Toy example for token tracking:
Top: Sequential revisions R1, R2, R3, and R4 of a dummy article from left to right. Color of underlining indicates origin revision. Bottom: The token tracking data after revision R4, showing all tokens that have so far appeared in the article, if currently deleted or not. Each token is linked to a unique token ID (note some identical string values for distinct tokens), the revision ID of original introduction, and each revision in which a token was possibly a) deleted and b) reinserted (both are lists). Arrows in the table indicate undo actions among revisions (not yet part of the API output).