Enh/thrustcurve api cache

[Monta120]

Pull request type

• Code changes (bugfix, features)

Checklist

• Tests for the changes have been added (if needed)
• Docs have been reviewed and added / updated
• Lint (black rocketpy/ tests/) has passed locally
• All tests (pytest tests -m slow --runslow) have passed locally
• CHANGELOG.md has been updated (if relevant)

Current behavior

Every call to GenericMotor.load_from_thrustcurve_api triggers a new download from the ThrustCurve API, even if the same motor has already been requested. This results in repeated network requests and slower test runs.

New behavior

• Introduce caching for ThrustCurve API requests to avoid multiple downloads for the same motor.
• Subsequent requests for the same motor now return cached data, improving efficiency and reducing network load.

Breaking change

• No

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 80.32%. Comparing base (9cf3dd4) to head (bd29470).
⚠️ Report is 8 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop     #881      +/-   ##
===========================================
+ Coverage    80.27%   80.32%   +0.05%     
===========================================
  Files          104      106       +2     
  Lines        12769    13001     +232     
===========================================
+ Hits         10250    10443     +193     
- Misses        2519     2558      +39     

☔ View full report in Codecov by Sentry.
📢 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.

[Gui-FernandesBR]

This way the cache will be deleted everytime you restart a python session. It doesn't solve 100% of the problem.

My idea is for us to create a .rocketpy_cache folder and save every download inside it. Before downloading a motor rom the API, we simply read it from file.

There should be options like "no_cache = Tre" to skip cache.

Please dedicate a few minutes to research how other python libraries tackle the same problem (caching downloads from rest API requests). I know OSMnx is a good example, but certainly there are many other libs doing the same thing.

Can you use these guidelines to improve this PR pls? I will mark it as a draft for now, but pls lemme know whener you finish it again so I can review it

---

Adittionaly, I think you have a problem with git tree, as I can see 43 commits on the commits section. I recommend you running the following: git rebase -i develop, then you select which commits you really need in this PR, and drop the others.

[Monta120]

Hi @Gui-FernandesBR , thank you for the feedback! I have implemented the changes you requested:

Persistent Caching: I updated the logic to use a persistent cache directory located at Path.home() / ".rocketpy_cache". The code now checks if the motor file exists locally before making a network request, persisting data across Python sessions.

no_cache Option: I added a no_cache=False argument to load_from_thrustcurve_api. Setting it to True forces a fresh download from the API.

I also ensured that the unit tests use a temporary directory for the cache (via monkeypatch) so they don't clutter the user's home folder during testing.

[Gui-FernandesBR]

Hi @Gui-FernandesBR , thank you for the feedback! I have implemented the changes you requested:

Persistent Caching: I updated the logic to use a persistent cache directory located at Path.home() / ".rocketpy_cache". The code now checks if the motor file exists locally before making a network request, persisting data across Python sessions.

no_cache Option: I added a no_cache=False argument to load_from_thrustcurve_api. Setting it to True forces a fresh download from the API.

I also ensured that the unit tests use a temporary directory for the cache (via monkeypatch) so they don't clutter the user's home folder during testing.

Amazing! You're a star @Monta120 !!

[Gui-FernandesBR]

LGTM, only need a final touch on the CHANGELOG file.

And let's wait for copilot's review as well.

Great work

[Copilot]

Pull request overview

This PR adds caching functionality to the ThrustCurve API integration to avoid redundant network requests when loading the same motor multiple times. The implementation uses a file-based cache stored in ~/.rocketpy_cache and introduces a no_cache parameter to allow bypassing the cache when needed.

Key Changes

• Introduced automatic caching of ThrustCurve API responses to avoid repeated downloads
• Added no_cache parameter to force fresh API downloads when needed
• Implemented comprehensive test coverage for caching behavior

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 8 comments.

File	Description
rocketpy/motors/motor.py	Adds cache directory initialization, implements caching logic in _call_thrustcurve_api, and adds no_cache parameter to load_from_thrustcurve_api
tests/unit/motors/test_genericmotor.py	Adds new test test_thrustcurve_api_cache to verify caching works correctly, including cache reads, writes, and bypass functionality
docs/user/motors/genericmotor.rst	Updates documentation to explain caching behavior and the no_cache parameter with examples
CHANGELOG.md	Documents the new caching feature and related changes

Comments suppressed due to low confidence (1)

rocketpy/motors/motor.py:2018

• The no_cache parameter is missing from the docstring. According to NumPy docstring conventions, all function parameters should be documented in the "Parameters" section. Add documentation for the no_cache parameter, including its type (bool), default value (False), and description.

    def load_from_thrustcurve_api(name: str, no_cache: bool = False, **kwargs):
        """
        Creates a Motor instance by downloading a .eng file from the ThrustCurve API
        based on the given motor name.

        Parameters
        ----------
        name : str
            The motor name according to the API (e.g., "Cesaroni_M1670" or "M1670").
            Both manufacturer-prefixed and shorthand names are commonly used; if multiple
            motors match the search, the first result is used.
        **kwargs :
            Additional arguments passed to the Motor constructor or loader, such as
            dry_mass, nozzle_radius, etc.

[Gui-FernandesBR]

@Monta120 pls mark all the comments as "solved" after solving it, so we can merge it directly

[Monta120]

I'm working on testing exception handling for cache operations to ensure above 80.27% coverage.