Re: Java Circle

From: Daniel A. Koepke (
Date: 04/09/99

On Fri, 9 Apr 1999, George Greer wrote:

> I suppose instead of counting spaces, you'd just check for > or < the
> current number of spaces?

That's what the test lexical analyzer is currently doing.  For the other
method, where the spacing must be consistent throughout a particular file,
a table of levels would need to be setup and we'd store the indentation
required for each level there.  But I don't see too much advantage with
doing that.

> Line continuations work as usual?

Line contituations should be placed within the ownership of the statement,

  long_function_call(blah, blah, ...
    blah, blah, blah, ...
    blah, blah, blah, ...

In addition, one can use a semi-colon to place multiple statements on the
same line, where the semi-colon works as a, "inheret the pervious
statements level," marker.

  int i;
  if (i == 10)
    i++; log(i + " is the value of 'i' now."); i--
  else // works much the same as...
    log(i + " is the value of 'i' now.")

Anyone else have any input on the matter?  I'm leaning towards indentation
indicating block level.  And if no-one cares, that's what I'll do.

BTW, this scripting language is low-level.  It is intended to write
commands, skills, special procedures, etc. in.  While it still does
support the writing of triggers, it isn't intended as a direct replacement
of DG Scripts.  The functionality will, of course, allow it to do so, but
I'm trying to keep the code clean enough that you can have DG Scripts
around, as well.  With a bit of work, I should have a generic script
engine that doesn't care about the language of the script it's calling, so
you can write stuff in perl or DG scripts or whatever.

Continuing the thread in a different direction, anyone have any thoughts
on the merit of byte compiling the language as opposed to interpreting it?


     | Ensure that you have read the CircleMUD Mailing List FAQ:  |
     |  |

This archive was generated by hypermail 2b30 : 12/15/00 PST