Compatibility

SharpMUSH does away with some of the more unique aspects of PennMUSH parser curiosities.

Unbalanced Parenthesis

In PennMUSH, the following is legal, while is not in SharpMUSH:

PennMUSH

think add(1,2
> 3

Instead, complete the parenthesis.

SharpMUSH

think add(1,2)
> 3

Forgiving Parenthesis

In PennMUSH, the following is legal, while is not in SharpMUSH:

PennMUSH

lsearch(all,eval,\[strmatch(##,%#)\])
> #1

Instead, escape all the contents. This is also needed for #lambda evaluation.

SharpMUSH

lsearch(all,eval,\[strmatch\(##\,%#\)\])
> #1

Full Recursion Parsing

In PennMUSH, TinyMUX, etc, you can do the following:

PennMUSH

&fn me=ucstr; think [v(fn)](foo)
> FOO

Instead, use callfn() to call the intended function.

SharpMUSH

&fn me=ucstr; think callfn(v(fn),foo)
> FOO

This is an awesome feature of MUSH-likes, but we explicitly do not support this, as it makes the parser non-deterministic, ambiguous and harder to maintain. Most existing Softcode does not rely on this behavior however.

Note

SharpMUSH introduces callfn() and @callcmd to replace this loss, which ensures most scenarios have a way to build the compatibility back in.

## Double-Evaluation

In PennMUSH, ## inside iter() performs double-evaluation, meaning the substitution result is evaluated again as MUSHcode:

PennMUSH

iter(\[if(true\,red\,white)\],##,)
> red

In SharpMUSH, ## is substituted as %iL (a literal reference), so the result is returned as-is without further evaluation:

SharpMUSH

iter(\[if(true\,red\,white)\],##,)
> [if(true,red,white)]

If double-evaluation is needed, use s() or objeval() to explicitly evaluate the result.