ghc/Database/HaskellDB/HDBRec.hs:93:0:
>     Illegal instance declaration for `RecCat (Record r1)
>                                              (Record r2)
>                                              (Record r3)'
>         (the Coverage Condition fails for one of the functional=20
> dependencies)
>     In the instance declaration for `RecCat (Record r1) (Record r2)=20
> (Record r3)'
> make[2]: *** [../build/ghc/Database/HaskellDB/HDBRec.o] Error 1
> make[1]: *** [all-ghc] Error 2
> make: *** [src] Error 2
>=20
>=20
> If anyone has any thoughts, I would greatly appreciate the help.
> /kel

This seems to be the release version of HaskellDB, which is quite=20
outdated. The current darcs version should build with 6.6 and not have=20
problems with Intel Macs. You can get it with:

darcs get --partial http://darcs.haskell.org/haskelldb/

/Bj=F6rn

[Haskelldb-users] Intel Mac Build Problems

[Kelly N. <ni...@ke...>]

Hi folks,

I haven't had very much luck building haskelldb on an intel mac. The error
that finally put an end to forward progress is copied below. I am running
ghc-6.6 and ./configure seemed to like my setup (although I did have to run
./configure --build i386-apple-macosx8.8.0, since config.guess couldn't make
out my mac). The summary given by ./configure was:


HaskellDB will support:
-----------------------

GHC:               yes
  HSQL:            yes
    ODBC:          no
    MySql:         yes
    SQLite:        no
    PostgreSQL:    no
  WxHaskell:       no
Hugs:              no
  HSQL:            no
    ODBC:          no
    MySql:         no
    SQLite:        no
    PostgreSQL:    no
  WxHaskell:       no

The actual error emitted during make.

/usr/local/bin/ghc -fglasgow-exts -fallow-overlapping-instances -package
hsql  -i../build/ghc -package-name haskelldb -c
../build/ghc/Database/HaskellDB/HDBRec.hs

../build/ghc/Database/HaskellDB/HDBRec.hs:90:0:
    Illegal instance declaration for `RecCat (RecCons f a r1)
                                             r2
                                             (RecCons f a r3)'
        (the Coverage Condition fails for one of the functional
dependencies)
    In the instance declaration for `RecCat (RecCons f a r1) r2 (RecCons f a
r3)'

../build/ghc/Database/HaskellDB/HDBRec.hs:93:0:
    Illegal instance declaration for `RecCat (Record r1)
                                             (Record r2)
                                             (Record r3)'
        (the Coverage Condition fails for one of the functional
dependencies)
    In the instance declaration for `RecCat (Record r1) (Record r2) (Record
r3)'
make[2]: *** [../build/ghc/Database/HaskellDB/HDBRec.o] Error 1
make[1]: *** [all-ghc] Error 2
make: *** [src] Error 2


If anyone has any thoughts, I would greatly appreciate the help.
/kel

[Haskelldb-users] PostgreSQL timestamp columns, with HDBC

[Matthew S. <ma...@we...>]

Hi,

I've been using the hdbc-postgresql driver, and there are problems with
pulling timestamp values out of the database as they arrive as strings
and then get interpreted as ints which causes errors.

So, attached is a darcs patch which solves this problem. It adds a
dependancy on parsedate, from

http://www.cs.chalmers.se/~bringert/darcs/parsedate/doc/api/System-Time-Parse.html

I'm not convinced my changes are in the right place - being in Types.hs
makes it common to all the HDBC drivers. Plus the format string is going
to vary between databases and installations, versions and configurations
of PostgreSQL.

On the other hand, it does at least work for me! Any comments gratefully
received. The patch should apply in the hdbc directory.

Matthew
-- 
Matthew Sackman

BOFH excuse #343:
The ATM board has run out of 10 pound notes.  We are having a whip
round to refill it, care to contribute ?

[Haskelldb-users] dbSpecToDatabase, CalendarTimeT and nullable?

[Marc W. <mar...@gm...>]

Hi.

I've noticed that this piece of code doesn't create a nullable field finished_date. Why?


[...]
--con :: (D.Database -> IO a) -> IO a
con = mysqlConnect ( MySQLOptions {server="localhost", db=dbName, uid="sadmin", pwd=""})

createDatabase :: IO ()
createDatabase= do
  con $ \a -> dbSpecToDatabase a dbinfo
  where tblTodoItems=
	     [ CInfo {cname="ID", descr = (IntT, True)}
	     , CInfo {cname="ID2", descr = (IntT, True)}
	     , CInfo {cname="finished_date", descr = (CalendarTimeT, True)} -- !!!! <<<<<<<<<<<<<<<<<<<<<< here
	      ]
	tblTodoAssoc = TInfo { tname="todos", cols=tblTodoItems }
	testopts = DBOptions {useBString = False}
	dbinfo::DBInfo
	dbinfo = DBInfo {dbname = "todos"  , opts = testopts, tbls = [tblTodoAssoc] }

------------------

Another question.
Is 
  con $ \db -> insert db todos ( (iD << (constant id))
                               # (iD2 << (constant Nothing))
                               # (finished_date << (constant Nothing))
                               )
the way to insert a new record? ( In hope it's used this way so I think yes.. )
Then let me ask the question wether it would be possible to create a function like

insertValues3 a b c = ( iD << a # iD2 << b # finished_date << c)
to use the insert query like this:
con $ \db -> insert db todosu( insertValues3 (constant id) (constant Nothing) (constant Nothing) )

because order and types are defined by the table description so there is no
need to repeat "iD <<", "iD2 <<" and "finished_date <<" again, right ?

Sincerely
	Marc Weber

Re: [Haskelldb-users] What about introducing a Connection type declaration?

[Marc W. <mar...@gm...>]

I tracked it down. The problem was that I couldn't use the same con type
for two different tables..
All the time I was only thinking about the return value.. which was IO
() in either case
The trick was to pass another con value for eaach table ;)

Marc

Re: [Haskelldb-users] What about introducing a Connection type declaration?

[Marc W. <mar...@gm...>]

I had/have some problems with types of haskelldb so I  tried to add
type annotations.

I've used Select query mainly yet. Thus my code looks like this:

con = mysqlConnect <params>

doSomethinng con = con $ \db -> query db $  table <table>

Then I wanted to add type annotations. Thus I took the type of
	mysqlConnect :: MonadIO m => MySQLOptions -> (Database -> m a) -> m a
and removed the connection settings yielding
	mysqlConnect :: MonadIO m => (Database -> m a) -> m a
which is very much to type/reead IMHO. That's why I proposed
introducnig:
newtype CConnection :: MonadIO m => (Database -> m a) -> m a

I'm struggling here now:

line 39  is equal to 56
line 39 works fine
line 56 doesn't. Why?
Isn't  both a IO monad (because of the print statements)
After commenting out 56 it compiles.


module Modules.ObjectTree where
import Debug.Trace
import Data.FunctorM
import DBUtils
import qualified DB.VT.Ezcontentobject_tree as EOT
import qualified DB.VT.Ezcontentobject as CO
import Database.HaskellDB.HDBRec
import Database.HaskellDB
import Database.HaskellDB.Query as Q
import Data.Tree
import Monad
import Control.Monad.Trans
import Maybe
import qualified List
 
instance FunctorM Tree where
  fmapM f (Node a forest) = do
    a' <- f a
    forest' <- mapM (fmapM f) forest
    return $ Node a' forest'
 
 
type ObjectTree a = Tree  (Record a)
 
truncTree 1 (Node a _) = Node a []
truncTree x (Node a forest) = Node a $ map (truncTree (x-1)) forest
 
oT con = do
  print "blah" -- because of this we  should  have a simple IO Monad ?
  lookupField con (CO.name) (CO.ezcontentobject)  (CO.xid) (constant (1 :: Int)) >>= print :: IO () -- <<<<<<<<<<<<<<<<<<<<<<<<< 39
  return "blah"
 
-- printObjectsAsTree :: MonadIO m => ((Database -> m a) -> m a) -> Int -> IO ()
printObjectsAsTree con startid= do --<<<<<<<<<<<<<<<< 4
  print "test"
  root <- liftM head  $ lRS (EOT.parent_node_id) (constant (startid :: Int))
  print root
  --showRS root >>= putStrLn
  node <- po root
  node_show <- fmapM showRS node
  return  $ drawTree node_show
  -- return "end"
  where lRS = lookupFieldRS con (EOT.ezcontentobject_tree)
	po root = let root_id = (root!(EOT.node_id) :: Int)
		  in do  print "dumm" -- IO Monad too ?
			 print (root!(EOT.node_id))
			 lookupField con (CO.name) (CO.ezcontentobject)  (CO.xid) (constant (1 :: Int)) >>= print --  <<<<<<<<<<<<<<    56
			 return $ Node root []
			 --childs <- lRS (EOT.parent_node_id) (constant root_id)
			 --mapM_ (\r -> r!(EOT.node_id)) childs >>= print
        showRS r = do -- name <- lookupField con (CO.name) (CO.ezcontentobject) (CO.xid) (constant 1) >>= print
		      return "ab" :: IO String
		      --return $ (show $ r!node_id) ++ " (" ++ (fromJust name) ++ " )"
 
-----------------------------------------------
|| Preprocessing executables for dbez-0.0...
|| Building dbez-0.0...
|| Chasing modules from: db_ez.hs
|| [1 of 6] Skipping  DBUtils          ( DBUtils.hs, dist/build/db_ez/db_ez-tmp/DBUtils.o )
|| [2 of 6] Skipping  DB.VT.Ezcontentobject_tree ( DB/VT/Ezcontentobject_tree.hs, dist/build/db_ez/db_ez-tmp/DB/VT/Ezcontentobject_tree.o )
|| [3 of 6] Skipping  DB.VT.Ezcontentobject ( DB/VT/Ezcontentobject.hs, dist/build/db_ez/db_ez-tmp/DB/VT/Ezcontentobject.o )
|| [4 of 6] Compiling Modules.ObjectTree ( Modules/ObjectTree.hs, dist/build/db_ez/db_ez-tmp/Modules/ObjectTree.o )
|| 
Modules/ObjectTree.hs|43| 0:
||     Couldn't match `DB.VT.Ezcontentobject_tree.Contentobject_id'
|| 	   against `DB.VT.Ezcontentobject.Contentclass_id'
||       Expected type: RecCons DB.VT.Ezcontentobject_tree.Contentobject_id
|| 			     (Maybe Int)
|| 			     vr
||       Inferred type: RecCons DB.VT.Ezcontentobject.Contentclass_id
|| 			     Int
|| 			     vr1
||     When using functional dependencies to combine
||       Database.HaskellDB.Database.GetRec (RecCons f (Expr a) er)
|| 					 (RecCons f a vr),
|| 	arising from the instance declaration at Imported from Database.HaskellDB.Database
||       Database.HaskellDB.Database.GetRec (RecCons DB.VT.Ezcontentobject_tree.Contentobject_id
|| 						  (Expr (Maybe Int))
													[...]
|| 																						(RecCons DB.VT.Ezcontentobject_tree.Sort_order
|| 																							 (Expr (Maybe Int))
|| 																							 RecNil))))))))))))))))
|| 					 (RecCons DB.VT.Ezcontentobject.Contentclass_id Int vr),
arising from use of `lookupFieldRS' at Modules/ObjectTree.hs|52| 14-26
||     When generalising the type(s) for `printObjectsAsTree'
_______________________________________________
Haskell-Cafe mailing list
Has...@ha...
http://www.haskell.org/mailman/listinfo/haskell-cafe

Re: [Haskelldb-users] What about introducing a Connection type declaration?

[Bjorn B. <bri...@cs...>]

On Aug 6, 2006, at 5:15 PM, Marc Weber wrote:

> eg mysqlConnect is of type:
> mysqlConnect :: MonadIO m =3D> MySQLOptions -> (Database -> m a) -> m =
a
>
> Typical usage looks like this, doesn't it?
>
>
> 	doSomething :: MonadIO m =3D> (Database -> m a) -> m a -> IO () =
-- <<
> 	main=3Ddo
> 		con <- mysqlConnect Options
> 		doSomething con ..
>
> Wouln't it be mucht clearer to use another typedef to write =20
> something like this ?
>
>
> 	newtype CConnection :: MonadIO m =3D> (Database -> m a) -> m a
> 	doSomething :: CConnection -> IO ()  -- <<  much shorter and =
clearer
> 	main=3Ddo
> 		con <- mysqlConnect Options
> 		doSomething con ..
>
> The first C migt indicate that it's a "connected" connection =20
> (parameters have already been applied)..
> Or does this already exist?
>
> I'd like to provide a patch if you consider it useful
>
> Greetings
> Marc Weber

Hi Marc,

I'm not sure that I completely understand your example. I think that =20
there is something wrong with the types.

Here's how I think a typical application looks with the current types:

options :: MysqlOptions
options =3D ...

doSomething :: Database -> IO ()
doSomething db =3D ...

main :: IO ()
main =3D mysqlConnect options doSomething


I probably misunderstood what you want to do though.

/Bj=F6rn

[Haskelldb-users] What about introducing a Connection type declaration?

[Marc W. <mar...@gm...>]

eg mysqlConnect is of type:
mysqlConnect :: MonadIO m => MySQLOptions -> (Database -> m a) -> m a

Typical usage looks like this, doesn't it?


	doSomething :: MonadIO m => (Database -> m a) -> m a -> IO () -- <<
	main=do
		con <- mysqlConnect Options
		doSomething con ..

Wouln't it be mucht clearer to use another typedef to write something like this ?


	newtype CConnection :: MonadIO m => (Database -> m a) -> m a
	doSomething :: CConnection -> IO ()  -- <<  much shorter and clearer
	main=do
		con <- mysqlConnect Options
		doSomething con ..

The first C migt indicate that it's a "connected" connection (parameters have already been applied)..
Or does this already exist?

I'd like to provide a patch if you consider it useful

Greetings
Marc Weber

Re: [Haskelldb-users] BoundedList ?

[Bjorn B. <bri...@cs...>]

On Jul 28, 2006, at 2:26 PM, Marc Weber wrote:

> I've tried understanding what the BoundedList.hs module is all =20
> about but
> I couldn' get it completely
>
> -- This inductive approach gives the complexity O(n) on the number =20
> of instances
> -- and very short type checking times compared to an O(n\^2) =20
> implementation.
>
> Where is the difference between [a] and (BoundedList a Nxx)?
> Is it only used for BoundedString by now?
> (to prevent data loss when copying from char(10) to char(4)?

Yes, BoundedList is only used for BoundedString in HaskellDB. The =20
purpose of BoundedString is, as you guessed, to avoid string =20
truncation. For more information, see our Haskell Workshop paper, =20
http://haskelldb.sourceforge.net/haskelldb.pdf

> Did I get it right that the internal representation is still an =20
> ordinary
> list ? (newtype BoundedList a n =3D BoundedList [a])?

Yes. BoundedList uses a phantom type to keep the string max length in =20=

the type of the list. The internal representation is a normal list.

> Looking at
>
> instance (Show a, Size n) =3D> Show (BoundedList a n) where
>     show l@(L xs) =3D show xs
>
> l@ is superfluous, isn't it?

Yes. There are probably lots of unused bindings in the HaskellDB =20
code. I guess we should compile with -Wall and fix all the warnings.

/Bj=F6rn

Re: [Haskelldb-users] HaskellDB and joins

[Bjorn B. <bri...@cs...>]

On Jul 28, 2006, at 11:45 AM, Daniel Larsson wrote:

> I started looking at HaskellDB recently, and it's quite =20
> interesting. One
> thing I would like to do, which I can't find any support for, is outer
> joins. Is this something that is planned to support, or have I missed
> something?

You are perfectly right that outer join support is missing. It's the =20
query language feature that I miss the most, and I would like to =20
implement it. The problem is that it seems a bit difficult to figure =20
out how to do it in the current version of HaskellDB, since with an =20
outer join, the fields form the second table can be NULL, even if =20
they are not nullable. This is one of the reasons why I believe that =20
HaskellDB could benefit from moving to a non-monadic query language =20
closer to that of CoddFish [1].

/Bj=F6rn

[1] http://lmf.di.uminho.pt/wiki/bin/view/PURe/CoddFish

[Haskelldb-users] BoundedList ?

[Marc W. <mar...@gm...>]

I've tried understanding what the BoundedList.hs module is all about but
I couldn' get it completely

-- This inductive approach gives the complexity O(n) on the number of instances
-- and very short type checking times compared to an O(n\^2) implementation.

Where is the difference between [a] and (BoundedList a Nxx)?
Is it only used for BoundedString by now?
(to prevent data loss when copying from char(10) to char(4)?

Did I get it right that the internal representation is still an ordinary
list ? (newtype BoundedList a n = BoundedList [a])?

Looking at

instance (Show a, Size n) => Show (BoundedList a n) where
    show l@(L xs) = show xs

l@ is superfluous, isn't it?

Marc

[Haskelldb-users] HaskellDB and joins

[Daniel L. <dan...@se...>]

I started looking at HaskellDB recently, and it's quite interesting. One
thing I would like to do, which I can't find any support for, is outer
joins. Is this something that is planned to support, or have I missed
something?