Migrate FuseArgs away from using Optparse

[lonetwin]

The current implementation of FuseArgs extends Optparse, which is scheduled to be deprecated in favor of argparse. Although there isn't a schedule for when it would be deprecated, it might be a good idea to reimplement this using argparse.

Furthermore, imho, the current implementation is a bit unwieldy and inflexible. For instance, I have spent way too longer than I expected and yet have not been able to figure out how to implement the command line to support something like:

my_fuse_cmd.py <required parameter> <required path parameter> <mountpoint> [fuse options]

If there is interest in migrating away from optparse, and nobody is already working on it, I would like to attempt doing this.

If it is felt that this isn't a worthwhile effort for any reason, could someone guide me on how to create a Fuse instance such that the command-line above is supported (using FuseArgs, that is -- I obviously could split up the parsing of args I am interested in and those that get passed to the parent Fuse class ...but that would be ugly).

[sdelafond]

@lonetwin this would totally be worthwhile, and I'd gladly accept a PR for it.

[lonetwin]

@sdelafond I'm happy to start working on this. However, I was wondering whether this project intents to retain python 2.x support (afaict, 2.7 is still being supported according to the package classifiers in setup.py)

[sdelafond]

Yes, I'd like to retain 2.x compatibility.

[lonetwin]

About 4 years after I initially request this, I am returning to actually implement it :).

I've submitted PR #95 to address this. The PR is split into distinct commits:

• The first that introduces the tests, without any modification to the code.
• The second that migrates the code, without any modification to the tests.

This ensures that we maintain compatibility with current usage. I could break these into 2 separate PRs if that would be easier to review / track.

Note that I haven't really checked for compatibility with python 2.x as requested in the last comment and tbh, I am not very motivated to do this. If however, this is going to be a deal-breaker to get it merged, depending on whether the PR itself looks good, I'll attempt it (hopefully before another 4 yrs :) )