Add a option for local search to fido client

[nicHoch]

No description provided.

[codecov]

Codecov Report

Attention: Patch coverage is 96.15385% with 4 lines in your changes missing coverage. Please review.

Project coverage is 75.31%. Comparing base (35c88f3) to head (5717a53).
Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
stixpy/net/client.py	95.52%	3 Missing ⚠️
stixpy/net/attrs.py	97.29%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #143      +/-   ##
==========================================
+ Coverage   73.56%   75.31%   +1.75%     
==========================================
  Files          32       32              
  Lines        2016     2111      +95     
==========================================
+ Hits         1483     1590     +107     
+ Misses        533      521      -12     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[samaloney]

Looks ok will need a change log e.g. https://github.com/TCDSolar/stixpy/blob/main/changelog/README.rst

[samaloney]

I think we should switch to anc, asp, ephemeris for level, datatype and product so remove aux etc?

I don't think we need to add so many files I know they are empty but should be able to just mock the list of file names similar to here but for a local client https://github.com/sunpy/radiospectra/blob/main/radiospectra/net/sources/tests/test_rstn_client.py

Also need to think about the versions thing what happens if you search for both min and latest specifically how if works with https://docs.sunpy.org/en/latest/topic_guide/extending_fido.html#search-attrs

It may be better as function which operates on the Fido.search result?

Not sure if the test fail is real or just a typo?

[ayshih]

Some (trivial) regex comments/suggestions

[ayshih]

• Did the changes to ql_filename an sci_filename actually affect matching? Your commit message implies that you made these changes for Windows, but I don't immediately see how that would be the case.
• The period in .fits should formally be escaped