@FORMAT (commas in string) inside @IF?

[vefatica]

I spent a couple hours figuring out a bug in a BTM and it boils down to this.

This works as expected.

This works differently. The comma between xxx and yyy is being seen as belonging to @IF but the ,yyy still contributes to @FORMAT's length.

Escaping that comma isn't perfect. It fixes the problem with @IF but the escape character itself contributes to @FORMAT's length even though it's not printed.

I figure the escape character shouldn't contribute to @FORMAT's length (and that that might be hard or impossible).

 

[rconn]

Escape processing is obviously done after variable substitution (it couldn't work the other way around).

 

[vefatica]

The comma between xxx and yyy is being seen as belonging to @IF.

What about that first observation. Doesn't the parser know that the comma between xxx and yyy is inside the %FORMAT? I thought (naively as usual) that when parsing nested variable functions the parser would know which ] ends which function.

 

[rconn]

WAD - TCC is doing exactly what you told it to.

The %@FORMAT[12,xxx,yyy] returns " xxx,yyy".

So %@IF is being passed the argument: "1==1, xxx,yyy,".

It's unclear to me what you are actually trying to do, or why you would think that @IF would know what a nested function was doing.