Gerwin Klein

Closing this, because I think this is best handled adding to the input, not by changing the scanning engine.

Closing this as the maco-in-char-classes feature seems to have alleviated the problem and opened #1026 as a tracking issue for the potential common-prefix optimisation mentioned in this issue.

I'll leave this open -- #765 does not fix it, it just provides specific dates for the jar file.

So I tried this and couldn't reproduce it. When I do `mvn generate-sources` or `mvn package` and get an "up to date" output from the plugin, the file modification time...

There is now a mechanism for leaving out the default SuppressWarnings annotation, so I think we can close this one.

While there are some other problems in the spec, we should at least get a useful error message, and in fact most of the variations here would actually make sense....

Closing this in favour of #1027. Users can still add their own `@Generated` annotation, so there is no need to force an additional runtime dependency on all users, and #764...

To answer my own question from above: this is the minimal spec, it does indeed generate a DFA with 79k states. This is likely the cause of the exception, i.e....

What is wrong in the table output is that the table is encoded as a Java string to circumvent java class size restrictions (other methods would have failed earlier). Java...

Ok, so after implementing it in #1043, and measuring performance, we do get a performance degradation of about 15% percent. I'm not entirely sure of the cause, but it looks...