- web-devel - Haskell.org

Parallel IO manager
by Kazu Yamamoto 13 Feb '13

13 Feb '13

Hello, As you may know, parallel IO manager has been merged to the master branch of GHC and the base library. If you compile your network code by this GHC with the "-threaded" option and you execute it with +RTS -Nx option where "x" is the number of cores, your network code can utilize the cores. No modification is necessary for your network code. No more prefork technique! The parallel IO manager is stable on Linux and FreeBSD. But it is a little bit unstable on Mac at this moment. If you are interested, please give a try. And let Andreas and me know if you have problems. --Kazu

2 1

WAI enhancement: request body size
by Michael Snoyman 12 Feb '13

12 Feb '13

I'd like to make an addition to the WAI protocol: a field in the Request data type to track the request body size. I would like this to be: requestBodyLength :: Maybe Word64 I think the choice of Word64 makes sense without further explanation. The perhaps surprising part is the use of `Maybe`. The reason for this is that in the case of a chunked request body, there's no way to know the size of the request body from the headers alone. So the logic for populating this field would go something like: - If content-length is supplied by the user, parse it and populate the field with Just content-length. - If the request is chunked, populate with Nothing. - Otherwise, there is a zero-sized request body, and populate with Just 0. One other idea would be to create a special datatype such as `data RequestBodyLength = ChunkedBody | KnownLength Word64`, but I don't believe it really adds much above the `Maybe Word64`. Any thoughts on this proposal? Note: This is related to Yesod issue: https://github.com/yesodweb/yesod/issues/468

4 4

ANNOUNCE: MFlow 0.2
by Alberto G. Corona 18 Jan '13

18 Jan '13

Hello. haskellers and, specially, web developpers. This is the second version of MFlow, a deep-first effort in the development of a platform for web applications at the higher level, by including as much haskell magic as possible. It is at the same time experimental and intended to be used in industry. I believe that haskell will not have its chance in the Web if it does not bring unique advantages and challenging paradigms beyond speed and type safety. Deep-first means that the effort is more in the addition higher level features rather than in being complete and bug free, with the conviction that feedback from real usage at the highest level is the best guide for development. Rather than to mimic other platforms in other languages, MFlow is as Haskell'ish as possible, and introduces new approaches like statefulness, event sourcing, back-execution, composable, active, self contained widgets and persistent STM. All of them collaborate to create a high level environment for web programming. This release adds bidings for WAI, blaze-html, stateful AJAX, active widgets, requirements, content management, multilanguage and URLs to pages inside stateful procedures. The package: [http://hackage.haskell.org/package/MFlow The entry in my blog, with the announcement and the philosophy behind http://haskell-web.blogspot.com.es/2013/01/announce-mflow-02.html The same, in other words: http://haskell-web.blogspot.com.es/2012/12/on-spirit-of-mflow-anatomy-of-wi… An online example, with some links that explain the new features. http://81.169.134.95:8081 Stateful, but virtually stateless, thanks to event sourcing http://haskell-web.blogspot.com/2013/01/stateful-but-stateless-at-last-than… Example of use of active widgets http://haskell-web.blogspot.com.es/2013/01/now-example-of-use-of-active-wid… Bug tracker https://github.com/agocorona/MFlow/issues Source repository git clone http://github.com/agocorona/MFlow Feedback is welcomed warmly Alberto.

3 3

Happstack events and timeouts
by Asafe Ribeiro 17 Jan '13

17 Jan '13

I've a timeline that is automatically updated when a user I follow post something new. I've already done the client-side code and some of the server-side as well. But I have a couple of issues that I cant figure out how to with happstack: How do I set a timeout for the request? I don't want to have more than one request for a same user. How can I stop all other request when I get a new one for a given user ?

2 2

ANN: Hakyll 4
by Jasper Van der Jeugt 16 Jan '13

16 Jan '13

After a few weeks of beta status [1], I've now released version 4.0 of the Hakyll static site generator library. I'm really glad with this release, as from what I've found and heard, it makes many things a lot easier. # Main changes - The important `Compiler` type has been changed from `Arrow` to `Monad`: this makes it much easier to write custom compilers, as most Haskellers are more familiar with monads. - The template stays superficially the same, but it has grown much more powerful and flexible underneath. - Early and fast access to metadata makes things such as tags and pagination much easier. - All items (images, css...) can now have metadata associated. Metadata can no longer be manipulated, and this immutability should reduce the number of encountered bugs. - A `check` command has been added. This allows you to check that all internal (or external) links are still alive. # Installation, migration In order to install Hakyll 4, grab it from Hackage: cabal update cabal install hakyll Here are some useful links: - Website: <http://jaspervdj.be/hakyll/> - Tutorial index: <http://jaspervdj.be/hakyll/tutorials.html> - Migration guide: <http://jaspervdj.be/hakyll/tutorials/hakyll-3-to-hakyll4-migration-guide.ht…> All feedback is welcome as always. [1]: https://groups.google.com/d/topic/hakyll/eRKWz8bVB0w/discussion Peace, Jasper

1 0

Esqueleto type trickery question
by Erik de Castro Lopo 24 Nov '12

24 Nov '12

Hi all, Below my .sig is a little Persist/Esqueleto program that works correctly. The getUserCount function does thr right thing and returns the number of rows in the User table. However, what I'd like is a generic function that returns the row count of any table, something vaguely like this: queryRowCount :: tableType -> SqlPersist IO Int64 queryRowCount tableName = do [Value x] <- select . from $ \(_ :: SqlExpr (Entity tableName)) -> return countRows return x Is there any way to do this? Its probably possible in Agda, but can it be made to work with GHC? Cheers, Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/ {-# LANGUAGE FlexibleContexts, GADTs, GeneralizedNewtypeDeriving, OverloadedStrings, QuasiQuotes, ScopedTypeVariables, TypeFamilies, TemplateHaskell #-} import Control.Monad (void) import Control.Monad.IO.Class (liftIO) import Database.Esqueleto import Database.Persist.Sqlite import Database.Persist.TH import GHC.Int (Int64) share [mkPersist sqlSettings, mkMigrate "migrateAll"] [persist| User firstName String lastName String emailAddress String UserName firstName lastName deriving Show Project name String description String deriving Show |] -- This works, but only for the User table. queryUserCount :: SqlPersist IO Int64 queryUserCount = do [Value x] <- select . from $ \(_ :: SqlExpr (Entity User)) -> return countRows return x main :: IO () main = withSqliteConn ":memory:" $ runSqlConn $ do void $ runMigrationSilent migrateAll void $ insert $ User "Fred" "Smith" "<fred.p.smith1972(a)gmail.com>" void $ insert $ User "Johne" "Jones" "<john.jones(a)gmail.com>" void $ insert $ User "Paul" "Peters" "<paul.p.peters(a)gmail.com>" queryUserCount >>= \c -> liftIO $ putStrLn $ "queryUserCount => " ++ show c

3 2

Esqueleto : SELECT COUNT(*) FROM table_name ;
by Erik de Castro Lopo 23 Nov '12

23 Nov '12

Hi all, Any idea how I would do the equivalent of: SELECT COUNT(*) FROM table_name ; with Esqueleto? I really just want the row count of the table. Erik -- ---------------------------------------------------------------------- Erik de Castro Lopo http://www.mega-nerd.com/

4 6

Updates to Julius interpolation
by Michael Snoyman 16 Nov '12

16 Nov '12

Hi all, Following some discussions on the issue both recently and many months ago, I've just released some updates to how Julius interpolation works. The basic idea is that we want to make interpolation of properly escaped JSON data the default, and the interpolation of unchecked Javascript code the exception to the rule. Concretely, take the following Julius template: let greeting = "Say \"Hello\"" :: String in [julius|alert("#{greeting}")|] In Julius 1.0, greeting is interpolated directly into your Javascript template, resulting in the output: alert("Say "Hello"") Importantly, the double quotes are *not* escaped. Instead, we would like `greeting` to be treated as JSON, the double quotes within the text to be escaped, and the surrounding double quotes to be automatically added. I've just released versions 1.0.2 and 1.1 of shakespeare-js (the package providing Julius), and it is now valid to say: let greeting = "Say \"Hello\"" :: String in [julius|alert(#{toJSON greeting})|] Resulting in: alert("Say \"Hello\"") What if you really want to spit out raw Javascript code? You can do so via the `rawJS` function: let somecode = "alert(\"Hello\")" :: String in [julius|#{rawJS somecode}|] I've set up this release system to try and maintain backwards compatibility, while simultaneously making it a compile-time error when you misuse the library. For the former, I've released shakespeare-js 1.0.2, which keeps the same interpolation semantics as 1.0.1, but adds in the rawJS function. You can safely add `rawJS` to virtually any existing #{} interpolation in Julius without any change in behavior. The difference with version 1.1 is that the old interpolations will no longer work. Under the surface, this is because I've removed the ToJavascript instances for String and Text. You can now only interpolate JSON values (by using toJSON) or raw Javascript (by using rawJS). Down the road, it will likely make sense to add in ToJavascript instances for common types which first convert to JSON. However, doing so in this first release would mean that old code would silently change its semantics. Hopefully you will now be able to upgrade to 1.0.2 without any breakage, and in the near future upgrade to 1.1 and get error messages where you need to insert rawJS calls. I am not yet updating the yesod-platform to use the new versions of these packages; I'd like to ask people to test things out and report any issues they run into so that we can make a properly stable yesod-platform release soon. Thanks, Michael

1 0

A string escape issue with Yesod Julius
by Dmitry Vyal 15 Nov '12

15 Nov '12

Greetings Recently I stumbled upon an unexpected escaping problem. I hoped Shakespeare templates would solve these forever, but looks like it's not the case. For example, look at this snippet: Prelude Text.Julius Data.Text.Lazy Data.Text> let t = "X\"Y\"Z"; j = [julius| str = "a#{t}bc"; |] in putStrLn $ Data.Text.Lazy.unpack $ renderJavascriptUrl (\u env -> Data.Text.pack "") j str = "aX"Y"Zbc"; Clearly we have have a problem here. Can something be done with it? Best wishes, Dmitry

3 3

Upcoming significant change to Cassius
by Michael Snoyman 31 Oct '12

31 Oct '12

Hi all, I've pushed some changes to Github regarding Cassius. Cassius is now simply a whitespace version of Lucius. In other words, when you run Cassius, braces are added in based on indentation level and the code is passed off to Lucius. Why the change? Twofold: 1. It only causes breakage for some incredibly strange corner cases (or so I believe). 2. It reduces the body of code we have to maintain. 3. And most importantly, it adds some really nice features to Cassius: nested blocks, variables, media queries, etc. I would appreciate if users of Cassius could take a little bit of time to pull the code from Github[1] and see how the changes affect them. I'm hoping to make the release next week. Michael [1] https://github.com/yesodweb/shakespeare

1 0