Using monadic validation with Digestive Functors and Snap - forms

I try quite a while now to wrap my head around how to use validation in a digestive functors form field, that requires access to another monad. To cut it short I have a digestive form like this
studentRegistrationForm :: Monad m => Form Text m StudentRegistrationData
studentRegistrationForm = StudentRegistrationData
<$> "school" .: choice schools Nothing
<*> "studentId" .: check studentIdErrMsg (not . T.null) (text Nothing)
<*> "firstName" .: check firstNameErrMsg (not . T.null) (text Nothing)
<*> "lastName" .: check lastNameErrMsg (not . T.null) (text Nothing)
<*> "email" .: check emailErrMsg (E.isValid . T.unpack) (text Nothing)
(studentId is basically the username)
and would like to use the function usernameExists of Snap.Snaplet.Auth to check if the entered username is unique.
Just for completeness, here is the corresponding data type:
data StudentRegistrationData = StudentRegistrationData
{ school :: School -- ^ school the student is enroled
, studentId :: Text -- ^ matriculation number of the student
, firstName :: Text -- ^ first name of the student
, lastName :: Text -- ^ last name of the student
, email :: Text -- ^ email for sending password
} deriving (Show)
I create my form in a handler like:
studentRegistrationHandler :: AppHandler ()
studentRegistrationHandler = do
(view, registrationData) <- runForm "form" SRF.studentRegistrationForm
maybe (showForm "registration" view) createUser registrationData
showForm :: String -> View Text -> AppHandler ()
showForm name view =
heistLocal (bindDigestiveSplices view) $ render template
where
template = BS.pack $ "student-" ++ name ++ "-form"
So the problem I have now is to understand how to access the state of the Auth snaplet inside the form. Is it passed already or do I have to passed it myself? Would the functions checkM respectively validateM in the Text.Digestive.Form help me there?
I have found several examples of how to use digestive functors and snap auth and session, like:
snap example
digestive functors tutorial
postgres example
Getting started with Snap-Auth
But none shows Snap.Snaplet.Auth and digestive functors working together directly, and I am still such a noob when it comes to monad transformers and lifting... maybe it is too easy for me to see. :(
I can upload a standalone example on github, which shows my problem if it helps to illustrate it. Any hints, pointers and suggestions are very welcome! :)
Hannes
add on: I created an example application demonstrating basic authentication functionality, you may have a look here: digestive-functors-snap-auth-example enjoy!

I haven't tried this out to see if everything type checks, but here's the general idea. You are correct that you want to use either checkM or validateM to do your monadic validation. The type signature for checkM is informative:
checkM :: Monad m => v -> (a -> m Bool) -> Form v m a -> Form v m a
This tells us that the validation function will need to have the type (a -> m Bool) and the m must be the same as the m in the form. This means that you need to change the type of your form to something like this:
studentRegistrationForm :: Form Text AppHandler StudentRegistrationData
Now let's write the validator. Since we plan on using the usernameExists function in our validator, we need to look at that type signature:
usernameExists :: Text -> Handler b (AuthManager b) Bool
This actually looks a lot like the (a -> m Bool) type signature that we need. In fact, it's an exact match because Handler b (AuthManager b) is a monad. But even though it matches the (a -> m Bool) pattern exactly doesn't mean we're done quite yet. When you run your form, you're in the AppHandler monad which is probably just a type alias for Handler App App where App is your application's top-level snaplet state type. So what we need to do is convert Handler b (AuthManager b) into Handler b b which will unify with Handler App App. The with function from the snaplet API is exactly what we need. This makes our validation function quite simple:
validUser :: Text -> Handler App App Bool
validUser = liftM not . with auth . usernameExists
With this, you can use checkM usernameErrMsg validUser just like you use check in the above code.

Related

Making custom Yesod Form: Could not deduce (Monad (FormInput m))

I am trying to make multi-file form input. I am using Handling a collection of data in a Yesod Form as a reference.
Here I am trying to make association list of field names to files.
multiFileInput :: Monad m => RenderMessage (HandlerSite m) FormMessage =>
[Text] -> FormInput m [(Text, FileInfo)]
multiFileInput = mapM $ secondM (ireq fileField) . (getFieldKey &&& id)
I get error:
Could not deduce (Monad (FormInput m))
arising from a use of ‘mapM’
But I don't know how to handle this. If I just add this as a constraint I have to propagade this constraint "(Monad (FormInput Handler))" up to a call site, where I don't know how to handle it. FormInput m is an instance of Monad, so I don't understand the issue.
fileInfos <- runInputPost $ multiKeyFileInput "files"
-> No instance for (Monad (FormInput Handler))
arising from a use of ‘multiKeyFileInput’
I will try to use runRequestBody instead, but it would be nice to understand the problem.
The FormInput data type was supposedly changed from Monad to Applicative, so you have to use traverse, which is an Applicative version of mapM.

How do you set the document title using Purescript?

After searching for some time I found in Pursuit the module DOM.HTML.History which has the data type DocumentTitle. This type could probably be used together with the function
replaceState ::
∀ e. Foreign -> DocumentTitle -> URL -> History -> Eff (history :: HISTORY | e) Unit
To change the document.title property of the page, however, I can't find examples showing how to call this function (e.g., where do I get the external Foreign data type?). Also, I'm not even sure if this function would do what I expect it to do...
In the unfortunate case that the Purescript team didn't include in their core API a way to change the document title, it's still possible to do so by making use of purescript's handy FFI mechanism.
Add these two files into your project:
Document.js
exports.setDocumentTitle =
function (title)
{
return function ()
{
window.document.title = title;
};
};
Document.purs
module Document
where
import Control.Monad.Eff (kind Effect, Eff)
import Data.Unit (Unit)
foreign import data DOCUMENT :: Effect
foreign import setDocumentTitle ::
∀ fx . String -> Eff (document :: DOCUMENT | fx) Unit
Now you can call setDocumentTitle as you would call Console's log function, except the effect would be DOCUMENT instead of CONSOLE, of course.
kazouas answer would look like this (in PS 0.12)
import Effect (Effect)
import Data.Unit (Unit)
foreign import setDocumentTitle :: String -> Effect Unit
Javascript remains the same.

How do you present any data type to the user with PureScript?

I want to make a very human-friendly development environment, and I'm considering using PureScript to provide the language part. I see that out of the box, Show doesn't work on records of things which are instances of Show:
log (show {a:5})
The 'Try PureScript!' (http://try.purescript.org/) compiler says:
No type class instance was found for
Prelude.Show { a :: Int
}
Is there a tool for generically printing any data structure, especially one containing records? Is there some type trickery that would support generically walking over the record to support my own class like present :: Present a => a -> Presentation? The problem is that I don't know what the types will be ahead of time. The user enters a record and I want to be able to present it. It seems that I'll have to patch the compiler to support this.
Records are disallowed in instance heads. For discussion and reasons, see this thread.They must be wrapped in data or newtype if we want to write instances for them.
However, there is a generics library and a deriving mechanism that lets us generate Show instances.
import Data.Generic
data Foo = Foo {a :: Int} | Bar {b :: String}
derive instance genericFoo :: Generic Foo
instance showFoo :: Show Foo where
show = gShow
Working with untyped data in PureScript is done using the purescript-foreign or the purescript-argonaut libraries. I'd suggest argonaut.
The representation of a record with unknown fields and unknown types for these fields would be: StrMap Json from the purescript-maps package. I'd suggest you take a look at the (not yet merged) documentation over here: https://github.com/hdgarrood/purescript-argonaut-core/blob/565c7e650c51c45570663cf1838ec9cfa307a9c7/README.md. I've also put together a little example, showing how to match on a heterogeneous array from JavaScript:
-- src/Main.purs
module Main where
import Prelude
import Control.Monad.Eff (Eff)
import Control.Monad.Eff.Console (CONSOLE, log)
import Data.Argonaut (foldJson, Json)
import Data.Foldable (traverse_)
newtype Presentation = Presentation String
unPresentation :: Presentation -> String
unPresentation (Presentation p) = p
instance showPresentation :: Show Presentation where
show = unPresentation
class Present a where
present :: a -> Presentation
instance presentInt :: Present Int where
present = Presentation <<< show
instance presentNumber :: Present Number where
present = Presentation <<< show
instance presentBoolean :: Present Boolean where
present = Presentation <<< show
instance presentString :: Present String where
present = Presentation
presentJson :: Json -> Presentation
presentJson =
foldJson
(const (Presentation "null"))
present
present
present
(const (Presentation "array"))
(const (Presentation "record"))
foreign import vals :: Array Json
main :: forall e. Eff ( console :: CONSOLE | e) Unit
main = traverse_ (log <<< show <<< presentJson) vals
And the corresponding js file:
// src/Main.js
// module Main
exports.vals = [1, 1.2, "hello", true, [1,2,3], {a: 3, b: "hi"}];
Running this program gives you:
> pulp run
* Building project in/home/creek/Documents/so-christopher-done
* Build successful.
1.0
1.2
hello
true
array
record
Yes, traceAny and related functions from purescript-debug. Here are a few examples: test/Main.purs#L22. I'd post the links to Pursuit, but it doesn't seem to have purescript-debug at the moment.

Passing records to ffi

When I pass a record to javascript, it works:
data Record = Record {
elem :: String
}
doSomethingForeign :: Record -> Fay ()
doSomethingForeign = ffi " callJsFun(%1) "
But when the function is not monomorphical, the record is not evaluated, one needs to do it manually:
class Passable a
instance Passable Record
instance Passable Text
doSomethingForeign' :: (Passable a) => a -> Fay ()
doSomethingForeign' = ffi " callJsFun(Fay$$_(%1)) "
This is the simple case, when the extra typing of Fay$$_ isn't that annoying, but if I pass more complex structures with type parameters to js, then adding just Fay$$_ won't solve it. I'd like to know the rule, when the evaluation to native js types is applied and where not.
The thunks will remain and type conversions won't happen if you have a type variable or Ptr X in the FFI, in contrast to a concrete type or Automatic a where the opposite applies.
I think what you want here is :: Passable a => Automatic a -> Fay () to force any thunks. It should be equivalent to separating this into two functions with a monomorphic argument. Using Automatic with a foreign type such as Text will only force the thunk and not do any type conversions.

State monad - adapt functions that only work with parts of the state?

I have a general state which is essentially a 3-tuple, and a number of functions which each concern themselves with parts of that state. I'm trying to work out a set of generic adapters for such functions, so that I can use them in a State monad pipeline.
This is possibly entirely wrongheaded; feel free to make that case.
I apologize in advance for mix of Java and pidgin Scala. I'm actually doing this in Java as a learning exercise, but nobody has time to read all that. I've elided a lot of uninteresting complexity for the sake of discussion; don't worry about the domain modeling.
The state in question is this:
ImportState(row:CsvRow, contact:Contact, result:ImportResult)
ImportResult is one of ADD, MERGE, or REJECT.
The functions I've defined are these:
def rowToContact: ImportRow => Contact
def findMergeCandidates: Contact => (Contact, List[Contact])
// merges, or declines to merge, setting the result
def merge: (Contact, List[Contact]) => (Contact, ImportResult)
def persist: Contact => ImportResult
def commitOrRollback: ImportState => ImportState
def notifyListener: ImportState => Nothing
The adapters I've defined so far are pretty simple, and deal with individual properties of ImportState:
def getRow: ImportState => ImportRow
def getContact: ImportState => Contact
def setRow(f: _ => ImportRow): ImportState => ImportState
def setContact(f: _ => Contact): ImportState => ImportState
def setResult(f: _ => ImportResult): ImportState => ImportState
The (broken) pipeline looks something like this (in Java):
State.<ImportState>init()
.map( setRow( constant(row) ) )
.map( setContact( getRow.andThen(rowToContact) ) )
.map( getContact.andThen(findMergeCandidates).andThen(merge) ) // this is where it falls apart
.map( setResult( getContact.andThen(persist) ) )
// ... lots of further processing of the persisted contact
.map(commitOrRollback)
.map(notifyListener);
The immediate problem is that merge returns a tuple (Contact, ImportResult), which I'd like to apply to two properties of the state (contact and result), while keeping the third property, row.
So far, I've come up with a couple of approaches to adaptation of merge that both suck:
Define some functions that pack and unpack tuples, and use them directly in the pipeline. This option is extremely noisy.
Define a one-off adapter for ImportState and merge. This option feels like giving up.
Is there a better way?
Your question is tagged Haskell - I'm hoping that means you can read Haskell, and not that someone saw 'monads' and added it. On that assumption, I'll be speaking Haskell in this answer, since it's the language I think in these days ;)
There's a useful concept called "functional lenses" with a couple Haskell library implementations. The core idea is that a "lens" is a pair of functions:
data Lens a b = Lens { extract :: (a -> b), update :: (a -> b -> a) }
This represents a functional way of getting and updating "parts" of a structure. With a type like this, you can write a function such as:
subState :: Lens a b -> State a t -> State b t
subState lens st = do
outer <- get
let (inner, result) = runState st (extract lens outer)
put (update lens outer inner)
return result
Translating that into Java sounds like an interesting (and possibly quite challenging) exercise!
Interesting I wrote this exact operation last night using fclabels:
withGame :: (r :-> r', s :-> s') -> GameMonad r' s' a -> GameMonad r s a
withGame (l1,l2) act = do
(r,s) <- (,) <$> askM l1 <*> getM l2
(a, s') <- liftIO $ runGame r s act
setM l2 s'
return a
GameMonad is a new type that is a monad transformer stack of state, reader, IO. I'm also using a bit of applicative functor style code don't let it put you off, it's pretty much the same as mokus.
Take a look at replacing the tuple approach with case classes. You get a lot for free in a structure that is virtually as easy to define, in particular a compiler generated copy method which allows you to create a copy of an instance, changing only the fields you want to change.