Safe Haskell	None
Language	Haskell98

Data.Repa.Scalar.Int

Contents

Reading from strings
Loading from buffers
Showing to strings
- Unpadded
- Padded
Storing to buffers
- Unpadded
- Padded

Description

Loading and storing integers directly from/to memory buffers.

The implementation uses native Haskell unboxed primitives. There there should not be any significant performance penalty relative to the standard implementations in other languages (like C).

Synopsis

readInt :: String -> Maybe Int
readIntFromByteString :: ByteString -> Maybe Int
loadInt :: Ptr Word8 -> Int -> b -> (Int -> Int -> b) -> b
loadInt# :: Addr# -> Int# -> (#Int#, Int#, Int##)
loadIntWith# :: Int# -> (Int# -> Int#) -> (#Int#, Int#, Int##)
showInt :: Int -> String
showIntToByteString :: Int -> ByteString
showIntPad :: Int -> Int -> String
showIntPadToByteString :: Int -> Int -> ByteString
storeInt :: Ptr Word8 -> Int -> IO Int
storeInt# :: Addr# -> Int# -> IO Int
storeIntWith# :: (Int# -> IO buf) -> (buf -> Int# -> Int# -> IO ()) -> Int# -> (buf -> Int# -> IO b) -> IO b
storeIntPad :: Ptr Word8 -> Int -> Int -> IO Int
storeIntPad# :: Addr# -> Int# -> Int# -> IO Int
storeIntPadWith# :: (Int# -> IO buf) -> (buf -> Int# -> Int# -> IO ()) -> Int# -> Int# -> (buf -> Int# -> IO b) -> IO b

Reading from strings

readInt :: String -> Maybe Int Source #

Read an Int from a String, or Nothing if this isn't one.

readIntFromByteString :: ByteString -> Maybe Int Source #

Read an Int from a ByteString, or Nothing if this isn't one.

Loading from buffers

loadInt Source #

Arguments

:: Ptr Word8	Buffer holding digits.
-> Int	Length of buffer.
-> b	On convert failure, return this value.
-> (Int -> Int -> b)	On convert success, given int read and number of chars.
-> b

Load an ASCII Int from a foreign buffer, returning the value and number of characters read.

This function is set to INLINE. It unboxes the pointer and integer then calls loadInt#, which is NOINLINE.

loadInt# Source #

Arguments

:: Addr#	Address of buffer holding digits
-> Int#	Length of buffer.
-> (#Int#, Int#, Int##)	Convert success?, value, length read.

Load an ASCII Int from a buffer, producing an unboxed tuple describing the result.

This function is set to NOINLINE, so it can be safely called from multiple places in the program.

loadIntWith# Source #

Arguments

:: Int#	Length of input buffer.
-> (Int# -> Int#)	Function to get a byte from the source.
-> (#Int#, Int#, Int##)	Convert success?, value, length read

Primitive Int loading function.

This function is set to INLINE, so you will get a new copy of it in the compiled program each time it is invoked. Consider providing an appropriate wrapper for your use case.

Showing to strings

Unpadded

showInt :: Int -> String Source #

Show an Int, allocating a new String.

showIntToByteString :: Int -> ByteString Source #

Show an Int, allocating a new ByteString.

Padded

showIntPad :: Int -> Int -> String Source #

Show an Int, allocating a new String.

showIntPadToByteString :: Int -> Int -> ByteString Source #

Show an Int, allocating a new ByteString.

Storing to buffers

Unpadded

storeInt Source #

Arguments

:: Ptr Word8	Pointer to output buffer.
-> Int	Int to store.
-> IO Int	Number of bytes written.

Store an ASCII Int into a buffer, producing the number of bytes written.

This functon is set to INLINE. It unboxes the pointer and integer then calls storeInt# which is NOINLINE.

storeInt# Source #

Arguments

:: Addr#	Address of output buffer.
-> Int#	Int to store.
-> IO Int	Number of bytes written.

Store an ASCII Int into a buffer, producing the number of bytes written.

This function is set to NOINLINE, so it can be safely called from multiple places in the program.

storeIntWith# Source #

Arguments

:: (Int# -> IO buf)	Function to allocate a new output buffer, given the length in bytes.
-> (buf -> Int# -> Int# -> IO ())	Function to write a byte to the buffer, given the index and byte value.
-> Int#	Int to store.
-> (buf -> Int# -> IO b)	Continuation for buffer and bytes written.
-> IO b

Primitive Int storing function.

This function is set to INLINE, so you will get a new copy of it in the compiled program each time it is invoked. Consider providing an appropriate wrapper for your use case.

Padded

storeIntPad Source #

Arguments

:: Ptr Word8	Pointer to output buffer.
-> Int	Int to store.
-> Int	Minimum number of digits.
-> IO Int	Number of bytes written.

Store an ASCII Int into a buffer, producing the number of bytes written.

storeIntPad# Source #

Arguments

:: Addr#	Address of output buffer.
-> Int#	Int to store.
-> Int#	Minimum number of digits.
-> IO Int	Number of bytes written.

Store an ASCII Int into a buffer, producing the number of bytes written.

This function is set to NOINLINE, so it can be safely called from multiple places in the program.

storeIntPadWith# Source #

Arguments

:: (Int# -> IO buf)	Function to allocate a new output buffer, given the length in bytes.
-> (buf -> Int# -> Int# -> IO ())	Function to write a byte to the buffer, given the index and byte value.
-> Int#	Int to store.
-> Int#	Pad out result to achieve at this many digits.
-> (buf -> Int# -> IO b)	Continuation for buffer and bytes written.
-> IO b

Like storeIntWith#, but add leading zeros to the front of the integer to pad it out to at least the given number of digits.