I think I finally can claim that I at least know how this project needs to be implemented (took me long enough, right). Right now implementation is still in complete disarray, but latest IR refactor seems to be just the right thing. Having said that, right now it is probably hard to make sense of all the questions and "open for design discussion" issues that I created, and swarm of bullet points in the readme.
This means I need (assuming I expect someone to contribute) to at least explain
[ ] General project architecture - it is not as difficult as it might seem, especially considering I probably deleted twice more code than exists now.
[ ] How exactly hcparse uses other libraries - clang, tree-sitter, boost::wave etc.
[ ] How final nim wrapper code is generated, including existing and planned boilerplate providers
Core part of the hcparse has a finishing line - I need to read the cxx or c code into IR. After that part is completed (implemented) there are only additional bugfixes. Things that come after have a much bigger potential for expansion, and at the same time easier to improve.
I think I finally can claim that I at least know how this project needs to be implemented (took me long enough, right). Right now implementation is still in complete disarray, but latest IR refactor seems to be just the right thing. Having said that, right now it is probably hard to make sense of all the questions and "open for design discussion" issues that I created, and swarm of bullet points in the readme.
This means I need (assuming I expect someone to contribute) to at least explain
Core part of the hcparse has a finishing line - I need to read the cxx or c code into IR. After that part is completed (implemented) there are only additional bugfixes. Things that come after have a much bigger potential for expansion, and at the same time easier to improve.