sforman
/
Thun
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Minor cleanup of the README file.

This commit is contained in:
Simon Forman 2018-04-22 09:16:48 -07:00
parent 669c86cbf1
commit 7c80a22851
2 changed files with 73 additions and 61 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

125
README
View File

@ -1,15 +1,16 @@
--------------------------------------------------

Thun
A dialect of Joy in Python.
v0.1.0
--------------------------------------------------
Copyright © 2014, 2015, 2017, 2018 Simon Forman
Copyright © 2014-2018 Simon Forman
This file is part of Thun
@ -29,7 +30,7 @@ Thun. If not see <http://www.gnu.org/licenses/>.
--------------------------------------------------
§ Introduction
§.1 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
@ -42,16 +43,16 @@ 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.
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.
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
@ -73,24 +74,27 @@ Joy and its deeper facets as well as how to program in it and several
interesting aspects. It's quite a treasure trove.
§ Installation
§.2 Installation
From PyPI in the usual way, e.g.:
pip install Thun
Or if you have downloaded the source, from the joypy directory:
Or if you have downloaded the source, from the top directory:
python ./setup.py install
Or you can run the module from the joypy directory (see below.)
To start a crude REPL:
Or you can run the package from the top directory. To start a crude REPL:
python -m joy
§ Basics of Joy
§.3 Documentation
The docs/ folder contains Jupyter notebooks, ... TODO
§.4 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
@ -98,27 +102,13 @@ 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.
A Joy expression is just a sequence (a.k.a. "list") of items. Sequences
intended as programs are called "quoted programs". 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.
§.4.1 Python Semantics
In general, where otherwise unspecified, the semantics of Thun are that
of the underlying Python. That means, for example, that integers are
@ -133,47 +123,65 @@ 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
§.4.2 Literals and Simple Functions
joy? 1 2 3
-> 3 2 1
. 1 2 3
1 . 2 3
1 2 . 3
1 2 3 .
joy? +
-> 5 1
1 2 3 <-top
joy? +
-> 6
joy? + +
1 2 3 . + +
1 5 . +
6 .
joy? 7
-> 7 6
6 <-top
joy? *
-> 42
joy? 7 *
6 . 7 *
6 7 . *
42 .
42 <-top
joy?
§ Simple Combinators
§.4.3 Combinators
The main loop is very simple as most of the action happens through what
are called "combinators": functions 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. They
work by changing the pending expression the interpreter is about to
execute. 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.
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
§.4.4 Definitions and More Elaborate Functions
§ Programming and Metaprogramming
§.4.5 Programming and Metaprogramming
§ Further Reading
§.4.6 Refactoring
--------------------------------------------------
This Implementation
§.5 This Implementation
Run with:
@ -203,10 +211,8 @@ joypy
`-- setup.py
--------------------------------------------------
References
§.6 References & Further Reading
Wikipedia entry for Joy:
@ -217,6 +223,9 @@ 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 -

7
joy/utils/brutal_hackery.py
View File

@ -56,8 +56,11 @@ function name! Hopefully they will discover this documentation.
def rename_code_object(new_name):
'''
If you want to wrap a function in another function and have the wrapped
function's name show up in the traceback, you must do this brutal
hackery to change the func.__code__.co_name attribute. See:
function's name show up in the traceback when an exception occurs in
the wrapper function, you must do this brutal hackery to change the
func.__code__.co_name attribute. Just functools.wraps() is not enough.
See:
https://stackoverflow.com/questions/29919804/function-decorated-using-functools-wraps-raises-typeerror-with-the-name-of-the-w