uv version bump to 0.4.7

[dirrao]

uv version bump from 0.4.1 to 0.4.7

[topherinternational]

Out of interest (I love uv) I took a quick look here and was able to reproduce the failing test locally. First the error:

+ uv pip install --python /usr/local/bin/python --resolution lowest-direct --upgrade --editable '.[google]'
  error: Failed to download and build `pandas==0.1`
    Caused by: Build backend failed to determine extra requires with `build_wheel()` with exit status: 1
  --- stdout:
  
  --- stderr:
  Traceback (most recent call last):
    File "<string>", line 14, in <module>
    File "/tmp/.tmpKLzrjb/builds-v0/.tmpp1FCdI/lib/python3.8/site-packages/setuptools/build_meta.py", line 332, in get_requires_for_build_wheel
      return self._get_build_requires(config_settings, requirements=[])
    File "/tmp/.tmpKLzrjb/builds-v0/.tmpp1FCdI/lib/python3.8/site-packages/setuptools/build_meta.py", line 302, in _get_build_requires
      self.run_setup()
    File "/tmp/.tmpKLzrjb/builds-v0/.tmpp1FCdI/lib/python3.8/site-packages/setuptools/build_meta.py", line 503, in run_setup
      super().run_setup(setup_script=setup_script)
    File "/tmp/.tmpKLzrjb/builds-v0/.tmpp1FCdI/lib/python3.8/site-packages/setuptools/build_meta.py", line 318, in run_setup
      exec(code, locals())
    File "<string>", line 5, in <module>
  ModuleNotFoundError: No module named 'numpy'

Repro steps for me (on an Intel OS X Mac w/ homebrew) were:

1. Create and activate a fresh Python 3.8 environment
2. Install uv
3. Run the uv command args from the trace above (removing the system python marg)

cd <airflow repo directory>
python3.8 -m venv venv38
source venv38/bin/activate
pip install uv==0.4.11
source venv38/bin/activate  
# ^^ this repeat seemed necessary to point to the freshly-installed uv instead of the homebrew/system uv
uv pip install  --resolution lowest-direct --upgrade --editable '.[google]'

[topherinternational]

I also quickly experimented with lower uv versions, it looks like uv 0.4.7 is the last version without this error (0.4.8 and above fail with the numpy error).