Add dark ecology support to read_vpts()

[PietrH]

Two ideas: read via arrow, or via guessing paths and vroom

TODO

• create base equivalent to stringr::str_pad
• Refactor to only use directory scanning, not {arrow}
• use bioRad:read_cajun() 🌶️ instead of vroom
• remove default arg for local dir
• rename dateinterval argument to rounded_interval
• rename path to directory
• add call argument
• add ellipsis before call
• implement call
• Include some files in package for testing, but add them to buildignore to exclude them from CRAN
• skip examples if no test files are present
• add tests
• drop stringr dep
• skip tests if no test files are present
• check coverage
• update NEWS

Ideas

• Make arrow dep optional and fall back on vroom
• Make fs dep optional or use base all the time
• Allow returning as tibble to simplify return_type argument in get_vpts
• consider duckdb over arrow? duckdbfs will not help because we don't have filesystem access.

Trouble

• Getting a local unarchived version of dark ecology vpts profiles is non trivial, a single year exceeds the inode quota of the ext4 partition I am using
• how to test this if you need a specific local filesystem?

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.67%. Comparing base (0dd2fc9) to head (4dc1ada).
⚠️ Report is 7 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #187      +/-   ##
==========================================
+ Coverage   92.45%   92.67%   +0.22%     
==========================================
  Files          28       30       +2     
  Lines        2054     2185     +131     
==========================================
+ Hits         1899     2025     +126     
- Misses        155      160       +5     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[bart1]

Also consider using bioRad::read_cajun instead of vroom as it avoids having to implement the parsing

[bart1]

I moved all path test to get_vpts_local so they work for both functions

One thing we might think about it if for dark ecology the return_type="tibble" should work

[bart1]

@YohanSassi to also support reading data from different sources the setup for local data changes a bit in the release from the development version. I'm not sure if you are using it but in that case I wanted to give you a heads up

[bart1]

Note, currently if no data has been found for a radar it is omitted from the result with a warning. As a result the return value of the following call might either be a list of vpts's or a vpts. That might be some what annoying, Should we try to always return a list when multiple radars are requested with only entries for the radars where data has been found?

get_vpts(
             c("bewid","bejab"),
             as.Date("2016-2-1"),
             path = normalizePath(local_dir)
           )

[bart1]

@PietrH @peterdesmet I think this is pretty much done, though I have not tested with large dataset. Only this comment is still open : #187 (comment) .

It would be good to have one extra pair of eyes.

[PietrH]

I had a look at your changes and have no notes. Thanks for picking this up so quickly Bart!