Re: [Haskell-cafe] Haskell, Queries and Monad Comprehension

Hello Torsten, well thank you for taking the time and answering that. It seems that the one thing I am good for, as far as contributions to this list go, is coaxing answers out of our Functional Pros. :) That is good news then, I was getting frustrated reading fantastic papers which eventually were nothing more than teasers. So, will Ferry then be usable from within Haskell? Best regards Günther Am 25.05.10 21:28, schrieb Torsten Grust:
Günther,
we are currently underway building the second version (as in: done right this time) of Ferry, a query compiler that
(1) accepts queries over ordered, nested lists, (2) compiles these queries into an intermediate algebraic form, then (3) emits (small bundles of) SQL queries that evaluate the input query over your off-the-shelf RDBMS.
We've used Ferry's first version to build new LINQ to SQL providers for Microsoft's LINQ as well as a new code generator for Phil Wadler's Links.
The Ferry compiler itself is built in Haskell. Surf to www.ferry-lang.org for more information, screencasts, papers, talks, and contact us for more details. We will be happy to share Ferry's Haskell code once in digestable shape (soon).
Cheers, --Torsten
On May 24, 2010, at 03:20 , Günther Schmidt wrote:
Hi all,
is there anybody currently using Haskell to construct or implement a query language?
I have read a couple of papers on "Monad Comprehension Calculus" and similar but none using Haskell nor any other existing programming language to build an actual implementation.
Most papers give some sort of "Pseudo code", but I couldn't find any meat.
Günther
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe

Hi Günther, On May 25, 2010, at 21:37 , Günther Schmidt wrote:
Hello Torsten,
well thank you for taking the time and answering that. It seems that the one thing I am good for, as far as contributions to this list go, is coaxing answers out of our Functional Pros. :)
That is good news then, I was getting frustrated reading fantastic papers which eventually were nothing more than teasers.
So, will Ferry then be usable from within Haskell?
Ferry is much more about the compilation of queries (over list-based data models) than the invention of some new query syntax. We thus are mostly after embeddings of queries into existing host languages. Such embeddings exist for -- C# (Ferry-based LINQ to SQL) -- Links (new Ferry-based SQL code generator) -- Ruby (in the works, provides a much richer and arguably more seamless embedding of queries over Ruby arrays as well as relational tables than does ARel or ActiveRevord's find_by* methods). There's some interesting connection between the compilation techniques employed by Ferry and Data Parallel Haskell. I talked to Simon Peyton Jones and he suggested to attempt a Haskell embedding. So, Haskell: not yet. But conceivable. Cheers, --Torsten
Am 25.05.10 21:28, schrieb Torsten Grust:
Günther,
we are currently underway building the second version (as in: done right this time) of Ferry, a query compiler that
(1) accepts queries over ordered, nested lists, (2) compiles these queries into an intermediate algebraic form, then (3) emits (small bundles of) SQL queries that evaluate the input query over your off-the-shelf RDBMS.
We've used Ferry's first version to build new LINQ to SQL providers for Microsoft's LINQ as well as a new code generator for Phil Wadler's Links.
The Ferry compiler itself is built in Haskell. Surf to www.ferry-lang.org for more information, screencasts, papers, talks, and contact us for more details. We will be happy to share Ferry's Haskell code once in digestable shape (soon).
Cheers, --Torsten
On May 24, 2010, at 03:20 , Günther Schmidt wrote:
Hi all,
is there anybody currently using Haskell to construct or implement a query language?
I have read a couple of papers on "Monad Comprehension Calculus" and similar but none using Haskell nor any other existing programming language to build an actual implementation.
Most papers give some sort of "Pseudo code", but I couldn't find any meat.
Günther
_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
-- | Prof. Dr. Torsten Grust torsten.grust@uni-tuebingen.de | | www-db.informatik.uni-tuebingen.de | | Database Systems - Universität Tübingen (Germany) |
participants (2)
-
Günther Schmidt
-
Torsten Grust