jwdt / csv_converter Goto Github PK

This will work in a similar way to the lambda handler, but instead of including the function code within the JSON, it will just reference a python function.

{ "New Header One": { "old_column": "Old Header One", "default": "DEFAULT FOR ONE", "lambda": "lambda a, b: str(a.get(b.get('old_column'))).lower()" }, "New Header Two": { "old_column": "Old Header Two", "default": "DEFAULT FOR TWO", "funlink": "custom_python_file_or_module/custom_function" } }

It would then split the string by the forward slash, and import it using:

https://docs.python.org/3/library/importlib.html

Finally, it would run the imported function, in a similar way to the lambda one, with the following variables:

input_line - the current INPUT line
previous_input_lines - all the previous input file lines
current_output_lines - the current output dictionary, not a copy, but the actual one, so these functions are able to view values from, and make changes to previous entries.
item - the current column item in the JSON config.

It would output both the individual item it is processing, as well as the current_output_lines, which is to be updated if there are any changes.