printf() format question?

PCM programmer · Joined: 06 Sep 2003 Posts: 21708

It's been buggy for a while. I did a few experiments just now,
and it looks like the "width" value must be greater than the
"precision" value to avoid the bug.
i.e., The value on the left side of the decimal point must be greater
than the value on the right side.

Example:
The following code displays: Voltage=1.344 volts

Ttelmah · Guest

Though printf is buggy, what is being tried in the original post, is 'asking for trouble'. The number in front of the DP, is the 'total field width', while the number after the DP is the number of digits after the decimal point. Most compilers are smart enugh, that if you ask for a 'field width' of one digit, and three digits after the DP, they increase the first number for you, when decimals are being printed. What happens with CCS at the moment is that the arithmetic 'wraps', giving 256-4 (three decimals, plus the DP itself) digits in front of the DP...
So though it is a fault, it is using the field definitions in a 'abnormal' way, that causes it. So long as you ensure that the total field width exceeds the digits required after the DP (and include space for the DP itself), the problem does not appear.

Best Wishes

Guest · Posted: Wed Oct 12, 2005 1:57 pm

Ahhh yes. I referred to Kernigan and Ritchie and found the description. The CCS manual only has one line devoted to this statement and doesn't fully explain it. I guess the period doesn't count in the number of digits.

Its been about 15 years since my one and only C class.

Thanks Ttelmah.