fix(models): preserve LiteLLM reasoning fragments

[he-yufeng]

Summary

• concatenate LiteLLM reasoning fragments without inserting synthetic newlines
• preserve the previous single-fragment behavior
• add regression coverage for multi-fragment reasoning text

Fixes #5645

To verify

• uv run --extra test python -m pytest tests\unittests\models\test_litellm.py -q
• uv run --extra dev pyink --check src\google\adk\models\lite_llm.py tests\unittests\models\test_litellm.py
• uv run python -m py_compile src\google\adk\models\lite_llm.py tests\unittests\models\test_litellm.py
• git diff --check

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.