You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

249 lines
10 KiB

{-# OPTIONS_HADDOCK hide #-}
14 years ago
{-# LANGUAGE NoMonomorphismRestriction, OverloadedStrings #-}
{-# LANGUAGE TupleSections #-}
14 years ago
module Network.Xmpp.Stream where
14 years ago
import Control.Applicative ((<$>), (<*>))
import qualified Control.Exception as Ex
import Control.Monad.Error
import Control.Monad.Reader
import Control.Monad.State.Strict
14 years ago
import qualified Data.ByteString as BS
import Data.Conduit
import qualified Data.Conduit.Internal as DCI
import Data.Conduit.List as CL
import Data.Maybe (fromJust, isJust, isNothing)
14 years ago
import Data.Text as Text
import Data.Void (Void)
import Data.XML.Pickle
import Data.XML.Types
14 years ago
import Network.Xmpp.Connection_
import Network.Xmpp.Pickle
import Network.Xmpp.Types
import Network.Xmpp.Marshal
14 years ago
import Text.Xml.Stream.Elements
import Text.XML.Stream.Parse as XP
-- import Text.XML.Stream.Elements
-- Unpickles and returns a stream element.
streamUnpickleElem :: PU [Node] a
-> Element
-> StreamSink a
streamUnpickleElem p x = do
case unpickleElem p x of
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
Left l -> throwError $ XmppOtherFailure -- TODO: Log: StreamXmlError (show l)
Right r -> return r
-- This is the conduit sink that handles the stream XML events. We extend it
-- with ErrorT capabilities.
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
type StreamSink a = ErrorT XmppFailure (Pipe Event Event Void () IO) a
-- Discards all events before the first EventBeginElement.
14 years ago
throwOutJunk :: Monad m => Sink Event m ()
14 years ago
throwOutJunk = do
next <- CL.peek
case next of
Nothing -> return () -- This will only happen if the stream is closed.
Just (EventBeginElement _ _) -> return ()
_ -> CL.drop 1 >> throwOutJunk
14 years ago
-- Returns an (empty) Element from a stream of XML events.
openElementFromEvents :: StreamSink Element
14 years ago
openElementFromEvents = do
lift throwOutJunk
hd <- lift CL.head
case hd of
Just (EventBeginElement name attrs) -> return $ Element name attrs []
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
_ -> throwError $ XmppOtherFailure
-- Sends the initial stream:stream element and pulls the server features. If the
-- server responds in a way that is invalid, an appropriate stream error will be
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
-- generated, the connection to the server will be closed, and a XmppFailure
-- will be produced.
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
startStream :: StateT Connection IO (Either XmppFailure ())
startStream = runErrorT $ do
state <- lift $ get
con <- liftIO $ mkConnection state
-- Set the `from' (which is also the expected to) attribute depending on the
-- state of the connection.
let expectedTo = case cState state of
ConnectionPlain -> if cJidWhenPlain state
then cJid state else Nothing
ConnectionSecured -> cJid state
case cHostName state of
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
Nothing -> throwError XmppOtherFailure -- TODO: When does this happen?
Just hostname -> lift $ do
pushXmlDecl
pushOpenElement $
14 years ago
pickleElem xpStream ( "1.0"
, expectedTo
14 years ago
, Just (Jid Nothing hostname Nothing)
, Nothing
, cPreferredLang state
14 years ago
)
response <- ErrorT $ runEventsSink $ runErrorT $ streamS expectedTo
case response of
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
Left e -> throwError e
-- Successful unpickling of stream element.
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
Right (Right (ver, from, to, id, lt, features))
| (unpack ver) /= "1.0" ->
closeStreamWithError con StreamUnsupportedVersion Nothing
| lt == Nothing ->
closeStreamWithError con StreamInvalidXml Nothing
-- If `from' is set, we verify that it's the correct one. TODO: Should we check against the realm instead?
| isJust from && (from /= Just (Jid Nothing (fromJust $ cHostName state) Nothing)) ->
closeStreamWithError con StreamInvalidFrom Nothing
| to /= expectedTo ->
closeStreamWithError con StreamUndefinedCondition (Just $ Element "invalid-to" [] []) -- TODO: Suitable?
| otherwise -> do
modify (\s -> s{ cFeatures = features
, cStreamLang = lt
, cStreamId = id
, cFrom = from
} )
return ()
-- Unpickling failed - we investigate the element.
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
Right (Left (Element name attrs children))
| (nameLocalName name /= "stream") ->
closeStreamWithError con StreamInvalidXml Nothing
| (nameNamespace name /= Just "http://etherx.jabber.org/streams") ->
closeStreamWithError con StreamInvalidNamespace Nothing
| (isJust $ namePrefix name) && (fromJust (namePrefix name) /= "stream") ->
closeStreamWithError con StreamBadNamespacePrefix Nothing
| otherwise -> ErrorT $ checkchildren con (flattenAttrs attrs)
where
-- closeStreamWithError :: MonadIO m => TMVar Connection -> StreamErrorCondition ->
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
-- Maybe Element -> ErrorT XmppFailure m ()
closeStreamWithError con sec el = do
liftIO $ do
withConnection (pushElement . pickleElem xpStreamError $
StreamErrorInfo sec Nothing el) con
closeStreams con
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
throwError XmppOtherFailure
checkchildren con children =
let to' = lookup "to" children
ver' = lookup "version" children
xl = lookup xmlLang children
in case () of () | Just (Nothing :: Maybe Jid) == (safeRead <$> to') ->
runErrorT $ closeStreamWithError con
StreamBadNamespacePrefix Nothing
| Nothing == ver' ->
runErrorT $ closeStreamWithError con
StreamUnsupportedVersion Nothing
| Just (Nothing :: Maybe LangTag) == (safeRead <$> xl) ->
runErrorT $ closeStreamWithError con
StreamInvalidXml Nothing
| otherwise ->
runErrorT $ closeStreamWithError con
StreamBadFormat Nothing
safeRead x = case reads $ Text.unpack x of
[] -> Nothing
[(y,_),_] -> Just y
flattenAttrs :: [(Name, [Content])] -> [(Name, Text.Text)]
flattenAttrs attrs = Prelude.map (\(name, content) ->
( name
, Text.concat $ Prelude.map uncontentify content)
)
attrs
where
uncontentify (ContentText t) = t
uncontentify _ = ""
14 years ago
-- Sets a new Event source using the raw source (of bytes)
-- and calls xmppStartStream.
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
restartStream :: StateT Connection IO (Either XmppFailure ())
restartStream = do
raw <- gets (cRecv . cHandle)
let newSource = DCI.ResumableSource (loopRead raw $= XP.parseBytes def)
(return ())
modify (\s -> s{cEventSource = newSource })
startStream
where
loopRead read = do
bs <- liftIO (read 4096)
if BS.null bs
then return ()
else yield bs >> loopRead read
14 years ago
-- Reads the (partial) stream:stream and the server features from the stream.
-- Returns the (unvalidated) stream attributes, the unparsed element, or
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
-- throwError throws a `XmppOtherFailure' (if something other than an element
-- was encountered at first, or if something other than stream features was
-- encountered second).
-- TODO: from.
streamS :: Maybe Jid -> StreamSink (Either Element ( Text
, Maybe Jid
, Maybe Jid
, Maybe Text
, Maybe LangTag
, ServerFeatures ))
streamS expectedTo = do
header <- xmppStreamHeader
case header of
Right (version, from, to, id, langTag) -> do
features <- xmppStreamFeatures
return $ Right (version, from, to, id, langTag, features)
Left el -> return $ Left el
where
xmppStreamHeader :: StreamSink (Either Element (Text, Maybe Jid, Maybe Jid, Maybe Text.Text, Maybe LangTag))
xmppStreamHeader = do
lift throwOutJunk
-- Get the stream:stream element (or whatever it is) from the server,
-- and validate what we get.
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
el <- openElementFromEvents -- May throw `XmppOtherFailure' if an
-- element is not received
14 years ago
case unpickleElem xpStream el of
Left _ -> return $ Left el
Right r -> return $ Right r
xmppStreamFeatures :: StreamSink ServerFeatures
xmppStreamFeatures = do
e <- lift $ elements =$ CL.head
case e of
Tweak failure approach I'm assuming and defining the following: 1. XMPP failures (which can occur at the TCP, TLS, and XML/XMPP layers (as a stream error or forbidden input)) are fatal; they will distrupt the XMPP session. 2. All fatal failures should be thrown (or similar) by `session', or any other function that might produce them. 3. Authentication failures that are not "XMPP failures" are not fatal. They do not necessarily terminate the stream. For example, the developer should be able to make another authentication attempt. The `Session' object returned by `session' might be useful even if the authentication fails. 4. We can (and should) use one single data type for fatal failures. (Previously, both StreamFailure and TlsFailure was used.) 5. We can catch and rethrow/wrap IO exceptions in the context of the Pontarius XMPP error system that we decide to use, making the error system more intuitive, Haskell-like, and more straight-forward to implement. Calling `error' may only be done in the case of a program error (a bug). 6. A logging system will remove the need for many of the error types. Only exceptions that seem likely to affect the flow of client applications should be defined. 7. The authentication functions are prone to fatal XMPP failures in addition to non-fatal authentication conditions. (Previously, `AuthStreamFailure' was used to wrap these errors.) I'm hereby suggesting (and implementing) the following: `StreamFailure' and `TlsFailure' should be joined into `XmppFailure'. `pullStanza' and the other Connection functions used to throw `IOException', `StreamFailure' and `TlsFailure' exceptions. With this patch, they have been converted to `StateT Connection IO (Either XmppFailure a)' computations. They also catch (some) IOException errors and wrap them in the new `XmppIOException' constructor. `newSession' is now `IO (Either XmppFailure Session)' as well (being capable of throwing IO exceptions). Whether or not to continue to a) wrap `XmppFailure' failures in an `AuthStreamFailure' equivalent, or, b) treat the authentication functions just like the other functions that may result in failure (Either XmppFailure a), depends on how Network.Xmpp.Connection.auth will be used. Since the latter will make `auth' more consistent, as well as remove the need for a wrapped (and special-case) "AuthFailure" type, I have decided to give the "b" approach a try. (The drawback being, of course, that authentication errors can not be accessed through the use of ErrorT. Whether or not this might be a problem, I don't really know at this point.) As the SASL code (and SaslM) depended on `AuthStreamFailure', it remains for internal use, at least for the time-being. `session' is now an ErrorT computation as well. Some functions have been updated as hacks, but this will be changed if we decide to move forward with this approach.
13 years ago
Nothing -> throwError XmppOtherFailure
Just r -> streamUnpickleElem xpStreamFeatures r
14 years ago
xpStream :: PU [Node] (Text, Maybe Jid, Maybe Jid, Maybe Text, Maybe LangTag)
xpStream = xpElemAttrs
(Name "stream" (Just "http://etherx.jabber.org/streams") (Just "stream"))
(xp5Tuple
14 years ago
(xpAttr "version" xpId)
(xpAttrImplied "from" xpPrim)
(xpAttrImplied "to" xpPrim)
(xpAttrImplied "id" xpId)
14 years ago
xpLangTag
)
14 years ago
-- Pickler/Unpickler for the stream features - TLS, SASL, and the rest.
xpStreamFeatures :: PU [Node] ServerFeatures
xpStreamFeatures = xpWrap
(\(tls, sasl, rest) -> SF tls (mbl sasl) rest)
(\(SF tls sasl rest) -> (tls, lmb sasl, rest))
(xpElemNodes
(Name
"features"
(Just "http://etherx.jabber.org/streams")
(Just "stream")
)
(xpTriple
(xpOption pickleTlsFeature)
(xpOption pickleSaslFeature)
(xpAll xpElemVerbatim)
)
)
where
pickleTlsFeature :: PU [Node] Bool
pickleTlsFeature = xpElemNodes "{urn:ietf:params:xml:ns:xmpp-tls}starttls"
(xpElemExists "required")
pickleSaslFeature :: PU [Node] [Text]
pickleSaslFeature = xpElemNodes
"{urn:ietf:params:xml:ns:xmpp-sasl}mechanisms"
(xpAll $ xpElemNodes
"{urn:ietf:params:xml:ns:xmpp-sasl}mechanism" (xpContent xpId))