Delete HTML table columns?

[rainmakr]

We have an agreement with another site to maintain some data for us. Because they won't make the data available via XML or give us DB accesss, I'm using CFHTTP to do some screen-scraping. However, I only want to use a few columns of the HTML table I am getting. I'm fairly new to CF and am not seeing a way that I can capture a specific number of columns (or delete a specific number of columns).

In other languages (maybe FoxPro?), there was a function that would tell me the position of the nth occurrence of a string, but I'm not seeing a similar function in CF. Can someone get me pointed in the right direction?

Thanks,
John

[usachrisk]

Sorry, no such table manipulation functions are available via ColdFusion. Giving you the data in HTML only is pretty weak, imo. But beyond that, if there's absolutely no way to get the data, you'll have to do some manipulation.

If I had to do this, I would probably start by stripping out everything but the table. Then, I'd use the List Functions to break the data down into rows, and then break the rows down into columns. Then do whatever you want with the data inside.

Good Luck!

[rainmakr]

Thanks, Chris.

I've already gotten the table extracted and all of the style IDs extracted and the missing TDs added...

Because I know the # of columns, I was hoping to find an instr() or similar function that would let me look for the 5th occurrence of the TD tag. Then I was going to do a Left() up until that point. Instead, it sounds like I'll have to set up a loop to move each cell, one at a time, 'X' number of times.

John

[creole]

There's a very simple way to find various things. To find the 5th occurence of a TD tag you'd just need to do this:

Find("<td",yourString [, start position])

Finds the first occurrence of a substring in a string, from a specified start position. The search is case-sensitive.

I heartily recommend the use of CfQuickDocs for ColdFusion reference.

[rainmakr]

There's a very simple way to find various things. To find the 5th occurence of a TD tag you'd just need to do this:

Find("<td",yourString [, start position])

But that doesn't find the fifth occurence, it only finds the first occurence. the start position doesn't help because the length of the cell data varies. Unless I'm missing something.

What I'll need to do is loop through the FIND() a specific number of times (in this case, 6), then utilize a LEFT() to grab the first six columns. I was wondering if it could be done without a loop.

Thanks,
John