-- Hoogle documentation, generated by Haddock
-- See Hoogle, http://www.haskell.org/hoogle/


-- | Distributed processing of changing tasks
--   
--   A framework for distributing tasks running on HDFS data using Cloud
--   Haskell. The goal is speedup through distribution on clusters using
--   regular hardware. This framework provides different, simple
--   workarounds to transport new code to other cluster nodes. See project
--   home / README.md for more information.
@package task-distribution
@version 0.1.0.2

module Control.Distributed.Task.Distribution.LogConfiguration

-- | Sets up hslogger.
initLogging :: Priority -> Priority -> FilePath -> IO ()

-- | Sets up hslogger with a part logfile, part stdout configuration.
initDefaultLogging :: String -> IO ()

module Control.Distributed.Task.Types.HdfsConfigTypes
type HdfsConfig = (String, Int)
type HdfsPath = String
type HdfsLocation = (HdfsConfig, HdfsPath)

module Control.Distributed.Task.Types.TaskTypes
type TaskInput = [ByteString]
type TaskResult = [ByteString]
type Task = TaskInput -> TaskResult


-- | Contains all node communication (using Cloud Haskell). This includes
--   distribution logic.
module Control.Distributed.Task.Distribution.TaskDistribution

-- | Start a slave listening on given hostname, port.
startSlaveNode :: NodeConfig -> IO ()

-- | Run a calculation on all accessible slaves. This is a low-level
--   method, look at the RunComputaiton module for a nicer interface.
executeDistributed :: NodeConfig -> TaskDef -> [DataDef] -> ResultDef -> ([TaskResult] -> IO ()) -> IO ()

-- | List all accessible slaves.
showSlaveNodes :: NodeConfig -> IO ()

-- | List all accessible slaves that have at least a single block of the
--   specified path stored physically.
showSlaveNodesWithData :: NodeConfig -> String -> IO ()

-- | Convenience method to stop all accessible slaves remotely.
shutdownSlaveNodes :: NodeConfig -> IO ()


-- | Defines a higher level interface to running calculations. Resolves
--   HDFS input paths. |
module Control.Distributed.Task.Distribution.RunComputation

-- | The definition of a distributed calculation.
data MasterOptions
MasterOptions :: String -> Int -> TaskSpec -> DataSpec -> ResultSpec -> MasterOptions

-- | the master hostname
_host :: MasterOptions -> String

-- | the master port
_port :: MasterOptions -> Int

-- | the task logic
_taskSpec :: MasterOptions -> TaskSpec

-- | which data to process
_dataSpecs :: MasterOptions -> DataSpec

-- | how to process the result
_resultSpec :: MasterOptions -> ResultSpec

-- | Task logic definition, most modes expect task mode support, see
--   RemoteExecutionSupport.
data TaskSpec

-- | build the given string as module remotely (restrictions apply)
SourceCodeSpec :: String -> TaskSpec

-- | run this binary as task
FullBinaryDeployment :: TaskSpec

-- | serialize the given function in the context of the given program, run
--   both as task (restrictions apply)
SerializedThunk :: (TaskInput -> TaskResult) -> TaskSpec

-- | only transport some of the generated object code and relink remotely
--   (restrictions apply) - the function here is ignored, it only forces
--   the compilation of the contained module
ObjectCodeModuleDeployment :: (TaskInput -> TaskResult) -> TaskSpec

-- | definition of input data
data DataSpec

-- | simple test data, the path is configured, amount of files can be
--   limited
SimpleDataSpec :: Int -> DataSpec

-- | use given HDFS as starting directory, descend a number of directories
--   from there and take all files starting with the filter prefix (if any
--   given)
HdfsDataSpec :: HdfsPath -> Int -> (Maybe String) -> DataSpec

-- | what to do with the result
data ResultSpec

-- | process all results with the given method
CollectOnMaster :: ([TaskResult] -> IO ()) -> ResultSpec

-- | store the results in HDFS, in the given directory(1), with the given
--   suffix (2), based on the input path.
StoreInHdfs :: String -> String -> ResultSpec

-- | do nothing, for testing purposes only
Discard :: ResultSpec

-- | Run a computation.
runMaster :: MasterOptions -> IO ()


-- | Catches expected entry points for full binary deployment / thunk
--   serialization. These modes deploy the itself as a program and are
--   called remote with different arguments, which is handled here. |
module Control.Distributed.Task.TaskSpawning.RemoteExecutionSupport

-- | Combines all defined task mode hooks.
withRemoteExecutionSupport :: (TaskInput -> TaskResult) -> IO () -> IO ()

-- | Provides support for fullbinary task mode.
withFullBinaryRemoteExecutionSupport :: (TaskInput -> TaskResult) -> IO () -> IO ()

-- | Provides support for serialized thunk task mode.
withSerializedThunkRemoteExecutionSupport :: IO () -> IO ()