I’ve worked on a similar tool in the past called Charlatan [1] (no I didn’t choose that name), but it implemented its own SQL-like language and was extensible to support any file type.
Its key difference with TextQL and similar alternatives is it works on streams instead of importing everything in a SQLite then querying it. It seems strange to read a whole file in memory then perform a `SELECT` query on it when you could just run that query while reading the file. That means a much lower memory footprint and faster execution, but on the other hand you can only use the subset of SQL that’s implemented.
Its key difference with TextQL and similar alternatives is it works on streams instead of importing everything in a SQLite then querying it. It seems strange to read a whole file in memory then perform a `SELECT` query on it when you could just run that query while reading the file. That means a much lower memory footprint and faster execution, but on the other hand you can only use the subset of SQL that’s implemented.
[1]: https://github.com/BatchLabs/charlatan#charlatan