Unconditionally store image path in POSIX notation in configuration

[mih]

[This PR sits on #242, only the last commit is unique and relevant here]

This establishes a fixed format for the configuration. Previously, the path was stored in platform conventions without any indication of the nature of that platform. This turned platform detection into guesswork. Configuration created on windows was not functional on other platforms.

Some of that guesswork is now implemented in a path normalization helper (_normalize_image_path()). It is executed on-read, in order to keep older dataset configuration functional for the cases

• we can tell the platform conventions, because the image is placed under .datalad\\environments\\ (the default)
• we can locate a matching element on the file system

Notably, this guesswork does not cover the case of an image that was configured on windows, and is hosted in a currently uninstalled subdataset. This seems rare enough of a case to not deserve the implied complexity of supporting it in a backward compatible way.

Closes #224

TODO:

• [x] run this on windows and confirm that it works (stores POSIX, and operates on platform paths)
• [x] adjust tests to simulate an old setup with windows paths

The missing lines in the coverage are

I am not planning on adding a test for this case (a path that is verified to be in windows convention by finding a matching filesystem item). If this is unacceptable, I am OK with removing support for this entirely.

[codeclimate[bot]]

Code Climate has analyzed commit cd49d66a and detected 0 issues on this pull request.

View more on Code Climate.

[codecov[bot]]

Codecov Report

Attention: 2 lines in your changes are missing coverage. Please review.

Comparison is base (c505e52) 94.69% compared to head (cd49d66) 94.70%. Report is 10 commits behind head on master.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #243 +/- ## ========================================== + Coverage 94.69% 94.70% +0.01% ========================================== Files 25 25 Lines 1093 1133 +40 ========================================== + Hits 1035 1073 +38 - Misses 58 60 +2 ``` | [Files](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad) | Coverage Δ | | |---|---|---| | [datalad\_container/containers\_add.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvY29udGFpbmVyc19hZGQucHk=) | `89.80% <100.00%> (+0.13%)` | :arrow_up: | | [datalad\_container/containers\_list.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvY29udGFpbmVyc19saXN0LnB5) | `95.74% <100.00%> (-0.81%)` | :arrow_down: | | [datalad\_container/containers\_remove.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvY29udGFpbmVyc19yZW1vdmUucHk=) | `97.43% <100.00%> (-0.07%)` | :arrow_down: | | [datalad\_container/containers\_run.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvY29udGFpbmVyc19ydW4ucHk=) | `84.84% <ø> (ø)` | | | [datalad\_container/tests/test\_containers.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvdGVzdHMvdGVzdF9jb250YWluZXJzLnB5) | `100.00% <100.00%> (ø)` | | | [datalad\_container/tests/test\_run.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvdGVzdHMvdGVzdF9ydW4ucHk=) | `100.00% <100.00%> (ø)` | | | [datalad\_container/utils.py](https://app.codecov.io/gh/datalad/datalad-container/pull/243?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=datalad#diff-ZGF0YWxhZF9jb250YWluZXIvdXRpbHMucHk=) | `92.68% <94.11%> (+6.96%)` | :arrow_up: |

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[mih]

Thanks for the reviews! Very helpful!