Right now, MacroParse language definition documents are treated to a janky ad-hoc analysis based on string manipulations. That's rather horrible in light of what this package is all about, don't you think? It seems to me that a much more respectable approach would be to take full advantage of MiniScan and MiniParse to generate a complete AST and then walk said AST (using appropriately-defined visitor objects) in order to generate the structures that will ultimately feed the table generators.
This approach should mean the MacroParse compiler can "eat its own dog-food" by relying on the provided frameworks for error reporting, etc. It should also make it easier to add or change features.
Right now, MacroParse language definition documents are treated to a janky ad-hoc analysis based on string manipulations. That's rather horrible in light of what this package is all about, don't you think? It seems to me that a much more respectable approach would be to take full advantage of MiniScan and MiniParse to generate a complete AST and then walk said AST (using appropriately-defined visitor objects) in order to generate the structures that will ultimately feed the table generators.
This approach should mean the MacroParse compiler can "eat its own dog-food" by relying on the provided frameworks for error reporting, etc. It should also make it easier to add or change features.