fix: cleanup test directories before running tests

[Edu92337]

Related to #760 (mitigation)

• Keep unconditional /data cleanup at start
• Make /logs cleanup conditional on --preserve-data-dir flag
• Prevents reusing old log state between test runs

Description and Notes

Currently, the cleanup of /logs directory happens at the END of test execution and only if tests pass. This causes:

• Failed tests leave corrupted log state for next run
• Tests may fail when run multiple times without manual cleanup

This PR provides a temporary mitigation by making /logs cleanup conditional on the --preserve-data-dir flag, while keeping /data unconditionally cleaned at start.

Note: This is a partial fix. Complete resolution will come in a follow-up PR after #742 is merged, where cleanup logic will be migrated to the Python runner for better control over test data management.

How to verify the changes you have done?

# Test 1: Run tests multiple times
./tests/prepare.sh
./tests/run.sh
./tests/run.sh  # Should work without manual cleanup

# Test 2: Preserve logs for debugging
./tests/run.sh --preserve-data-dir
ls -la $FLORESTA_TEMP_DIR/logs  # Logs preserved

Next Steps

After #742 is merged, I plan to submit a follow-up PR to:

• Migrate cleanup logic to the Python test runner
• Provide more granular control over test data management
• Fully resolve [Bug] Functional tests failing if you dont clean datadir #760

Contributor Checklist

• I've followed the contribution guidelines
• I've verified one of the following:
  • Ran just pcc (skipped - changes are bash-only, validated manually)
  • Ran just lint-features '-- -D warnings' && cargo test --release (not applicable)
  • Confirmed CI passed on my fork (will run after PR creation)
• I've linked any related issue(s) in the sections above

Manual validation performed:

• ✅ Bash syntax check (bash -n)
• ✅ Logic tested with simulations
• ✅ --preserve-data-dir flag tested
• ✅ Consecutive test runs verified

[moisesPompilio]

Please squash the commits.

[jaoleal]

As stated in #760 (comment)

I'll add the cleanup logic to run.sh at the start of execution,
respecting the --preserve-data-dir flag.
After #742 is merged, I can help migrate this cleanup logic to the Python runner if
needed. What do you think?
Will submit a PR soon

What would be fixing #760 is the follow-up pr that you mentioned. This should be a mitigation PR, including in the description that this fixes #760 will close the issue.

Also,

CACK

[Davidson-Souza]

Why did you remove the comment?

[Davidson-Souza]

Need to delete data as well, no?