Complete.Org: Mailing Lists: Archives: freeciv-dev: October 2001:
[Freeciv-Dev] Re: Server Overhaul (was Re: Re: Split patch)
Home

[Freeciv-Dev] Re: Server Overhaul (was Re: Re: Split patch)

[Top] [All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index] [Thread Index]
To: rf13@xxxxxxxxxxxxxxxxxxxxxx
Cc: Arien Malec <arien_malec@xxxxxxxxx>, Justin Moore <justin@xxxxxxxxxxx>, Freeciv Developers <freeciv-dev@xxxxxxxxxxx>
Subject: [Freeciv-Dev] Re: Server Overhaul (was Re: Re: Split patch)
From: Daniel L Speyer <dspeyer@xxxxxxxxxxx>
Date: Mon, 22 Oct 2001 17:26:36 -0400 (EDT)

The syntax proposed here is extremely similar to what I implemented
(http://arch.freeciv.org/freeciv-dev-200110/msg00816.html).  I did support
use of '=' as whitespace (yes, "set=mincitydist 3\n" would be legitimate,
but so what?), though whitespace compression (i.e. use of " = " or even
"  ") is waiting for the next version (which will be along Real Soon
Now).  The only difference is I ommitted automatic name setting (also
intended for future addition) and I used braces in stead of
parenthesis.  Actually, it's quite possible to support both.

I guess I have a set of features to work on when I next have free
time.  Does my parser look generally acceptable (with the changes stated
here)?

I really don't know the server code, but it looks like setting up the
parser may be a large semi-independant task.  So, unless there's reason
otherwise, I'll keep working on that and hopefully the server can be made
ready to receive it simultaneously.

--Daniel Speyer
"May the /src be with you, always"


On Mon, 22 Oct 2001, Raimar Falke wrote:

> On Mon, Oct 22, 2001 at 12:44:05PM -0700, Arien Malec wrote:
> > Here's another BNF diagram for the proposed language.
> > 
> > TODO:
> > 
> > 1) Attribute value (int, string, int list, string list, directory, object
> > identifier, object identifier list, more?)
> > 2) help, show, control commands
> > 
> > Some notes:
> > 
> > 1) Objects have object idenfiers, distinct from name_orig
> > 2) All commands are newline terminated BUT a syntactically correct unclosed
> > paren will force the parser into multi-line mode until the paren is closed
> > (Compare bash, etc.)
> > 
> > Examples:
> > 
> 
> > set foo "bar"
> 
> Is 'set foo="bar"' also possible?
> 
> > with advance.railroad (
> >   req1 road_building,
> >   req2 steam_engine,
> >   name "Railroad")
> > # Note that paren must be on the same line as the start of the command
> 
> > # Note that road_building is an identifier, not a name
> 
> So we really need this? Shouldn't "steam_engine" be
> "advance.steam_engine"? What about lists like:
> 
> with advance.railroad (
>   req [road_building, steam_engine]
>   name "Railroad")
> ?
> 
> What is the purpose of the "with" command? Setting? If yes, can we use
> "=" for this?
> 
> > create unittype.armor
> > # Create a new unittype, with default values (i.e., warrior)
> > # Default name is identifier, with underscores replaced by space, and
> > # words with init caps
> 
> I can also think of objects which are "incomplete". So a "create
> unittype.armor" is incomplete until you set certain key properties (it
> may also possible that all properties are key properties).
> 
> > create unittype.god (
> >   attack = 100000,
> 
> >   defense 100000,
> 
> I would like to require "=" always.
> 
> >   vision_range = 1000,
> >   move_rate = 1000,
> >   hitpoints 100000,
> >   build_cost = 100000000,
> >   pop_cost = 10000)
> 
>       Raimar
> 
> -- 
>  email: rf13@xxxxxxxxxxxxxxxxx
>  "Python 2.0 beta 1 is now available [...]. There is a long list of new 
>   features since Python 1.6, released earlier today. We don't plan on 
>   any new releases in the next 24 hours."
>     -- Jeremy Hylton at Slashdot
> 
> 



[Prev in Thread] Current Thread [Next in Thread]