Grand Document Model Refactoring [was #7808]
The XML editor's WYSIWYM currently uses a rather suboptimal data model: An improved version of Vex' original data model that is kept synchronized with the WST XML editor's data model. This is especially error prone because of the two content buffers with their two character offset schemes (WST: Source code positions, Vex: whitespace normalized text plus \0s as tag placeholder) that need to be synchronized.
We should evaluate a leaner data model that might use more of the WST Structured Text framework and gets rid of the content buffer synchronization. I.e. the Vex layer should be as thin as possible.