Question

I have a text field in an xib file. In a method in the .m file, I can print the contents of the text field, but I cannot get those contents converted to a float. The text field is formatted with commas, as in 123,456,789. Below is the code snippet, where datacellR2C2 is the textfield.

float originalValue2 = originalValue2 = [datacellR2C2.text  floatValue];
NSLog(@"datacellR2C2 as text --> %@ <---\n",datacellR2C2.text);  // this correctly shows the value in datacellR2C2
NSLog(@"originalValue2 = %f  <--\n", originalValue2);  // this incorrectly returns the value 1.0 

I would appreciate any suggestions for a fix or a direction where I should look for the problem.

Answer 1

In the declaration for -floatValue a comment is shown:

/* The following convenience methods all skip initial space characters
(whitespaceSet) and ignore trailing characters. NSScanner can be used
for more “exact” parsing of numbers.
*/

Ergo, commas cause truncation, because they are trailing chars. Even the string you provided (123,456,789) only prints 123.000, because that’s all -floatValue sees.

//test
NSString *string = @"123,456,789";
float originalValue2 = [string  floatValue];
NSLog(@"datacellR2C2 as text --> %@ <---\n",string);  // this correctly shows the value in datacellR2C2
NSLog(@"originalValue2 = %f  <--\n", originalValue2);

//log
2012-07-07 22:16:15.913 [5709:19d03] datacellR2C2 as text --> 123,456,789 <---
2012-07-07 22:16:15.916 [5709:19d03] originalValue2 = 123.000000  <--

Just get rid of them with a simple +stringByReplacingOccurrencesOfString:withString:, and remove those trailing commas:

//test
NSString *string = @"123,456,789";
NSString *cleanString = [string stringByReplacingOccurrencesOfString:@"," withString:@""];
float originalValue2 = [cleanString  floatValue];
NSLog(@"datacellR2C2 as text --> %@ <---\n",cleanString);  // this correctly shows the value in datacellR2C2
NSLog(@"originalValue2 = %f  <--\n", originalValue2);

//log
2012-07-07 22:20:20.737 [5887:19d03] datacellR2C2 as text --> 123456789 <---
2012-07-07 22:20:20.739 [5887:19d03] originalValue2 = 123456792.000000  <--

By the way, a float is rounding that string up to an even number, use double precision instead.