Add hook? and documentation on how to add custom extractors

[vanto]

I didn't find any mechanism to add custom extractors to beanhub-extract and beanhub cli. It would be great to have some clarification in the docs whether this is possible or not and if, how. Using a full qualified module name would be one possible solution. Or is the intended way to use the plaid CSV format as intermediate format?

[fangpenlin]

Hey @vanto, thanks for the feedbacks. You're totally right, there's no way to import with a custom extractor at this moment. I have the plan in mind that I would like to add it at somepoint. But usually I would like to wait until people ask to avoid building too much not needed features. Now since you have asked, let's see what I can do for you.

First idea is to add a new extractor called csv for extracting data from generic CSV files. The column will just follow the naming defined in the transactions object here. In that way, as long as you have a tool converting any other transactions into the csv extractor's format, you can use it from there. This approach is very easy, I can get it done very quickly.

Another approach is to allow a third-party plugin or loading extractor from an external module. It takes a little bit more effort for me to think about how and where the theird-party extractor should be loaded and used. But it's not too hard anyway.

Do you have any preference which way? Or, I can actually do both if it's needed. I think both usecase are useful regardless.

[fangpenlin]

Hey @vanto, I just added the CSV extractor with #3. Please let me know if this addresses your needs. If not, I am already thinking about making it possible to use a custom extractor for the import operations.