1. Py 2.4 Update
I've posted the latest version of Py to my web page, at:
http://www.lanset.com/dcuny/py.htm
It's available in Euphoria source code, bound executable, or as a compiled C
executable. New features include:
- Namespace completely rewritten
- Command line has syntax coloring mini-editor
- Improved support for dictionaries (associated lists)
- Dot notation for dictionaries
- Catches uninitialized variables
- Lots of bugfixes
As usual, feedback is appreciated!
-- David Cuny
2. Re: Py 2.4 Update
On 25 Oct 2000, at 18:05, Cuny, David@DSS wrote:
> I've posted the latest version of Py to my web page, at:
>
> http://www.lanset.com/dcuny/py.htm
Since Py is written in Ox, how much is Ox changed to implement a feature in Py?
Kat
3. Re: Py 2.4 Update
Kat wrote:
> Since Py is written in Ox, how much is Ox changed
> to implement a feature in Py?
I've only done bugfixes to Ox, so it's remained generic. I'll try to post an
update in a couple of days to the Euphoria site (and my own), stripped of
the Py demo.
The main problem I've had with Ox has been handling '.', since I had
declared that it was a valid identifier character. This is was a shortcoming
of my grammar, not Ox itself.
The number parsing code has also been beefed up. Ox got confused when it saw
stuff like this:
[1..2]
and parsed it as:
'[' '1.' '.2' ']'
It also correctly parses numbers like '10e9'.
Ox turns out to be a lot slicker than I thought it would be, especially once
I coded it to detect grammar ambiguities. I never would have been able to
throw Py together so quickly without it.
-- David Cuny
4. Re: Py 2.4 Update
On Thu, 26 Oct 2000 10:59:54 -0700, Cuny, David at DSS <David.Cuny at
DSS.CA.GOV>
wrote:
>The number parsing code has also been beefed up. Ox got confused when it
saw
>stuff like this:
>
> [1..2]
>
>and parsed it as:
>
> '[' '1.' '.2' ']'
>
>It also correctly parses numbers like '10e9'.
David:
Have you ever looked at my mixedlib.e it has a function in
it that works exactly as the "C" strtok function.
It can get the next token 0r substring, in a string delimited
by any character from a second string.
char *strtok(char *string1, const char *string2)
If you look at any "C" reference book you will see how it is
used. Also you can look at my code because I used it to parse
some strings.
It is very dynamic you can pass it a string once and keep
getting tokens from it while changing delimiters between each
token. At any time you can change the string that you are parsing.
The first call to it you pass the string1 pointer ( string to parse )
and string2 the pointer to delimiter string to isolate the token.
If in the second call to it you set the string1 pointer to
NULL then it will return the next token in string1 using the delimiters
in string2 ( which can be the same or different ). When it has no
more tokens in string it returns null
Anytime you can you can pass it another string1 to parse.
This is "C" function is used by many parser.
Bernie
5. Re: Py 2.4 Update
Bernie wrote:
> Have you ever looked at my mixedlib.e it has a
> function in it that works exactly as the "C" strtok
> function.
No, prior to you mentioning, I hadn't been aware of it. I tend to get in the
bad habit of building everything from scratch, rather than relying on other
libraries.
You've certainly done a ton of work on the library - I'm very impressed! I
don't have any immediate use for it, but I'll certainly file this away for
future reference. It looks like it would be quite useful for implementing C
structures that Jiri & Co. have been asking for.
Thanks!
-- David Cuny
6. Re: Py 2.4 Update
On Thu, 26 Oct 2000 12:54:28 -0700, Cuny, David at DSS <David.Cuny at
DSS.CA.GOV>
wrote:
>future reference. It looks like it would be quite useful for implementing C
>structures that Jiri & Co. have been asking for.
David:
Structures are already implemented in it.
Bernie
