detect number of times a character is in a string.

[orange gold]

I have a string with alot of text..
I want to detect how many times @"hello" is in the string... I know how to detect if it is or isn't but how do I detect the number of times it appears in the string?

[Duncan C]

Originally Posted by orange gold

I have a string with alot of text..
I want to detect how many times @"hello" is in the string... I know how to detect if it is or isn't but how do I detect the number of times it appears in the string?

I would suggest writing a method countOccurrencesOfString:inString:

You could write it by calling rangeOfStringptions:range: repeatedly. You'd set up an NSRange of the entire string to start with. Then if the first call found any occurrences, you'd increment a count, calculate a new NSRange that would start at the next character after the first occurrence, and keep searching while more occurrences are found.

Something like this:

- (int) countOccurrencesOfString: (NSString*) searchString 
  inString (String*) targetString;
{
  BOOL found;
  int count = 0;
  NSRange searchRange, foundRange;
  searchRange.location = 0;
  searchRange.length = [searchString length];
  foundRange = [targetString rangeOfString: searchString 
    options: NSLiteralSearch 
    range: searchRange.location];
  found = (foundRange.location != NSNotFound);
  while (found)
  {
    count++;
    searchRange.location = foundRange.location + foundRange.length;
    searchRange.length = [searchString length] - searchRange.location;
    if (searchRange.length < [searchString length])
      break;
    foundRange = [targetString rangeOfString: searchString 
      options: NSLiteralSearch 
      range: searchRange.location];
    found = (foundRange.location != NSNotFound);
  }
  return count;
}

Note that the code above is completely untested. I did not even check to see if it compiles. There may be syntax errors, logic errors, etc. Testing and debugging it is an "exercise for the reader."

[orange gold]

Thanks! Since I posted this I have recieved code from another online coder... This is what he said, it seems to work great

NSRegularExpression *regex = [NSRegularExpression regularExpressionWithPattern:@"\\bhello\\b" options:NSRegularExpressionCaseInsensitive error:NULL];
NSUInteger numberOfMatches = [regex numberOfMatchesInString:someString options:0 range:NSMakeRange(0, [string length])];

Thankyou for posting your code too though, it is neat to see other ways to do it... you build a whole counting engine.. very nice

Thanks again!

[Duncan C]

Originally Posted by orange gold

Thanks! Since I posted this I have recieved code from another online coder... This is what he said, it seems to work great

NSRegularExpression *regex = [NSRegularExpression regularExpressionWithPattern:@"\\bhello\\b" options:NSRegularExpressionCaseInsensitive error:NULL];
NSUInteger numberOfMatches = [regex numberOfMatchesInString:someString options:0 range:NSMakeRange(0, [string length])];

Thankyou for posting your code too though, it is neat to see other ways to do it... you build a whole counting engine.. very nice

Thanks again!

As predicted, my code had a few errors in it.

The regex-based version is more powerful. It will handle cases like searching for "the" when the string contains "their". My code would count "their" as containing "the". The regex version you posted will only match whole words.

[orange gold]

regex contains "options" (e.g. case-sensitive, alphanumeric only, etc.) I'm sure there is one for that too... I hope /: hhahah Because mine may need to involve that soon. Haha

[Duncan C]

Originally Posted by orange gold

regex contains "options" (e.g. case-sensitive, alphanumeric only, etc.) I'm sure there is one for that too... I hope /: hhahah Because mine may need to involve that soon. Haha

Both the code I posted and the regex string counting can do case sensitive/case insensitive, ignore diacritical marks, etc.

regex-based searching has lots more options however. You can set up a regular expression so a string only matches if it is a whole word. That' what the example you posted does. By enclosing your word between "/b" entries ("/bhello/b") it makes sure the word hello has a word break on either side.

You can also create regular expressions that will take characters from a list of characters. You could match a string that had an a, b, or c as the first character, a 1, 2, 3 as the second character, and an x, y, or z as the last character, for example.

All that would be a pain in the butt to do manually with code.