time-series-analytics: Fix for functional tests workflow (open-edge-platform#1740)

This PR fixes the functional tests workflow for the time-series-analytics project by updating the test execution process to properly prepare the Docker environment before running the full test suite.

Changes:

- Modified the pytest execution sequence to run a specific test first, load environment variables, import Docker images into k3s, and then execute the full test suite
- Changed test pattern variable reference from quoted to unquoted

Signed-off-by: Pooja Kumbharkar <pooja.kumbharkar@intel.com>

Author: pooja-intel

.github/workflows/timeseries-weekly-functional-tests.yaml

@@ -41,7 +41,10 @@ jobs:
         echo "Running function tests"
         pip3 install -r requirements.txt
         rm -rf /tmp/test_report/report.html
-        pytest -q -vv --self-contained-html --html=/tmp/test_report/report.html "${TEST_PATTERN}"
+        pytest -v test_docker.py::test_input_endpoint
+        source ../docker/.env
+        docker save ${DOCKER_REGISTRY}${TIME_SERIES_ANALYTICS_IMAGE}:${IMAGE_SUFFIX:-latest}${WEEKLY_BUILD_DATE:+-${WEEKLY_BUILD_DATE}}${WEEKLY_BUILD_DATE:+-weekly} | sudo k3s ctr images import -
+        pytest -q -vv --self-contained-html --html=/tmp/test_report/report.html ${TEST_PATTERN}
     - name: Upload HTML  test report to Github
       uses: actions/upload-artifact@v4
       with: