It would be interesting to compare the data stored in the refactoring histories of LTK and the data published by UDC.
It turns out that the command ID's in the UDC data don't match the ones in the refactoring histories. Moreover, the UDC data gives the same command ID to multiple refactoring ID's. For example, the UDC data uses a single command ID for different kinds of rename refactoring but it records different ID's for different kinds of extract refactoring.
Note that the UDC data capture the initiated commands while the Eclipse refactoring histories capture completed refactorings that haven't been undone.
It would be interesting to compare the data stored in the refactoring histories of LTK and the data published by UDC.
It turns out that the command ID's in the UDC data don't match the ones in the refactoring histories. Moreover, the UDC data gives the same command ID to multiple refactoring ID's. For example, the UDC data uses a single command ID for different kinds of rename refactoring but it records different ID's for different kinds of extract refactoring.
Note that the UDC data capture the initiated commands while the Eclipse refactoring histories capture completed refactorings that haven't been undone.