Bill Roper's Journal

There are some approaches in programming that simply don't scale.

One of them, I am now reminded after running a bunch of timing tests, is reading large XML data using DOM. My only consolation is that I was not the one who decided that using DOM was a good idea.

I have suggested using a nice serial forward-only XML reader.

And that someone else can fix up the code to do so. :)

I was wrong. My friend actually came back and put the fence posts into their holes while I was inside working and keeping track of the toilet installation. Now we just need to concrete them in place...

Meanwhile, I've been cleaning up some of my XML generation so that everything is reading and writing XML that looks like it was written with some set of guiding principles. The problem comes when you get to the point of XML documentation that says "Do it however suits you."

Well, yes, I can do that. I just want to avoid religious arguments once I set the format. :)

I have been frantically coding this week to try to fix a problem with formula translation as we move from the old account numbering regime to an account naming regime, because a function that I had hacked together that depended on the specific format of the old account numbers no longer worked. I had been trying to produce a solution that would require reverse parsing the byte stream that results from parsing the formula text, but that became increasingly unwieldy -- actually, it was more just a matter of it was a lot of work -- so I set that aside and went to a simpler solution using the parser with a flag setting that told it to correct the formula text on the fly so I could store the fixed version.

I think that works, but something went wrong when I tried to test it. So I pulled up the trusty audit dialog in the Excel add-in version of our product to see what the formula was in the debugger.

It immediately failed.

What the--?

A bit of debugging later, I discovered that this time they had remembered to encode the string with the quotation marks on the way in, but not to decode it on the way out from the XML format.

To quote the late Casey Stengel, "Can't anybody here play this game?"

I have now coded the XML retrieval to use one of my utility functions that automatically decodes the string output. I had to write a new version of the function, because the one that I had used boost::shared_ptr, while the function I was working with had used std::auto_ptr, but that was a quick substitution.

I don't really care which one we use, but it would be nice if we made up our minds...

Now to see if this damned thing works...