My guess is that this is a spoofax bug.
When I’m editing a somewhat larger file in my own defined language (read: more time is needed to analyze the ast), I sometimes get the following message in the console:
Internal error evaluating editor-analyse (CancellationException; see error log)
The trace following this error differs from time to time.

The error occurs when I have changed something in the editor and hit a button (or backspace) right after that. That is, changing the code in the editor whíle the editor is analyzing (in the bottom right corner of Eclipse it says: “Analyzing updates to: …”.
The error trace will differ each time this happens, because it depends on the point at which the previous editor-analyze is currently (at the time of the call to editor-analyze) working.

Because it says CancellationException, my guess is that the previous call to editor-analyze will be canceled, causing the error to pop-up. Canceling the previous call to editor-analyze seems OK to me, because analyzing an outdated AST is meaningless. Maybe the CancellationException can be caught and not reported?

Submitted on 2 December 2010 at 15:05

It looks like this is a duplicate of Spoofax/241, which was fixed in the 0.5.3 “unstable” release. See the FAQ about updating to this release before it’s officially released.