README.html

<title>Extensible Haskell front-end for the Programatica project -
        directory base</title>

<h1>Extensible Haskell front-end for the Programatica project</h1>

<h2>What is in this directory (base)</h2>

<p>
This directory contains the code that deals with the base language, i.e.,
more or less Haskell 98. It is structured to allow most of the code to
be reused without changes in extended versions of the language.

<!--
<h3>Test applications to run</h3>

Things are still rather transient, but here is what's available at the moment.

<dl>
<dt><code>tstModules</code> <var>options</var> <var>command</var> <var>modules</var>
<dd>Various tests of the front-end code, including production of cross
    reference info used by <code>hs2html</code>.
    Depending on the <var>command</var>, output
    appears on stdout or in files in the subdirectory <code>hi</code>.
<dt><code>tstTI</code> <var>options</var> <var>modules</var>
<dd>Tests the type checker. Pretty-prints a decorated syntax tree on stdout.
</dl>

More detailed descriptions are available on request.
<p>
There are also some additional, obsolete test programs.
-->

<h3>Obsolete subdirectories</h3>

<p>
<code>parse</code>, <code>SA</code>, <code>TC</code>

<h3>Active subdirectories</h3>

<dl>
<dt><a href="AST/"><code>AST</code></a>
<dd>nonrecursive data structures for the abstract syntax
    (reusable in extensions),

<dt><code>syntax</code>
<dd>knot-tying recursive definitions for the base language
    (not reusable in extensions),

<dt><a href="parse2/"><code>parse2</code></a>
<dd>Happy parser (not reusable), lexer, lexer generator.
    The lexer is described in
    <a href="http://www.cse.ogi.edu/~hallgren/Talks/LHiH/">A Lexer for
     Haskell in Haskell</a>.
<dt><a href="Modules/"><code>Modules</code></a>
<dd>implementation of Haskell's module system, as described in the
    paper <a href="/~diatchki/hsmod/"><cite>A Formal Specification for the
    Haskell&nbsp;98 Module System</cite></a>,

<dt><a href="defs/"><code>defs</code></a>
<dd>defined names, free names, mapping over names, scoping information,
    infix operator reassociation
    (both reusable parts and knot-tying parts for the base language),

<dt><a href="TI/"><code>TI</code></a>
<dd>type checker (both reusable parts and knot-tying parts for the base language),

<dt><code>pretty</code>
<dd>library of pretty printing combinators,

<dt><a href="lib/"><code>lib</code></a>
<dd>general library modules (does not depend on anything else here),

<dt><a href="transforms/"><code>transforms</code></a>
<dd>Source level transformations (e.g., pattern match simplification).
    Generation of code for derived instances.

<dt><code>tests</code>
<dd>test input,
<dt><a href="tests/HaskellLibraries"><code>tests/HaskellLibraries</code></a>
<dd>Haskell 98 Prelude and Standard Libraries, with some minor changes to
    make them go through our type checker. (This is now an important part
    of our tools and doesn't belong in the test subdirectory anymore!)

</dl>
<!--
<h3>A test run</h3>

When you have compiled <code>tstTI</code>, you can try
<pre>
  cd tests/HaskellLibraries
  ../../tstTI . +RTS -K5M
</pre>

This will type check the Prelude and the standard libraries, and output
annotated code on stdout. This takes about 40s on turing, 20s on kokotni.
<small>
(The flags <code>+RTS -K5M</code> prevents the pretty printer from running out
of stack space...)</small>
-->
<hr>
<address>Authors: various</address>
<address>Contact: <a href="http://www.cse.ogi.edu/~hallgren/">TH</address>