# One source, lots of Brainfuck [duplicate]

Write a Brainfuck interpreter that can interpret multiple Brainfuck programs embedded within the same source. The syntax is the same except that the instructions must be prefixxed with the number/ID of the targetted program. Example:

1st program

+++[->+<]


2nd program

+>++>+++.


Mashup

1+1+1+2+2>1[1-2+2+2>1>1+2+1<2+1]2+2.


What is required:

• That it is theoretically able to interpret an unlimited number programs embedded within the same source.
• That the instruction order does not matter.
• That the code does not use features of a language that would natively solve this problem or natively interpret Brainfuck. Pretty much, don't cheat. I'll leave this to your discretion.

What is not:

• Unlimited Brainfuck storage. Only a minimum of 1000 Brainfuck cells per is imposed.
• It is not required that the input and output specify which bf program is doing the io. It'd be a cool feature tho, so if you do implement this you get -15 bytes.

You can do postfix if you want, just specify in that case. This is code-golf, so the smallest source in bytes (after the the -15 byte bonus) wins.

• Could you be more specific as to how input/output is handled? – Shelvacu Aug 20 '14 at 1:12
• shelvacu: It's the brainfuck io that I was referring to. – user2958652 Aug 20 '14 at 2:14
• How will answers be scored? code-golf, popularity-context, or code-challenge? Based on the mention of byte-count bonuses, I'm guessing code-golf, but please tag it as such. – Bob Aug 20 '14 at 2:33
• Are the ID's only numeric? – Οurous Aug 20 '14 at 6:10
• I thought it was distinct enough, but fair enough. – user2958652 Aug 20 '14 at 12:01