|Atomik Xport SE: Reference > Chapter 6 Working with Rulesets > 6.6 Boxes tab||<< >>|
When an image appears at the very start an article in a series of articles, it can present a particular problem for basic box ordering. The extraction system can place the image at the end of the preceding article, if your content-model allows it.
A related problem is that, if your content model allows an article to start with an image, this means any image in the body of an article could potentially trigger a new article.
Both of these scenarios would lead to undesirable results in your XML.
By comparing the middle of images, rather than the top-left corner, images will be extracted later than identically aligned text boxes, and avoid these pitfalls. From Atomik Xport 4.3, this is the default behaviour of the extraction system. You can set this comparison point to ‘middle’ or ‘top-left’, for images, in the Box comparisons section of the Advanced Box Options.
The default behaviour of Atomik Xport is to follow threaded (chained) boxes, and extract their entire text thread as if the text were contained in the same box. However, there may be situations where you wish to treat every box containing threaded text as if it were separate, and instead compare it with other boxes on the page. Unchecking the ‘Follow text chains’ checkbox will allow your documents to be processed in this way.
Checking this option will ensure that text boxes which appear on the pasteboard are considered for extraction if they are threaded (chained) to boxes on the page.
Note: With this option unticked, any box which appears on the pasteboard will still be extracted if it overlaps the page.