Adding 0 at starting of int value in c - c

Question

I am in WinCE7 and to get the current time, I am using GetLocalTime(&systemTime);. This function gives the value of current time. Now if the milliseconds is 81, it displays it as 81 due to which the error occurs when I subtract two time values. For ex: time1 : 12:34:13:851 & time2: 12:34:14:81. Now I need to subtract seconds and milliseconds. So using sprintf, I am extracting seconds and milliseconds and putting them in time1 & time2 :
sprintf(time1,"%d.%d",systemTime.wSeconds,systemTime.wMilliseconds)
sprintf(time2,"%d.%d",systemTime.wSeconds,systemTime.wMilliseconds)
I am converting time1 & time2 into float using atof.Now time1 is 13.851 and time2 is 14.81. The milliseconds of time2 is actually 081 but it displays 81 so while subtracting it consider it as 810 which gives wrong values.
time2--> 14.810 14.081
time1--> 13.851 13.851
-------- ---------
result 0.959(wrong) 0.23(correct)
So to remove this error I thought of counting the digits of milliseconds and if it is 2 then add 0 at starting. So I did:
double digits = (floor (log10 (abs (milliseconds))) + 1); //calculate digits
if(digits == 2) //if milliseconds contains 2 digits, we need to add 0 at starting
{
sprintf(newMS,"0%d",milliseconds); //adding 0 to milliseconds
finalMilliseconds = atoi(newMS); //newMS is in char so converting it into integer and storing the value in finalMilliseconds
}
The problem occurs here. Lets say milliseconds = 18, so newMS = 018 but finalMilliseconds is again 18.
Please suggest any other way of conversion or any other way of adding 0 at starting

Answer 1

According to the documentation of SYSTEMTIME from MSDN:
It is not recommended that you add and subtract values from the
SYSTEMTIME structure to obtain relative times. Instead, you should
Convert the SYSTEMTIME structure to a FILETIME structure.
Copy the resulting FILETIME structure to a ULARGE_INTEGER structure.
Use normal 64-bit arithmetic on the ULARGE_INTEGER value.
The example here will give you some idea on how to get started.

Answer 2

It seems to me the simplest solution is to borrow what you need.
You already have integers. If you're subtracting two systemTime values, t2 from t1, say,
if( t1.wMilliseconds < t2.wMilliseconds ) {
t1.wMilliseconds += 1000;
t1.wSeconds--;
}
Or, just perform the subtraction. If the result's wMlliseconds is negative, adjust as above.
Take care to ensure t1 > t2, though. You don't want -1.25 = 0.0 - 0.75.
Instead of putzing with strings, if you want a float, make one:
double time1 = t1.wSeconds + 0.001 * t1.wMilliseconds;
C does the conversion for you. It's faster, more direct, and less error-prone than going through strings.

Answer 3

Another way of dealing with these marvels of lost leading or trailing zeroes (found in time and longlat), is to right pad the string with say four zeroes .i.e your newMS+"0000", and take the leftmost four characters.
You then have a number (as text) ranging from "0000" to "9990".
Put a "1" in front of it, then you can easily and unambiguously convert to an integer between 10000 and 19990.
Then you can add and subtract as you like.
Clumsy? Yes indeed :) But I have had to do weird tricks like this when GPS longitude readings go from 11.59(funny numbers) to 12.00