From voldermort at hotmail.com  Tue Sep  2 07:27:25 2014
From: voldermort at hotmail.com (Jeremy)
Date: Tue, 2 Sep 2014 07:27:25 +0000 (UTC)
Subject: [Haskell-beginners] Stackage latest snapshot
Message-ID: <loom.20140902T092509-732@post.gmane.org>

I would like to point my cabal file to the latest Stackage snapshot. I seem
to recall that there is a link which will automatically resolve to this, but
I can no longer find it.

Does this link still exist? Is there a good reason for not doing so? I
understand that this will introduce a little instability, but I'm OK dealing
with that in return for not getting stuck on the same package versions forever.


From jcb at iteris.com  Tue Sep  2 18:19:23 2014
From: jcb at iteris.com (Jeff C. Britton)
Date: Tue, 2 Sep 2014 18:19:23 +0000
Subject: [Haskell-beginners] Haskeline and forkIO
In-Reply-To: <87ppfkbt6k.fsf@pmade.com>
References: <4129696F7B970B4FA7C343A1587763E711B479A3@greyhound.iteris.com>
 <ltk6o5$vaf$1@ger.gmane.org>
 <4129696F7B970B4FA7C343A1587763E711B47C7E@greyhound.iteris.com>
 <87ppfkbt6k.fsf@pmade.com>
Message-ID: <4129696F7B970B4FA7C343A1587763E711B47EAF@greyhound.iteris.com>

Thanks Peter,

That suggestion works.
I will have to continue learning more about Monads.

--Jeff


-----Original Message-----
From: Beginners [mailto:beginners-bounces at haskell.org] On Behalf Of Peter Jones
Sent: Thursday, August 28, 2014 7:11 AM
To: beginners at haskell.org
Subject: Re: [Haskell-beginners] Haskeline and forkIO

"Jeff C. Britton" <jcb at iteris.com> writes:
> loop :: InputT IO ()
> loop = do
>     maybeLine <- getInputLine "Enter a file to compress> "
>     case maybeLine of
>       Nothing -> return ()      -- user entered EOF
>       Just "" -> return ()      -- treat no name as "want to quit" 
>       Just path -> do
>             return (runWorker path)
>             loop

The other issue you're having is because `runWorker path` is an `IO ()` value but at the point where you use it in the code the type system wants an `InputT IO ()`.  To try to satisfy the type system you used `return` to build a `InputT IO (IO ())` value, but that doesn't actually work (as you've noticed).  Since `InputT` is a transformer you have an extra layer to work through and so need to *lift* your `IO ()` value into the `InputT IO` layer.  Try this:

    -- Add this import
    import Control.Monad.IO.Class
    
    loop :: InputT IO ()
    loop = do
        maybeLine <- getInputLine "Enter a file to compress> "
        case maybeLine of
          Nothing -> return ()      -- user entered EOF
          Just "" -> return ()      -- treat no name as "want to quit" 
          Just path -> do
                liftIO (runWorker path)
                loop
    

You can think of `liftIO` as having this signature (in this context):

    liftIO :: IO () -> InputT IO ()
    
--
Peter Jones, Founder, Devalot.com
Defending the honor of good code

_______________________________________________
Beginners mailing list
Beginners at haskell.org
http://www.haskell.org/mailman/listinfo/beginners

From sgf.dma at gmail.com  Tue Sep  2 19:32:35 2014
From: sgf.dma at gmail.com (Dmitriy Matrosov)
Date: Tue, 02 Sep 2014 23:32:35 +0400
Subject: [Haskell-beginners] Check constructor's field numeric value at
	compile time
Message-ID: <54061B53.7060106@gmail.com>

Hi.

I have a type

data Interval	= Seconds Float
		| MicroSeconds Int

The Float field of data constructor Seconds should be >= 1, and Int 
field of constructor MicroSeconds should be in the range from 0 to 1000000.

How can i write this constraints so they're checked at compile time, not 
at runtime?

--
     Dmitriy Matrosov

From alexandrelucchesi at gmail.com  Tue Sep  2 20:58:24 2014
From: alexandrelucchesi at gmail.com (Alexandre Lucchesi)
Date: Tue, 2 Sep 2014 17:58:24 -0300
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <54061B53.7060106@gmail.com>
References: <54061B53.7060106@gmail.com>
Message-ID: <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>

You can't do that within the data type declaration.

In order to add such constraints you should define "constructor functions"
and apply the validation there, i.e.:

newSecondsInterval :: Float -> Maybe Interval
newSecondsInterval n
  | n >= 1     = Just $ Seconds n
  | otherwise = Nothing

Note that the "Maybe" type is only a way to handle the case where the
supplied value is invalid. Also, in order to provide encapsulation and
ensure no one is going to create a new "Interval" by using "Seconds" and
"MicroSeconds", you should hide these value constructors in the export list
of your module.


On Tue, Sep 2, 2014 at 4:32 PM, Dmitriy Matrosov <sgf.dma at gmail.com> wrote:

> Hi.
>
> I have a type
>
> data Interval   = Seconds Float
>                 | MicroSeconds Int
>
> The Float field of data constructor Seconds should be >= 1, and Int field
> of constructor MicroSeconds should be in the range from 0 to 1000000.
>
> How can i write this constraints so they're checked at compile time, not
> at runtime?
>
> --
>     Dmitriy Matrosov
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>



-- 
alexandre lucchesi

*Perfection is achieved, not when there is nothing more to add, but when
there is nothing left to take away!*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140902/d1f58cbf/attachment.html>

From julian.birch at gmail.com  Tue Sep  2 22:02:48 2014
From: julian.birch at gmail.com (Julian Birch)
Date: Tue, 2 Sep 2014 23:02:48 +0100
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>
References: <54061B53.7060106@gmail.com>
 <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>
Message-ID: <CAB0TuzCBwMy5XPv4uqQSZUZQGTpLq6uT8AzdCa38HZBiMJMOBA@mail.gmail.com>

Also, if you want your mind blown, check out Idris.

On Tuesday, September 2, 2014, Alexandre Lucchesi <
alexandrelucchesi at gmail.com> wrote:

> You can't do that within the data type declaration.
>
> In order to add such constraints you should define "constructor functions"
> and apply the validation there, i.e.:
>
> newSecondsInterval :: Float -> Maybe Interval
> newSecondsInterval n
>   | n >= 1     = Just $ Seconds n
>   | otherwise = Nothing
>
> Note that the "Maybe" type is only a way to handle the case where the
> supplied value is invalid. Also, in order to provide encapsulation and
> ensure no one is going to create a new "Interval" by using "Seconds" and
> "MicroSeconds", you should hide these value constructors in the export list
> of your module.
>
>
> On Tue, Sep 2, 2014 at 4:32 PM, Dmitriy Matrosov <sgf.dma at gmail.com
> <javascript:_e(%7B%7D,'cvml','sgf.dma at gmail.com');>> wrote:
>
>> Hi.
>>
>> I have a type
>>
>> data Interval   = Seconds Float
>>                 | MicroSeconds Int
>>
>> The Float field of data constructor Seconds should be >= 1, and Int field
>> of constructor MicroSeconds should be in the range from 0 to 1000000.
>>
>> How can i write this constraints so they're checked at compile time, not
>> at runtime?
>>
>> --
>>     Dmitriy Matrosov
>> _______________________________________________
>> Beginners mailing list
>> Beginners at haskell.org
>> <javascript:_e(%7B%7D,'cvml','Beginners at haskell.org');>
>> http://www.haskell.org/mailman/listinfo/beginners
>>
>
>
>
> --
> alexandre lucchesi
>
> *Perfection is achieved, not when there is nothing more to add, but when
> there is nothing left to take away!*
>


-- 
Sent from an iPhone, please excuse brevity and typos.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140902/bfae26be/attachment.html>

From ky3 at atamo.com  Wed Sep  3 03:04:10 2014
From: ky3 at atamo.com (Kim-Ee Yeoh)
Date: Wed, 3 Sep 2014 10:04:10 +0700
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>
References: <54061B53.7060106@gmail.com>
 <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>
Message-ID: <CAPY+ZdRcp1=_GDXrGOt5OtkwD-1ndk+=nsYH82Laa88KAmpqVw@mail.gmail.com>

On Wed, Sep 3, 2014 at 3:58 AM, Alexandre Lucchesi <
alexandrelucchesi at gmail.com> wrote:

> In order to add such constraints you should define "constructor functions"
> and apply the validation there, i.e.:


To OP: Yes, the searchable term is "smart constructors". However, it has
been used at least once on this list to mean something different, so look
out.

-- Kim-Ee
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/adfa3d0f/attachment.html>

From alexandrelucchesi at gmail.com  Wed Sep  3 03:07:10 2014
From: alexandrelucchesi at gmail.com (Alexandre Lucchesi)
Date: Wed, 3 Sep 2014 00:07:10 -0300
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <CAPY+ZdRcp1=_GDXrGOt5OtkwD-1ndk+=nsYH82Laa88KAmpqVw@mail.gmail.com>
References: <54061B53.7060106@gmail.com>
 <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>
 <CAPY+ZdRcp1=_GDXrGOt5OtkwD-1ndk+=nsYH82Laa88KAmpqVw@mail.gmail.com>
Message-ID: <CAGqBpeUaPeUPYWKjhjRqdHYEnt-s=48uzgqTnVpqN=SmHQCdPA@mail.gmail.com>

Hmm.. Nice to know about that!

Thank you. :)


On Wed, Sep 3, 2014 at 12:04 AM, Kim-Ee Yeoh <ky3 at atamo.com> wrote:

>
> On Wed, Sep 3, 2014 at 3:58 AM, Alexandre Lucchesi <
> alexandrelucchesi at gmail.com> wrote:
>
>> In order to add such constraints you should define "constructor
>> functions" and apply the validation there, i.e.:
>
>
> To OP: Yes, the searchable term is "smart constructors". However, it has
> been used at least once on this list to mean something different, so look
> out.
>
> -- Kim-Ee
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>


-- 
alexandre lucchesi

*Perfection is achieved, not when there is nothing more to add, but when
there is nothing left to take away!*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/a15f537c/attachment.html>

From allbery.b at gmail.com  Wed Sep  3 03:07:31 2014
From: allbery.b at gmail.com (Brandon Allbery)
Date: Tue, 2 Sep 2014 23:07:31 -0400
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <CAPY+ZdRcp1=_GDXrGOt5OtkwD-1ndk+=nsYH82Laa88KAmpqVw@mail.gmail.com>
References: <54061B53.7060106@gmail.com>
 <CAGqBpeVraecCSseqQW1M4cYRv9O2J+g0BXCXZFSTsJ7xSQMCYQ@mail.gmail.com>
 <CAPY+ZdRcp1=_GDXrGOt5OtkwD-1ndk+=nsYH82Laa88KAmpqVw@mail.gmail.com>
Message-ID: <CAKFCL4We5Vdavt0FJKxLPSytuq-VW4-jw-nF7jjvBGyY1tovYA@mail.gmail.com>

On Tue, Sep 2, 2014 at 11:04 PM, Kim-Ee Yeoh <ky3 at atamo.com> wrote:
>
> On Wed, Sep 3, 2014 at 3:58 AM, Alexandre Lucchesi <
> alexandrelucchesi at gmail.com> wrote:
>
>> In order to add such constraints you should define "constructor
>> functions" and apply the validation there, i.e.:
>
>
> To OP: Yes, the searchable term is "smart constructors". However, it has
> been used at least once on this list to mean something different, so look
> out.
>

Also note that this will not actually check at compile time; to do that you
need type level stuff that ghc is still finding its way through.

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b at gmail.com                                  ballbery at sinenomine.net
unix, openafs, kerberos, infrastructure, xmonad        http://sinenomine.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140902/87d73d10/attachment-0001.html>

From ky3 at atamo.com  Wed Sep  3 03:07:53 2014
From: ky3 at atamo.com (Kim-Ee Yeoh)
Date: Wed, 3 Sep 2014 10:07:53 +0700
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <54061B53.7060106@gmail.com>
References: <54061B53.7060106@gmail.com>
Message-ID: <CAPY+ZdQcNhUoh39Xh6A7wNQAj_KBDM4hF8Zdoan=fGRBsWs0iw@mail.gmail.com>

On Wed, Sep 3, 2014 at 2:32 AM, Dmitriy Matrosov <sgf.dma at gmail.com> wrote:

> How can i write this constraints so they're checked at compile time, not
> at runtime?


Smart constructors won't check them at compile time but it does the next
best thing, which is validate at the earliest possible moment.



-- Kim-Ee
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/9f6c7331/attachment.html>

From michael at snoyman.com  Wed Sep  3 04:50:09 2014
From: michael at snoyman.com (Michael Snoyman)
Date: Wed, 3 Sep 2014 07:50:09 +0300
Subject: [Haskell-beginners] Stackage latest snapshot
In-Reply-To: <loom.20140902T092509-732@post.gmane.org>
References: <loom.20140902T092509-732@post.gmane.org>
Message-ID: <CAKA2Jg+YwhoNfC1wEMaRS4OnxNcpUr7PaPAY_Nvaf9_oLN7xDA@mail.gmail.com>

On Tue, Sep 2, 2014 at 10:27 AM, Jeremy <voldermort at hotmail.com> wrote:

> I would like to point my cabal file to the latest Stackage snapshot. I seem
> to recall that there is a link which will automatically resolve to this,
> but
> I can no longer find it.
>
> Does this link still exist? Is there a good reason for not doing so? I
> understand that this will introduce a little instability, but I'm OK
> dealing
> with that in return for not getting stuck on the same package versions
> forever.
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>


The preparing your system for Stackage page[1] has those auto-updated links
available (e.g., GHC 7.8 exclusive is [2]). You already identified the
downside with this: you may magically be upgraded to a new version of a
package. If you really want to do this, you can, but my recommendation is
to manually decide to upgrade your Stackage snapshot consciously instead of
letter an arbitrary Jenkins build server make the decision for you.

Michael

[1]
https://github.com/fpco/stackage/wiki/Preparing-your-system-to-use-Stackage
[2] http://www.stackage.org/alias/fpcomplete/unstable-ghc78-exclusive
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/935540b3/attachment.html>

From voldermort at hotmail.com  Wed Sep  3 07:52:24 2014
From: voldermort at hotmail.com (Peter)
Date: Wed, 3 Sep 2014 07:52:24 +0000 (UTC)
Subject: [Haskell-beginners] Stackage latest snapshot
References: <loom.20140902T092509-732@post.gmane.org>
 <CAKA2Jg+YwhoNfC1wEMaRS4OnxNcpUr7PaPAY_Nvaf9_oLN7xDA@mail.gmail.com>
Message-ID: <loom.20140903T081134-646@post.gmane.org>

Michael Snoyman <michael <at> snoyman.com> writes:

> The preparing your system for Stackage page[1] has those auto-updated
links available (e.g., GHC 7.8 exclusive is [2]).

Thank you. I read the page at least 3 times, but missed that because the
links redirected to the latest snapshot, and the Stackage server only has
static snapshot links. Perhaps it should be made clearer that these links
will redirect to the current snapshot, and add them to
http://www.stackage.org/snapshots as well?

> You already identified the downside with this: you may magically be
upgraded to a new version of a package. If you really want to do this, you
can, but my recommendation is to manually decide to upgrade your Stackage
snapshot consciously instead of letter an arbitrary Jenkins build server
make the decision for you.

I'm using Stackage as a QA'd Hackage. It sounds like you see it more as a
server-side cabal freeze.


From michael at snoyman.com  Wed Sep  3 08:02:55 2014
From: michael at snoyman.com (Michael Snoyman)
Date: Wed, 3 Sep 2014 11:02:55 +0300
Subject: [Haskell-beginners] Stackage latest snapshot
In-Reply-To: <loom.20140903T081134-646@post.gmane.org>
References: <loom.20140902T092509-732@post.gmane.org>
 <CAKA2Jg+YwhoNfC1wEMaRS4OnxNcpUr7PaPAY_Nvaf9_oLN7xDA@mail.gmail.com>
 <loom.20140903T081134-646@post.gmane.org>
Message-ID: <CAKA2Jg+riW8hBATxTnpHPFYHZsZ-HLtqo=hoAp=U990Kdp6WnA@mail.gmail.com>

On Wed, Sep 3, 2014 at 10:52 AM, Peter <voldermort at hotmail.com> wrote:

> Michael Snoyman <michael <at> snoyman.com> writes:
>
> > The preparing your system for Stackage page[1] has those auto-updated
> links available (e.g., GHC 7.8 exclusive is [2]).
>
> Thank you. I read the page at least 3 times, but missed that because the
> links redirected to the latest snapshot, and the Stackage server only has
> static snapshot links. Perhaps it should be made clearer that these links
> will redirect to the current snapshot, and add them to
> http://www.stackage.org/snapshots as well?
>
>
That would be a good idea if I wanted people to directly use those URLs,
which I think is generally a bad idea. The lack of explicitness about the
presence of a redirect there is a feature, not a bug ;).


> > You already identified the downside with this: you may magically be
> upgraded to a new version of a package. If you really want to do this, you
> can, but my recommendation is to manually decide to upgrade your Stackage
> snapshot consciously instead of letter an arbitrary Jenkins build server
> make the decision for you.
>
> I'm using Stackage as a QA'd Hackage. It sounds like you see it more as a
> server-side cabal freeze.
>
>
I'm not sure what distinction you're going for in this. Imagine the
following sequence of events:

* You set up your system to use Stackage, with September 1's snapshot being
active at the time.
* You install foo-1.1, with bar-1.1 depending on it.
* A new Stackage snapshot comes out, providing foo-1.2.
* You now try to install something depending on foo, and depending on the
alignment of the stars, cabal will either download a new foo or use the
already installed one.

In other words, the usage you seem to be implying is highly
non-deterministic, which I'd be very careful about using.

Michael
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/66866ace/attachment.html>

From voldermort at hotmail.com  Wed Sep  3 08:37:47 2014
From: voldermort at hotmail.com (Peter)
Date: Wed, 3 Sep 2014 08:37:47 +0000 (UTC)
Subject: [Haskell-beginners] Stackage latest snapshot
References: <loom.20140902T092509-732@post.gmane.org>
 <CAKA2Jg+YwhoNfC1wEMaRS4OnxNcpUr7PaPAY_Nvaf9_oLN7xDA@mail.gmail.com>
 <loom.20140903T081134-646@post.gmane.org>
 <CAKA2Jg+riW8hBATxTnpHPFYHZsZ-HLtqo=hoAp=U990Kdp6WnA@mail.gmail.com>
Message-ID: <loom.20140903T102915-550@post.gmane.org>

Michael Snoyman <michael <at> snoyman.com> writes:

> > You already identified the downside with this: you may magically be
> upgraded to a new version of a package. If you really want to do this, you
> can, but my recommendation is to manually decide to upgrade your Stackage
> snapshot consciously instead of letter an arbitrary Jenkins build server
> make the decision for you.
> I'm using Stackage as a QA'd Hackage. It sounds like you see it more as a
> server-side cabal freeze.
> 
> 
> I'm not sure what distinction you're going for in this. Imagine the
following sequence of events:
> 
> * You set up your system to use Stackage, with September 1's snapshot
being active at the time.
> 
> * You install foo-1.1, with bar-1.1 depending on it.
> 
> * A new Stackage snapshot comes out, providing foo-1.2.
> 
> * You now try to install something depending on foo, and depending on the
alignment of the stars, cabal will either download a new foo or use the
already installed one.
> 
> In other words, the usage you seem to be implying is highly
non-deterministic, which I'd be very careful about using.

There are 3 levels of stability:

 1. Hackage, where any package may fail to compile, even by itself.
 2. Stackage latest, where at any one time, all packages will compile
together. New packages won't get in until they meet this stability criteria.
 3. Stackage snapshot, where all packages will compile together, and never
change.

All of these have legitimate uses. I prefer developing with 2, to avoid
being stuck with the same package versions until I change the snapshot, and
I'm prepared to fix up the example you gave manually if need be.

I guess it's similar to the difference between yum, which automatically
updates the package list on a fairly frequent basis, and apt-get, which
requires an explicit update. I personally prefer yum's approach (as do Red
Hat / Fedora), but I haven't seen any Debian folks calling to change apt.


From kc1956 at gmail.com  Wed Sep  3 13:01:18 2014
From: kc1956 at gmail.com (KC)
Date: Wed, 3 Sep 2014 06:01:18 -0700
Subject: [Haskell-beginners] Does Haskell work on Free BSD?
Message-ID: <CAMLKXy=3eyPmpqMBz4VDff+o-heZtSSOi-yO=PccK1m1hqTymg@mail.gmail.com>

Does Haskell work on Free BSD?

If not which is the better Linux to use?

-- 

--

Sent from an expensive device which will be obsolete in a few months! :D
Casey
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/7d87d80d/attachment.html>

From mwm at mired.org  Wed Sep  3 13:12:01 2014
From: mwm at mired.org (Mike Meyer)
Date: Wed, 3 Sep 2014 08:12:01 -0500
Subject: [Haskell-beginners] Does Haskell work on Free BSD?
In-Reply-To: <CAMLKXy=3eyPmpqMBz4VDff+o-heZtSSOi-yO=PccK1m1hqTymg@mail.gmail.com>
References: <CAMLKXy=3eyPmpqMBz4VDff+o-heZtSSOi-yO=PccK1m1hqTymg@mail.gmail.com>
Message-ID: <CAD=7U2CV74hptTkHhm3ij1q7_KajKU1089rZLKRLwJy97Guvng@mail.gmail.com>

Haskell works fine on FreeBSD. Both ghc and hugs are available from the
package system. Or you can install the Haskell Platform
(hs-haskell-platform) from it. There are also a variety of Haskell software
packages for development as well as end users.


On Wed, Sep 3, 2014 at 8:01 AM, KC <kc1956 at gmail.com> wrote:

> Does Haskell work on Free BSD?
>
> If not which is the better Linux to use?
>
> --
>
> --
>
> Sent from an expensive device which will be obsolete in a few months! :D
> Casey
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/3db4ada1/attachment.html>

From mlists at pmade.com  Wed Sep  3 16:24:55 2014
From: mlists at pmade.com (Peter Jones)
Date: Wed, 03 Sep 2014 10:24:55 -0600
Subject: [Haskell-beginners] Haskeline and forkIO
References: <4129696F7B970B4FA7C343A1587763E711B479A3@greyhound.iteris.com>
 <ltk6o5$vaf$1@ger.gmane.org>
 <4129696F7B970B4FA7C343A1587763E711B47C7E@greyhound.iteris.com>
 <87ppfkbt6k.fsf@pmade.com>
 <4129696F7B970B4FA7C343A1587763E711B47EAF@greyhound.iteris.com>
Message-ID: <87egvs7ju0.fsf@pmade.com>

"Jeff C. Britton" <jcb at iteris.com> writes:
> That suggestion works.
> I will have to continue learning more about Monads.

I should have also mentioned that if you enable GHC warnings it should
tell you that having the `return` before `loop` is discarding the value
given to it.

> -----Original Message-----
> From: Beginners [mailto:beginners-bounces at haskell.org]
> Sent: Thursday, August 28, 2014 7:11 AM
> To: beginners at haskell.org
> Subject: Re: [Haskell-beginners] Haskeline and forkIO
>
> "Jeff C. Britton" <jcb at iteris.com> writes:
>> loop :: InputT IO ()
>> loop = do
>>     maybeLine <- getInputLine "Enter a file to compress> "
>>     case maybeLine of
>>       Nothing -> return ()      -- user entered EOF
>>       Just "" -> return ()      -- treat no name as "want to quit" 
>>       Just path -> do
>>             return (runWorker path)
>>             loop
>
> The other issue you're having is because `runWorker path` is an `IO
> ()` value but at the point where you use it in the code the type
> system wants an `InputT IO ()`.  To try to satisfy the type system you
> used `return` to build a `InputT IO (IO ())` value, but that doesn't
> actually work (as you've noticed).  Since `InputT` is a transformer
> you have an extra layer to work through and so need to *lift* your `IO
> ()` value into the `InputT IO` layer.  Try this:
>
>     -- Add this import
>     import Control.Monad.IO.Class
>     
>     loop :: InputT IO ()
>     loop = do
>         maybeLine <- getInputLine "Enter a file to compress> "
>         case maybeLine of
>           Nothing -> return ()      -- user entered EOF
>           Just "" -> return ()      -- treat no name as "want to quit" 
>           Just path -> do
>                 liftIO (runWorker path)
>                 loop
>     
>
> You can think of `liftIO` as having this signature (in this context):
>
>     liftIO :: IO () -> InputT IO ()

-- 
Peter Jones, Founder, Devalot.com
Defending the honor of good code


From neeraj2608 at gmail.com  Wed Sep  3 16:51:50 2014
From: neeraj2608 at gmail.com (Neeraj Rao)
Date: Wed, 3 Sep 2014 12:51:50 -0400
Subject: [Haskell-beginners] Haskeline and forkIO
In-Reply-To: <87egvs7ju0.fsf@pmade.com>
References: <4129696F7B970B4FA7C343A1587763E711B479A3@greyhound.iteris.com>
 <ltk6o5$vaf$1@ger.gmane.org>
 <4129696F7B970B4FA7C343A1587763E711B47C7E@greyhound.iteris.com>
 <87ppfkbt6k.fsf@pmade.com>
 <4129696F7B970B4FA7C343A1587763E711B47EAF@greyhound.iteris.com>
 <87egvs7ju0.fsf@pmade.com>
Message-ID: <CAMtPA6aniDjp9Y1bQ_e0gvscXDCYSx8BakNZZpoKYjcS4xOuQQ@mail.gmail.com>

As a Haskell beginner, I would also like to recommend HLint as a very
useful tool to catch these kinds of things.
On Sep 3, 2014 12:25 PM, "Peter Jones" <mlists at pmade.com> wrote:

> "Jeff C. Britton" <jcb at iteris.com> writes:
> > That suggestion works.
> > I will have to continue learning more about Monads.
>
> I should have also mentioned that if you enable GHC warnings it should
> tell you that having the `return` before `loop` is discarding the value
> given to it.
>
> > -----Original Message-----
> > From: Beginners [mailto:beginners-bounces at haskell.org]
> > Sent: Thursday, August 28, 2014 7:11 AM
> > To: beginners at haskell.org
> > Subject: Re: [Haskell-beginners] Haskeline and forkIO
> >
> > "Jeff C. Britton" <jcb at iteris.com> writes:
> >> loop :: InputT IO ()
> >> loop = do
> >>     maybeLine <- getInputLine "Enter a file to compress> "
> >>     case maybeLine of
> >>       Nothing -> return ()      -- user entered EOF
> >>       Just "" -> return ()      -- treat no name as "want to quit"
> >>       Just path -> do
> >>             return (runWorker path)
> >>             loop
> >
> > The other issue you're having is because `runWorker path` is an `IO
> > ()` value but at the point where you use it in the code the type
> > system wants an `InputT IO ()`.  To try to satisfy the type system you
> > used `return` to build a `InputT IO (IO ())` value, but that doesn't
> > actually work (as you've noticed).  Since `InputT` is a transformer
> > you have an extra layer to work through and so need to *lift* your `IO
> > ()` value into the `InputT IO` layer.  Try this:
> >
> >     -- Add this import
> >     import Control.Monad.IO.Class
> >
> >     loop :: InputT IO ()
> >     loop = do
> >         maybeLine <- getInputLine "Enter a file to compress> "
> >         case maybeLine of
> >           Nothing -> return ()      -- user entered EOF
> >           Just "" -> return ()      -- treat no name as "want to quit"
> >           Just path -> do
> >                 liftIO (runWorker path)
> >                 loop
> >
> >
> > You can think of `liftIO` as having this signature (in this context):
> >
> >     liftIO :: IO () -> InputT IO ()
>
> --
> Peter Jones, Founder, Devalot.com
> Defending the honor of good code
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140903/7a807ade/attachment.html>

From martin.drautzburg at web.de  Sun Sep  7 12:14:22 2014
From: martin.drautzburg at web.de (martin)
Date: Sun, 07 Sep 2014 14:14:22 +0200
Subject: [Haskell-beginners] Continuations vs. Values
Message-ID: <540C4C1E.1010204@web.de>

Hello all,

in the continuation monad, I can turn a value into a Cont by means of return. Conversely I can convert a Cont into a
value by passing id as the continuation. This dualism of Cont and plain values is well known.

The bind operator kind of(!) takes a value a and produces a new value b. This is all wrapped in Cont stuff, so the
result of bind is not really a plain a, but a Cont which takes a function from b -> r. But this is just the same dualism
as above.

Most Cont examples I've seen could much more easily be implemented with just values or calling functions on values. Then
again there is "a poor man's concurrency" which I don't get at all. In between there is a huge gap of (my) understanding.

Can someone help me cross this gap? There must be certain things I can do with Conts which I cannot do easily with
values and functions.

From johnw at newartisans.com  Sun Sep  7 12:57:09 2014
From: johnw at newartisans.com (John Wiegley)
Date: Sun, 07 Sep 2014 14:57:09 +0200
Subject: [Haskell-beginners] Continuations vs. Values
In-Reply-To: <540C4C1E.1010204@web.de> (martin's message of "Sun, 07 Sep 2014
 14:14:22 +0200")
References: <540C4C1E.1010204@web.de>
Message-ID: <m238c3a8re.fsf@newartisans.com>

>>>>> martin  <martin.drautzburg at web.de> writes:

> Can someone help me cross this gap? There must be certain things I can do
> with Conts which I cannot do easily with values and functions.

Any function can be turned into a function accepting a continuation:

         f :: a -> b

    f_cont :: a -> (b -> c) -> c

The Cont construction abstracts this pattern into its own type:

    f_Cont :: a -> Cont c b

The advantage is that we now have a Functor instance over 'b', rather than
over 'c' (for Functor (->) applied to f_cont).

I cover Cont is more detail here, with a reference to green threading at the
very bottom:

    https://www.fpcomplete.com/user/jwiegley/understanding-continuations

John

From dav.vire+haskell at gmail.com  Mon Sep  8 14:31:54 2014
From: dav.vire+haskell at gmail.com (David Virebayre)
Date: Mon, 8 Sep 2014 16:31:54 +0200
Subject: [Haskell-beginners] Check constructor's field numeric value at
 compile time
In-Reply-To: <CAPY+ZdQcNhUoh39Xh6A7wNQAj_KBDM4hF8Zdoan=fGRBsWs0iw@mail.gmail.com>
References: <54061B53.7060106@gmail.com>
 <CAPY+ZdQcNhUoh39Xh6A7wNQAj_KBDM4hF8Zdoan=fGRBsWs0iw@mail.gmail.com>
Message-ID: <CAM_wFVsRV8BjC5=2raaFMDoq3tAexTc0PFFhPNUUUMhAMP3dbg@mail.gmail.com>

2014-09-03 5:07 GMT+02:00 Kim-Ee Yeoh <ky3 at atamo.com>:

>
> On Wed, Sep 3, 2014 at 2:32 AM, Dmitriy Matrosov <sgf.dma at gmail.com>
> wrote:
>
>> How can i write this constraints so they're checked at compile time, not
>> at runtime?
>
>
> Smart constructors won't check them at compile time but it does the next
> best thing, which is validate at the earliest possible moment.
>


In between compile time and runtime, you probably could use refinements
types; see LiquidHaskell (http://goto.ucsd.edu:8090/index.html,
https://github.com/ucsd-progsys/liquidhaskell/ )

David.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140908/b9340c2e/attachment.html>

From rdesfo at zoho.com  Mon Sep  8 18:03:49 2014
From: rdesfo at zoho.com (ryan)
Date: Mon, 08 Sep 2014 14:03:49 -0400
Subject: [Haskell-beginners] hpc doctest
Message-ID: <540DEF85.4040407@zoho.com>

Hello,

Are there any known issues with doctest and hpc? 

Cabal test runs correctly, but hpc doesn't seem to see anything.  I have
tried to make some of the test fail on purpose so see it hpc would pick
it up, since there is additional output from cabal test when there is a
failure, but it doesn't help. 

--
-- test file
--
module Doctests where

import Test.DocTest

main :: IO ()
main = doctest ["-isrc", "src/Ch1.hs"]

--
-- command line
--
$ cabal test
Building DiscreteMathematicsUsingAComputer-0.1.0.0...
Preprocessing library DiscreteMathematicsUsingAComputer-0.1.0.0...
In-place registering DiscreteMathematicsUsingAComputer-0.1.0.0...
Preprocessing test suite 'doctests' for
DiscreteMathematicsUsingAComputer-0.1.0.0...
Warning: output was redirected with -o, but no output will be generated
because there is no Main module.
Running 1 test suites...
Test suite doctests: RUNNING...
Test suite doctests: PASS
Test suite logged to:
dist/test/DiscreteMathematicsUsingAComputer-0.1.0.0-doctests.log
Writing: hpc_index.html
Writing: hpc_index_fun.html
Writing: hpc_index_alt.html
Writing: hpc_index_exp.html
Test coverage report written to dist/hpc/html/doctests/hpc_index.html
1 of 1 test suites (1 of 1 test cases) passed.
Writing: hpc_index.html
Writing: hpc_index_fun.html
Writing: hpc_index_alt.html
Writing: hpc_index_exp.html
Package coverage report written to
dist/hpc/html/DiscreteMathematicsUsingAComputer-0.1.0.0/hpc_index.html

Ryan


From gale at sefer.org  Tue Sep  9 18:53:56 2014
From: gale at sefer.org (Yitzchak Gale)
Date: Tue, 9 Sep 2014 21:53:56 +0300
Subject: [Haskell-beginners] [Haskell-cafe] 2014 haskell cabal update
	hangs on mac
In-Reply-To: <1397DF7E-BDB9-46F8-B7D8-4887681F91C2@dc.uba.ar>
References: <CALh-VV5Qob3MCh3_srM=LfDXDE30wUwXPEpdTWKyXUiuyyHqUA@mail.gmail.com>
 <59D8EF48-B854-407B-A067-C87DCDF1C9B7@gmail.com>
 <CALh-VV6gLG3iuE=0CMKp8XygNt2XsN9ngyXzrgX=wYwt0Te6Qg@mail.gmail.com>
 <1397DF7E-BDB9-46F8-B7D8-4887681F91C2@dc.uba.ar>
Message-ID: <CAOrUaLY74A_ZyDh8733L8SnS4DCoHpDq9hHBfJL6VYAikedVGQ@mail.gmail.com>

Daniel Gor?n wrote:
> I had the same problem. It turned out to be that for some unknown reason I ended up with:
>
> /Library/Haskell/ghc-7.8.3-x86_64/bin/cabal -> cabal.wrap
> /Library/Haskell/ghc-7.8.3-x86_64/bin/cabal.real -> cabal.wrap
> /Library/Haskell/ghc-7.8.3-x86_64/bin/cabal.wrap
>
> where cabal.wrap is a shell script that checks if a new cabal.config
> should be created and then calls? cabal.real. Replacing cabal.real with a binary version of cabal solves it.
>
> I don?t know why this happened, I suspect that it might occur if you had ghc 7.8.2 already installed...

I believe this is a bug in the scripts activate-hs and uninstall-hs
provided by the Haskell Platform installer for Mac OS X.
I have reported it here:

https://github.com/haskell/haskell-platform/issues/147

-Yitz

From martin.drautzburg at web.de  Wed Sep 10 18:06:26 2014
From: martin.drautzburg at web.de (martin)
Date: Wed, 10 Sep 2014 20:06:26 +0200
Subject: [Haskell-beginners] How to add a "method" to a record
Message-ID: <54109322.1070901@web.de>

Hello all

if I have a record like

	data Foo pl = Foo {
	            label :: String,
	            payload :: pl
	        }

how can I create a similar type where I can populate label so it is not a plain string, but a function which operates on
payload? Something like

	label (Foo pl) = show pl



From toad3k at gmail.com  Wed Sep 10 18:16:02 2014
From: toad3k at gmail.com (David McBride)
Date: Wed, 10 Sep 2014 14:16:02 -0400
Subject: [Haskell-beginners] How to add a "method" to a record
In-Reply-To: <54109322.1070901@web.de>
References: <54109322.1070901@web.de>
Message-ID: <CAN+Tr42sByVOLxrsQZyfRuFGJqTrHU6UHLpOtCwOh97PvfMVEg@mail.gmail.com>

Is this what you are looking for?

data Foo pl = Foo {
  label :: pl -> String,
  payload :: pl
}

On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de> wrote:

> Hello all
>
> if I have a record like
>
>         data Foo pl = Foo {
>                     label :: String,
>                     payload :: pl
>                 }
>
> how can I create a similar type where I can populate label so it is not a
> plain string, but a function which operates on
> payload? Something like
>
>         label (Foo pl) = show pl
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140910/f29ce44d/attachment.html>

From julian.birch at gmail.com  Wed Sep 10 18:30:16 2014
From: julian.birch at gmail.com (Julian Birch)
Date: Wed, 10 Sep 2014 19:30:16 +0100
Subject: [Haskell-beginners] How to add a "method" to a record
In-Reply-To: <CAN+Tr42sByVOLxrsQZyfRuFGJqTrHU6UHLpOtCwOh97PvfMVEg@mail.gmail.com>
References: <54109322.1070901@web.de>
 <CAN+Tr42sByVOLxrsQZyfRuFGJqTrHU6UHLpOtCwOh97PvfMVEg@mail.gmail.com>
Message-ID: <CAB0TuzBXqb7bLiRb2WhbsK_FQQzS_NCmROipO2s88jO9b-vTQA@mail.gmail.com>

Bear in mind you can just create functions at the top level that operate on
your data structure.  You only need the function to be a member of the
record if the function itself changes, which is relatively rare.

Say you need a Foo and a Bar, and they both have labels, implemented in
different ways, then you can use a typeclass to achieve your goals.

On Wednesday, September 10, 2014, David McBride <toad3k at gmail.com> wrote:

> Is this what you are looking for?
>
> data Foo pl = Foo {
>   label :: pl -> String,
>   payload :: pl
> }
>
> On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de
> <javascript:_e(%7B%7D,'cvml','martin.drautzburg at web.de');>> wrote:
>
>> Hello all
>>
>> if I have a record like
>>
>>         data Foo pl = Foo {
>>                     label :: String,
>>                     payload :: pl
>>                 }
>>
>> how can I create a similar type where I can populate label so it is not a
>> plain string, but a function which operates on
>> payload? Something like
>>
>>         label (Foo pl) = show pl
>>
>>
>> _______________________________________________
>> Beginners mailing list
>> Beginners at haskell.org
>> <javascript:_e(%7B%7D,'cvml','Beginners at haskell.org');>
>> http://www.haskell.org/mailman/listinfo/beginners
>>
>
>

-- 
Sent from an iPhone, please excuse brevity and typos.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140910/9592b852/attachment.html>

From cma at bitemyapp.com  Wed Sep 10 18:49:56 2014
From: cma at bitemyapp.com (Christopher Allen)
Date: Wed, 10 Sep 2014 13:49:56 -0500
Subject: [Haskell-beginners] How to add a "method" to a record
In-Reply-To: <CAB0TuzBXqb7bLiRb2WhbsK_FQQzS_NCmROipO2s88jO9b-vTQA@mail.gmail.com>
References: <54109322.1070901@web.de>
 <CAN+Tr42sByVOLxrsQZyfRuFGJqTrHU6UHLpOtCwOh97PvfMVEg@mail.gmail.com>
 <CAB0TuzBXqb7bLiRb2WhbsK_FQQzS_NCmROipO2s88jO9b-vTQA@mail.gmail.com>
Message-ID: <CADnndOrsKTeODCMmJkXRNjbEh2+u_2_WfsaE2RyaPB5jiKX-ng@mail.gmail.com>

I'd agree with Julian Birch, except to say that you shouldn't make a
typeclass unless overloading is really needed and you can define laws on
the behavior thereof.

A fairly major (IMO) inconvenience with embedding functions in records that
have non-function data they operate on is that you won't have a Show
instance. Also, you'll be forced to extract the method and then apply it.
In doing so, to make things nicer you'll probably pull out an "extract and
apply" function. At that point, you're better off lifting the varying
implementations into top-level functions and making a wrapper that decides
which runs where based on the *data* in the record.

That is, including enough information in the record to decide which
function applied.

And this is without having broached whether or not you plan to be able to
serialize your data or not.


On Wed, Sep 10, 2014 at 1:30 PM, Julian Birch <julian.birch at gmail.com>
wrote:

> Bear in mind you can just create functions at the top level that operate
> on your data structure.  You only need the function to be a member of the
> record if the function itself changes, which is relatively rare.
>
> Say you need a Foo and a Bar, and they both have labels, implemented in
> different ways, then you can use a typeclass to achieve your goals.
>
>
> On Wednesday, September 10, 2014, David McBride <toad3k at gmail.com> wrote:
>
>> Is this what you are looking for?
>>
>> data Foo pl = Foo {
>>   label :: pl -> String,
>>   payload :: pl
>> }
>>
>> On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de> wrote:
>>
>>> Hello all
>>>
>>> if I have a record like
>>>
>>>         data Foo pl = Foo {
>>>                     label :: String,
>>>                     payload :: pl
>>>                 }
>>>
>>> how can I create a similar type where I can populate label so it is not
>>> a plain string, but a function which operates on
>>> payload? Something like
>>>
>>>         label (Foo pl) = show pl
>>>
>>>
>>> _______________________________________________
>>> Beginners mailing list
>>> Beginners at haskell.org
>>> http://www.haskell.org/mailman/listinfo/beginners
>>>
>>
>>
>
> --
> Sent from an iPhone, please excuse brevity and typos.
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140910/82e5daea/attachment.html>

From corentin.dupont at gmail.com  Wed Sep 10 18:50:18 2014
From: corentin.dupont at gmail.com (Corentin Dupont)
Date: Wed, 10 Sep 2014 20:50:18 +0200
Subject: [Haskell-beginners] How to add a "method" to a record
In-Reply-To: <CAB0TuzBXqb7bLiRb2WhbsK_FQQzS_NCmROipO2s88jO9b-vTQA@mail.gmail.com>
References: <54109322.1070901@web.de>
 <CAN+Tr42sByVOLxrsQZyfRuFGJqTrHU6UHLpOtCwOh97PvfMVEg@mail.gmail.com>
 <CAB0TuzBXqb7bLiRb2WhbsK_FQQzS_NCmROipO2s88jO9b-vTQA@mail.gmail.com>
Message-ID: <CAEyhvmopJGyv1B7TpScgKN9mZVVMXYu42hWDUfyS=vbskdonZw@mail.gmail.com>

If the field "label" can be deduced from "payload", I recommend not to
include it in your structure, because that would be redundant.

Here how you could write it:

data Foo pl = Foo { payload :: pl}

labelInt :: Foo Int -> String
labelInt (Foo a) = "Int payload:" ++ (show a)

labelString :: Foo String -> String
labelString (Foo a) = "String payload" ++ a

You are obliged to define two separate label function, because "Foo Int"
and "Foo String" are two completly separate types.
Alternatively, if you want only one label function, use a type sum and
pattern match on it:

data T = TS String | TI Int
data Foo2 = Foo2 { payload2 :: T}

label :: Foo2 -> String
label (Foo2 (TI a)) = "Int payload:" ++ (show a)
label (Foo2 (TS a)) = "String payload:" ++ a



On Wed, Sep 10, 2014 at 8:30 PM, Julian Birch <julian.birch at gmail.com>
wrote:

> Bear in mind you can just create functions at the top level that operate
> on your data structure.  You only need the function to be a member of the
> record if the function itself changes, which is relatively rare.
>
> Say you need a Foo and a Bar, and they both have labels, implemented in
> different ways, then you can use a typeclass to achieve your goals.
>
>
> On Wednesday, September 10, 2014, David McBride <toad3k at gmail.com> wrote:
>
>> Is this what you are looking for?
>>
>> data Foo pl = Foo {
>>   label :: pl -> String,
>>   payload :: pl
>> }
>>
>> On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de> wrote:
>>
>>> Hello all
>>>
>>> if I have a record like
>>>
>>>         data Foo pl = Foo {
>>>                     label :: String,
>>>                     payload :: pl
>>>                 }
>>>
>>> how can I create a similar type where I can populate label so it is not
>>> a plain string, but a function which operates on
>>> payload? Something like
>>>
>>>         label (Foo pl) = show pl
>>>
>>>
>>> _______________________________________________
>>> Beginners mailing list
>>> Beginners at haskell.org
>>> http://www.haskell.org/mailman/listinfo/beginners
>>>
>>
>>
>
> --
> Sent from an iPhone, please excuse brevity and typos.
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140910/28fc1a81/attachment-0001.html>

From martin.drautzburg at web.de  Fri Sep 12 17:05:11 2014
From: martin.drautzburg at web.de (martin)
Date: Fri, 12 Sep 2014 19:05:11 +0200
Subject: [Haskell-beginners] How to add a "method" to a record
In-Reply-To: <CAEyhvmopJGyv1B7TpScgKN9mZVVMXYu42hWDUfyS=vbskdonZw@mail.gmail.com>
References: <54109322.1070901@web.de>
 <CAN+Tr42sByVOLxrsQZyfRuFGJqTrHU6UHLpOtCwOh97PvfMVEg@mail.gmail.com>
 <CAB0TuzBXqb7bLiRb2WhbsK_FQQzS_NCmROipO2s88jO9b-vTQA@mail.gmail.com>
 <CAEyhvmopJGyv1B7TpScgKN9mZVVMXYu42hWDUfyS=vbskdonZw@mail.gmail.com>
Message-ID: <541327C7.6040104@web.de>

Am 09/10/2014 08:50 PM, schrieb Corentin Dupont:
> If the field "label" can be deduced from "payload", I recommend not to include it in your structure, because that would
> be redundant.
> 
> Here how you could write it:
> 
> data Foo pl = Foo { payload :: pl}
> 
> labelInt :: Foo Int -> String
> labelInt (Foo a) = "Int payload:" ++ (show a)
> 
> labelString :: Foo String -> String
> labelString (Foo a) = "String payload" ++ a
> 
> You are obliged to define two separate label function, because "Foo Int" and "Foo String" are two completly separate types.

This is exactly my problem: Someone will use this type an define the type of pl. How can I know what type she'll use?
What I'd like to express is that whoever creates a concrete type should also provide the proper label function.

> 
>         On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de> wrote:
> 
>             Hello all
> 
>             if I have a record like
> 
>                     data Foo pl = Foo {
>                                 label :: String,
>                                 payload :: pl
>                             }
> 
>             how can I create a similar type where I can populate label so it is not a plain string, but a function which
>             operates on
>             payload? Something like
> 
>                     label (Foo pl) = show pl
> 


From jcb at iteris.com  Tue Sep 16 02:01:30 2014
From: jcb at iteris.com (Jeff C. Britton)
Date: Tue, 16 Sep 2014 02:01:30 +0000
Subject: [Haskell-beginners] ghci and randomRs
Message-ID: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>

I am trying to use randomRs at the ghci prompt like so

Prelude System.Random> take 10 $ (randomRs (1, 6) newStdGen)

but I get the following error
<interactive>:16:12:
    Could not deduce (RandomGen (IO StdGen))
      arising from a use of `randomRs'
    from the context (Random a, Num a)
      bound by the inferred type of it :: (Random a, Num a) => [a]
      at <interactive>:16:1-37
    In the second argument of `($)', namely
      `(randomRs (1, 6) newStdGen)'
    In the expression: take 10 $ (randomRs (1, 6) newStdGen)
    In an equation for `it': it = take 10 $ (randomRs (1, 6) newStdGen)


I have tried a variety of options, like wrapping it in a "do" or adding type annotations.
Nothing seems to work.



-----Original Message-----
From: Beginners [mailto:beginners-bounces at haskell.org] On Behalf Of martin
Sent: Friday, September 12, 2014 10:05 AM
To: The Haskell-Beginners Mailing List - Discussion of primarily beginner-level topics related to Haskell
Subject: Re: [Haskell-beginners] How to add a "method" to a record

Am 09/10/2014 08:50 PM, schrieb Corentin Dupont:
> If the field "label" can be deduced from "payload", I recommend not to 
> include it in your structure, because that would be redundant.
> 
> Here how you could write it:
> 
> data Foo pl = Foo { payload :: pl}
> 
> labelInt :: Foo Int -> String
> labelInt (Foo a) = "Int payload:" ++ (show a)
> 
> labelString :: Foo String -> String
> labelString (Foo a) = "String payload" ++ a
> 
> You are obliged to define two separate label function, because "Foo Int" and "Foo String" are two completly separate types.

This is exactly my problem: Someone will use this type an define the type of pl. How can I know what type she'll use?
What I'd like to express is that whoever creates a concrete type should also provide the proper label function.

> 
>         On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de> wrote:
> 
>             Hello all
> 
>             if I have a record like
> 
>                     data Foo pl = Foo {
>                                 label :: String,
>                                 payload :: pl
>                             }
> 
>             how can I create a similar type where I can populate label so it is not a plain string, but a function which
>             operates on
>             payload? Something like
> 
>                     label (Foo pl) = show pl
> 

_______________________________________________
Beginners mailing list
Beginners at haskell.org
http://www.haskell.org/mailman/listinfo/beginners

From allbery.b at gmail.com  Tue Sep 16 02:05:52 2014
From: allbery.b at gmail.com (Brandon Allbery)
Date: Mon, 15 Sep 2014 22:05:52 -0400
Subject: [Haskell-beginners] ghci and randomRs
In-Reply-To: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
References: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
Message-ID: <CAKFCL4XvyX_PpGAkYqB47zRn8YdKXAGKhPmugVSc6yKGE_5mZQ@mail.gmail.com>

On Mon, Sep 15, 2014 at 10:01 PM, Jeff C. Britton <jcb at iteris.com> wrote:

> Prelude System.Random> take 10 $ (randomRs (1, 6) newStdGen)


I think you need to read http://www.vex.net/~trebla/haskell/IO.xhtml.
You're trying to operate directly on an IO "program" (newStdGen) as if it
were a value.

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b at gmail.com                                  ballbery at sinenomine.net
unix, openafs, kerberos, infrastructure, xmonad        http://sinenomine.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140915/9c1e0cd1/attachment.html>

From toad3k at gmail.com  Tue Sep 16 02:27:02 2014
From: toad3k at gmail.com (David McBride)
Date: Mon, 15 Sep 2014 22:27:02 -0400
Subject: [Haskell-beginners] ghci and randomRs
In-Reply-To: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
References: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
Message-ID: <CAN+Tr41=8h2pdJOBzePv+XjGHyw-FE5TgMMrB+DFwcfZ6dr=vg@mail.gmail.com>

The simple answer is with do notation:

main = do
  g <- newStdGen
  print $ randomRs (1,2) g

Or without do notation, something like:

newStdGen >>= print . take 10 . randomRs (1,2)

On Mon, Sep 15, 2014 at 10:01 PM, Jeff C. Britton <jcb at iteris.com> wrote:

> I am trying to use randomRs at the ghci prompt like so
>
> Prelude System.Random> take 10 $ (randomRs (1, 6) newStdGen)
>
> but I get the following error
> <interactive>:16:12:
>     Could not deduce (RandomGen (IO StdGen))
>       arising from a use of `randomRs'
>     from the context (Random a, Num a)
>       bound by the inferred type of it :: (Random a, Num a) => [a]
>       at <interactive>:16:1-37
>     In the second argument of `($)', namely
>       `(randomRs (1, 6) newStdGen)'
>     In the expression: take 10 $ (randomRs (1, 6) newStdGen)
>     In an equation for `it': it = take 10 $ (randomRs (1, 6) newStdGen)
>
>
> I have tried a variety of options, like wrapping it in a "do" or adding
> type annotations.
> Nothing seems to work.
>
>
>
> -----Original Message-----
> From: Beginners [mailto:beginners-bounces at haskell.org] On Behalf Of martin
> Sent: Friday, September 12, 2014 10:05 AM
> To: The Haskell-Beginners Mailing List - Discussion of primarily
> beginner-level topics related to Haskell
> Subject: Re: [Haskell-beginners] How to add a "method" to a record
>
> Am 09/10/2014 08:50 PM, schrieb Corentin Dupont:
> > If the field "label" can be deduced from "payload", I recommend not to
> > include it in your structure, because that would be redundant.
> >
> > Here how you could write it:
> >
> > data Foo pl = Foo { payload :: pl}
> >
> > labelInt :: Foo Int -> String
> > labelInt (Foo a) = "Int payload:" ++ (show a)
> >
> > labelString :: Foo String -> String
> > labelString (Foo a) = "String payload" ++ a
> >
> > You are obliged to define two separate label function, because "Foo Int"
> and "Foo String" are two completly separate types.
>
> This is exactly my problem: Someone will use this type an define the type
> of pl. How can I know what type she'll use?
> What I'd like to express is that whoever creates a concrete type should
> also provide the proper label function.
>
> >
> >         On Wed, Sep 10, 2014 at 2:06 PM, martin <
> martin.drautzburg at web.de> wrote:
> >
> >             Hello all
> >
> >             if I have a record like
> >
> >                     data Foo pl = Foo {
> >                                 label :: String,
> >                                 payload :: pl
> >                             }
> >
> >             how can I create a similar type where I can populate label
> so it is not a plain string, but a function which
> >             operates on
> >             payload? Something like
> >
> >                     label (Foo pl) = show pl
> >
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140915/2ee704f1/attachment.html>

From jcb at iteris.com  Tue Sep 16 17:43:04 2014
From: jcb at iteris.com (Jeff C. Britton)
Date: Tue, 16 Sep 2014 17:43:04 +0000
Subject: [Haskell-beginners] ghci and randomRs
In-Reply-To: <CAN+Tr41=8h2pdJOBzePv+XjGHyw-FE5TgMMrB+DFwcfZ6dr=vg@mail.gmail.com>
References: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
 <CAN+Tr41=8h2pdJOBzePv+XjGHyw-FE5TgMMrB+DFwcfZ6dr=vg@mail.gmail.com>
Message-ID: <4129696F7B970B4FA7C343A1587763E711B5751B@greyhound.iteris.com>

Ok, I got it now.  I was misunderstanding how the REPL was interacting with the IO Monad.
I had once tried
do { g <- newStdGen; take 10 $ randomRs (1,6) g  }

but I actually needed this
do { g <- newStdGen; return . take 10 $ randomRs (1,6) g  }

Thanks,
Jeff

From: Beginners [mailto:beginners-bounces at haskell.org] On Behalf Of David McBride
Sent: Monday, September 15, 2014 7:27 PM
To: The Haskell-Beginners Mailing List - Discussion of primarily beginner-level topics related to Haskell
Subject: Re: [Haskell-beginners] ghci and randomRs

The simple answer is with do notation:
main = do
  g <- newStdGen
  print $ randomRs (1,2) g
Or without do notation, something like:
newStdGen >>= print . take 10 . randomRs (1,2)

On Mon, Sep 15, 2014 at 10:01 PM, Jeff C. Britton <jcb at iteris.com<mailto:jcb at iteris.com>> wrote:
I am trying to use randomRs at the ghci prompt like so

Prelude System.Random> take 10 $ (randomRs (1, 6) newStdGen)

but I get the following error
<interactive>:16:12:
    Could not deduce (RandomGen (IO StdGen))
      arising from a use of `randomRs'
    from the context (Random a, Num a)
      bound by the inferred type of it :: (Random a, Num a) => [a]
      at <interactive>:16:1-37
    In the second argument of `($)', namely
      `(randomRs (1, 6) newStdGen)'
    In the expression: take 10 $ (randomRs (1, 6) newStdGen)
    In an equation for `it': it = take 10 $ (randomRs (1, 6) newStdGen)


I have tried a variety of options, like wrapping it in a "do" or adding type annotations.
Nothing seems to work.



-----Original Message-----
From: Beginners [mailto:beginners-bounces at haskell.org<mailto:beginners-bounces at haskell.org>] On Behalf Of martin
Sent: Friday, September 12, 2014 10:05 AM
To: The Haskell-Beginners Mailing List - Discussion of primarily beginner-level topics related to Haskell
Subject: Re: [Haskell-beginners] How to add a "method" to a record

Am 09/10/2014 08:50 PM, schrieb Corentin Dupont:
> If the field "label" can be deduced from "payload", I recommend not to
> include it in your structure, because that would be redundant.
>
> Here how you could write it:
>
> data Foo pl = Foo { payload :: pl}
>
> labelInt :: Foo Int -> String
> labelInt (Foo a) = "Int payload:" ++ (show a)
>
> labelString :: Foo String -> String
> labelString (Foo a) = "String payload" ++ a
>
> You are obliged to define two separate label function, because "Foo Int" and "Foo String" are two completly separate types.

This is exactly my problem: Someone will use this type an define the type of pl. How can I know what type she'll use?
What I'd like to express is that whoever creates a concrete type should also provide the proper label function.

>
>         On Wed, Sep 10, 2014 at 2:06 PM, martin <martin.drautzburg at web.de<mailto:martin.drautzburg at web.de>> wrote:
>
>             Hello all
>
>             if I have a record like
>
>                     data Foo pl = Foo {
>                                 label :: String,
>                                 payload :: pl
>                             }
>
>             how can I create a similar type where I can populate label so it is not a plain string, but a function which
>             operates on
>             payload? Something like
>
>                     label (Foo pl) = show pl
>

_______________________________________________
Beginners mailing list
Beginners at haskell.org<mailto:Beginners at haskell.org>
http://www.haskell.org/mailman/listinfo/beginners
_______________________________________________
Beginners mailing list
Beginners at haskell.org<mailto:Beginners at haskell.org>
http://www.haskell.org/mailman/listinfo/beginners

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140916/2958a84b/attachment.html>

From fa-ml at ariis.it  Tue Sep 16 19:05:49 2014
From: fa-ml at ariis.it (Francesco Ariis)
Date: Tue, 16 Sep 2014 21:05:49 +0200
Subject: [Haskell-beginners] ghci and randomRs
In-Reply-To: <4129696F7B970B4FA7C343A1587763E711B5751B@greyhound.iteris.com>
References: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
 <CAN+Tr41=8h2pdJOBzePv+XjGHyw-FE5TgMMrB+DFwcfZ6dr=vg@mail.gmail.com>
 <4129696F7B970B4FA7C343A1587763E711B5751B@greyhound.iteris.com>
Message-ID: <20140916190549.GA14803@x60s.casa>

On Tue, Sep 16, 2014 at 05:43:04PM +0000, Jeff C. Britton wrote:
> Ok, I got it now.  I was misunderstanding how the REPL was interacting with the IO Monad.
> I had once tried
> do { g <- newStdGen; take 10 $ randomRs (1,6) g  }
>
> but I actually needed this
> do { g <- newStdGen; return . take 10 $ randomRs (1,6) g  }
>
> Thanks,
> Jeff

I am sure you have already figured it out, but in case you didn't,
this works too:

?> :m +System.Random
?> g <- newStdGen
?> take 10 $ randomRs (1,6) g
[3,3,5,1,6,6,3,3,1,6]
?>

(and keeps IO actions apart from pure code)

From exitconsole at gmail.com  Tue Sep 16 21:11:02 2014
From: exitconsole at gmail.com (=?UTF-8?B?RMOhbmllbCBBcmF0w7M=?=)
Date: Tue, 16 Sep 2014 23:11:02 +0200
Subject: [Haskell-beginners] What am I reinventing here
Message-ID: <CAHvKd2Knv3A=1PUyZVBaonBfsKLrS8stHrC==cbuAuuB1uNVcQ@mail.gmail.com>

Hey guys,

I was working through 99 Haskell Problems a while back and while
refactoring my solution to Problem 9 I couldn't help feeling I was
reinventing something basic. My thinking was, this problem can be
generalized to letting a function recursively process arbitrary chunks
of a given data structure in several passes until the whole data is
consumed while some function is extracting some information in an
accumulator.

So I wrote a mini-typeclass like so:

class Exhaustible e where
    isEmpty :: e -> Bool
instance Exhaustible [a] where
    isEmpty = null

exhaust :: Exhaustible e => (e -> (b, e)) -> (a -> b -> a) -> a -> e -> a
exhaust f g z xs
    | isEmpty xs = z
    | otherwise  = let (val, rest) = f xs in
                   exhaust f g (g z val) rest

Now the solution to Problem 9 is easy to define in terms of "exhaust".

Then I thought, "Why am I treating empty as a special value again?",
and wrote this instead:

stabilize :: Eq a => (a -> (b, a)) -> (c -> b -> c) -> c -> a -> (c, a)
stabilize f g z d
    | d == d'   = (z', d')
    | otherwise = stabilize f g z' d'
     where (val, d') = f d
           z' = g z val

This looks both really useful and extremely basic to me, so I'm
wondering what the standard way to do the same thing might be. I
looked at Foldable + foldr but the idea there seems to be to process a
data structure _one_ element at a time.

So my question is, is there an official implementation for this
combinator I called "stabilize"? Also, any tips on improving the
definitions and comments on how situations like Problem 9 should be
approached are welcome.

Daniel

From raabe at froglogic.com  Tue Sep 16 22:02:13 2014
From: raabe at froglogic.com (Frerich Raabe)
Date: Wed, 17 Sep 2014 00:02:13 +0200
Subject: [Haskell-beginners] What am I reinventing here
In-Reply-To: <CAHvKd2Knv3A=1PUyZVBaonBfsKLrS8stHrC==cbuAuuB1uNVcQ@mail.gmail.com>
References: <CAHvKd2Knv3A=1PUyZVBaonBfsKLrS8stHrC==cbuAuuB1uNVcQ@mail.gmail.com>
Message-ID: <19acfad4798068f14dd0db1eac89ea4b@roundcube.froglogic.com>

On 2014-09-16 23:11, D?niel Arat? wrote:
  Then I thought, "Why am I treating empty as a special value again?",
> and wrote this instead:
> 
> stabilize :: Eq a => (a -> (b, a)) -> (c -> b -> c) -> c -> a -> (c, a)
> stabilize f g z d
>     | d == d'   = (z', d')
>     | otherwise = stabilize f g z' d'
>      where (val, d') = f d
>            z' = g z val
> 
> This looks both really useful and extremely basic to me, so I'm
> wondering what the standard way to do the same thing might be.

I think you reinvented the 'unfoldr' function here:

   unfoldr :: (b -> Maybe (a, b)) -> b -> [a]

It applies the given function to the second argument until the function
returns Nothing, which correponds to your "d == d'" guard. You could use it
to solve Problem 9 like

   pack :: Eq a => [a] -> [[a]]
   pack = unfoldr (\xs -> if null xs then Nothing else Just (span (== head xs) 
xs))

I think this looks like it does two things at once:
> I looked at Foldable + foldr but the idea there seems to be to process a
> data structure _one_ element at a time.

It's true that a fold processes a data structure one element at a time,
but you also have access to the accumulator, i.e. you can consider what
you processed so far. I think you could solve Problem 9 using a foldr
like

   pack :: Eq a => [a] -> [[a]]
   pack = foldr go []
     where
       go x acc@(a:as) = if x == head a then (x:a):as else [x]:acc
       go x _          = [[x]]

- Frerich


From jcb at iteris.com  Tue Sep 16 23:55:16 2014
From: jcb at iteris.com (Jeff C. Britton)
Date: Tue, 16 Sep 2014 23:55:16 +0000
Subject: [Haskell-beginners] ghci and randomRs
In-Reply-To: <20140916190549.GA14803@x60s.casa>
References: <4129696F7B970B4FA7C343A1587763E711B564F3@greyhound.iteris.com>
 <CAN+Tr41=8h2pdJOBzePv+XjGHyw-FE5TgMMrB+DFwcfZ6dr=vg@mail.gmail.com>
 <4129696F7B970B4FA7C343A1587763E711B5751B@greyhound.iteris.com>
 <20140916190549.GA14803@x60s.casa>
Message-ID: <4129696F7B970B4FA7C343A1587763E711B57589@greyhound.iteris.com>

Thanks Francesco.
This actually helped me.

--Jeff

-----Original Message-----
From: Beginners [mailto:beginners-bounces at haskell.org] On Behalf Of Francesco Ariis
Sent: Tuesday, September 16, 2014 12:06 PM
To: beginners at haskell.org
Subject: Re: [Haskell-beginners] ghci and randomRs

On Tue, Sep 16, 2014 at 05:43:04PM +0000, Jeff C. Britton wrote:
> Ok, I got it now.  I was misunderstanding how the REPL was interacting with the IO Monad.
> I had once tried
> do { g <- newStdGen; take 10 $ randomRs (1,6) g  }
>
> but I actually needed this
> do { g <- newStdGen; return . take 10 $ randomRs (1,6) g  }
>
> Thanks,
> Jeff

I am sure you have already figured it out, but in case you didn't, this works too:

?> :m +System.Random
?> g <- newStdGen
?> take 10 $ randomRs (1,6) g
[3,3,5,1,6,6,3,3,1,6]
?>

(and keeps IO actions apart from pure code) _______________________________________________
Beginners mailing list
Beginners at haskell.org
http://www.haskell.org/mailman/listinfo/beginners

From dontdieych at gmail.com  Wed Sep 17 09:17:00 2014
From: dontdieych at gmail.com (Dontdie YCH)
Date: Wed, 17 Sep 2014 18:17:00 +0900
Subject: [Haskell-beginners] How can I replace content of tag in xml file
	with haxml.
Message-ID: <CAPtM4gmxPy4t9EBuHvHDTe6Fi0_HHJ24fN12S5Ap1aMQLrz1ew@mail.gmail.com>

I'm trying to replace `CurrentURI` tag's contents using HaXml.

from,

```
<?xml version="1.0"?>
<s:Envelope xmlns:s="http://schemas.xmlsoap.org/soap/envelope/"
s:encodingStyle="http://schemas.xmlsoap.org/soap/encoding/">
  <s:Body>
    <u:SetAVTransportURI xmlns:u="urn:schemas-upnp-org:service:AVTransport:1">
      <InstanceID>0</InstanceID>
      <CurrentURI>AAAAAAAAAAAA</CurrentURI>
      <CurrentURIMetaData/>
    </u:SetAVTransportURI>
  </s:Body>
</s:Envelope>
```

to,

```
<?xml version="1.0"?>
<s:Envelope xmlns:s="http://schemas.xmlsoap.org/soap/envelope/"
s:encodingStyle="http://schemas.xmlsoap.org/soap/encoding/">
  <s:Body>
    <u:SetAVTransportURI xmlns:u="urn:schemas-upnp-org:service:AVTransport:1">
      <InstanceID>0</InstanceID>
      <CurrentURI>BBBBBBBBBBBBBB</CurrentURI>
      <CurrentURIMetaData/>
    </u:SetAVTransportURI>
  </s:Body>
</s:Envelope>
```

I had look around, hxt, xml-light, and haxml. I found `onContents`
function from `Text.XML.HaXml.Wrappers` module and It looks like way
to go.

```
 ?: xml <- readFile "SetAVTransportURI.xml" :: IO String
 ?: let doc = xmlParse "dumb" xml
 ?: doc
Document (Prolog (Just (XMLDecl "1.0" Nothing Nothing)) [] Nothing [])
[] (Elem (N "s:Envelope") [(N
"xmlns:s",http://schemas.xmlsoap.org/soap/envelope/),(N
"s:encodingStyle",http://schemas.xmlsoap.org/soap/encoding/)] [CString
False "\n  " file dumb  at line 2 col 125,CElem (Elem (N "s:Body") []
[CString False "\n    " file dumb  at line 3 col 11,CElem (Elem (N
"u:SetAVTransportURI") [(N
"xmlns:u",urn:schemas-upnp-org:service:AVTransport:1)] [CString False
"\n      " file dumb  at line 4 col 79,CElem (Elem (N "InstanceID") []
[CString False "0\n      " file dumb  at line 5 col 19]) file dumb  at
line 5 col 7,CString False "\n      " file dumb  at line 6 col
20,CElem (Elem (N "CurrentURI") [] [CString False "\n" file dumb  at
line 7 col 19,CMisc (Comment " #EXTM3U ") file dumb  at line 8 col
1,CMisc (Comment " #EXTINF:5850,\52628\49437\53945\49440\50689\54868
\47609\47329\44284\44053 140909 HDTV x264 720p AC3 \50864\47532\47568
\45433\51020 &#45; nature boy ") file dumb  at line 9 col 1,CMisc
(Comment " http://192.168.219.2:8200/MediaItems/587.mkv ") file dumb
at line 10 col 1,CMisc (Comment " #EXTM3U ") file dumb  at line 11 col
1,CMisc (Comment "
#EXTINF:6411,The.Adventures.of.Tintin.2011.1080p.BluRay.x264.YIFY ")
file dumb  at line 12 col 1,CString False
"http://192.168.219.2:8200/MediaItems/165.mp4\n      " file dumb  at
line 13 col 1]) file dumb  at line 7 col 7,CString False "\n      "
file dumb  at line 14 col 20,CElem (Elem (N "CurrentURIMetaData") []
[]) file dumb  at line 15 col 7,CString False "\n    " file dumb  at
line 15 col 28]) file dumb  at line 4 col 5,CString False "\n  " file
dumb  at line 16 col 27]) file dumb  at line 3 col 3,CString False
"\n" file dumb  at line 17 col 12]) [Comment "\nOrignal header from
captured packets.\n\nPOST /smp_18_ HTTP/1.1\nHost:
192.168.219.6:7676\nSOAPAction:
\"urn:schemas-upnp-org:service:AVTransport:1#SetAVTransportURI\"\nAccept-Language:
en-us;q=1, en;q=0.5\nAccept-Encoding: gzip\nContent-Type: text/xml;
charset=\"utf-8\"\nUser-Agent: gupnp-av-cp GUPnP/0.20.12
DLNADOC/1.50\nConnection: Keep-Alive\nContent-Length: 1405\n"]

 ?: onContent (Text.XML.HaXml.Combinators.deep $ tag "CurrentURI") doc
 -- this result is not what I want.
Document (Prolog (Just (XMLDecl "1.0" Nothing Nothing)) [] Nothing [])
[] (Elem (N "CurrentURI") [] [CString False "\n" file dumb  at line 7
col 19,CMisc (Comment " #EXTM3U ") file dumb  at line 8 col 1,CMisc
(Comment " #EXTINF:5850,\52628\49437\53945\49440\50689\54868
\47609\47329\44284\44053 140909 HDTV x264 720p AC3 \50864\47532\47568
\45433\51020 &#45; nature boy ") file dumb  at line 9 col 1,CMisc
(Comment " http://192.168.219.2:8200/MediaItems/587.mkv ") file dumb
at line 10 col 1,CMisc (Comment " #EXTM3U ") file dumb  at line 11 col
1,CMisc (Comment "
#EXTINF:6411,The.Adventures.of.Tintin.2011.1080p.BluRay.x264.YIFY ")
file dumb  at line 12 col 1,CString False
"http://192.168.219.2:8200/MediaItems/165.mp4\n      " file dumb  at
line 13 col 1]) [Comment "\nOrignal header from captured
packets.\n\nPOST /smp_18_ HTTP/1.1\nHost:
192.168.219.6:7676\nSOAPAction:
\"urn:schemas-upnp-org:service:AVTransport:1#SetAVTransportURI\"\nAccept-Language:
en-us;q=1, en;q=0.5\nAccept-Encoding: gzip\nContent-Type: text/xml;
charset=\"utf-8\"\nUser-Agent: gupnp-av-cp GUPnP/0.20.12
DLNADOC/1.50\nConnection: Keep-Alive\nContent-Length: 1405\n"]
 ?:
```

How can I write filter to replace `CurrentURI` tag?

1. find 'CurrentURI' tag
2. replace content with some string
3. return modified whole document not just specific tag

Or, Where can I found tutorial or example for haxml? Yes, I'm reading
now haxml document. But needs some overview document.

Any advice would be appreciated.

Thanks.

From tjarhad at gmail.com  Thu Sep 18 22:23:50 2014
From: tjarhad at gmail.com (Tushar Jarhad)
Date: Fri, 19 Sep 2014 03:53:50 +0530
Subject: [Haskell-beginners] Parse Json tweets using Aeson
Message-ID: <CAH5gX+49=N2B39th0=o1_ytfFvBYC+czhRWcEs_HGiULt2MyoQ@mail.gmail.com>

{-# LANGUAGE OverloadedStrings #-}
{-# LANGUAGE DeriveGeneric #-}

import Data.Aeson
import GHC.Generics
import qualified Data.ByteString.Lazy as B

data  Tweet  =  Tweet {
                        id_str :: String,
                        text   :: String
                      }deriving Generic


instance FromJSON Tweet


main = do
  input <- B.readFile "1.json"
  let mm = decode input :: Maybe Tweet
  case mm of
    Nothing -> print "error parsing JSON"
    Just m -> (putStrLn.greet) m

greet m = (show.id_str) m


This program is fine if 1.json file has 1 tweet
if 1.json file has multiple tweets then it fails
so how we can parse multiple tweets
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140919/fd82005d/attachment.html>

From mdorman at ironicdesign.com  Fri Sep 19 08:40:55 2014
From: mdorman at ironicdesign.com (Michael Alan Dorman)
Date: Fri, 19 Sep 2014 04:40:55 -0400
Subject: [Haskell-beginners] Parse Json tweets using Aeson
In-Reply-To: <CAH5gX+49=N2B39th0=o1_ytfFvBYC+czhRWcEs_HGiULt2MyoQ@mail.gmail.com>
 (Tushar Jarhad's message of "Fri, 19 Sep 2014 03:53:50 +0530")
References: <CAH5gX+49=N2B39th0=o1_ytfFvBYC+czhRWcEs_HGiULt2MyoQ@mail.gmail.com>
Message-ID: <87oaucuhns.fsf@ironicdesign.com>

Tushar Jarhad <tjarhad at gmail.com> writes:
> let mm = decode input :: Maybe Tweet

You have told the program you expect one tweet---anything else is, by
definition, an error. The program takes you at your word.

So if you want it to parse multiple tweets, tell it that you expect a
*list* of tweets.

> let mm = decode input :: Maybe [Tweet]

Or are you saying that your file simply has multiple tweets appended one
after another?  If so, that file does not represent a valid JSON
structure.

I feel sure there's a way to run the parse incrementally---taking a
complete data structure at a time---but I don't know it offhand.

Mike.

From mdorman at ironicdesign.com  Fri Sep 19 08:51:24 2014
From: mdorman at ironicdesign.com (Michael Alan Dorman)
Date: Fri, 19 Sep 2014 04:51:24 -0400
Subject: [Haskell-beginners] Parse Json tweets using Aeson
In-Reply-To: <87oaucuhns.fsf@ironicdesign.com> (Michael Alan Dorman's message
 of "Fri, 19 Sep 2014 04:40:55 -0400")
References: <CAH5gX+49=N2B39th0=o1_ytfFvBYC+czhRWcEs_HGiULt2MyoQ@mail.gmail.com>
 <87oaucuhns.fsf@ironicdesign.com>
Message-ID: <87k350uh6b.fsf@ironicdesign.com>

Michael Alan Dorman <mdorman at ironicdesign.com> writes:
> I feel sure there's a way to run the parse incrementally---taking a
> complete data structure at a time---but I don't know it offhand.

Since I felt like I *should* know this offhand, I decided to figure it
out.  It's very easy---Aeson provides an Attoparsec parser, so you can,
in fact, do parsing-with-leftovers with virtually zero effort:

    import Data.Aeson
    import Data.Attoparsec.ByteString
    import Data.ByteString

    var :: ByteString
    var = "{\"Some\" : \"Text\"}Leftover"

    main = print $ parse json var

This produces a value representing the simple Object, as well as
leftovers---so you could, presumably, do a fold to pull out as many JSON
items as can be parsed, even if the overall file isn't a valid JSON
structure in itself.

Mike.

From james at jtoll.com  Sat Sep 20 01:48:36 2014
From: james at jtoll.com (James Toll)
Date: Fri, 19 Sep 2014 20:48:36 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
Message-ID: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>

Hi,

I?m trying to install Hakyll within a cabal sandbox.  The issue I?m running into is that the current version of Pandoc doesn?t build because of this bug:

https://github.com/jgm/pandoc/issues/1590

The bug is a simple fix, and is fixed on github, but not in the version on hackage.  I?d like to use ?cabal get? to download the source, manually fix the source, then compile.  The question I have is how do I pick up the installation process after fixing the source?

So far this is what I?ve done:

$ mkdir hakyll
$ cabal update
$ cd hakyll
$ cabal sandbox init --sandbox .
$ cabal install -j hakyll
[this is the spot where pandoc fails to install]
$ cabal get pandoc-1.13.1
$ cd pandoc-1.13.1/
[patch the source]

At this point I need to resume the compilation and installation process.  How do I do that?  Do I move back up to the parent directory and then issue the 'cabal install -j hakyll? command again?

Thanks,


James

From allbery.b at gmail.com  Sat Sep 20 01:54:29 2014
From: allbery.b at gmail.com (Brandon Allbery)
Date: Fri, 19 Sep 2014 21:54:29 -0400
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
Message-ID: <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>

On Fri, Sep 19, 2014 at 9:48 PM, James Toll <james at jtoll.com> wrote:

> At this point I need to resume the compilation and installation process.
> How do I do that?  Do I move back up to the parent directory and then issue
> the 'cabal install -j hakyll? command again?


"cabal install -j" without a package name should install the package in the
current directory as it is. Specifying a package name will cause that
package to be downloaded from Hackage, which in this case you don't want.

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b at gmail.com                                  ballbery at sinenomine.net
unix, openafs, kerberos, infrastructure, xmonad        http://sinenomine.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140919/6332cfe0/attachment.html>

From james at jtoll.com  Sat Sep 20 02:03:26 2014
From: james at jtoll.com (James Toll)
Date: Fri, 19 Sep 2014 21:03:26 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
Message-ID: <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>

On Sep 19, 2014, at 8:54 PM, Brandon Allbery <allbery.b at gmail.com> wrote:

> On Fri, Sep 19, 2014 at 9:48 PM, James Toll <james at jtoll.com> wrote:
> At this point I need to resume the compilation and installation process.  How do I do that?  Do I move back up to the parent directory and then issue the 'cabal install -j hakyll? command again?
> 
> "cabal install -j" without a package name should install the package in the current directory as it is. Specifying a package name will cause that package to be downloaded from Hackage, which in this case you don't want.

Thank you for your suggestion.  I think you're saying that I should ?cabal install -j? from within the pandoc-1.13.1 directory to build pandoc.  But then I will still need to build hakyll, so do I cd up to the hakyll parent directory and cabal install -j from there as well?

Thanks,


James

From allbery.b at gmail.com  Sat Sep 20 02:06:13 2014
From: allbery.b at gmail.com (Brandon Allbery)
Date: Fri, 19 Sep 2014 22:06:13 -0400
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
Message-ID: <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>

On Fri, Sep 19, 2014 at 10:03 PM, James Toll <james at jtoll.com> wrote:

> Thank you for your suggestion.  I think you're saying that I should ?cabal
> install -j? from within the pandoc-1.13.1 directory to build pandoc.  But
> then I will still need to build hakyll, so do I cd up to the hakyll parent
> directory and cabal install -j from there as well?


Just "cabal install -j hakyll" as usual. It'll see the installed pandoc
and, as long as the version is compatible, not reinstall it. It should give
you a warning if it's incompatible. In that case, you'll need to ask the
hakyll folks and possibly have to install whatever package requires it
manually in the same way you are installing pandoc, but in this case you
would just be editing the cabal file to fix the version.

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b at gmail.com                                  ballbery at sinenomine.net
unix, openafs, kerberos, infrastructure, xmonad        http://sinenomine.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140919/c9a575ce/attachment.html>

From james at jtoll.com  Sat Sep 20 04:01:57 2014
From: james at jtoll.com (James Toll)
Date: Fri, 19 Sep 2014 23:01:57 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
 <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
Message-ID: <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>

On Sep 19, 2014, at 9:06 PM, Brandon Allbery <allbery.b at gmail.com> wrote:
> On Fri, Sep 19, 2014 at 10:03 PM, James Toll <james at jtoll.com> wrote:
> Thank you for your suggestion.  I think you're saying that I should ?cabal install -j? from within the pandoc-1.13.1 directory to build pandoc.  But then I will still need to build hakyll, so do I cd up to the hakyll parent directory and cabal install -j from there as well?
> 
> Just "cabal install -j hakyll" as usual. It'll see the installed pandoc and, as long as the version is compatible, not reinstall it. It should give you a warning if it's incompatible. In that case, you'll need to ask the hakyll folks and possibly have to install whatever package requires it manually in the same way you are installing pandoc, but in this case you would just be editing the cabal file to fix the version.

Unfortunately this isn?t working.  The pandoc build appears to work with the patched source, but then the hakyll build fails.  If I?m reading this properly, it appears that pandoc is not getting installed in the sandbox.  Then it appears that hakyll  is not finding the successful pandoc build (probably because it wasn?t installed to the sandbox) and is trying to build pandoc again, using un-patched code and failing with the same original error.  Here is the output, truncated a bit during the pandoc build.  Any suggestions on how to fix this?  Thanks.

James

[james ~/src/haskell/sandboxes/hakyll]$ cd pandoc-1.13.1/
[james ~/src/haskell/sandboxes/hakyll/pandoc-1.13.1]$ cabal install -j 
Resolving dependencies...
In order, the following will be installed:
pandoc-1.13.1 -network-uri (reinstall)
Warning: Note that reinstalls are always dangerous. Continuing anyway...
Configuring pandoc-1.13.1...
Building pandoc-1.13.1...
Installed pandoc-1.13.1
Updating documentation index /Users/james/Library/Haskell/share/doc/index.html
[james ~/src/haskell/sandboxes/hakyll/pandoc-1.13.1]$ cd ..
[james ~/src/haskell/sandboxes/hakyll]$ cabal install -j hakyll
Resolving dependencies...
Notice: installing into a sandbox located at
/Users/james/src/haskell/sandboxes/hakyll
Configuring pandoc-1.13.1...
Building pandoc-1.13.1...
Failed to install pandoc-1.13.1
Build log ( /Users/james/src/haskell/sandboxes/hakyll/logs/pandoc-1.13.1.log ):
[1 of 1] Compiling Main             ( /var/folders/jl/h90f3bxs7hxb389cvjdc6k5w0000gn/T/pandoc-1.13.1-84931/pandoc-1.13.1/dist/dist-sandbox-8fe59d82/setup/setup.hs, /var/folders/jl/h90f3bxs7hxb389cvjdc6k5w0000gn/T/pandoc-1.13.1-84931/pandoc-1.13.1/dist/dist-sandbox-8fe59d82/setup/Main.o )
Linking /var/folders/jl/h90f3bxs7hxb389cvjdc6k5w0000gn/T/pandoc-1.13.1-84931/pandoc-1.13.1/dist/dist-sandbox-8fe59d82/setup/setup ...
Configuring pandoc-1.13.1...
Building pandoc-1.13.1...
Preprocessing library pandoc-1.13.1...
[ 1 of 67] Compiling Text.Pandoc.Asciify
   etc . . . 
[19 of 67] Compiling Text.Pandoc.Shared ( src/Text/Pandoc/Shared.hs, dist/dist-sandbox-8fe59d82/build/Text/Pandoc/Shared.o )

src/Text/Pandoc/Shared.hs:808:55: Not in scope: ?toChunks?
cabal: Error: some packages failed to install:
hakyll-4.5.5.0 depends on pandoc-1.13.1 which failed to install.
pandoc-1.13.1 failed during the building phase. The exception was:
ExitFailure 1
pandoc-citeproc-0.4.0.1 depends on pandoc-1.13.1 which failed to install.



From allbery.b at gmail.com  Sat Sep 20 04:05:45 2014
From: allbery.b at gmail.com (Brandon Allbery)
Date: Sat, 20 Sep 2014 00:05:45 -0400
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
 <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
 <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>
Message-ID: <CAKFCL4XiGHU=5QyciAg1N8RAjbgqpk0C+wwGKeWUGVsoJya5mw@mail.gmail.com>

On Sat, Sep 20, 2014 at 12:01 AM, James Toll <james at jtoll.com> wrote:

> [james ~/src/haskell/sandboxes/hakyll/pandoc-1.13.1]$ cabal install -j
> Resolving dependencies...
> In order, the following will be installed:
>

This sounds like it did not indeed install in the sandbox (compare the
notice after "Resolving dependencies" in the hakyll install). Unfortunately
I don't know what follows from here; it sounds like a cabal bug to me, but
maybe being in the patched pandoc source directory keeps it from seeing the
sandbox, in which case maybe there's an option to point to a specific
sandbox for installs? (I still don't have a working Haskell install on my
new machine, and am not sure at this point when I'll be able to get to it,
sigh.)

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b at gmail.com                                  ballbery at sinenomine.net
unix, openafs, kerberos, infrastructure, xmonad        http://sinenomine.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140920/1a6298f1/attachment-0001.html>

From james at jtoll.com  Sat Sep 20 05:04:35 2014
From: james at jtoll.com (James Toll)
Date: Sat, 20 Sep 2014 00:04:35 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <CAKFCL4XiGHU=5QyciAg1N8RAjbgqpk0C+wwGKeWUGVsoJya5mw@mail.gmail.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
 <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
 <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>
 <CAKFCL4XiGHU=5QyciAg1N8RAjbgqpk0C+wwGKeWUGVsoJya5mw@mail.gmail.com>
Message-ID: <778FC1FC-25AA-4276-A03D-10EBF03D5B7B@jtoll.com>

On Sep 19, 2014, at 11:05 PM, Brandon Allbery <allbery.b at gmail.com> wrote:

> On Sat, Sep 20, 2014 at 12:01 AM, James Toll <james at jtoll.com> wrote:
> [james ~/src/haskell/sandboxes/hakyll/pandoc-1.13.1]$ cabal install -j
> Resolving dependencies...
> In order, the following will be installed:
> 
> This sounds like it did not indeed install in the sandbox (compare the notice after "Resolving dependencies" in the hakyll install). Unfortunately I don't know what follows from here; it sounds like a cabal bug to me, but maybe being in the patched pandoc source directory keeps it from seeing the sandbox, in which case maybe there's an option to point to a specific sandbox for installs? (I still don't have a working Haskell install on my new machine, and am not sure at this point when I'll be able to get to it, sigh.)


Thank you for the help. I wonder if I need to set this up as a shared sandbox so that pandoc gets installed to the sandbox in the parent directory.  Kind of like the first example from the ?Advanced usage? section of:

http://coldwa.st/e/blog/2013-08-20-Cabal-sandbox.html  

Anyway, thank you for pointing me in the right direction.

James

From james at jtoll.com  Sat Sep 20 06:06:09 2014
From: james at jtoll.com (James Toll)
Date: Sat, 20 Sep 2014 01:06:09 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <778FC1FC-25AA-4276-A03D-10EBF03D5B7B@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
 <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
 <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>
 <CAKFCL4XiGHU=5QyciAg1N8RAjbgqpk0C+wwGKeWUGVsoJya5mw@mail.gmail.com>
 <778FC1FC-25AA-4276-A03D-10EBF03D5B7B@jtoll.com>
Message-ID: <52BCCC1F-E5B2-4A26-8D1D-982E29D9103B@jtoll.com>

On Sep 20, 2014, at 12:04 AM, James Toll <james at jtoll.com> wrote:

> On Sep 19, 2014, at 11:05 PM, Brandon Allbery <allbery.b at gmail.com> wrote:
> 
>> On Sat, Sep 20, 2014 at 12:01 AM, James Toll <james at jtoll.com> wrote:
>> [james ~/src/haskell/sandboxes/hakyll/pandoc-1.13.1]$ cabal install -j
>> Resolving dependencies...
>> In order, the following will be installed:
>> 
>> This sounds like it did not indeed install in the sandbox (compare the notice after "Resolving dependencies" in the hakyll install). Unfortunately I don't know what follows from here; it sounds like a cabal bug to me, but maybe being in the patched pandoc source directory keeps it from seeing the sandbox, in which case maybe there's an option to point to a specific sandbox for installs? (I still don't have a working Haskell install on my new machine, and am not sure at this point when I'll be able to get to it, sigh.)
> 
> 
> Thank you for the help. I wonder if I need to set this up as a shared sandbox so that pandoc gets installed to the sandbox in the parent directory.  Kind of like the first example from the ?Advanced usage? section of:
> 
> http://coldwa.st/e/blog/2013-08-20-Cabal-sandbox.html  

Creating a shared sandbox ended up working.  I don?t know it that?s the proper way to do this kind of thing.  For reference, here is what I did to get Hakyll installed with the broken Pandoc dependency:

$ mkdir hakyll
$ cd hakyll
$ cabal sandbox init --sandbox .
$ cabal install -j hakyll
[compilation fails at Pandoc]
$ cabal get pandoc-1.13.1
$ cd pandoc-1.13.1/
$ cabal sandbox init --sandbox ../
[make trivial source change]
$ cabal install -j
$ cd ..
$ cabal install -j hakyll

I?ve not tested the Hakyll installation yet to see if it works properly, but the installation succeeded.

James


From benjaminfjones at gmail.com  Sat Sep 20 06:09:52 2014
From: benjaminfjones at gmail.com (Benjamin Jones)
Date: Fri, 19 Sep 2014 23:09:52 -0700
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <52BCCC1F-E5B2-4A26-8D1D-982E29D9103B@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
 <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
 <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>
 <CAKFCL4XiGHU=5QyciAg1N8RAjbgqpk0C+wwGKeWUGVsoJya5mw@mail.gmail.com>
 <778FC1FC-25AA-4276-A03D-10EBF03D5B7B@jtoll.com>
 <52BCCC1F-E5B2-4A26-8D1D-982E29D9103B@jtoll.com>
Message-ID: <CALfiL_KzLP9+O9CqzeYt3_EoVhShdxshkZFa3TPFto5s84Sj2g@mail.gmail.com>

I ran into this same build problem myself the other day. I believe that
forcing pandoc to build with the https flag set is a workaround until the
next hackage release. You can add this on the command line with:

    $ cabal install pandoc -fhttps

or in the cabal.config file:

    constraints: pandoc +https

--
Benjamin Jones

On Fri, Sep 19, 2014 at 11:06 PM, James Toll <james at jtoll.com> wrote:

> On Sep 20, 2014, at 12:04 AM, James Toll <james at jtoll.com> wrote:
>
> > On Sep 19, 2014, at 11:05 PM, Brandon Allbery <allbery.b at gmail.com>
> wrote:
> >
> >> On Sat, Sep 20, 2014 at 12:01 AM, James Toll <james at jtoll.com> wrote:
> >> [james ~/src/haskell/sandboxes/hakyll/pandoc-1.13.1]$ cabal install -j
> >> Resolving dependencies...
> >> In order, the following will be installed:
> >>
> >> This sounds like it did not indeed install in the sandbox (compare the
> notice after "Resolving dependencies" in the hakyll install). Unfortunately
> I don't know what follows from here; it sounds like a cabal bug to me, but
> maybe being in the patched pandoc source directory keeps it from seeing the
> sandbox, in which case maybe there's an option to point to a specific
> sandbox for installs? (I still don't have a working Haskell install on my
> new machine, and am not sure at this point when I'll be able to get to it,
> sigh.)
> >
> >
> > Thank you for the help. I wonder if I need to set this up as a shared
> sandbox so that pandoc gets installed to the sandbox in the parent
> directory.  Kind of like the first example from the ?Advanced usage?
> section of:
> >
> > http://coldwa.st/e/blog/2013-08-20-Cabal-sandbox.html
>
> Creating a shared sandbox ended up working.  I don?t know it that?s the
> proper way to do this kind of thing.  For reference, here is what I did to
> get Hakyll installed with the broken Pandoc dependency:
>
> $ mkdir hakyll
> $ cd hakyll
> $ cabal sandbox init --sandbox .
> $ cabal install -j hakyll
> [compilation fails at Pandoc]
> $ cabal get pandoc-1.13.1
> $ cd pandoc-1.13.1/
> $ cabal sandbox init --sandbox ../
> [make trivial source change]
> $ cabal install -j
> $ cd ..
> $ cabal install -j hakyll
>
> I?ve not tested the Hakyll installation yet to see if it works properly,
> but the installation succeeded.
>
> James
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140919/1795f9e8/attachment.html>

From nadirsampaoli at gmail.com  Sat Sep 20 06:52:29 2014
From: nadirsampaoli at gmail.com (Nadir Sampaoli)
Date: Sat, 20 Sep 2014 08:52:29 +0200
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <52BCCC1F-E5B2-4A26-8D1D-982E29D9103B@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <CAKFCL4VO--93_ESWGNwipR_2o+ZFDT-suYgwUjypBTGwU7DFVA@mail.gmail.com>
 <03C825EB-E697-4F67-A7B5-A054E8A93A56@jtoll.com>
 <CAKFCL4VoYFs6K7g7it6gv9W9x=q1CV973_xDq1f0b_Z2s5w_dQ@mail.gmail.com>
 <383C4CFD-D581-408C-BB03-4EF49C843AEF@jtoll.com>
 <CAKFCL4XiGHU=5QyciAg1N8RAjbgqpk0C+wwGKeWUGVsoJya5mw@mail.gmail.com>
 <778FC1FC-25AA-4276-A03D-10EBF03D5B7B@jtoll.com>
 <52BCCC1F-E5B2-4A26-8D1D-982E29D9103B@jtoll.com>
Message-ID: <CAFYwTdSdhm44ZCpSmvxqhh7AS1CqdU5hqnMsS07w7UZnhaiz0w@mail.gmail.com>

Hello,

> > I wonder if I need to set this up as a shared sandbox so that pandoc
gets installed to the sandbox in the parent directory.  Kind of like the
first example from the ?Advanced usage? section of:
> >
> > http://coldwa.st/e/blog/2013-08-20-Cabal-sandbox.html
>
> Creating a shared sandbox ended up working.  I don?t know it that?s the
proper way to do this kind of thing.

For the moment Benjamin's solution seems the best way to go but, just for
the record, I think the command you're looking for might be `cabal sandbox
add-source ./your-local-patched-pandoc/` and then `install` (see
http://www.haskell.org/cabal/users-guide/installing-packages.html under
"Sandboxes: Basic usage" or probably the very same article you referred to,
which seems the source for the wiki).

--
Nadir
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140920/62d6257f/attachment.html>

From vlatko.basic at gmail.com  Sat Sep 20 07:43:12 2014
From: vlatko.basic at gmail.com (Vlatko Basic)
Date: Sat, 20 Sep 2014 09:43:12 +0200
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
Message-ID: <541D3010.9050100@gmail.com>


Get (patched) package into a separate dir
$ cabal sandbox add-source /my/patched/library # Add a new add-source dependency
$ cabal install --dependencies-only            # Install it into the sandbox
$ cabal build                                  # Build the local package

This way cabal first searches in the add-source dir for any dependency, and 
builds it from there. You can add several dirs with patched packages

You can/might have to add constraints in local cabal.config file (in sandboxed 
dir) to force cabal to use the latest version of the package

If a file in any add-source dir is changed, cabal will rebuild it automatically

Take a look at cabal user guide
http://www.haskell.org/cabal/users-guide/installing-packages.html

vlatko

-------- Original Message  --------
Subject: [Haskell-beginners] Patch source, then pickup compile
From: James Toll <james at jtoll.com>
To: haskell-beginners <beginners at haskell.org>
Date: 20.09.2014 03:48

> Hi,
>
> I?m trying to install Hakyll within a cabal sandbox.  The issue I?m running into is that the current version of Pandoc doesn?t build because of this bug:
>
> https://github.com/jgm/pandoc/issues/1590
>
> The bug is a simple fix, and is fixed on github, but not in the version on hackage.  I?d like to use ?cabal get? to download the source, manually fix the source, then compile.  The question I have is how do I pick up the installation process after fixing the source?
>
> So far this is what I?ve done:
>
> $ mkdir hakyll
> $ cabal update
> $ cd hakyll
> $ cabal sandbox init --sandbox .
> $ cabal install -j hakyll
> [this is the spot where pandoc fails to install]
> $ cabal get pandoc-1.13.1
> $ cd pandoc-1.13.1/
> [patch the source]
>
> At this point I need to resume the compilation and installation process.  How do I do that?  Do I move back up to the parent directory and then issue the 'cabal install -j hakyll? command again?
>
> Thanks,
>
>
> James
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>

From james at jtoll.com  Sat Sep 20 16:58:05 2014
From: james at jtoll.com (James Toll)
Date: Sat, 20 Sep 2014 11:58:05 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <541D3010.9050100@gmail.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <541D3010.9050100@gmail.com>
Message-ID: <0796DBCD-62E4-4378-9883-B67031A26C49@jtoll.com>

On Sep 20, 2014, at 2:43 AM, Vlatko Basic <vlatko.basic at gmail.com> wrote:
> Get (patched) package into a separate dir
> $ cabal sandbox add-source /my/patched/library # Add a new add-source dependency
> $ cabal install --dependencies-only            # Install it into the sandbox
> $ cabal build                                  # Build the local package
> 
> This way cabal first searches in the add-source dir for any dependency, and builds it from there. You can add several dirs with patched packages
> 
> You can/might have to add constraints in local cabal.config file (in sandboxed dir) to force cabal to use the latest version of the package

Thanks to you, Nadir, and Benjamin for the follow-up emails explaining the proper way to handle this.  From your emails, it sounds like instead of using a shared sandbox, I should have added the Pandoc source using 'cabal sandbox add-source'.  What I?m wondering now is what are the implications of the way I managed to get it to work?  And whether I should go back and re-do it the way you?ve described.

I do have one concern though.  The part where you mention that if any of the source files are changed, then cabal will rebuild.  I don?t want or need that functionality given that this is supposed to be a temporary work around until the fixed version of Pandoc is in Hackage.

> If a file in any add-source dir is changed, cabal will rebuild it automatically

If possible, I might prefer to just delete the pandoc-1.13.1 source subdirectory entirely, as long as it doesn?t mess up the current hakyll installation. Is that possible?  If not, in the future, once the version of Pandoc in Hackage has been updated, is it then possible to remove the added source?

Thanks,


James

From nadirsampaoli at gmail.com  Sat Sep 20 17:08:24 2014
From: nadirsampaoli at gmail.com (Nadir Sampaoli)
Date: Sat, 20 Sep 2014 19:08:24 +0200
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <0796DBCD-62E4-4378-9883-B67031A26C49@jtoll.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <541D3010.9050100@gmail.com>
 <0796DBCD-62E4-4378-9883-B67031A26C49@jtoll.com>
Message-ID: <CAFYwTdRZjppo220jkRgqL-U6ExAEgp_OTjjza5yrdQ9+0NBskQ@mail.gmail.com>

Il 20/set/2014 18:58 "James Toll" <james at jtoll.com> wrote:
>
> I do have one concern though.  The part where you mention that if any of
the source files are changed, then cabal will rebuild.  I don?t want or
need that functionality given that this is supposed to be a temporary work
around until the fixed version of Pandoc is in Hackage.

According to the wiki ("sandboxes: advanced usage" section) you should use
`cabal sandbox add-source --snapshot` "which disables the change tracking".

>
> If possible, I might prefer to just delete the pandoc-1.13.1 source
subdirectory entirely, as long as it doesn?t mess up the current hakyll
installation. Is that possible?  If not, in the future, once the version of
Pandoc in Hackage has been updated, is it then possible to remove the added
source?
To remove an added source I think the correct sandbox subcommand is
`delete-source`. See the --help for more info.

--
Nadir
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140920/40401c28/attachment.html>

From james at jtoll.com  Sat Sep 20 17:41:28 2014
From: james at jtoll.com (James Toll)
Date: Sat, 20 Sep 2014 12:41:28 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <CAFYwTdRZjppo220jkRgqL-U6ExAEgp_OTjjza5yrdQ9+0NBskQ@mail.gmail.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <541D3010.9050100@gmail.com> <0796DBCD-62E4-4378-9883-B67031A26C49@jtoll.com>
 <CAFYwTdRZjppo220jkRgqL-U6ExAEgp_OTjjza5yrdQ9+0NBskQ@mail.gmail.com>
Message-ID: <03F81764-63A2-477B-95F6-B2BA3EC4C6DA@jtoll.com>


On Sep 20, 2014, at 12:08 PM, Nadir Sampaoli <nadirsampaoli at gmail.com> wrote:

> 
> Il 20/set/2014 18:58 "James Toll" <james at jtoll.com> wrote:
> >
> > I do have one concern though.  The part where you mention that if any of the source files are changed, then cabal will rebuild.  I don?t want or need that functionality given that this is supposed to be a temporary work around until the fixed version of Pandoc is in Hackage.
> 
> According to the wiki ("sandboxes: advanced usage" section) you should use `cabal sandbox add-source --snapshot` "which disables the change tracking".
> 
> >
> > If possible, I might prefer to just delete the pandoc-1.13.1 source subdirectory entirely, as long as it doesn?t mess up the current hakyll installation. Is that possible?  If not, in the future, once the version of Pandoc in Hackage has been updated, is it then possible to remove the added source?
> To remove an added source I think the correct sandbox subcommand is `delete-source`. See the --help for more info.

Yes, thank you.  The ??snapshot' was in the documentation.  I should have noticed it.

The 'delete-source' doesn?t appear to be in the ?cabal sandbox ?help? page, although 'cabal sandbox list-sources? supposedly mentions it once you have local build trees.  I'll see if that ends up being the case.

Thank you.

James


From james at jtoll.com  Sun Sep 21 01:26:40 2014
From: james at jtoll.com (James Toll)
Date: Sat, 20 Sep 2014 20:26:40 -0500
Subject: [Haskell-beginners] Patch source, then pickup compile
In-Reply-To: <CAFYwTdRZjppo220jkRgqL-U6ExAEgp_OTjjza5yrdQ9+0NBskQ@mail.gmail.com>
References: <25BB30C7-FABA-4EB0-884F-179F7D326694@jtoll.com>
 <541D3010.9050100@gmail.com> <0796DBCD-62E4-4378-9883-B67031A26C49@jtoll.com>
 <CAFYwTdRZjppo220jkRgqL-U6ExAEgp_OTjjza5yrdQ9+0NBskQ@mail.gmail.com>
Message-ID: <E439BC06-2BF3-46E8-88F8-6D20C9B0DF72@jtoll.com>


On Sep 20, 2014, at 12:08 PM, Nadir Sampaoli <nadirsampaoli at gmail.com> wrote:
> Il 20/set/2014 18:58 "James Toll" <james at jtoll.com> wrote:
> >
> > I do have one concern though.  The part where you mention that if any of the source files are changed, then cabal will rebuild.  I don?t want or need that functionality given that this is supposed to be a temporary work around until the fixed version of Pandoc is in Hackage.
> 
> According to the wiki ("sandboxes: advanced usage" section) you should use `cabal sandbox add-source --snapshot` "which disables the change tracking".
> 
> >
> > If possible, I might prefer to just delete the pandoc-1.13.1 source subdirectory entirely, as long as it doesn?t mess up the current hakyll installation. Is that possible?  If not, in the future, once the version of Pandoc in Hackage has been updated, is it then possible to remove the added source?
> To remove an added source I think the correct sandbox subcommand is `delete-source`. See the --help for more info.


I wish I could say that I got this to work, but I?ve had no success.  The problem I?m having is that I can?t even get 'cabal sandbox add-source --snapshot? to work.  This is what I?m trying to do, and the output:

$ mkdir hakyll
$ cd hakyll
$ cabal sandbox init --sandbox .
$ cabal install -j hakyll
[compilation fails]
$ cabal get pandoc-1.13.1
$ cabal sandbox add-source --snapshot pandoc-1.13.1/
cabal: Error: Could not find module: Text.Pandoc.Data with any suffix:
["gc","chs","hsc","x","y","ly","cpphs","hs","lhs?]

This last command creates the ?snapshots' directory within the hakyll directory.  Snapshots contains an empty subdirectory called ?pandoc-1.13.1-tmp?. If I then try a different slightly different path to pandoc, it fails, but differently:

$ cabal sandbox add-source --snapshot ./pandoc-1.13.1/
cabal: /Users/james/src/haskell/sandboxes/hakyll/snapshots/pandoc-1.13.1: does
not exist

If I delete the ?snapshots? directory, it fails with the first error where it can?t find the module.

What am I doing wrong?  I?ve tried many different iterations of this.  Executing the command from within the pandoc directory, renaming the directory, plus many other variations.  Nothing has worked.  How is this supposed to work.  It just doesn?t seem like it should be this hard.

Thanks.

James



From derek.mcloughlin at gmail.com  Sun Sep 21 14:02:55 2014
From: derek.mcloughlin at gmail.com (Derek McLoughlin)
Date: Sun, 21 Sep 2014 15:02:55 +0100
Subject: [Haskell-beginners] Using a monadic operation in a list
	comprehension.
Message-ID: <CAAw9fmkLRmuiDRKpTd8ng2u=cxCZDuadRCVm90ras2JZet-TmQ@mail.gmail.com>

Hi,

I'm trying to get Graham Hutton's "Programming with Effects" article
(http://www.cs.nott.ac.uk/~gmh/monads) working with Data.Tree instead
of the Tree he provides.

My code is here:

https://gist.github.com/derekmcloughlin/95ab690e0c07c5a7221e

The compiler is complaining about the use of

   lab <- mlabel c

in the list comprehension:

Couldn't match expected type ?[Tree (a, Int)]?
                with actual type ?ST (Tree (a, Int))?

This makes sense, but is there any way to execute the monad operation
"mlabel" within the list comprehension? I've tried the
MonadComprehensions extension but still get an error.

Derek.

From defigueiredo at ucdavis.edu  Sun Sep 21 17:05:43 2014
From: defigueiredo at ucdavis.edu (Dimitri DeFigueiredo)
Date: Sun, 21 Sep 2014 11:05:43 -0600
Subject: [Haskell-beginners] what space leak could happen here?
Message-ID: <541F0567.4060205@ucdavis.edu>

Hi All,

I was looking at the definition of intersperse on

http://hackage.haskell.org/package/base-4.7.0.1/docs/src/Data-List.html#intersperse

And found this

-- | The 'intersperse' function takes an element and a list and
-- \`intersperses\' that element between the elements of the list.
-- For example,
--
-- > intersperse ',' "abcde" == "a,b,c,d,e"

intersperse             :: a -> [a] -> [a]
intersperse _   []      = []
intersperse sep (x:xs)  = x : prependToAll sep xs


-- Not exported:
-- We want to make every element in the 'intersperse'd list available
-- as soon as possible to avoid space leaks. Experiments suggested that
-- a separate top-level helper is more efficient than a local worker.
prependToAll            :: a -> [a] -> [a]
prependToAll _   []     = []
prependToAll sep (x:xs) = sep : x : prependToAll sep xs



I don't understand why we need to "make every element in the 
'intersperse'd list available as soon as possible to avoid space leaks." 
Could somebody shed some light on what could cause a space leak here? In 
particular, would this simpler version cause problems?

intersperse :: a -> [a] -> [a]
intersperse _ []  = []
intersperse _ [x] = [x]
intersperse a (x: xs@(y:ys)) = x : a : intersperse a xs



Thanks,


Dimitri



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140921/2f4d7959/attachment.html>

From ky3 at atamo.com  Sun Sep 21 18:38:36 2014
From: ky3 at atamo.com (Kim-Ee Yeoh)
Date: Mon, 22 Sep 2014 01:38:36 +0700
Subject: [Haskell-beginners] Using a monadic operation in a list
	comprehension.
In-Reply-To: <CAAw9fmkLRmuiDRKpTd8ng2u=cxCZDuadRCVm90ras2JZet-TmQ@mail.gmail.com>
References: <CAAw9fmkLRmuiDRKpTd8ng2u=cxCZDuadRCVm90ras2JZet-TmQ@mail.gmail.com>
Message-ID: <CAPY+ZdRiYOkBtKfZ5E0g6Sr=iSkPS5RW10xj-Si2bzW+FKVUkw@mail.gmail.com>

On Sun, Sep 21, 2014 at 9:02 PM, Derek McLoughlin <
derek.mcloughlin at gmail.com> wrote:

> I'm trying to get Graham Hutton's "Programming with Effects" article
> (http://www.cs.nott.ac.uk/~gmh/monads) working with Data.Tree instead
> of the Tree he provides.
>

Hi Derek,

This is a lovely problem that's inspired a bunch of interesting papers
recently.

Here's a stab at the compilation error you're facing:

http://www.atamo.com/blog/how-to-solve-a-tricky-monad-problem-1/

Enjoy  :)

-- Kim-Ee
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140922/5b88c000/attachment.html>

From chaddai.fouche at gmail.com  Sun Sep 21 19:20:04 2014
From: chaddai.fouche at gmail.com (=?UTF-8?B?Q2hhZGRhw68gRm91Y2jDqQ==?=)
Date: Sun, 21 Sep 2014 21:20:04 +0200
Subject: [Haskell-beginners] what space leak could happen here?
In-Reply-To: <541F0567.4060205@ucdavis.edu>
References: <541F0567.4060205@ucdavis.edu>
Message-ID: <CANfjZRZ-=iFhvEE79ah+HjmxNZO9NB53o=eZgGZ3io61UbpoCw@mail.gmail.com>

On Sun, Sep 21, 2014 at 7:05 PM, Dimitri DeFigueiredo <
defigueiredo at ucdavis.edu> wrote:

> I don't understand why we need to "make every element in the
> 'intersperse'd list available as soon as possible to avoid space leaks."
> Could somebody shed some light on what could cause a space leak here? In
> particular, would this simpler version cause problems?
>

This is delicate but basically your version needs to force the second
element of the list (to check if it's []) before it produce the first
element of the result while the base version doesn't, check with "take 1 .
intersperse ',' $ 'a' : undefined".

The leak can occur when you only need the first element of the result to
start producing but the second one is expensive to produce and store, the
base version avoid the problem by being maximally lazy.
-- 
Jeda?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140921/9a2117a2/attachment.html>

From defigueiredo at ucdavis.edu  Sun Sep 21 20:12:04 2014
From: defigueiredo at ucdavis.edu (Dimitri DeFigueiredo)
Date: Sun, 21 Sep 2014 14:12:04 -0600
Subject: [Haskell-beginners] what space leak could happen here?
In-Reply-To: <CANfjZRZ-=iFhvEE79ah+HjmxNZO9NB53o=eZgGZ3io61UbpoCw@mail.gmail.com>
References: <541F0567.4060205@ucdavis.edu>
 <CANfjZRZ-=iFhvEE79ah+HjmxNZO9NB53o=eZgGZ3io61UbpoCw@mail.gmail.com>
Message-ID: <541F3114.4010308@ucdavis.edu>

Thanks!

I don't quite understand the last sentence, though. So, it is a leak 
just because it needs to evaluate an element that may not be required? I 
thought a space leak implied we would be changing the "Big O" asymptotic 
order of space usage. How would the extra evaluation would change that?

Thanks again,

Dimitri

On 21/09/14 13:20, Chadda? Fouch? wrote:
> On Sun, Sep 21, 2014 at 7:05 PM, Dimitri DeFigueiredo 
> <defigueiredo at ucdavis.edu <mailto:defigueiredo at ucdavis.edu>> wrote:
>
>     I don't understand why we need to "make every element in the
>     'intersperse'd list available as soon as possible to avoid space
>     leaks." Could somebody shed some light on what could cause a space
>     leak here? In particular, would this simpler version cause problems?
>
>
> This is delicate but basically your version needs to force the second 
> element of the list (to check if it's []) before it produce the first 
> element of the result while the base version doesn't, check with "take 
> 1 . intersperse ',' $ 'a' : undefined".
>
> The leak can occur when you only need the first element of the result 
> to start producing but the second one is expensive to produce and 
> store, the base version avoid the problem by being maximally lazy.
> -- 
> Jeda?
>
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140921/5e52412e/attachment.html>

From derek.mcloughlin at gmail.com  Sun Sep 21 20:55:50 2014
From: derek.mcloughlin at gmail.com (Derek McLoughlin)
Date: Sun, 21 Sep 2014 21:55:50 +0100
Subject: [Haskell-beginners] Using a monadic operation in a list
	comprehension.
In-Reply-To: <CAPY+ZdRiYOkBtKfZ5E0g6Sr=iSkPS5RW10xj-Si2bzW+FKVUkw@mail.gmail.com>
References: <CAAw9fmkLRmuiDRKpTd8ng2u=cxCZDuadRCVm90ras2JZet-TmQ@mail.gmail.com>
 <CAPY+ZdRiYOkBtKfZ5E0g6Sr=iSkPS5RW10xj-Si2bzW+FKVUkw@mail.gmail.com>
Message-ID: <CAAw9fmkeJHo_pMvMjLpEhN0vY1kaEZ6M_qsL0mS=xWNyx+vq1A@mail.gmail.com>

Thanks very much for your clear explanation! Worked perfectly. You're
right about the ST - I got very confused at the start with
Control.Monad.ST. I'll be removing Graham's ST in favour of the
'normal' State monad.

On 21 September 2014 19:38, Kim-Ee Yeoh <ky3 at atamo.com> wrote:
>
> On Sun, Sep 21, 2014 at 9:02 PM, Derek McLoughlin
> <derek.mcloughlin at gmail.com> wrote:
>>
>> I'm trying to get Graham Hutton's "Programming with Effects" article
>> (http://www.cs.nott.ac.uk/~gmh/monads) working with Data.Tree instead
>> of the Tree he provides.
>
>
> Hi Derek,
>
> This is a lovely problem that's inspired a bunch of interesting papers
> recently.
>
> Here's a stab at the compilation error you're facing:
>
> http://www.atamo.com/blog/how-to-solve-a-tricky-monad-problem-1/
>
> Enjoy  :)
>
> -- Kim-Ee
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>

From phil at cnphil.com  Mon Sep 22 04:10:45 2014
From: phil at cnphil.com (Phil Xiaojun Hu)
Date: Mon, 22 Sep 2014 12:10:45 +0800
Subject: [Haskell-beginners] what space leak could happen here?
In-Reply-To: <541F3114.4010308@ucdavis.edu>
References: <541F0567.4060205@ucdavis.edu>
 <CANfjZRZ-=iFhvEE79ah+HjmxNZO9NB53o=eZgGZ3io61UbpoCw@mail.gmail.com>
 <541F3114.4010308@ucdavis.edu>
Message-ID: <20140922041045.GA23800@narita.takau.net>

On Sun, Sep 21, 2014 at 02:12:04PM -0600, Dimitri DeFigueiredo wrote:
> Thanks!
> 
> I don't quite understand the last sentence, though. So, it is a leak just
> because it needs to evaluate an element that may not be required? I thought
> a space leak implied we would be changing the "Big O" asymptotic order of
> space usage. How would the extra evaluation would change that?

Think of a network socket as the source of your list, and the first
character is all you need. If you need to evaluate the second character
to get the first, you could be stuck on the network socket for some
time waiting for the second one to come.

And I think the reason it would break some "Big O"s is that most
complexity analysis in Haskell are amortized bounds and it uses the fact
that Haskell is lazy and won't evaluate something until we actually need
it.

--
Phil Xiaojun Hu
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140922/00bc6e38/attachment-0001.sig>

From allbery.b at gmail.com  Mon Sep 22 04:19:19 2014
From: allbery.b at gmail.com (Brandon Allbery)
Date: Mon, 22 Sep 2014 00:19:19 -0400
Subject: [Haskell-beginners] what space leak could happen here?
In-Reply-To: <20140922041045.GA23800@narita.takau.net>
References: <541F0567.4060205@ucdavis.edu>
 <CANfjZRZ-=iFhvEE79ah+HjmxNZO9NB53o=eZgGZ3io61UbpoCw@mail.gmail.com>
 <541F3114.4010308@ucdavis.edu>
 <20140922041045.GA23800@narita.takau.net>
Message-ID: <CAKFCL4W9dvOJManR7G2CAHV1ijXci4xos8ThmOz+RjveNntb4w@mail.gmail.com>

On Mon, Sep 22, 2014 at 12:10 AM, Phil Xiaojun Hu <phil at cnphil.com> wrote:

> Think of a network socket as the source of your list, and the first
> character is all you need. If you need to evaluate the second character
> to get the first, you could be stuck on the network socket for some
> time waiting for the second one to come.
>

That's more of a time leak. I think the space leak aspect comes from the
fact that you've evaluated a second list element that was not requested,
may never be used, and will be held in the heap until all shared consumers
of the list go out of scope. (It may of course also be a time leak if
computing that second element is expensive.)

-- 
brandon s allbery kf8nh                               sine nomine associates
allbery.b at gmail.com                                  ballbery at sinenomine.net
unix, openafs, kerberos, infrastructure, xmonad        http://sinenomine.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140922/ac44bccc/attachment.html>

From miroslav.karpis at gmail.com  Tue Sep 23 11:04:44 2014
From: miroslav.karpis at gmail.com (Miroslav Karpis)
Date: Tue, 23 Sep 2014 13:04:44 +0200
Subject: [Haskell-beginners] fresh install - cabal version confusion
Message-ID: <D047206C.B79E%miroslav.karpis@gmail.com>

Hi, please can you help me with following? I?m having some issues with a
Haskell upgrade. 
Before upgrade I have uninstalled Haskell, removed directory + removed .ghc
and .cabal files. After I install haskell and run cabal ?V I get the output
below. Confusing is the versions cabal returns: cabal-install version
1.18.0.5 using version 1.18.1.4? I have tried to run also cabal update, what
installed 1.20.0.3, but I still got the same output. What does it mean?

I?m running on Mac.


**********************************************************************
=== Configuration for cabal has been written to
    /Users/miro/.cabal/config

=== Executables will be installed in:
    /Users/miro/Library/Haskell/bin

    You may wish to place this on your PATH by adding the following
    line to your ~/.bash_profile:

    export PATH="$HOME/Library/Haskell/bin:$PATH"

=== When documentation is built, a master index to all documentation
    will be placed in:

    /Users/miro/Library/Haskell/doc/index.html

    You may wish to bookmark that file once it gets built (after the
    first cabal install).

**********************************************************************

Downloading the latest package list from hackage.haskell.org
Note: there is a new version of cabal-install available.
To upgrade, run: cabal install cabal-install
cabal-install version 1.18.0.5
using version 1.18.1.4 of the Cabal library


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140923/3a7474a9/attachment.html>

From miroslav.karpis at gmail.com  Tue Sep 23 11:10:36 2014
From: miroslav.karpis at gmail.com (Miroslav Karpis)
Date: Tue, 23 Sep 2014 13:10:36 +0200
Subject: [Haskell-beginners] FW: fresh install - cabal version confusion
In-Reply-To: <D047206C.B79E%miroslav.karpis@gmail.com>
References: <D047206C.B79E%miroslav.karpis@gmail.com>
Message-ID: <D04720FE.B7A1%miroslav.karpis@gmail.com>

So I found out that cabal and cabal-install are 2 different modules ;-)

Would it be not cleaner with this:
> cabal-install version 1.18.0.5
> Cabal version 1.18.1.4

Instead of what we currently get?
> cabal-install version 1.18.0.5
> using version 1.18.1.4 of the Cabal library


Cheers Miro


From:  Miroslav Karpis <miroslav.karpis at gmail.com>
Date:  Tuesday 23 September 2014 13:04
To:  Haskell-Beginners Haskell <Beginners at haskell.org>
Subject:  fresh install - cabal version confusion

Hi, please can you help me with following? I?m having some issues with a
Haskell upgrade. 
Before upgrade I have uninstalled Haskell, removed directory + removed .ghc
and .cabal files. After I install haskell and run cabal ?V I get the output
below. Confusing is the versions cabal returns: cabal-install version
1.18.0.5 using version 1.18.1.4? I have tried to run also cabal update, what
installed 1.20.0.3, but I still got the same output. What does it mean?

I?m running on Mac.


**********************************************************************
=== Configuration for cabal has been written to
    /Users/miro/.cabal/config

=== Executables will be installed in:
    /Users/miro/Library/Haskell/bin

    You may wish to place this on your PATH by adding the following
    line to your ~/.bash_profile:

    export PATH="$HOME/Library/Haskell/bin:$PATH"

=== When documentation is built, a master index to all documentation
    will be placed in:

    /Users/miro/Library/Haskell/doc/index.html

    You may wish to bookmark that file once it gets built (after the
    first cabal install).

**********************************************************************

Downloading the latest package list from hackage.haskell.org
Note: there is a new version of cabal-install available.
To upgrade, run: cabal install cabal-install
cabal-install version 1.18.0.5
using version 1.18.1.4 of the Cabal library


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140923/4abf9ec5/attachment.html>

From gmizrahi22 at gmail.com  Tue Sep 23 11:13:45 2014
From: gmizrahi22 at gmail.com (Gil Mizrahi)
Date: Tue, 23 Sep 2014 14:13:45 +0300
Subject: [Haskell-beginners] fresh install - cabal version confusion
In-Reply-To: <D047206C.B79E%miroslav.karpis@gmail.com>
References: <D047206C.B79E%miroslav.karpis@gmail.com>
Message-ID: <CAKAQuQsjZxgr40xmPZUEVs7uDxRW30EPQVjcLcdKinJzcUqD8Q@mail.gmail.com>

Hi,

You probably want to add the path in which the new cabal was installed to
the $PATH variable in your ~/.bashrc or ~/.bash_profile file. So add the
following line:
    export PATH="$HOME/Library/Haskell/bin:$PATH"
at the end of either ~/.bash_profile or ~/.bashrc and restart your terminal.

This will tell your shell where to find the new cabal you installed with
cabal install.

On Tue, Sep 23, 2014 at 2:04 PM, Miroslav Karpis <miroslav.karpis at gmail.com>
wrote:

> Hi, please can you help me with following? I?m having some issues with a
> Haskell upgrade.
> Before upgrade I have uninstalled Haskell, removed directory + removed
> .ghc and .cabal files. After I install haskell and run cabal ?V I get the
> output below. Confusing is the versions cabal returns: *cabal-install
> version 1.18.0.5 using version 1.18.1.4*? I have tried to run also cabal
> update, what installed 1.20.0.3, but I still got the same output. What does
> it mean?
>
> I?m running on Mac.
>
>
> **********************************************************************
> === Configuration for cabal has been written to
>     /Users/miro/.cabal/config
>
> === Executables will be installed in:
>     /Users/miro/Library/Haskell/bin
>
>     You may wish to place this on your PATH by adding the following
>     line to your ~/.bash_profile:
>
>     export PATH="$HOME/Library/Haskell/bin:$PATH"
>
> === When documentation is built, a master index to all documentation
>     will be placed in:
>
>     /Users/miro/Library/Haskell/doc/index.html
>
>     You may wish to bookmark that file once it gets built (after the
>     first cabal install).
>
> **********************************************************************
>
> Downloading the latest package list from hackage.haskell.org
> Note: there is a new version of cabal-install available.
> To upgrade, run: cabal install cabal-install
> cabal-install version 1.18.0.5
> using version 1.18.1.4 of the Cabal library
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140923/2e9d46d8/attachment.html>

From ky3 at atamo.com  Thu Sep 25 19:07:13 2014
From: ky3 at atamo.com (Kim-Ee Yeoh)
Date: Fri, 26 Sep 2014 02:07:13 +0700
Subject: [Haskell-beginners] Using a monadic operation in a list
	comprehension.
In-Reply-To: <CAAw9fmkeJHo_pMvMjLpEhN0vY1kaEZ6M_qsL0mS=xWNyx+vq1A@mail.gmail.com>
References: <CAAw9fmkLRmuiDRKpTd8ng2u=cxCZDuadRCVm90ras2JZet-TmQ@mail.gmail.com>
 <CAPY+ZdRiYOkBtKfZ5E0g6Sr=iSkPS5RW10xj-Si2bzW+FKVUkw@mail.gmail.com>
 <CAAw9fmkeJHo_pMvMjLpEhN0vY1kaEZ6M_qsL0mS=xWNyx+vq1A@mail.gmail.com>
Message-ID: <CAPY+ZdQciAP1AYYGhPs7q8mDyOYAc2eUAivUy+8=CQ_hFKjivQ@mail.gmail.com>

On Mon, Sep 22, 2014 at 3:55 AM, Derek McLoughlin <
derek.mcloughlin at gmail.com> wrote:

> Thanks very much for your clear explanation! Worked perfectly.


Thanks for taking the time to say this :)

I hope I didn't belabor the mismatched effects bit. Combining effects is a
wide open field of research if you look at papers like Extensible Effects,
Monad Coproducts, etc.

Point being, there just might be a way to make your code as originally
written compile and work as you expect it to. Ah, the joys of life at the
bleeding edge!

-- Kim-Ee
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140926/9f28f7c0/attachment.html>

From miroslav.karpis at gmail.com  Fri Sep 26 14:34:32 2014
From: miroslav.karpis at gmail.com (Miro Karpis)
Date: Fri, 26 Sep 2014 16:34:32 +0200
Subject: [Haskell-beginners] aeson json paring
Message-ID: <CAJnnbxF_wrS=8AZQfQXdEN4O7tD98YuORhYV9R8+EWyAtBk__Q@mail.gmail.com>

Hi, I'm trying to run an example from the eason documentation
<https://hackage.haskell.org/package/aeson-0.6.1.0/docs/Data-Aeson.html#g:3>
:

?> do result <- decode "{\"name\":\"Dave\",\"age\":2}"
       flip parseMaybe result $ \obj -> do
         age <- obj .: "age"
         name <- obj .: "name"
         return (name ++ ": " ++ show (age*2))


I made a function:

jsonTest = do
	result <- decode "{\"name\":\"Dave\",\"age\":2}"
	flip parseMaybe result $ \obj -> do
		age <- obj .: "age"
		name <- obj .: "name"
		return (name ++ ": " ++ show (age*2))

which can not compile:
hh.hs:35:41:
    No instance for (Show a0) arising from a use of ?show?
    The type variable ?a0? is ambiguous
....
...


Please what am I doing wrong?

Cheers, Miro
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140926/f1394940/attachment.html>

From raphaelsimeon at gmail.com  Fri Sep 26 14:41:35 2014
From: raphaelsimeon at gmail.com (=?UTF-8?Q?Rapha=C3=ABl_Mongeau?=)
Date: Fri, 26 Sep 2014 10:41:35 -0400
Subject: [Haskell-beginners] aeson json paring
In-Reply-To: <CAJnnbxF_wrS=8AZQfQXdEN4O7tD98YuORhYV9R8+EWyAtBk__Q@mail.gmail.com>
References: <CAJnnbxF_wrS=8AZQfQXdEN4O7tD98YuORhYV9R8+EWyAtBk__Q@mail.gmail.com>
Message-ID: <CAMTQk-FavSr0rxcgTyNV0HhmCu5mDW4b7WiyW4+bZ5d=DVbHFA@mail.gmail.com>

The key here is "The type variable ?a0? is ambiguous".

You need to add a type signature to age so the compiler know if age is of
type In,Float or Double ...

:)

2014-09-26 10:34 GMT-04:00 Miro Karpis <miroslav.karpis at gmail.com>:

> Hi, I'm trying to run an example from the eason documentation
> <https://hackage.haskell.org/package/aeson-0.6.1.0/docs/Data-Aeson.html#g:3>
> :
>
> ?> do result <- decode "{\"name\":\"Dave\",\"age\":2}"
>        flip parseMaybe result $ \obj -> do
>          age <- obj .: "age"
>          name <- obj .: "name"
>          return (name ++ ": " ++ show (age*2))
>
>
> I made a function:
>
> jsonTest = do
> 	result <- decode "{\"name\":\"Dave\",\"age\":2}"
> 	flip parseMaybe result $ \obj -> do
> 		age <- obj .: "age"
> 		name <- obj .: "name"
> 		return (name ++ ": " ++ show (age*2))
>
> which can not compile:
> hh.hs:35:41:
>     No instance for (Show a0) arising from a use of ?show?
>     The type variable ?a0? is ambiguous
> ....
> ...
>
>
> Please what am I doing wrong?
>
> Cheers, Miro
>
>
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
>


-- 
Viva Cila
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140926/0f8002eb/attachment.html>

From fuuzetsu at fuuzetsu.co.uk  Sat Sep 27 04:20:11 2014
From: fuuzetsu at fuuzetsu.co.uk (Mateusz Kowalczyk)
Date: Sat, 27 Sep 2014 05:20:11 +0100
Subject: [Haskell-beginners] aeson json paring
In-Reply-To: <CAJnnbxF_wrS=8AZQfQXdEN4O7tD98YuORhYV9R8+EWyAtBk__Q@mail.gmail.com>
References: <CAJnnbxF_wrS=8AZQfQXdEN4O7tD98YuORhYV9R8+EWyAtBk__Q@mail.gmail.com>
Message-ID: <54263AFB.5040208@fuuzetsu.co.uk>

On 09/26/2014 03:34 PM, Miro Karpis wrote:
> Hi, I'm trying to run an example from the eason documentation
> <https://hackage.haskell.org/package/aeson-0.6.1.0/docs/Data-Aeson.html#g:3>
> :
> 
> ?> do result <- decode "{\"name\":\"Dave\",\"age\":2}"
>        flip parseMaybe result $ \obj -> do
>          age <- obj .: "age"
>          name <- obj .: "name"
>          return (name ++ ": " ++ show (age*2))
> 
> 
> I made a function:
> 
> jsonTest = do
> 	result <- decode "{\"name\":\"Dave\",\"age\":2}"
> 	flip parseMaybe result $ \obj -> do
> 		age <- obj .: "age"
> 		name <- obj .: "name"
> 		return (name ++ ": " ++ show (age*2))
> 
> which can not compile:
> hh.hs:35:41:
>     No instance for (Show a0) arising from a use of ?show?
>     The type variable ?a0? is ambiguous
> ....
> ...
> 
> 
> Please what am I doing wrong?
> 
> Cheers, Miro
> 

As pointed out in the other response, you'll need a type signature
somewhere to indicate what type your ?age? should be.

I'd like to point out something else however. You appear to be using
hard tabs (?\t? character): please don't do this! Haskell is
alignment-sensitive and you'll run into problems due to this unless you
*very* strictly only insert tabs. FTR GHC treats one \t as 8 spaces.
There are beginners on daily basis in #haskell wondering why their code
doesn't compile just to discover tabs once they share it through
lpaste.net or similar service.

[1] is the standard reference for style, especially see the ?Tabs? section.

So, please, set your editor to only insert spaces.

[1]: http://urchin.earth.li/~ian/style/haskell.html

-- 
Mateusz K.

From lists at webconect.ch  Sat Sep 27 15:57:08 2014
From: lists at webconect.ch (Elias Diem)
Date: Sat, 27 Sep 2014 17:57:08 +0200
Subject: [Haskell-beginners] HXT: encoding problem
Message-ID: <20140927155708.GA6547@webconect.local>

Hi guys

I have got the following haskell program:

------------------------------------------------------
import Text.XML.HXT.Core

main = do
    xml <- readFile "test_data-small.xml"
    let doc = readString config xml
    res <- runX . xshow $
        doc
        >>>
        getChildren >>> isElem >>> hasName "contacts"
        >>>
        deep isText
    mapM_ putStrLn res

config =
    [ withParseHTML                 no
    , withWarnings                  yes
    , withInputEncoding             utf8
    , withOutputEncoding            utf8
    , withValidate                  yes
    ]
------------------------------------------------------

The file 'test_data-small.xml' contains the following data:

------------------------------------------------------
<?xml version='1.0' encoding='UTF-8' ?>

<contacts>

<person>
    <name>
        <firstname>Max</firstname>
        <lastname>M?ller</lastname>
    </name>
</person>

</contacts>
------------------------------------------------------

Note the umlaut in the lastname!

If I run the program, I get the following error:

------------------------------------------------------
error: UTF-8 encoding error at input position 127: ValueOutOfBounds
------------------------------------------------------

Any help is appreciated. Thanks.

-- 
Greetings
Elias



From derek.mcloughlin at gmail.com  Sat Sep 27 16:13:56 2014
From: derek.mcloughlin at gmail.com (Derek McLoughlin)
Date: Sat, 27 Sep 2014 17:13:56 +0100
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <20140927155708.GA6547@webconect.local>
References: <20140927155708.GA6547@webconect.local>
Message-ID: <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>

I just ran this (OS/X + Platform 2014 + hxt 9.3.1.7) and it worked perfectly.

Are you sure that the XML file is actually saved with UTF-8 encoding?
Can you attach it?

On 27 September 2014 16:57, Elias Diem <lists at webconect.ch> wrote:
> Hi guys
>
> I have got the following haskell program:
>
> ------------------------------------------------------
> import Text.XML.HXT.Core
>
> main = do
>     xml <- readFile "test_data-small.xml"
>     let doc = readString config xml
>     res <- runX . xshow $
>         doc
>         >>>
>         getChildren >>> isElem >>> hasName "contacts"
>         >>>
>         deep isText
>     mapM_ putStrLn res
>
> config =
>     [ withParseHTML                 no
>     , withWarnings                  yes
>     , withInputEncoding             utf8
>     , withOutputEncoding            utf8
>     , withValidate                  yes
>     ]
> ------------------------------------------------------
>
> The file 'test_data-small.xml' contains the following data:
>
> ------------------------------------------------------
> <?xml version='1.0' encoding='UTF-8' ?>
>
> <contacts>
>
> <person>
>     <name>
>         <firstname>Max</firstname>
>         <lastname>M?ller</lastname>
>     </name>
> </person>
>
> </contacts>
> ------------------------------------------------------
>
> Note the umlaut in the lastname!
>
> If I run the program, I get the following error:
>
> ------------------------------------------------------
> error: UTF-8 encoding error at input position 127: ValueOutOfBounds
> ------------------------------------------------------
>
> Any help is appreciated. Thanks.
>
> --
> Greetings
> Elias
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners

From lists at webconect.ch  Sat Sep 27 17:21:34 2014
From: lists at webconect.ch (Elias Diem)
Date: Sat, 27 Sep 2014 19:21:34 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
Message-ID: <20140927172134.GA6848@webconect.local>

Hi Derek

On 2014-09-27,  Derek McLoughlin wrote:

> I just ran this (OS/X + Platform 2014 + hxt 9.3.1.7) and 
> it worked perfectly.

Good. Thanks.

> Are you sure that the XML file is actually saved with 
> UTF-8 encoding?

I *think* so. Vim tells me that it's UTF-8. I will double 
check.

> Can you attach it?

Here it is.

-- 
Greetings
Elias


-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_data-small.xml
Type: application/xml
Size: 181 bytes
Desc: not available
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140927/81b50920/attachment.xml>

From lists at webconect.ch  Sat Sep 27 17:41:21 2014
From: lists at webconect.ch (Elias Diem)
Date: Sat, 27 Sep 2014 19:41:21 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <20140927172134.GA6848@webconect.local>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927172134.GA6848@webconect.local>
Message-ID: <20140927174121.GA6950@webconect.local>

On 2014-09-27,  Elias Diem wrote:

> I *think* so. Vim tells me that it's UTF-8. I will double 
> check.

I just double checked. I'm 99% sure now that it is indeed 
UTF-8.

-- 
Greetings
Elias



From lists at webconect.ch  Sat Sep 27 17:53:12 2014
From: lists at webconect.ch (Elias Diem)
Date: Sat, 27 Sep 2014 19:53:12 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
Message-ID: <20140927175312.GA7102@webconect.local>

On 2014-09-27,  Derek McLoughlin wrote:

> I just ran this (OS/X + Platform 2014 + hxt 9.3.1.7) and 
> it worked perfectly.

My version of HXT is 9.2.2.

I run Debian GNU/Linux stable.

-- 
Greetings
Elias



From derek.mcloughlin at gmail.com  Sat Sep 27 21:12:37 2014
From: derek.mcloughlin at gmail.com (Derek McLoughlin)
Date: Sat, 27 Sep 2014 22:12:37 +0100
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <20140927175312.GA7102@webconect.local>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927175312.GA7102@webconect.local>
Message-ID: <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>

That file ran fine for me.

I also tested it on a Cloud9 installation with GHC 7.6.3 and HXT 9.3
and it ran fine.

Also Ubuntu 14.04, GHC 7.6.3 and HXT 9.3 worked fine.

What's your default locale in Debian?

On my Mac and test Ubuntu box, it's:
LANG="en_IE.UTF-8"
LC_COLLATE="en_IE.UTF-8"
LC_CTYPE="en_IE.UTF-8"
...
all values = "C.UTF-8"

On my Cloud9 instance:

LANG=C
LANGUAGE=
LC_CTYPE="C.UTF-8"
...
all values = "C.UTF-8"

On 27 September 2014 18:53, Elias Diem <lists at webconect.ch> wrote:
> On 2014-09-27,  Derek McLoughlin wrote:
>
>> I just ran this (OS/X + Platform 2014 + hxt 9.3.1.7) and
>> it worked perfectly.
>
> My version of HXT is 9.2.2.
>
> I run Debian GNU/Linux stable.
>
> --
> Greetings
> Elias
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners

From lists at webconect.ch  Sun Sep 28 07:57:28 2014
From: lists at webconect.ch (Elias Diem)
Date: Sun, 28 Sep 2014 09:57:28 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927175312.GA7102@webconect.local>
 <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>
Message-ID: <20140928075728.GA4505@webconect.local>

Hi Derek

Thanks for your help so far.

On 2014-09-27,  Derek McLoughlin wrote:

> That file ran fine for me.

Ok.

> I also tested it on a Cloud9 installation with GHC 7.6.3 and HXT 9.3
> and it ran fine.
> 
> Also Ubuntu 14.04, GHC 7.6.3 and HXT 9.3 worked fine.

I will test it later this day on another computer as well.

> What's your default locale in Debian?
> 
> On my Mac and test Ubuntu box, it's:
> LANG="en_IE.UTF-8"
> LC_COLLATE="en_IE.UTF-8"
> LC_CTYPE="en_IE.UTF-8"
> ...
> all values = "C.UTF-8"
> 
> On my Cloud9 instance:
> 
> LANG=C
> LANGUAGE=
> LC_CTYPE="C.UTF-8"
> ...
> all values = "C.UTF-8"

LANG=en_US.UTF-8
LANGUAGE=en_US:en
LC_COLLATE=
LC_CTYPE=

I haven't got any environment variables starting with LC 
defined.

-- 
Greetings
Elias



From lists at webconect.ch  Sun Sep 28 12:11:35 2014
From: lists at webconect.ch (Elias Diem)
Date: Sun, 28 Sep 2014 14:11:35 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <20140928075728.GA4505@webconect.local>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927175312.GA7102@webconect.local>
 <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>
 <20140928075728.GA4505@webconect.local>
Message-ID: <20140928121135.GA22868@webconect.local>

On 2014-09-28,  Elias Diem wrote:

> I will test it later this day on another computer as well.

I just tested it on another Linux box. And it works!! What 
could be the problem?

I noticed that on the other box I use HXT 9.3.1.1. Maybe 
that is solving the problem.

-- 
Greetings
Elias



From toad3k at gmail.com  Sun Sep 28 15:20:52 2014
From: toad3k at gmail.com (David McBride)
Date: Sun, 28 Sep 2014 11:20:52 -0400
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <20140928121135.GA22868@webconect.local>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927175312.GA7102@webconect.local>
 <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>
 <20140928075728.GA4505@webconect.local>
 <20140928121135.GA22868@webconect.local>
Message-ID: <CAN+Tr42Sc16smm_6ihu87nuY1jFuC0uPzfjjgvWZ0av2Rd5TDQ@mail.gmail.com>

I've had issues like this before where it had to do with the locale
settings on my machine at the time.  It is subtle and annoying but it will
cause various haskell functions that read, like hGetContents to flip out if
they see a character that is not readable by the locale you have set.  It
has to be something to do with that.

On Sun, Sep 28, 2014 at 8:11 AM, Elias Diem <lists at webconect.ch> wrote:

> On 2014-09-28,  Elias Diem wrote:
>
> > I will test it later this day on another computer as well.
>
> I just tested it on another Linux box. And it works!! What
> could be the problem?
>
> I noticed that on the other box I use HXT 9.3.1.1. Maybe
> that is solving the problem.
>
> --
> Greetings
> Elias
>
>
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140928/640d2ee6/attachment.html>

From chaddai.fouche at gmail.com  Mon Sep 29 04:46:08 2014
From: chaddai.fouche at gmail.com (=?UTF-8?B?Q2hhZGRhw68gRm91Y2jDqQ==?=)
Date: Mon, 29 Sep 2014 06:46:08 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <20140928121135.GA22868@webconect.local>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927175312.GA7102@webconect.local>
 <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>
 <20140928075728.GA4505@webconect.local>
 <20140928121135.GA22868@webconect.local>
Message-ID: <CANfjZRbyeOWpABmkpm8B7XYWUK=SjQEyQuY9TE4MpunGmVyF0Q@mail.gmail.com>

On Sun, Sep 28, 2014 at 2:11 PM, Elias Diem <lists at webconect.ch> wrote:

> On 2014-09-28,  Elias Diem wrote:
>
> > I will test it later this day on another computer as well.
>
> I just tested it on another Linux box. And it works!! What
> could be the problem?
>
>
readString is documented as not doing any decoding, so you're dependent on
your readFile doing it right for you, but that depends on your locale !
You could set your IO system input encoding yourself to avoid the problem
but it seems simpler to use "readDocument" provided by Hxt instead since
that'll read the file with your specified input encoding.

-- 
Jeda?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140929/95c6265a/attachment.html>

From dontdieych at gmail.com  Mon Sep 29 09:27:00 2014
From: dontdieych at gmail.com (Dontdie YCH)
Date: Mon, 29 Sep 2014 18:27:00 +0900
Subject: [Haskell-beginners] Is there a way to add package to .cabal when I
	import some module?
Message-ID: <CAPtM4gnFvsoYWFpJO7BBUNuj546Ucrijo1jDo8xgCQq-LdK7zg@mail.gmail.com>

Is there a way to add package to .cabal when I import new module?

From raymond at kestrel.nmt.edu  Mon Sep 29 15:52:43 2014
From: raymond at kestrel.nmt.edu (David Raymond)
Date: Mon, 29 Sep 2014 09:52:43 -0600
Subject: [Haskell-beginners] Unformatted binary IO
Message-ID: <21545.32843.148976.521710@gargle.gargle.HOWL>


I am trying to read and write to/from files or stdout/stdin that have
an ascii header followed by unformatted float32 binary data.  (The
files are created using a C program.)  The ascii header first needs to
be parsed to understand the structure of the floating point data.  I
have solved the parsing problem, but getting the unformatted binary
float data into a series of float (or double) immutable, unboxed
vectors has defeated me so far.  The binary package doesn't help as
far as I can see, as the binary format it reads and writes has some
control information at the beginning that doesn't exist in the format
I am reading.

This is easy in C, but seems to be hard in Haskell unless I am missing
something.

Any suggestions?

-- 
David J. Raymond
Prof. of Physics
New Mexico Tech
http://www.physics.nmt.edu/~raymond/index.html

From chaddai.fouche at gmail.com  Mon Sep 29 17:35:03 2014
From: chaddai.fouche at gmail.com (=?UTF-8?B?Q2hhZGRhw68gRm91Y2jDqQ==?=)
Date: Mon, 29 Sep 2014 19:35:03 +0200
Subject: [Haskell-beginners] Unformatted binary IO
In-Reply-To: <21545.32843.148976.521710@gargle.gargle.HOWL>
References: <21545.32843.148976.521710@gargle.gargle.HOWL>
Message-ID: <CANfjZRYyt4DL9xzvBW3s3JE_mveT2=EwSnfGVvv_4=Sogidu0Q@mail.gmail.com>

This is easy in Haskell too and binary provide everything you need. It's
not like you're _forced_ to use the Vector instance to construct a Vector
from your values, you can simply repeat (with replicateM) the get for the
Float (or Double) instance instead and build a Vector from the resulting
list (if you do it properly it will probably fuse into a tight loop anyway).

I don't understand your problem ?

On Mon, Sep 29, 2014 at 5:52 PM, David Raymond <raymond at kestrel.nmt.edu>
wrote:

>
> I am trying to read and write to/from files or stdout/stdin that have
> an ascii header followed by unformatted float32 binary data.  (The
> files are created using a C program.)  The ascii header first needs to
> be parsed to understand the structure of the floating point data.  I
> have solved the parsing problem, but getting the unformatted binary
> float data into a series of float (or double) immutable, unboxed
> vectors has defeated me so far.  The binary package doesn't help as
> far as I can see, as the binary format it reads and writes has some
> control information at the beginning that doesn't exist in the format
> I am reading.
>
> This is easy in C, but seems to be hard in Haskell unless I am missing
> something.
>
> Any suggestions?
>
> --
> David J. Raymond
> Prof. of Physics
> New Mexico Tech
> http://www.physics.nmt.edu/~raymond/index.html
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140929/b2b15fd1/attachment.html>

From byorgey at gmail.com  Mon Sep 29 23:43:00 2014
From: byorgey at gmail.com (Brent Yorgey)
Date: Mon, 29 Sep 2014 19:43:00 -0400
Subject: [Haskell-beginners] Unformatted binary IO
In-Reply-To: <21545.32843.148976.521710@gargle.gargle.HOWL>
References: <21545.32843.148976.521710@gargle.gargle.HOWL>
Message-ID: <CAH0njdtOz=oLvK8NXYwk0X2GDH68dk8C+sAdMcZqkocmrvr-eg@mail.gmail.com>

I am not sure I have quite understood your problem, but perhaps the
data-binary-ieee754 package may be of use?

-Brent


I am trying to read and write to/from files or stdout/stdin that have
an ascii header followed by unformatted float32 binary data.  (The
files are created using a C program.)  The ascii header first needs to
be parsed to understand the structure of the floating point data.  I
have solved the parsing problem, but getting the unformatted binary
float data into a series of float (or double) immutable, unboxed
vectors has defeated me so far.  The binary package doesn't help as
far as I can see, as the binary format it reads and writes has some
control information at the beginning that doesn't exist in the format
I am reading.

This is easy in C, but seems to be hard in Haskell unless I am missing
something.

Any suggestions?

--
David J. Raymond
Prof. of Physics
New Mexico Tech
http://www.physics.nmt.edu/~raymond/index.html
_______________________________________________
Beginners mailing list
Beginners at haskell.org
http://www.haskell.org/mailman/listinfo/beginners
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140929/f9eb54c2/attachment.html>

From lists at webconect.ch  Tue Sep 30 07:23:45 2014
From: lists at webconect.ch (Elias Diem)
Date: Tue, 30 Sep 2014 09:23:45 +0200
Subject: [Haskell-beginners] HXT: encoding problem
In-Reply-To: <CANfjZRbyeOWpABmkpm8B7XYWUK=SjQEyQuY9TE4MpunGmVyF0Q@mail.gmail.com>
References: <20140927155708.GA6547@webconect.local>
 <CAAw9fm=2o-DhjpkAH5oWkPutccf1Eu2Ww9=0NJvkrhRkAVp9QQ@mail.gmail.com>
 <20140927175312.GA7102@webconect.local>
 <CAAw9fmnQwMWqXxAxq49sQmGYjHLhioONrzG=LT6VX0ZpPXFmcA@mail.gmail.com>
 <20140928075728.GA4505@webconect.local>
 <20140928121135.GA22868@webconect.local>
 <CANfjZRbyeOWpABmkpm8B7XYWUK=SjQEyQuY9TE4MpunGmVyF0Q@mail.gmail.com>
Message-ID: <20140930072345.GA4565@webconect.local>

Hi Jeda?

On 2014-09-29,  Chadda? Fouch? wrote:

> readString is documented as not doing any decoding, so you're dependent on
> your readFile doing it right for you, but that depends on your locale !
> You could set your IO system input encoding yourself to avoid the problem
> but it seems simpler to use "readDocument" provided by Hxt instead since
> that'll read the file with your specified input encoding.

I use readDocument now as sugested and it works. Thanks for 
the explanation.

Thanks to the others too!

-- 
Greetings
Elias



From raymond at kestrel.nmt.edu  Tue Sep 30 10:56:41 2014
From: raymond at kestrel.nmt.edu (David Raymond)
Date: Tue, 30 Sep 2014 04:56:41 -0600
Subject: [Haskell-beginners] Unformatted binary IO
In-Reply-To: <CAH0njdtOz=oLvK8NXYwk0X2GDH68dk8C+sAdMcZqkocmrvr-eg@mail.gmail.com>
References: <21545.32843.148976.521710@gargle.gargle.HOWL>
 <CAH0njdtOz=oLvK8NXYwk0X2GDH68dk8C+sAdMcZqkocmrvr-eg@mail.gmail.com>
Message-ID: <21546.35945.707901.833368@gargle.gargle.HOWL>

Brent Yorgey writes:
 > I am not sure I have quite understood your problem, but perhaps the
 > data-binary-ieee754 package may be of use?
 > 
 > -Brent

This package looks interesting.  However, I am a true Haskell beginner
and I am having a hard time figuring out how the various Binary packages
are actually used to read and write files.  Any help or pointers to
tutorials would be greatly appreciated.

Dave

-- 
David J. Raymond
Prof. of Physics
New Mexico Tech
http://www.physics.nmt.edu/~raymond/index.html

From francesquini at gmail.com  Tue Sep 30 15:33:51 2014
From: francesquini at gmail.com (Emilio De Camargo Francesquini)
Date: Tue, 30 Sep 2014 12:33:51 -0300
Subject: [Haskell-beginners] Different behaviors between GHC 7.6.3 and 7.8.3
Message-ID: <CANN71N6u8+ptZiNRg-twx7z2n5WQyBe9r393sSS0rGy_ZfChtQ@mail.gmail.com>

Hello everyone,

I'm quite intrigued about these different behaviors between GHCs 7.6.3 and
7.8.3:


GHCi, version 7.6.3: http://www.haskell.org/ghc/  :? for help
Loading package ghc-prim ... linking ... done.
Loading package integer-gmp ... linking ... done.
Loading package base ... linking ... done.
Prelude> let x = 3 + 4
Prelude> :print x
x = (_t1::Integer)
Prelude> :force x
x = 7
Prelude> print _t1
7
Prelude>


GHCi, version 7.8.3: http://www.haskell.org/ghc/  :? for help
Loading package ghc-prim ... linking ... done.
Loading package integer-gmp ... linking ... done.
Loading package base ... linking ... done.
Prelude> let x = 3 + 4
Prelude> :print x
x = (_t1::Num a => a)
Prelude> :force x
x = _
Prelude> print _t1
ghc: panic! (the 'impossible' happened)
  (GHC version 7.8.3 for x86_64-unknown-linux):
    tcTyVarDetails a{tv atm} [tv]

Please report this as a GHC bug:  http://www.haskell.org/ghc/reportabug


Is it really a bug or is it something I really shouldn't be doing? I've
found this (fixed) bug (https://ghc.haskell.org/trac/ghc/ticket/8557), with
a similar error output, but it does not seem to be the same case.

Thanks

Emilio
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140930/dc8c4ba3/attachment.html>

From thomas at koch.ro  Tue Sep 30 20:15:57 2014
From: thomas at koch.ro (Thomas Koch)
Date: Tue, 30 Sep 2014 22:15:57 +0200
Subject: [Haskell-beginners] How to pass configuration to a function further
	down the call stack?
Message-ID: <201409302215.58309.thomas@koch.ro>

Hi,

I wrote my first Haskell 'project', a test suite for my personal SIEVE mail 
filter script using dovecots 'sieve-test' tool:

https://github.com/thkoch2001/sieve-test-hs

The file Network/Sieve/Test.hs might actually be interesting as a standalone 
library. The most important function is

assertMailActions :: Mail -> Actions -> IO ()

'Actions' is a type that describes the result of running a SIEVE filter over 
the passed mail, e.g. store in a certain folder, discard, reject, ...

Somewhere the external process 'sieve-test' is called. At this point I'd like 
to make certain things configurable:

- the path and name of the executable
- the parameters for the test-sieve executable
- the name of the sieve script
- the place where to store the temporary mail files
- whether or not to remove the temporary mail files

Does anybody has an idea where I could start to look for a solution?

Thank you,

Thomas Koch

From cma at bitemyapp.com  Tue Sep 30 21:23:58 2014
From: cma at bitemyapp.com (Christopher Allen)
Date: Tue, 30 Sep 2014 16:23:58 -0500
Subject: [Haskell-beginners] How to pass configuration to a function
 further down the call stack?
In-Reply-To: <201409302215.58309.thomas@koch.ro>
References: <201409302215.58309.thomas@koch.ro>
Message-ID: <CADnndOo4+Kkh=wHwG7ah_jScng7Zo8sHFFU5LgrxC3oZ-9EZXQ@mail.gmail.com>

First, define a datatype unifying all this config information.

Then there are two simple, common solutions.

If only one or two functions need to be passed the Config data, then just
pass it as an argument. If use of the config data is going to be prolific,
you might want to consider Reader.

http://stackoverflow.com/questions/14178889/reader-monad-purpose
http://book.realworldhaskell.org/read/programming-with-monads.html
https://hackage.haskell.org/package/transformers-0.4.1.0/docs/Control-Monad-Trans-Reader.html

You could engage in deeper magic than this, but it's not worth it for your
use-case.

Hope this helps.

-- Chris


On Tue, Sep 30, 2014 at 3:15 PM, Thomas Koch <thomas at koch.ro> wrote:

> Hi,
>
> I wrote my first Haskell 'project', a test suite for my personal SIEVE mail
> filter script using dovecots 'sieve-test' tool:
>
> https://github.com/thkoch2001/sieve-test-hs
>
> The file Network/Sieve/Test.hs might actually be interesting as a
> standalone
> library. The most important function is
>
> assertMailActions :: Mail -> Actions -> IO ()
>
> 'Actions' is a type that describes the result of running a SIEVE filter
> over
> the passed mail, e.g. store in a certain folder, discard, reject, ...
>
> Somewhere the external process 'sieve-test' is called. At this point I'd
> like
> to make certain things configurable:
>
> - the path and name of the executable
> - the parameters for the test-sieve executable
> - the name of the sieve script
> - the place where to store the temporary mail files
> - whether or not to remove the temporary mail files
>
> Does anybody has an idea where I could start to look for a solution?
>
> Thank you,
>
> Thomas Koch
> _______________________________________________
> Beginners mailing list
> Beginners at haskell.org
> http://www.haskell.org/mailman/listinfo/beginners
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.haskell.org/pipermail/beginners/attachments/20140930/cdc97e05/attachment.html>