Anyone who has worked with XML tables knows comparing is tricky – there is a trade-off between capturing structural changes and showing the content changes in a way that is useful for the reader. This release of XML Compare includes a completely new approach to table comparison that is focused on showing content changes in as much detail as possible. Those used to seeing row duplications in their table results will be pleased to know that these have been dramatically reduced. In the vast majority of cases, changes are now shown at the cell level, leading to a more intuitive table result.
Handling spans is much improved with the results far more easy to display and understand when reviewed. Previously, row duplication often occurred when table changes included a modification to a span. The following examples show the new, superior table results for both row and column spanning changes.
Old Result
New Result
Old Result
New Result
Our new approach delivers a more detailed result for cells and rows, highlighting the change that matters. In some cases, the old algorithm duplicated the entire table to display changes. In the majority of cases, that ‘fallback’ position is no longer used and we get a great result that is easy to read.
Old Result
New Result
One of the most fundamental changes to understand in a table is column moves, additions or deletions. Due to the way that tables are represented in XML, this is not always straightforward. The new result makes more sense and the simple representation makes the change more immediately apparent to the viewer.
Input A
Input B
Old Result
New Result
Now that columns are handled more intuitively, it’s possible to change the way that they are compared. By default, column order is significant i.e. if column order changes, that is flagged up in the result (as in the example above). However, if column order doesn’t matter, configure the comparison to ignore column order and it will align them differently in the result.
Input A
Input B
Column order is significant
Column order is ignored
When ignoring the order of columns, you may want to control the way in which columns are aligned. By default, they are aligned based on their content. However, it is possible to set up column ‘keys’ to control how the alignment takes place. The following keying modes are available:
AUTO – the content of each column is used to find the best alignment
COLNAME – columns are only aligned if they have the same colname attribute on their cells
POSITION – columns are aligned purely by position i.e. column 1 in A aligns with column 1 in B etc.
It is also possible to manually provide custom keys for each column by adding processing instructions into the table. See the user documentation for more details.
Well, we will be removing a lot of classes that were deprecated in 11, some customers may still be using these so please install your update on a test server. For a full list of classes and guide explaining which classes/methods customers should use instead please check out our updated documentation area.
These table enhancements will rolled out to our other products shortly, including DITA and DocBook, and will also support xHTML tables, so subscribe now for the latest updates. We are always keen to hear your feedback so update your XML Compare today simply by downloading from MyDelta. If you are new to DeltaXML request an evaluation trial licence today and our technical experts we will be happy to onboard you.
Sign up to the DeltaXML newsletter to get the latest on DeltaXML and other company news.
WHERE CHANGE MATTERS
From ISO Quality Services Ltd.
Editing advanced XML document comparison pipeline configurations
Scroll to top
There is more detail about this in a paper presented at the Balisage conference in August: https://www.balisage.net/Proceedings/vol27/html/LaFontaine01/BalisageVol27-LaFontaine01.html