Arihiro Yoshida

> 1. Is it okay if the maps for extra data are a fixed size? I don't think it is good. It would be better for users to be freed...

For a fixed arity node, we just embed a fixed length array as a member in the node structure. We need only few children in most cases (ex. arithmetic operators,...

I've reflected upon the customizable payload on the AST nodes. Your solution is 'map'. I think the customization can be realized by introducing `%ast_value` directive that specifies the data type...

To introduce the AST node types of the corresponding rules, can you share your concrete implementation image? Are there as many `pcc_ast_node_???_t` types as rules? Or is there one type...

PackCC does not have a functionality like _"import"_. Thanks for the suggestion. I understand its usability, but currently I'm reluctant to realize it since it brings some complexity in PackCC's...

> I see in the TinyC example that a second data structure is passed around in `auxil` that looks like it holds the AST. Is this the intended approach for...

Thanks for your interesting report. PackCC requires more memory to support left-recursive grammars, which cannot be handled by most PEG parser generators. By the way, ~~~ /usr/bin/time cat c99-mouse.pp.c |...

That makes sense. Left-recursive grammars can be detected before generating parsers. To tell the truth, I'm not so interested in a parser generator without left-recursive grammar support, but I add...

Thank you for the valuable information. > It looks quite useful for ctags users. Yes, I think so, too. I understand what they want to do, but the algorithm is...

You overrate me :-) I might challenge it when I have time.