Revise MRS 1d residual fringe correction to run automatically and populate x1d extension

[stscijgbot-jp]

Issue JP-3790 was created on JIRA by David Law:

After considerable discussion within the MIRI team, we would like to revise the current implementation of the 1d residual fringe correction within the extract_1d step.

At present, ifu_rfcorr is an optional argument that users can provide to apply the 1d correction.  This correction is off by default as it can remove periodic features in some cases with molecular bands, but this also means that the vast majority of users need to rerun both cube building (to create per-band cubes) and spectral extraction in order to get clean spectra.

The requested change is to have 1d residual fringe correction always run by default, but to provide two versions of the extracted spectra.  The first copy, in the usual EXTRACT1D 1st extension, will be the normal spectrum with no defringing applied.  The defringed spectra (science, background, etc) will be in a new 2nd extension with some sensible FITS EXTNAME.  Maybe RFCORR1D?

This way, users can simply choose for themselves which to use without rerunning the pipeline, or requiring additional MAST data products.

A few details:

• It may be worth retaining the ifu_rfcorr option flag in extract_1d for a little while, but when set to TRUE simply log a statement informing users of the change in behaviour.  This at least won't break previous code.
• At the same time, the MIRI team will deliver a parameter reference file update that changes the default Spec3 cubes to use per-band cubes for which residual fringe correction works well.  This will have a significant impact on DMS and MAST.
• We'll presumably need an updated extract1d datamodel to handle the new extension?  Not sure of the details here.