Home

Awesome

pywhitespace

by Phillip Bradbury, licensed under GPLv2

Whitespace mailing list

Posted 2004-08-03 12:27:27 +0100

I just saw your site and I think whitespace is quite clever ;)

Unfortunately I don't have enough room on my computer for GHC, so I ported the code to Python.

http://www.cs.newcastle.edu.au/~c3018900/pywhitespace.tar.bz2

this has everything from the 0.3 Haskell source, minus the haskell, plus some python code and two more examples (chr.ws and ord.ws, for ASCII->Char and vice versa)

the main.py has a #! line, so if you're in linux you can just go

chmod o+x main.py
./main.py program.ws

on Windows you have to go

C:\Python23\python main.py program.ws

or just

python main.py program.ws

if python is in your path.

I'm not sure if I implemented Ref correctly - the tutorial doesn't say whether it's 0-based or 1-based, and I don't know whether Haskell is, either. I coded it as if it was 1-based, ie Ref 1 is the same as Dup. Slide should be fine however.

It's got quite nice error reporting (for invalid commands such as [LF][LF][SP], and things like pop from an empty stack etc) and the code is quite simple. It also has the ability to (by uncommenting a couple of lines of code) to dump the code to screen readably (uncomment Input.py:57) or to run a "trace" through the code (uncomment second-last line of Program.py)

It's a little slow though - it takes about 20-25 seconds to run hanoi.ws with 10 discs on my old P150 (I don't have a faster computer to test it on, unfortunately...)

But anyways, it was fun to make!

Feel free to host this on your site (please credit me, Phillip Bradbury)

Comments are very welcome

=====

Phlip

Christmas comes but once per 31.6887nHz cycle

Whitespace mailing list, update

Posted 2007-12-10 10:07:39 +0000

Hey, if you're reading through the archive history, and updating the website, I should probably mention that the link to my python ws interpreter is broken... I've reuploaded it at http://www.mrphlip.com/pywhitespace.tar.bz2

It's changed a little since my post back in August '04... in particular, I have a non-pathetic computer now, so I could run the Haskell version and compare... turns out, I implemented Ref wrong (Ref 0 should behave the same as Dup, not Ref 1), and that's fixed in the above link.

Enjoy!

--

Phlip