This exposes SPL operators in the com.ibm.streamsx.avro
toolkit as Python methods.
Package is organized using standard packaging to upload to PyPi.
The package is uploaded to PyPi in the standard way:
cd package
python setup.py sdist bdist_wheel upload -r pypi
Note: This is done using the ibmstreams
account at pypi.org and requires .pypirc
file containing the credentials in your home directory.
Package details: https://pypi.python.org/pypi/streamsx.avro
Documentation is using Sphinx and can be built locally using:
cd package/docs
make html
or
ant doc
and viewed using
firefox package/docs/build/html/index.html
The documentation is also setup at readthedocs.io
.
Documentation links:
To change the version information of the Python package, edit following files:
- ./package/docs/source/conf.py
- ./package/streamsx/avro/__init__.py
When the development status changes, edit the classifiers in
- ./package/setup.py
When the documented sample must be changed, change it here:
- ./package/streamsx/avro/__init__.py
- ./package/DESC.txt
When using local build (e.g. not forcing remote build), then you need to specifiy the toolkit location, for example:
export STREAMSX_AVRO_TOOLKIT=<PATH_TO_AVRO_TOOLKIT>/com.ibm.streamsx.avro
Package can be tested with TopologyTester using the Streaming Analytics service.
Run the test with:
ant test-sas
or
cd package
python3 -u -m unittest streamsx.avro.tests.test_avro.TestStreamingAnalytics
For using the toolkit from the build service (force_remote_build) run the test with:
Run the test with:
ant test-sas-remote
or
cd package
python3 -u -m unittest streamsx.avro.tests.test_avro.TestStreamingAnalyticsRemote
Package can be tested with TopologyTester using a local and running Streams domain. Make sure that the streams environment is set, the domain and instance is running and the environment variables: STREAMS_USERNAME STREAMS_PASSWORD are setup.
Run the test with:
ant test
or
cd package
python3 -u -m unittest streamsx.avro.tests.test_avro.TestDistributed