This is an implementation of effectful, memory-constrained bytestrings (byte
streams) and functions for streaming bytestring manipulation, adequate for
The implementation follows the details of 'Data.ByteString.Lazy' and
'Data.ByteString.Lazy.Char8' in unrelenting detail, omitting only transparently
non-streaming operations like 'reverse'. It is just a question of replacing the
lazy bytestring type:
> data ByteString = Empty | Chunk Strict.ByteString ByteString
with the /minimal/ effectful variant:
> data ByteString m r = Empty r | Chunk Strict.ByteString (ByteString m r) | Go
(m (ByteString m r))
(Constructors are necessarily hidden in internal modules in both the 'Lazy' and
That's it. As a lazy bytestring is implemented internally by a sort of list of
strict bytestring chunks, a streaming bytestring is simply implemented as a
/producer/ or /generator/ of strict bytestring chunks. Most operations are
defined by simply adding a line to what we find in 'Data.ByteString.Lazy'.
Something like this alteration of type is of course obvious and mechanical,
once the idea of an effectful bytestring type is contemplated and lazy io is
rejected. Indeed it seems that this is the proper expression of what was
intended by lazy bytestrings to begin with. The documentation, after all, reads
- "A key feature of lazy ByteStrings is the means to manipulate large or
unbounded streams of data without requiring the entire sequence to be resident
in memory. To take advantage of this you have to write your functions in a lazy
streaming style, e.g. classic pipeline composition. The default I/O chunk size
is 32k, which should be good in most circumstances."
... which is very much the idea of this library: the default chunk size for
'hGetContents' and the like follows 'Data.ByteString.Lazy'; operations like
'lines' and 'append' and so on are tailored not to increase chunk size.
The present library is thus nothing but /lazy bytestring done right/.
The authors of 'Data.ByteString.Lazy' must have supposed that the directly
monadic formulation of such their type would necessarily make things slower.
This appears to be a prejudice. For example, passing a large file of short
lines through this benchmark transformation
> Lazy.unlines . map (bs -> "!" <> Lazy.drop 5 bs) . Lazy.lines >
Streaming.unlines . S.maps (bs -> chunk "!" >> Streaming.drop 5 bs) .
gives pleasing results like these
> $ time ./benchlines lazy >> /dev/null > real 0m2.097s > ... > $ time
./benchlines streaming >> /dev/null > real 0m1.930s
For a more sophisticated operation like
> Lazy.intercalate "!n" . Lazy.lines > Streaming.intercalate "!n" .
we get results like these:
> time ./benchlines lazy >> /dev/null > real 0m1.250s > ... > time ./benchlines
streaming >> /dev/null > real 0m1.531s
The pipes environment would express the latter as
> Pipes.intercalates (Pipes.yield "!n") . view Pipes.lines
meaning almost exactly what we mean above, but with results like this
> time ./benchlines pipes >> /dev/null > real 0m6.353s
The difference, however, is emphatically not intrinsic to pipes; it is just
that this library depends the 'streaming' library, which is used in place of
'free' to express the
"perfectly streaming"> splitting and iterated division or "chunking" of byte
These concepts belong to the ABCs of streaming; 'lines' is just a textbook
example, and it is of course handled correctly in 'Data.ByteString.Lazy'.
But the concepts are /catastrophically mishandled/ in /all/ streaming io
libraries other than pipes. Already the 'enumerator' and 'iteratee' libraries
were completely defeated by 'lines': see e.g. the 'enumerator' implementation
splitWhen and lines>. This will concatenate strict text forever, if that's what
is coming in. The rot spreads from there. It is just a fact that in all of the
general streaming io frameworks other than pipes,it becomes torture to express
elementary distinctions that are transparently and immediately contained in any
idea of streaming whatsoever.
Though, as was said above, we barely alter signatures in 'Data.ByteString.Lazy'
more than is required by the types, the point of view that emerges is very much
that of 'pipes-bytestring' and 'pipes-group'. In particular we have these
> Lazy.splitAt :: Int -> ByteString -> (ByteString, ByteString) >
Streaming.splitAt :: Int -> ByteString m r -> ByteString m (ByteString m r) >
Pipes.splitAt :: Int -> Producer ByteString m r -> Producer ByteString m
(Producer ByteString m r)
> Lazy.lines :: ByteString -> [ByteString] > Streaming.lines :: ByteString m r
-> Stream (ByteString m) m r > Pipes.lines :: Producer ByteString m r -> FreeT
(Producer ByteString m) m r
where the 'Stream' type expresses the sequencing of 'ByteString m _' layers
with the usual 'free monad' sequencing.
Interoperation with 'pipes-bytestring' uses this isomorphism:
> Streaming.ByteString.unfoldrChunks Pipes.next :: Monad m => Producer
ByteString m r -> ByteString m r > Pipes.unfoldr Streaming.ByteString.nextChunk
:: Monad m => ByteString m r -> Producer ByteString m r
Interoperation with 'io-streams' is thus:
> IOStreams.unfoldM Streaming.ByteString.unconsChunk :: ByteString IO () -> IO
(InputStream ByteString) > Streaming.ByteString.reread IOStreams.read ::
InputStream ByteString -> ByteString IO ()
and similarly for other rational streaming io libraries.
Problems and questions about the library can be put as issues on the github
page, or mailed to the <https://groups.google.com/forum/#!forum/haskell-pipes
A tutorial module is in the works;
<https://gist.github.com/michaelt/6c6843e6dd8030e95d58 here>, for the moment,
is a sequence of simplified implementations of familiar shell utilities.
The same programs are implemented at the end of the excellent
io-streams tutorial>. It is generally much simpler; in some case simpler than
what you would write with lazy bytestrings.
<https://gist.github.com/michaelt/2dcea1ba32562c091357 Here> is a simple GET
request that returns a byte stream. .