OneAndOneIs2

Fri, May 02, 2008

K&R's reverse Polish notation calculator

Section 4.3, pages 76-78

I realise that this is a long shot, but: Can anyone tell me why the buffer used by getch() and ungetch() is an array of 100 chars?

Because no matter how I follow the logic, I can't see any way that you'll ever have more than one char going into the buffer: getop() only puts a maximum of one char into the buffer at the end of its run, and the first thing it does when it's called is to take a char from the buffer if there is one.

Am I missing something, or is this array here for future-proof flexibility or something?

2 comments • Categories: Omni, Programming

Comments:

Comment from: herd [Visitor]

· http://erduman.de

I'd guess at historical reasons:
Back in 1969, when the keyboard controllers buffer was a whopping 8 bytes and dumb terminals were connected with a classy 300 baud, software buffering a whole line made a lot more sense than today...

Permalink 02/05/08 @ 21:50

Comment from: oneandoneis2 [Member] · http://geekblog.oneandoneis2.org/

Buffering a whole line would make sense in that situation, but makes none in this context: It *never* gets more than a single character in the buffer, and yet the buffer is a 100-character array.

I can only assume it's to make the code more portable/versatile for future use elsewhere..

Permalink 03/05/08 @ 21:07