Support for reading ASCII line lists for WLS

[awhoward]

Currently, we're using .npy-formatted line lists for ThAr. They are specified in configs/wls_auto.cfg:

red_linelist = /data/reference_fits/kpfMaster_ThArLines20221005_red.npy
green_linelist = /data/reference_fits/kpfMaster_ThArLines20230221_green.npy

These numpy files contain per-order dictionaries of the wavelengths and associated pixels, as shown below.

@shalverson proposes that we use an ASCII-formatted line list instead. See Issue https://github.com/Keck-DataReductionPipelines/KPF-Pipeline/issues/746 for a sample file. Note that this file only contains the wavelengths of known good lines. I assume that Sam is proposing that the code use an initial guess wavelength solution to infer the associated order (and initial guess pixel) for each line. Is that right, @shalverson. If this is all correct, the revised code should read in the ASCII line list and create a dictionary like the ones referenced above (which presumably are used later in the code?).

The motivation for all of this is that ASCII line lists are commonly used in the community and reported in the literature. We'd like to be able to quickly test line lists from various sources.

Who would like to take this on? @aaronhouseholder, if you have the bandwidth, that would be helpful. If not, that's okay, but let us know.

Related issue: https://github.com/Keck-DataReductionPipelines/KPF-Pipeline/issues/746

[aaronhouseholder]

I think @bjfultn said that he would work on this.

[bjfultn]

I was going to write up a tool to convert between ASCII and the format the pipeline wants but if we want to transition the pipeline to using ASCII then this will require a bit more work.