Using jsonschema as a "validation backend" in another library

[Paul Walsh]

Hi,

I'm looking to use JSON Schema as a validation backend in a library I'm working on. 

The library takes a data source (CSV, JSON) and runs it through a pipeline to *validate* and *transform* the data. 

I have standard error interface for the pipeline (so I wouldn't be using the jsonschema errors directly in what is returned to the calling code).

Specifically, I want to be able to call validate per row/object/item of data, as I'm iterating over a stream of data, and doing transforms as it moves from one processor to the next (the JSON schema processor being one of the processors).

So I'd like something like:

for row in data:

    valid, row, errors = validate_row_against_schema(row)

Instead of the normal jsonschema.validate function.

I'm just getting to know the code here, but perhaps someone here can point me in the right direction.

[nick...@gmail.com]

If I may chime in with a small idea, you could try work asynchronously and make the pipeline into an iterator that yields the validator each row/object/item of data.
As iter_errors also seems to be an iterator or at least, you could feed from that the other way and raise errors immidiately.
I would appear to me that some kind of solution like this would require comparatively small modifications.

//Nicklas