<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html><head><title>Using the York Binary library</title></head>
<body bgcolor='#ffffff'>
<table><tr><td width=500>
<center><h1>Using the York Binary library in nhc13</h1></center>
<hr>
This document describes the York Binary library. (See also the
<a href="BinArray.html">BinArray</a> library for an example of the
use of Binary to build other abstractions.)
<hr>
<h3>The York Binary library</h3>
<pre>
module Binary where
data BinPtr a = ...
data BinLocation = Memory | File FilePath BinIOMode
data BinIOMode = RO | RW | WO
data BinHandle = ...
stdmem :: BinHandle
openBin :: BinLocation -> IO BinHandle
freezeBin :: BinHandle -> IO () -- changes BinIOMode to RO
closeBin :: BinHandle -> IO ()
copyBin :: BinHandle -> BinLocation -> IO BinHandle
isEOFBin :: BinHandle -> IO Bool
seekBin :: BinHandle -> BinPtr a -> IO ()
tellBin :: BinHandle -> IO (BinPtr a)
class Binary a where
put :: BinHandle -> a -> IO (BinPtr a)
get :: BinHandle -> IO a
putAt :: BinHandle -> BinPtr a -> a -> IO ()
getAt :: BinHandle -> BinPtr a -> IO a
getFAt :: BinHandle -> BinPtr a -> a
</pre>
<hr>
<h3>Programming model</h3>
<p>
Both in-heap data compression and binary I/O can be achieved using the
York <tt>Binary</tt> library. The basic model is rather like file
I/O: binary data resides in a separate space which is accessed only
through a <tt>BinHandle</tt> acting like a buffering file descriptor.
Each item of binary data lies at a particular position within the
space, the position being denoted by a <tt>BinPtr</tt>. Data can be
written and read sequentially just as with ordinary files. Also, like
ordinary files, we allow random-access reading and writing. However,
the particular beauty of this scheme is the ability to engage in
<em>pure, lazy,</em> random-access reading when a <tt>BinHandle</tt> is
in the appropriate <tt>RO</tt> (read-only) mode. (A <tt>BinHandle</tt>
which is already open for writing can be changed to <tt>RO</tt> mode
with the <tt>freezeBin</tt> call.)
<p>
<tt>BinHandle</tt>s do not just denote files - they can also refer to
areas of heap memory. One such area is available by default - called
<tt>stdmem</tt> - but new areas can be opened in just the
same way as files. They are opened in the default mode <tt>RW</tt>.
Binary heap areas grow automatically to fit the data placed in them, and,
like files, they are naturally garbage-collected when they are no longer
in use. (The <tt>closeBin</tt> operation is an explicit means to close
a file or discard some memory.)
<p>
The <tt>Binary</tt> class is derivable for any datatype defined in a
program except functions. (Please note however that cyclic or infinite
values will cause the compressing function to diverge.)
<p>
The class member functions come in two varieties, one for sequential
access, the other for random access. A <tt>BinHandle</tt> contains a
hidden state, including the <em>current position</em> in the file or
memory. Understanding the notion of the current position is important
for using the sequential operations correctly. <tt>put</tt> and
<tt>get</tt> always start reading or writing from the current position.
All operations <em>including the random-access ones</em>, when they
return, set the current position to the end of the value which has
just been read or written.
<table><tr>
<td colspan=2><tt>put :: BinHandle -> a -> IO (BinPtr a)</tt></td>
</tr><tr>
<td width=200 valign=top><tt>put bh x</tt></td>
<td valign=top>writes a binary representation of the ordinary value
<tt>x</tt> sequentially at the current position, returning a pointer
to the beginning of the value.
Where later sequential reading is sufficient, the return value of
<tt>put</tt> can be discarded. When random-access is required, the
return value of <tt>put</tt> can be used as the positional argument of
<tt>getAt</tt> and <tt>putAt</tt>.</td>
</tr><tr>
<td colspan=2><tt>get :: BinHandle -> IO a</tt></td>
</tr><tr>
<td width=200 valign=top><tt>get bh</tt></td>
<td valign=top>reads a binary representation sequentially from the current
position, returning the ordinary representation of the value.</td>
</tr><tr>
<td colspan=2><tt>putAt :: BinHandle -> a -> BinPtr a -> IO ()</tt></td>
</tr><tr>
<td width=200 valign=top><tt>putAt bh p x</tt></td>
<td valign=top>writes a binary representation of the ordinary value
<tt>x</tt> at the position <tt>p</tt>, returning nothing. The pointer
<tt>p</tt> might have been obtained as the result of an earlier
<tt>put</tt> operation, or it may been read from a binary stream via
a <tt>get</tt> operation, or indeed it may have been calculated.</td>
</tr><tr>
<td colspan=2><tt>getAt :: BinHandle -> BinPtr a -> IO a</tt></td>
</tr><tr>
<td width=200 valign=top><tt>getAt bh p</tt></td>
<td valign=top>reads a binary representation from the position <tt>p</tt>,
returning the ordinary representation of the value.</td>
</tr><tr>
<td colspan=2><tt>getFAt :: BinHandle -> BinPtr a -> a</tt></td>
</tr><tr>
<td width=200 valign=top><tt>getFAt bh p</tt></td>
<td valign=top>is a pure, lazy, version of the <tt>getAt</tt> method, which
can only be used on "frozen" <tt>BinHandle</tt>s.</td>
</tr></table>
<hr>
<h3>Transferring bits in bulk</h3>
<p>
The easiest way to transfer bits in bulk is with the <tt>copyBin</tt>
operation. It takes an active <tt>BinHandle</tt> and copies its entire
contents into the given <tt>BinLocation</tt>, returning a fresh
<tt>BinHandle</tt> denoting the copy.
<p>
A completely different method exists to transfer individual binary
values in bulk. By recording the size of each binary value explicitly
(which costs a certain amount of space), we can transfer just the
bits belonging to that value, rather than the entire <tt>BinHandle</tt>.
(This method predates the introduction of <tt>BinHandle</tt>s, so it
may turn out to be less space-efficient, less quick, and more complex than
the newer <tt>copyBin</tt> method. Tell us which method you prefer.)
<p>
There are "sized" variations of two of the binary operations:
<pre>
data SizedBin a = SB Size BinHandle (BinPtr a)
sizedPut :: Binary a => BinHandle -> a -> IO (SizedBin a)
sizedGetFAt :: Binary a => SizedBin a -> a
</pre>
The main purpose of the sized operations is to enable efficient bulk
transfer of binary data between two <tt>BinHandle</tt>s (typically
between memory and file, although it can equally well occur between
two files or two memory spaces). The sized operations do not implement
this bulk transfer themselves - however bulk transfer is efficient
only when the amount of data is known beforehand.
<p>
Hence there is a standard instance of class <tt>Binary</tt> for the type
<tt>SizedBin a</tt>: the operations <tt>put</tt> and <tt>get</tt>, when
used on sized binary values, effect the bulk transfer of bits from one
<tt>BinHandle</tt> to another. It should be noted that <tt>get</tt>ting
a sized binary value automatically allocates a fresh <tt>BinHandle</tt>
in memory, since there is no other way of specifying a destination for a
transfer in this direction. This fresh <tt>BinHandle</tt> (enclosed
anonymously within the <tt>SizedBin a</tt> value) is also
frozen into <tt>RO</tt> mode after the transfer, ready for a later
<tt>sizedGetFAt</tt> operation.
<hr>
<h3>Defining your own compression</h3>
<p>
If you want to play with defining your own instances of <tt>Binary</tt>, have
a look at some of the instances for standard types like Int and Lists
in <tt>src/prelude/Binary/Instances.hs</tt> to see how things work.
One need define only three of the member functions - the others are
defined in terms of them.
<p>
The lower-level tools used in defining instances are:
<pre>
class Binary a where
...
unsafeGetAt :: BinHandle -> BinPtr a -> (a,BinPtr b)
getBits :: BinHandle -> Int -> BinPtr a -> IO Int
putBits :: BinHandle -> Int -> Int -> IO (BinPtr a)
getAux :: BinHandle -> Int -> BinPtr a -> (Int,BinPtr a)
(<<) :: ((a->b),c) -> (c->(a,d)) -> (b,d)
</pre>
<hr>
<h3>Read and write modes</h3>
<p>
A file <tt>BinHandle</tt> can be opened in one of three modes: read-only
(<tt>RO</tt>), write-only (<tt>WO</tt>), or read-write (<tt>RW</tt>).
A memory <tt>BinHandle</tt> is always opened in <tt>RW</tt> mode, but
may be changed to <tt>RO</tt> mode by the <tt>freezeBin</tt> operation.
These modes differ from those of ordinary textual files:
<dl>
<dt>A binary operation never raises an I/O exception.</dt>
<dd>When in <tt>RO</tt> mode, the operations <tt>put</tt> and <tt>putAt</tt>
will not fail, but nor will they alter the file/memory.
<dd>When in <tt>WO</tt> mode, the operations <tt>get</tt> and <tt>getAt</tt>
will return odd values, not corresponding to the real file/memory.
<dd>Encountering EOF in <tt>RO</tt> mode does not raise an error -
reading beyond the end of the file/memory will simply return zeros.
However, the operation <tt>isEOFBin</tt> can be used to test the
condition.
<dd>The <tt>getFAt</tt> operation <em>will</em> give a runtime error if
the file/memory is not in <tt>RO</tt> mode, but since this error does
not arise within the I/O monad. it cannot be trapped.
<dt>In <tt>RW</tt> mode, interleaving read and write operations is safe.</dt>
<dd>The semantics of <tt>RW</tt> mode is that the file/memory is
<em>overwritten</em>. In other words, you can write just a single
bit in the middle of the file/memory if you want to - everything else
will stay the same. In particular, unlike <tt>WO</tt> mode, a file is
not truncated when you open it.
</dl>
<p>
An example program which uses binary I/O in <tt>RW</tt> mode is
<a href="../examples/ZooQuiz.hs">ZooQuiz.hs</a>.
<hr>
<p>
The latest updates to these pages are available on the WWW from
<a href="http://www.cs.york.ac.uk/fp/nhc13/">
<tt>http://www.cs.york.ac.uk/fp/nhc13/</tt></a>
<p>
1998.03.26<br>
<a href="http://www.cs.york.ac.uk/fp/">
York Functional Programming Group</a><br>
Malcolm Wallace <[email protected]>
</td></tr></table>
</body></html>
|
