Normalize dataset column names

[machristie]

Currently, the column names in uploaded source and sink files must exactly match expected values. For example, the "fieldCap" column of uploaded sink files must exactly be named fieldCap (MtCO2). Some tools like ArcGIS don't like parentheses in these column headers, so this is an additional burden for users to convert headers back and forth when moving between tools like ArcGIS and the web portal.

One improvement would be to more flexibly look for column headers and normalize them to specific values when generating GeoJSON. Continuing the example, the upload code could look for a column header with "fieldCap" in it and update it to fieldCap (MtCO2).

In addition, we might as well check that all expected headers are present as a way of validating the uploaded data.

[machristie]

I found a couple of interesting differences between column names for sources and sinks

• For source files, the 'name' column is called NAME but is Name in sink files
• For source files, the 'fixO&M' column is called fixO&M ($M/y) but is fixO&M ($M/yr) in sink files

[machristie]

here are the column name normalization rules: https://github.com/SciGaP/simccs-maptool/blob/dev/simccs_maptool/serializers.py#L331-L368