pyparsing VS cinje

After over a year since the last release of pyparsing, I've bundled up all the bug-fixes and changes, and they are now released as pyparsing 3.1.0. Visit this link for the details.

Look into "parser combinators" for building an interpreter. There's a few ones out there, but PyParsing is one I've seen around that looks pretty nifty.

I havent found a way to specify an exact character match in pyparsing - https://github.com/pyparsing/pyparsing/discussions/443

STDOUT: Lark or pyparsing

Since it is not mentioned in the article: Python users may also want to check out pyparsing [0]. It is slightly different from Parsec/FParsec (for instance, it ignores all whitespace by default), but I think it is a really good project.

[0]: https://github.com/pyparsing/pyparsing/

Here is the page of all the new changes and features in pyparsing 3.0.

Your pyparsing parser might benefit from a tune-up. This page has some performance tips: https://github.com/pyparsing/pyparsing/wiki/Performance-Tips.

If your SQL has fairly complex structure, you will need a full blown SQL parser. If your statements are mostly simple select, you can get pretty close with Pyparsing, here is an example.

Looks good. I'm not sure that a union is the way to go. When I've thought about creating something like this in the past I've mostly considered it from the angle of starting with a different preprocessor, such as https://github.com/cxxxr/lisp-preprocessor, and using it to generate iterator interfaces and the like. While it may sound absolutely undesirable I've also seen PHP used to great effect in a similar way, treating files like templates with inline PHP. I've wondered about a similar solution using the file encoding support in Python, similar to how https://github.com/marrow/cinje functions.