In the three-way compare case ranges of differing lines are highlighted and framed with different colors to show whether the difference is an incoming, outgoing, or conflicting change. The TextMergeViewer supports the notion of a current "differing range" and provides toolbar buttons to navigate from one range to the next (or previous).

If there is a current "differing range" and the underlying document is editable the TextMergeViewer enables actions in context menu and toolbar to copy a range from one side to the other side, thereby performing a merge operation.

In addition to a line-by-line comparison the TextMergeViewer uses a token based compare on differing lines. The token compare is activated when navigating into a range of differing lines. At first the lines are selected as a block. When navigating into this block the token compare shows for every line the differing token by selecting them.

The TextMergeViewer's default token compare works on characters separated by whitespace. If a different strategy is needed (for example, Java tokens in a Java-aware merge viewer), clients can create their own token comparators by implementing the ITokenComparator interface and overriding the TextMergeViewer.createTokenComparator factory method).

Access to the TextMergeViewer's model is by means of an IMergeViewerContentProvider. Its getXContent methods must return either an IDocument, an IDocumentRange, or an IStreamContentAccessor. In the IDocumentRange case the TextMergeViewer works on a subrange of a document. In the IStreamContentAccessor case a document is created internally and initialized from the stream.

A TextMergeViewer can be used as is. However clients may subclass to customize the behavior. For example a MergeTextViewer for Java would override the configureTextViewer method to configure the TextViewer for Java source code, the createTokenComparator method to create a Java specific tokenizer.