Hi everyone
Can anyone let me know how to convert a string to int.

Thx kevin

There is a built-in RobotC function called atoi().

So it would work like this?

@MHTS: What value is returned by atoi() if the string argument sSrc or pzSrc doesn't strictly contain decimal digits? If there's a decimal place, is the number rounded down to an int?

atoi is exactly what it said "ascii to integer". If you have a floating point number, then there is atof.

If you attempt to use atoi on a non-integer string (meaning, on the string "Hello", for example), atoi will NOT convert the individual letters to their ASCII equivalents. It will, however, partially convert a string that starts with an integer and has non-integer characters mixed in. For example:



This code results in the output:



The reason for this is that atoi checks the first character for a 0-9 integer value and if found, converts it to an integer. When it comes across a non-integer character it stops converting the characters over.

In the case of Test 1, the decimal '.' is the non-integer value that stops atoi, so only the integers before it (reading left to right) are converted. Test 2 starts with a non-integer character so no characters are converted nor assigned to Test2. Test 3 stops converting when it reaches the 'H' character in "Hello".

Note that Test 1 can be fully converted with all of the values intact if using the atof (ASCII to float) function instead:





Because the float data type is only reliable for up to ~7 decimal digits, ROBOTC is able to correctly convert the first four decimal digits (1, 2, 3, and 4), the decimal point, and the next four decimal digits (5, 6, 7, and . Anything beyond that may or may not calculate correctly, so please be aware of these limitations.

Last but certainly not least, you can also use atoi/atof with character arrays:


Which will return:


Again, unless you specify the amount of decimal places to display in the writeDebugStreamLine, ROBOTC will display the highest resolution it can for the float (8 decimal digits).

We have a full listing of all of the Math commands on our Wiki that you may want to take a look at as well.

That reminded me a question: does RobotC support "double"? I assume "float" in RobotC is 32-bit and "double" is 64-bit. In most scenarios float is fine, but in some case, we may need double.

At this point ROBOTC only supports 32 bit data types; while we do not have any plans to support 64 bit data types, we do appreciate the feedback and will take it into consideration for future development.

Out of personal curiosity, what would you be using the double types for?

Like I said, in most scenarios, I only need float. But for scenarios where a floating point number is involved in a complex expression and it is multiplied with a big number. Precision may be necessary.

Why the fancy function????

It's called a type cast, it's not a function. You can read about casting here: http://en.wikipedia.org/wiki/Type_conversion

You're basically forcing a (temporary) conversion between one type of variable to another.

= Xander

ThANKS!