[prep CDF-25295] 🫠 Introduce new columns (#1860)

# Description
**Context**: This is part of the `alpha` migration tooling in Toolkit,
for overview [see
here](https://cognitedata.atlassian.net/wiki/spaces/~170890702/pages/5455773697/Migration+Tooling+in+Toolkit).
This is step 3 of that process.

This PR extends the columns of the input format that the migrate command
expects to take new columns such that the user can provide the ingestion
and consumption views they would like to use for the instances. We do
not yet use these columns, so the user cannot set them and expect any
different behavior.

A side-effect of this change is what is described below. The reading of
the csv file becomes more robust to order of columns, and improved error
message.

## Changelog

- [x] Patch
- [ ] Skip

## cdf

### Improved

- In the `cdf migrate timeseries/files` commands the mapping file can
now have columns in an arbitrary order and will ignore irrelevant
columns instead of raising an error.

## templates

No changes.


