Test Mount API

[daviewales]

WIP branch adding tests to Mount API for various backends.

[buhtz]

Great. I am most of the days offline until next week. But i am on it

[daviewales]

Note that it's still WIP. I've added high-level tests for the 'local' backend, as it's the simplest. I'll work on adding tests for the remaining backends.

[buhtz]

• I updated the branch to the latest dev.
• I kind of recreated the new encode.py file so that its git history (from encfstools.py) is preserved. Because of a Microsoft GitHub bug you wont see this history at GitHub. But in your local branch you should see its history back to year 2012/13 with git log --follow encode.py (the --follow is important!).
• Add encode.py to the list of files that will be checked for all linter rules.
• Some minor adjustments.

I really like your tests. I can read and understand them very well. This is high quality.
I will dive a bit deeper. But for today I am finished. So you can work on your branch if you want to. I'll be back later.

[buhtz]

I don't understand the purpose of MountWithLocalEncFS.test_unconfigured_mount(). What is the difference between "unconfigured" and "configured" in context of EncFS?

[daviewales]

Unconfigured corresponds to the first run, before the BiT configuration or the EncFS directory is initialised.

Configured corresponds to later runs, where there is a preexisting configuration and EncFS directory.

[buhtz]

Unconfigured corresponds to the first run, before the BiT configuration or the EncFS directory is initialised.

Configured corresponds to later runs, where there is a preexisting configuration and EncFS directory.

There is this out-commented line in test_mount.py

# unconfigured_mount = mount.Mount(cfg=cfg, tmp_mount=True)

To my unerstanding cfg here is configured for local-encfs. So how can it be unconfigured. I still don't get it, sorry.

I don't know that the desired test result could be for test_unconfigured_mount(). But I also don't understand all the mount behavior of BIT. Have not touched that part of the code base ever.

If you can be more specific with your questions and needs I might be able to help.

[daviewales]

I've had a chance to look at this. 'Unconfigured' in this case refers to the EncFS configuration, rather than the BackInTime configuration. Yes, the BackInTime configuration is reused, but the EncFS mount has not yet been created. Perhaps I should rename it to test_uninitialised_mount()?

Edit:

Specifically, an initialised backup location will contain the file .encfs6.xml, while an uninitialised backup location will not.

[buhtz]

"uninitialized" sounds good to me.

[buhtz]

Hello David,
Can you please give me a short update about this PR?
Regards,
Christian

[daviewales]

I haven't had much chance to work on the lately. The latest commit is work in progress, as the test I'm adding still needs some work to make it pass. In theory, it is supposed to create a new EncFS mount, and confirm that it was created.

[buhtz]

Hello David,
in the last days, I was quit active with refactoring. Have not done this if I had knew that you are working currently.

I am finished now and hope I have not interfered your work to much with conflicts.

Best,
Christian

[daviewales]

Don't worry about me! Conflicts are expected with a slow moving branch like this. (And there weren't many.)

[daviewales]

I'm getting a R1732 lint warning when creating a temporary directory, but I can't seem to disable the warning with noqa.

[buhtz]

Which linter?
If pylint use "#pylint: disable=R1732"

Or follow the error and use "with"

Sorry. I am at holiday just using a touch device