Kohsuke KAWAGUCHI wrote:
> But we need a code to examine a start tag and pick the right sub
> automaton, which is already a departure from the plain-vanilla automaton
> algorithm. That's why I wrote
>
> | So my understanding is that your proposed restrictions are not for
> | making the automaton-based validation possible, but for making a simple
> | workaround for them.
I see what you mean. Yes, this is not a simple tree automaton. On the other
hand, I think that this algorithm is useful especially for table-driven
validators. If we disallow overlapping of tag names, the code for examining
start tags only requies two lists of tag names.
Cheers,
Makoto
