Pip install complexities (just Kamodo now)

[sapols]

It's not currently possible to install every PyHC package in a single pip call since a few of them require special treatment. Here's what we're doing at the time of this writing:

# Install PyHC packages with pip (SpacePy, Kamodo, and pySPEDAS separately)
RUN pip install --no-cache-dir numpy==1.24.3 && \
    pip install --no-cache-dir spacepy --no-build-isolation && \
    pip install --use-pep517 --retries 5 --no-cache-dir -r /app/requirements.txt && \
    pip install --no-cache-dir git+https://github.com/nasa/Kamodo.git && \
    pip install --no-cache-dir pytplot-mpl-temp && \
    pip install --no-cache-dir pyspedas

Causes:

1. SpacePy fails to install if numpy isn't already installed and you don't use the --no-build-isolation flag
2. Kamodo is being installed from GitHub not PyPI
3. pySPEDAS fails to install if pytplot-mpl-temp isn't installed first

[sapols]

1. Might no longer be an issue after the SpacePy 0.5.0 release.

[sapols]

1. is indeed no longer an issue after SpacePy 0.5.0.

3. pySPEDAS installs fine on its own now and will pull in the correct pytplot-mpl-temp

So the only package getting special treatment now is Kamodo which gets installed from GitHub:

# Install PyHC packages with pip (Kamodo separately)
RUN pip install --use-pep517 --retries 5 --no-cache-dir -r /app/requirements.txt && \
    pip install --no-cache-dir git+https://github.com/nasa/Kamodo.git

[sapols]

I'll note that I swapped the order of the installs to prevent Kamodo's dependencies from potentially clobbering others':

# Install PyHC packages with pip (Kamodo separately)
RUN pip install --no-cache-dir git+https://github.com/nasa/Kamodo.git && \
    pip install --use-pep517 --retries 5 --no-cache-dir -r /app/requirements.txt

And I'll note we get this warning about pip being outdated for each package we process:

Notice:  A new release of pip is available: 23.0.1 -> 24.3.1
Notice:  To update, run: pip install --upgrade pip

It'd be cool to actually upgrade pip to avoid those and instead print something useful like:

Processed package <name> (x/70)

UPDATE: I've added print statements so the GitHub Actions output includes: Processing package <name> (x/y). We still get pip notices between those though.