new python bindings: SqliteReportBuilder.from_pyreport()

[matt-codecov]

fixes codecov/engineering-team#2315

commit titles describe the changes. the result is python bindings like:

from codecov_rs.report import SqliteReportBuilder

# Create a new database file in a temp dir
report_builder = SqliteReportBuilder.from_pyreport("report_json.json", "chunks.txt")
print(report_builder.filepath())
# /var/folders/96/92bch5jn5llfz6gcvvchfy100000gn/T/codecov-W5jDQymDQHNKitVH

# Use a specified filepath
report_builder = SqliteReportBuilder.from_pyreport("report_json.json", "chunks.txt", "db.sqlite")
print(report_builder.filepath())
# /Users/matt/dev/codecov/codecov-rs/db.sqlite

immediate next steps:

• review and then flip the repo to public
• CI glue to publish to pypi

there is some awkwardness to sort out due to python expecting to access SqliteReportBuilder functionality (db inserts) and SqliteReport functionality (.totals(), .merge(), converting back to pyreport) through a single handle. this PR punts that issue down the line but i am thinking about it

[codecov]

Codecov Report

Attention: Patch coverage is 92.50000% with 6 lines in your changes missing coverage. Please review.

Project coverage is 98.53%. Comparing base (dc4903d) to head (72e3eac).
Report is 1 commits behind head on main.

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
bindings/src/error.rs	0.00%	6 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #36      +/-   ##
==========================================
- Coverage   98.62%   98.53%   -0.09%     
==========================================
  Files          21       22       +1     
  Lines        6751     6776      +25     
==========================================
+ Hits         6658     6677      +19     
- Misses         93       99       +6     

Components	Coverage Δ
core	98.61% <100.00%> (-0.01%)	⬇️
bindings	79.31% <76.00%> (-20.69%)	⬇️
python	100.00% <100.00%> (ø)

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

[Swatinem]

lgtm, though this brings up an interesting question:

should the output path/file be the full responsibility of the caller?
In the sense that the API does not have the new fallback, and always requires the caller to provide a file handle?

related to that, I believe the rust tests are currently leaking temporary files, as it does not look like you are using the tempfile crate thus far.

[matt-codecov]

should the output path/file be the full responsibility of the caller?

is there a reason to force the responsibility to one side or the other? why not allow the caller to supply a file if they want to manage it and fall back to a tempfile created by the library if they don't care?

related to that, I believe the rust tests are currently leaking temporary files, as it does not look like you are using the tempfile crate thus far

i think all of the tests use tempfiles/tempdirs. python uses NamedTemporaryFile and rust uses tempfile::TempDir (example) which delete themselves on drop

the ::new() constructor will leak temp files though, which is somewhat intentional but i am torn. it feels a little strong for our library to say "we will create a file but if you want to keep it you either need to keep us in scope forever or copy it" but a little weak to say "we're fine throwing potentially gigabytes of garbage in your temp dir". what do you think @Swatinem?

[Swatinem]

Thats why I would suggest delegating that responsibility to the caller. That "resolves" that question, along with slimming down the API surface and the potential for surprises.

[matt-codecov]

@Swatinem i've removed ::new() for now and made an outpath required. if that proves tedious i'll revisit