NSScanner behavior - iphone

Question

I am very new to iOS development. I am trying to parse a simple csv file that has about 10 lines separated by commas. I am using the code below but not able understand why NSScanner, when parsing the fields (fields in the code below) does not go to the next string after the comma. I have to execute the line
[fields scanCharactersFromSet:fieldCharSet intoString:nil];
to make it go past the delimiter. However, I don't have to do the same thing for lines - NSScanner automatically sets the position to the next line past the newline. In both cases I am using the same method - [lines scanUpToCharactersFromSet:intoString] Is there something I am not understanding?
Here is the test file I am trying to parse:
Name,Location,Number,Units
A,AA,4,mm
B,BB,3.5,km
C,CC,10.2,mi
D,DD,2,mm
E,EE,6,in
F,FF,2.8,m
G,GG,3.7,km
H,HH,4.3,mm
I,II,4,km
Here is my code:
-(void)parseFile {
NSCharacterSet *lineCharSet = [NSCharacterSet newlineCharacterSet];
NSCharacterSet *fieldCharSet = [NSCharacterSet characterSetWithCharactersInString:self.separator];
// import the file
NSStringEncoding *encoding = nil;
NSError *error = nil;
NSString *data = [[NSString alloc] initWithContentsOfURL:self.absoluteURL usedEncoding:encoding error:&error];
NSString *line,*field;
NSScanner *lines = [NSScanner scannerWithString:data];
while (![lines isAtEnd]) {
[lines scanUpToCharactersFromSet:lineCharSet intoString:&line];//automatically sets to next line - why?
NSLog(#"%#\n",line);
NSScanner *fields = [NSScanner scannerWithString:line];
while (![fields isAtEnd]) {
[fields scanUpToCharactersFromSet:fieldCharSet intoString:&field];
[fields scanCharactersFromSet:fieldCharSet intoString:nil]; //have to do this otherwise will not go to next symbol
NSLog(#"%#\n", field);
}
}
}

Answer 1

That's just the way NSScanner works. When you use scanUpToCharactersFromSet:intoString:, it scans characters up to but not including the characters in the set. If you want it to move past characters in the set, you have two options:
Make it scan those characters. You are doing this now using scanCharactersFromSet:intoString:. Another way you could do it is [fields scanString:self.separator intoString:nil].
Tell the scanner that the separator character is to be skipped, using setCharactersToBeSkipped:. However, this will make it hard for you to detect empty fields.
The scanner's default set of characters-to-be-skipped includes the newline. That's why your outer scanner skips the newline.
You could do this entirely using componentsSeparatedByString:, instead of using NSScanner. Example:
-(void)parseFile {
NSString *data = [[NSString alloc] initWithContentsOfURL:self.absoluteURL usedEncoding:encoding error:&error];
for (NSString *line in [data componentsSeparatedByString:#"\n"]) {
if (line.length == 0)
continue;
NSLog(#"line: %#", line);
for (NSString *field in [line componentsSeparatedByString:self.separator]) {
NSLog(#" field: %#", field);
}
}
}