you need
file_types not supported_extensions.
It really would go a lot faster if you just read the builtin plugin for pdf input instead of trying to get chatgpt to do it for you. Just look at the existing PDFInput class and override the appropriate methods/fields.
|