229 lines
7.3 KiB
Plaintext
229 lines
7.3 KiB
Plaintext
--------------------------------------------------
|
|
|
|
|
|
Thun
|
|
|
|
A dialect of Joy in Python.
|
|
|
|
|
|
--------------------------------------------------
|
|
|
|
|
|
Copyright © 2014, 2015, 2017, 2018 Simon Forman
|
|
|
|
This file is part of Thun
|
|
|
|
Thun is free software: you can redistribute it and/or modify it under the
|
|
terms of the GNU General Public License as published by the Free Software
|
|
Foundation, either version 3 of the License, or (at your option) any later
|
|
version.
|
|
|
|
Thun is distributed in the hope that it will be useful, but WITHOUT ANY
|
|
WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A
|
|
PARTICULAR PURPOSE. See the GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License along with
|
|
Thun. If not see <http://www.gnu.org/licenses/>.
|
|
|
|
|
|
--------------------------------------------------
|
|
|
|
|
|
§ Introduction
|
|
|
|
Joy is a programming language created by Manfred von Thun that is easy to
|
|
use and understand and has many other nice properties. This Python
|
|
package implements an interpreter for a dialect of Joy that attempts to
|
|
stay very close to the spirit of Joy but does not precisely match the
|
|
behaviour of the original version(s) written in C. The main difference
|
|
between Thun and the originals, other than being written in Python, is
|
|
that it works by the "Continuation-Passing Style".
|
|
|
|
As I study Joy I find that it is very aptly named. It is clear, concise,
|
|
and ameniable to advanced techniques for constructing bug-free software.
|
|
|
|
Developed by Manfred von Thun, don't know much about him, not much on
|
|
the web about Joy and von Thun (Von Thun?) See references below.
|
|
|
|
Because it has desirable properties (concise, highly factored) the
|
|
programming process changes, the ratio of designing to writing code
|
|
shifts in favor of design. The documentation becomes extensive while
|
|
the code shrinks to stable bodies of small well-factored incantations
|
|
that are highly expressive, much like mathematical papers consist of
|
|
large bodies of exposition interlaced with mathematical formula that
|
|
concisely and precisely express the meaning of the text.
|
|
|
|
The time and attention of the programmer shifts from thinking about the
|
|
language to thinking in the language, and the development process feels
|
|
more like deriving mathematical truths than like writing ad-hoc
|
|
solutions.
|
|
|
|
I hope that this package is useful in the sense that it provides an
|
|
additional joy interpreter (the binary in the archive from La Trobe seems
|
|
to run just fine on my modern Linux machine!) But I also hope that you
|
|
can read and understand the Python code and play with the implementation
|
|
itself.
|
|
|
|
The best source (no pun intended) for learning about Joy is the
|
|
information made available at the website of La Trobe University (see the
|
|
references section below for the URL) which contains source code for the
|
|
original C interpreter, Joy language source code for various functions,
|
|
and a great deal of fascinating material mostly written by Von Thun on
|
|
Joy and its deeper facets as well as how to program in it and several
|
|
interesting aspects. It's quite a treasure trove.
|
|
|
|
|
|
§ Installation
|
|
|
|
From PyPI in the usual way, e.g.:
|
|
|
|
pip install Thun
|
|
|
|
Or if you have downloaded the source, from the joypy directory:
|
|
|
|
python ./setup.py install
|
|
|
|
Or you can run the module from the joypy directory (see below.)
|
|
|
|
To start a crude REPL:
|
|
|
|
python -m joy
|
|
|
|
|
|
§ Basics of Joy
|
|
|
|
Joy is stack-based. There is a main stack that holds data items:
|
|
integers, floats, strings, functions, and sequences or quotes which hold
|
|
data items themselves.
|
|
|
|
23 1.8 'a string' "another" dup [21 18 /] [1 [2 [3]]]
|
|
|
|
A Joy expression is just a sequence of items, also called lists.
|
|
Sequences intended as programs are called "quoted programs". The
|
|
evaluation proceeds by iterating through the terms in the expression,
|
|
putting all literals onto the main stack and executing functions as they
|
|
are encountered. Functions receive the current stack and return the next
|
|
stack.
|
|
|
|
The main loop is very simple as most of the action happens through what
|
|
are called "combinators", which accept quoted programs on the stack and
|
|
run them in various ways. These combinators factor specific patterns
|
|
that provide the effect of control-flow in other languages (such as ifte
|
|
which is like if..then..else..) Combinators receive the current
|
|
expession in addition to the stack and return the next expression.
|
|
In Joy control-flow is done by combinators that expect quoted programs
|
|
on the stack and execute them in various ways. In Thun they work by
|
|
changing the pending expression that the interpreter is about to execute.
|
|
In concrete terms, the combinators could work by making recursive calls
|
|
to the interpreter and all intermediate state would be held in the call
|
|
stack of the implementation language, in this joy implementation they
|
|
work instead by changing the pending expression and intermediate state
|
|
is put there.
|
|
|
|
In general, where otherwise unspecified, the semantics of Thun are that
|
|
of the underlying Python. That means, for example, that integers are
|
|
unbounded (whatever your machine can handle), strings cannot be added to
|
|
integers but can be multiplied, Boolean True and False are effectively
|
|
identical to ints 1 and 0, empty sequences are considered False, etc.
|
|
|
|
Nothing is done about Python exceptions currently, although it would be
|
|
possible to capture the stack and expression just before the exception
|
|
and build a robust and flexible error handler. Because they are both
|
|
just datastructures, you could immediately retry them under a debugger,
|
|
or edit either or both of the stack and expression. All state is in one
|
|
or the other.
|
|
|
|
§ Literals and Simple Functions
|
|
|
|
joy? 1 2 3
|
|
-> 3 2 1
|
|
|
|
joy? +
|
|
-> 5 1
|
|
|
|
joy? +
|
|
-> 6
|
|
|
|
joy? 7
|
|
-> 7 6
|
|
|
|
joy? *
|
|
-> 42
|
|
|
|
joy?
|
|
|
|
|
|
§ Simple Combinators
|
|
|
|
joy? 23 [0 >] [dup --] while
|
|
|
|
-> 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
|
|
|
|
|
|
TODO:
|
|
|
|
§ Definitions and More Elaborate Functions
|
|
Refactoring
|
|
|
|
§ Programming and Metaprogramming
|
|
|
|
§ Further Reading
|
|
|
|
|
|
--------------------------------------------------
|
|
|
|
|
|
This Implementation
|
|
|
|
Run with:
|
|
|
|
python -m joy
|
|
|
|
joypy
|
|
|-- COPYING - license
|
|
|-- README - this file
|
|
|
|
|
|-- archive - info on Joy
|
|
| |-- Joy-Programming.zip
|
|
| `-- README
|
|
|
|
|
|-- docs - Various Examples and Demos
|
|
| |-- * - Jupyter Notebooks on Thun and supporting modules
|
|
| `-- README - Table of Contents
|
|
|
|
|
|-- joy
|
|
| |-- joy.py - main loop, REPL
|
|
| |-- library.py - Functions, Combinators, Definitions
|
|
| |-- parser.py - convert text to Joy datastructures
|
|
| |
|
|
| `-- utils
|
|
| |-- pretty_print.py - convert Joy datastructures to text
|
|
| `-- stack.py - work with stacks
|
|
|
|
|
`-- setup.py
|
|
|
|
|
|
--------------------------------------------------
|
|
|
|
|
|
References
|
|
|
|
|
|
Wikipedia entry for Joy:
|
|
https://en.wikipedia.org/wiki/Joy_%28programming_language%29
|
|
|
|
|
|
Homepage at La Trobe University:
|
|
http://www.latrobe.edu.au/humanities/research/research-projects/past-projects/joy-programming-language
|
|
|
|
|
|
Misc...
|
|
|
|
Stack based - literals (as functions) - functions - combinators -
|
|
Refactoring and making new definitions - traces and comparing
|
|
performance - metaprogramming as programming, even the lowly integer
|
|
range function can be expressed in two phases: building a specialized
|
|
program and then executing it with a combinator - ?Partial evaluation?
|
|
- ?memoized dynamic dependency graphs? - algebra
|
|
|