Feature

ColdFusion 8 has a load of awesome image manipulation functionality. So much so, in fact, that it will take several articles to discuss it in any sort of decent way.

In Part 1 of this article (CFDJ, Vol. 9, issue 5), we spent the entire post just exploring all the ways in which ColdFusion 8 can read in and write out/save images. If you haven't read that already, do so now because I won't be explaining the reading/writing of images going forward (I'm assuming that you already understand all the crazy, awesome combinations of ways in which this can be done).

Just as with the reading and writing, image manipulation can be done through both the ColdFusion 8 CFImage tag and through the ColdFusion 8 image manipulation functions. In Part 2, I am going to focus on all the stuff that you can accomplish with just the CFImage tag (and a few of the functions). The CFImage tag provides a very small subset of all that can be done, so I figured it would make for a good primer on manipulation.

In addition to the reading and writing of images, the ColdFusion 8 CFImage tag provides these image manipulation functions:

• Adding a border
• Resize image
• Rotate image
• Creating a CAPTCHA image

Before we get into any of it, I am just going to start off by saying that I believe that 90% of the image manipulation that people want to accomplish is merely the resizing of uploaded images to create thumbnails (come on, let's be honest). If that's the case, you are going to be tickled pink with excitement. CFImage resizes images fast and with a furious amount of quality. I'm not that surprised, after all, this is Adobe we're talking about - the ruler of the photo and image editing empire.

To help present these image manipulation features a little more clearly, I've created a function called AddImageInfo(). This user-defined ColdFusion function takes in a ColdFusion image object and writes the width and height of the image to the top-left corner of the graphic. This will help us see how the size of the image changes as we manipulate it. The algorithm of the AddImageInfo() function goes beyond the scope of this CFImage article but, hopefully, seeing some of this in Listing 1 will get you more excited about the ColdFusion 8 image editing possibilities.
Even though it's not entirely relevant to this article, there is something I would like to go into a tad bit of detail about (regarding the above UDF) because at first it was throwing me for a loop. Once I get the ColdFusion image object passed in the AddImageInfo() function, I create a copy of it using the ImageCopy() method. I do this because I don't want to corrupt the image that was passed in as it was passed in by reference (not by value).

At first, I thought I would just copy the image using the CFImage tag:

<!---
Copy the image and overwrite the passed in argument
value. Since ColdFusion images are passed around by
reference, this should create a COPY of the image and
then save that new image reference over the one that
was passed in.
--->
<cfimage
action="read"
source="#ARGUMENTS.Image#"
name="ARGUMENTS.Image"
/>

Notice that we are reading in the passed-in ColdFusion image object and then overwriting the ARGUMENTS.Image value. This should have created a new variable value, thereby detaching this reference from the image that was passed in. However, ColdFusion is doing something very strange here - CFImage[ action=read ] does not copy the image to a new variable; it seems to be copying the image data back into the existing variable. To see this in action, take a look at this UDF in Listing 2.

Notice that I am supposedly overwriting the passed in argument with a duplicate of it. Okay, now let's put it into action in Listing 3.
Notice that we are altering the image as we write it to the browser. Since we creating a copy of the image in the UDF, each outputted image should only have one line on it. However, when we run the above code, we get the images shown in Figure 1.

Notice that each AlterImage() call updated the original image data such that each subsequent call built on top of the previous one. Clearly, our CFImage[ action=read ] is not overwriting the variable, ARGUMENTS.Image. This got me thinking - is this complicated by the fact that it was a function or something? I figured I would try this without using a UDF (see Listing 4).

Notice here that we are reading in the image but storing it into a new variable name (objImage versus objNewImage). If this works the same way as the preceding example, objImage and objNewImage should point to the same image data. Yet, when we run the above code, we get the images shown in Figure 2.

In this case, the two variables, objImage and objNewImage, point to independent copies of the image. We know this because, if they pointed to the same image, a border applied to one would show up in both WriteToBrowser tags.

This is very strange behavior, and it feels to me like a bug in the way CFImage handles variable names that overwrite themselves. Anyway, just be careful when reading one image into another image as it might not always do what you expect it to. Just so you know, to make sure I wasn't crazy, I tried something similar with query-based function where the I ran a query that overwrites itself:

<cfquery name="ARGUMENTS.Query" dbtype="query">
SELECT
*
FROM
Query
</cfquery>