How can I loop over a string and get each character?
What you are trying to do is iterate over the characters in a string value. In ColdFusion, there are several ways to accomplish this. Some are fairly low level while others are more complex. I will quickly discuss three ways here.
First, as always though, let's just build a string to test with:
<!--- Set up a string to iterate over. ---> <cfset strText = "Oh man! That Libby is one cool cat!" />
Now, for our testing, we will iterate over each letter and in a bracket, output both the index of the character and the character itself.
ColdFusion Mid() Function Method
Probably the simplest and most commonly used method is to loop over the Length of a string and get the character at a given index:
<!--- Loop over the length of the string. ---> <cfloop index="intChar" from="1" to="#Len( strText )#" step="1"> <!--- Get the character at the given index. The ColdFusion Mid() function takes the string, the starting index and the number of characters to return. In this case, we only want one character, so the length is one. ---> <cfset strChar = Mid( strText, intChar, 1 ) /> <!--- Output the character and the index. ---> [#intChar#:#strChar#] </cfloop>
This gives us the following output:
[1:O] [2:h] [3: ] [4:m] [5:a] [6:n] [7:!] [8: ] [9:T] [10:h] [11:a] [12:t] [13: ] [14:L] [15:i] [16:b] [17:b] [18:y] [19: ] [20:i] [21:s] [22: ] [23:o] [24:n] [25:e] [26: ] [27:c] [28:o] [29:0] [30:l] [31: ] [32:c] [33:a] [34:t] [36:!]
ColdFusion / Java Byte Array Method
This one takes the complexity up a notch. We are going to convert the string to a byte array where each index holds the byte representation of a single character in the string. Then, instead of looping over the length of the string, we loop over the length of the byte array.
<!--- Get the byte array from the text string. Here we are directly calling the underlying Java methods of the string object. ---> <cfset arrBytes = strText.GetBytes() /> <!--- Loop over the byte array as if it were standard ColdFusion array object. ---> <cfloop index="intChar" from="1" to="#ArrayLen( arrBytes )#" step="1"> <!--- Get the character at the given index. The bytes are ascii representations of the number. We have to get the Char() of those values. ---> <cfset strChar = Chr( arrBytes[ intChar ] ) /> <!--- Output character and the index. ---> [#intChar#:#strChar#] </cfloop>
This gives us the same exact output as the first method. As you can see, even though we are returning a Java byte array for the string (which is traditionally indexed at zero), we can treat it as if it were a standard ColdFusion array and start the index at one. ColdFusion rocks for some nice automatic conversions.
Java String Character Iterator Method
This takes it up one more notch. This creates an instance of the Java string character iterator, java.text.StringCharacterIterator, whose sole purpose in life is to do exactly what you are trying to do. We initialize the iterator with the given string, then we keep asking it for the next character until it is done iterating:
<!--- Create a string character iterator. Pass it the string as a constructor parameter so that it can create its own internal representation of the string. ---> <cfset objIterator = CreateObject( "java", "java.text.StringCharacterIterator" ).Init( strText ) /> <!--- We want to keep looping until the current character is the DONE character. ---> <cfloop condition="objIterator.Current() NEQ objIterator.DONE"> <!--- Get the current character. ---> <cfset strChar = objIterator.Current() /> <!--- Output the current character. ---> [#objIterator.GetIndex()#:#strChar#] <!--- Get the iterator to move onto the next character. ---> <cfset objIterator.Next() /> </cfloop>
This gives us "basically" the same output. Sort of. It gives us the same characters but at different indexes:
[0:O] [1:h] [2: ] [3:m] [4:a] [5:n] [6:!] [7: ] [8:T] [9:h] [10:a] [11:t] [12: ] [13:L] [14:i] [15:b] [16:b] [17:y] [18: ] [19:i] [20:s] [21: ] [22:o] [23:n] [24:e] [25: ] [26:c] [27:o] [28:o] [29:l] [30: ] [31:c] [32:a] [33:t] [35:!]
As you can see, this starts at index zero, not one. This is because Java in zero based and we are using it in a completely Java way. In the second example we created a Java array but used it like it was a ColdFusion array and hence the automatic index translation. In this case, however, the iteration is done internally to the Java StringCharacterIterator instance and hence, no automatic conversion.
So that's three different ways to iterate over a string. Which one to use? I guess it depends on what you are trying to do? Each of them is going to have pros and cons. I have not done any specific testing, but I can guess that the byte array method (example 2) is the fastest since iterating over an array is always fast. I don't know how the character iterator object does things, so it might be faster or slower than the Mid() methodology. But of course, it depends on how simple or complex you want to be. If it's simple, go with the Mid() methodology... it's just easy. But if you have a ton of data to go through, you might consider the slightly more complicated byte array method.
I, personally, have never used the character iterator in practice. I suppose that you would use it if you need to pass it off to another black boxed algorithm that is expecting something that implements the CharacterIterator interface. But I doubt you are doing that.
Would the byte array method work with UTF-8?
Interesting question! I am not sure off-hand.
Is it possible to take an XML file, read it and convert it to a single byte array? Also is it possilbe to determine the total filesize of that array?
In the underlying Java of ColdFusion, you can take a string and convert it to a byte array by simply calling .GetTypes() on it. This returns data of type byte.
The file size, I assume would just be the length of the array? If each index is one byte, the file size is directly mirrored by the length of the array?? I think.
If you want to get the length with unicode characters, then you need to use GetBytes("UTF8") instead of GetBytes().
Then you can do either Len(ArrBytes) or ArrayLen(ArrBytes) to get the number of bytes.
Thanks for the tip. I deal so little with extended characters, I have to admit that unicode is a real weakness in my understanding.
Not much strength either ;). But recently I found a bug in Coldfusion where Coldfusion forces IIS to close a connection even if you have Keep-Alive enabled. This is due to Coldfusion not passing a "Content-Length" response header as it should.
I have to calculate the document size and then return it in Content-Length, but I realized that I had unicode characters and GetBytes() counted them as 1 byte instead of 3 or 4.
So GetBytes("Utf8") will count your document correctly, whether your document has Unicode characters or not.
Many thanks for the information. It was very helpful for me.
Just used some of the above code. It's still good after 4+ years. Just wanted to thank you. Amazing how you take the time to provide these awesome posts.
My pleasure good man :)
I want to loop over a string to determine if there are ASCII characters ....ex:char(124) %20
am i missing a way to determine this via the iterations noted above?
Hey Ben, I've read / learned / used a bunch off of your posts and haven't thanked you yet - so first off thanks!
Second thing, I was looking for ways to iterate over a string's characters because I was trying to trim multiple consecutive blank spaces in the middle of a string from a data feed before storing it in my database. These work great - but was curious if you knew of another (better) method to use for that specific purpose? If not i'll just use the byte array method.