Let users pass max_threads as an argument to _avif.AvifDecoder

[yit-b]

Although the Image.open API doesn't yet let you pass arguments nicely to the underlying decoder, it would be really nice to be able to configure the decoder's max_threads option via this plugin when I know that the image I am reading is an AVIF. Example:

decode_threads.py

import pillow_avif
from pillow_avif import _avif
from PIL import Image

if __name__ == "__main__":

    from pillow_avif import _avif

    orig = Image.open("tests/images/flower.jpg")
    orig.save("flower.avif", quality=80)

    with open("flower.avif", "rb") as f:
        decoder = _avif.AvifDecoder(f.read(), "auto", "auto", 4) # 4 is max_threads
        width, height, n_frames, mode, icc, exif, xmp = decoder.get_info()
        data, timescale, tsp_in_ts, dur_in_ts = decoder.get_frame(0)

        img = Image.frombytes(mode=mode, size=(width, height), data=data)
        img.save("foobar.jpg")

The default behavior is unchanged, max_threads is still 0 when you open an avif via Image.open(). Only when you drop down a layer can you configure this.

As with #49, maybe we should consider a more reasonable default value for the decoder's max_threads for the same reasons.

Reproduce:

conda activate foobar # This is some new empty conda environment
conda install -c conda-forge pillow 'libavif>=1.0.2' aom 'python=3.10.*' pillow
cd pillow-avif-plugin
pip install --no-deps .
python decode_threads.py

[yit-b]

@fdintino gentle bump