fix: validate calls polling arguments like goals does

[cnpierrepapi]

Summary

Closes #30. The other half of it, waiting on the result instead of the status, already landed in 1.0.

calls.wait_for_result passed interval_seconds straight to time.sleep. Zero meant polling GET /v2/calls/{id} with no delay until the timeout, and nan or a negative number blew up inside time.sleep without saying which argument was wrong. goals already has _validate_polling_seconds for exactly this, so both call paths use it now.

create_and_wait checks before the POST, not after. @cuentapraces07-ops caught this on the issue: checking only inside wait_for_result means a bad interval_seconds still places the call first. That's the same order goals.run_and_wait already uses.

16 new tests, the same 8 bad values for each method, each checking the error names the argument and that no request went out. All 16 fail on main and pass here.

Heads up that #45 edits the calle.errors import line right above the one this adds, so whichever lands second needs a one-line rebase.

Checklist

• I kept this change within the supported server SDK scope.
• I did not add browser/client-side patterns that expose CALL-E API keys.
• I did not include private collaboration links or unconfirmed public repository references.
• I updated tests, examples, or docs when behavior changed.
• I updated the changelog when the change affects package users.
• I ran the relevant local checks.

Local checks

I don't have uv on this machine, so I ran the steps from validate.sh directly in a venv rather than the script itself:

pytest -q                               81 passed
ruff check .    (locked 0.15.15)        All checks passed
mypy src/calle                          no issues in 6 source files
scripts/check_public_repo_hygiene.py    passed