Use sha256 hash of extra coordinates for cache dir name, pepify

[kephale]

To address #31 (comment)

[hanslovsky]

Can you push without formatting changes? That will make it easier to review the actual changes.

[ctrueden]

I'm a fan of the pepification. I agree with @hanslovsky that doing it as a separate commit would be best.

Here is how I typically tease such things apart using git:

git reset HEAD^ # undo the commit
git stash
do-the-pepification # so the unchanged code gets pepified
git commit -a -m 'Pepify!'
git checkout 'stash^{tree}' -- . && git stash drop # force exactly what you stashed back into the working copy
git commit -a -m 'My awesome change, clean this time'

Alternately, @kephale, if you don't have time... just tell us the command to do the pepification? Then I can do the above process myself.

[kephale]

Okey doke, @hanslovsky @ctrueden all done.

@ctrueden thank you for the commands! That was super useful

[kephale]

FWIW, the pepify is with https://github.com/google/yapf which is attached to the python mode in my emacs

[hanslovsky]

Thanks @kephale
I have a pretty busy schedule right now but will try to make some time this weekend to look over it.

[kephale]

Sure, no rush. It is just 4 lines, but I'm pip installing the branch now anyway, so this doesn't block me anymore.

[hanslovsky]

Can you share an example path in the ~/.jgo cache? I would expect something like

~/.jgo/${groupId}/${artifactId}/${version}/e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855

Is that correct?

As a user, can I see what the coordinate is for a workspace/hash? I think it would make sense to add a file to the workspace that shows coordinate or coordinate[1:]?
This file may also be helpful in the (extremely unlikely) case of a hash collision - jgo could throw an error instead of executing the wrong Java application.

[kephale]

.jgo/sc.fiji/bigdataviewer-vistools/1.0.0-beta-29-SNAPSHOT+eba447da3454eafae1d5fd4a362d3457b84d0c5a20db7e40767c390d1dab6cba

This solved my use case, and I was able to move on to the work it blocked. Saving the coordinate to file makes sense, will put it on the TODO.

[kephale]

@hanslovsky okey doke. I adjusted the path and added the coordinates file as well:

$ cat ~/.jgo/net.imagej/imagej/2.1.0/ad412f409ae00c86c7c3f165d0d787ed4dd7a42191903e4f0256bebee4b4fa45/coordinates.txt 
net.imagej:imagej:2.1.0
graphics.scenery:scenery:937ba10
org.jetbrains.kotlin:kotlin-stdlib-jdk8:1.4.20
org.jetbrains.kotlinx:kotlinx-coroutines-core:1.3.9
sc.fiji:bigdataviewer-core:10.1.1-SNAPSHOT
sc.iview:sciview:f4dd286

[hanslovsky]

Looks good to me, will leave it to @ctrueden to merge it.

[kephale]

bump

[kephale]

bump

[cthoyt]

I would suggest using black to deterministically reformat the whole package. I could send a PR for that

Update: see #63

[cthoyt]

@ctrueden now that this is merged, do you think you can mint a new release to PyPI?

[ctrueden]

do you think you can mint a new release to PyPI?

@cthoyt I pushed jgo 1.0.2 to PyPI. Please let me know how it works for you!