gstreasy 0.1.0

Makes using GStreamer in python easy!

A re-imagining of gstreamer-python. I was going to just work on a fork of that repo but there was so much I thought should be changed. Some of the new features:

• Auto detect caps if in gst-launch command
• Auto detect and configure appsink and appsrc if in command
• Support for appsink and appsrc in same pipeline
• Faster Gst.Sample -> ndarray

Example Usage

Also check out the user_code.py script for an appsrc example.

Simple pipeline without an appsink element:

simple_cmd = "videotestsrc num-buffers=60 ! autovideosink"
with GstPipeline(simple_cmd) as pipeline:
    print("Running simple pipeline")

Pipeline with an appsink element:

appsink_cmd = "videotestsrc num-buffers=60 ! appsink emit-signals=true sync=false"
with GstPipeline(appsink_cmd) as pipeline:
    while pipeline:
        buffer = pipeline.pop()
        if buffer:
            type(buffer.data)  # np.ndarray

Pipeline using tee element and multiple sinks:

tee_cmd = '''
    videotestsrc num-buffers=60 ! tee name=t
    t. ! queue ! video/x-raw,format=RGB,framerate=60/1
       ! appsink emit-signals=true sync=false
    t. ! queue ! video/x-raw,format=RGB,framerate=60/1
       ! jpegenc ! avimux
       ! filesink location=recording.mp4
'''
with GstPipeline(tee_cmd) as pipeline:
    while pipeline:
        buffer = pipeline.pop()
        # do whatever you want with the buffer's ndarray
    # Meanwhile recording.mp4 is being written

Develop

All dev tasks can be handled with the run.sh script but it just wraps standard tools if you can't/don't want use it.

• python -m pip install -e .[dev,doc] to install deps
• tox to run tests for py3.7 and py3.10.
• flake8 and mypy for linting
• pdoc -d google src/gstreasy for online docs