Parsing an Autopatch Response String?

String arguments are your friend indeed! String manipulation is probably one of my favorite things to do. (Smiling is also my favorite . . . ok - j/k)

Let's say the incoming string is CL2I3O5T, in theory (I just woke up because I couldn't sleep, so my brain is a bit fuzzy . . . and this completely untested) this could be accomplished in just a few commands.

I prefer to chip away at strings and avoid using LEFT, RIGHT and MID_STRINGS as they leave the original string intact, and it's just not as flexible since you're telling it to looking for only 'X' number of characters - I prefer that to instead look for things that are always present - in this case and 'I' and an 'O'. Let's pretend you have an 8x8 - your code would work fine. However, if you had an 18x18 - Autopatch does not return padded values (i.e. 01), so using LEFT/RIGHT/MID_STRING the way you have on an input value of 12 would return a '1' - definitely not what you're looking for.

There's no rule or anything that says you can't use string functions within string functions, and if fact I find it quite helpful. For instance if you want to remove just one word or group of words, you could do this:

// cXML = <somewhere><something>hello!</something><somewhere>
// cNODE = 'something'
cXML_TEMP = 
	REMOVE_STRING(cXML,"'</',cNODE,'>'",FIND_STRING(cXML,"'<',cNODE,'>'",1))
// cXML_TEMP = <something>hello!</something>

Removing portions of the string is helpful in getting where you need to go. Here's what I'd do with a returning string (which is "CL2I3O5T") from the Autopatch.

STRING:
{
	// FIND END OF ROUTE DATA
	IF (FIND_STRING(DATA.TEXT,'T',1)) // T = TAKE
	{
		STACK_VAR INTEGER nLEVEL;
		STACK_VAR INTEGER nINPUT;
		STACK_VAR INTEGER nOUTPUT;
		STACK_VAR CHAR    cBUFFER[12]
		
		// Remove string grabs data ENDING IN "T" starting at the position of "C"
		// I've never tried this before, and it's all theory, but I'd assume this would
		// ensure the first and full & proper string you are looking for in DATA.TEXT is grabbed
		cBUFFER = REMOVE_STRING(DATA.TEXT,'T',FIND_STRING(DATA.TEXT,'C',1));
		
		// cBUFFER = CL2I3O5T
		REMOVE_STRING(cBUFFER,'CL',1);
		
		// cBUFFER = 2I3O5T
		// Removed portion will be '2I', so ATOI will grab the 2
		nLEVEL = ATOI(REMOVE_STRING(cBUFFER,'I',1));
		
		// cBUFFER = 3O5T
		// Removed portion will be '3O', so ATOI will grab the 3
		nINPUT = ATOI(REMOVE_STRING(cBUFFER,'O',1));
		
		// cBUFFER = 5T
		// ATOI will grab the 5
		nOUTPUT= ATOI(cBUFFER);

		// Do whatever you want with your values now.
}		

Hope this helps! And if I'm totally wrong - it'd be nice if someone showed me where I am in error . . . okay - back to bed!

That function , SPLIT_STRING doesn't seem to be supported in the NetLinx.AXI. You'd have to copy the function out of the Tech Note and paste it into your program.

jjames wrote: »

How about using AMX's SPLIT_STRING function (technote 662)?

Just a warning, I found a bug in split_string that I found hard to believe wasn't tested for so I am not sure how heavily these functions were tested when they were conceived.
Paul

a_riot42 wrote: »

Just a warning, I found a bug in split_string that I found hard to believe wasn't tested for so I am not sure how heavily these functions were tested when they were conceived.
Paul

What bug is that?

I know that REPLACE_STRING (TN661) will spin you into an infinite loop if you try to replace one character with two or more of the same character. Oops.