feat: support coverage ci for python sdk

[mangoGoForward]

Signed-off-by: mango [email protected]

• What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)
  feature

• What is the current behavior? (You can also link to an open issue here)
  fixed test: add coverage report for python sdk #1971

• What is the new behavior (if this is a feature change)?
  None

[codecov]

Codecov Report

Merging #2025 (68a7b7e) into main (b78fb44) will increase coverage by 0.06%.
The diff coverage is n/a.

@@             Coverage Diff              @@
##               main    #2025      +/-   ##
============================================
+ Coverage     75.60%   75.67%   +0.06%     
  Complexity      347      347              
============================================
  Files           612      615       +3     
  Lines        116926   116739     -187     
  Branches       1022     1015       -7     
============================================
- Hits          88404    88337      -67     
+ Misses        28313    28193     -120     
  Partials        209      209              

Impacted Files	Coverage Δ
src/codec/field_codec.h	16.66% <0.00%> (-28.93%)	⬇️
src/rpc/rpc_client.h	85.13% <0.00%> (-4.06%)	⬇️
hybridse/src/udf/udf_test.h	88.67% <0.00%> (-3.78%)	⬇️
src/catalog/tablet_catalog.cc	70.77% <0.00%> (-3.00%)	⬇️
hybridse/src/codec/fe_row_codec.cc	78.49% <0.00%> (-0.32%)	⬇️
hybridse/src/codegen/udf_ir_builder.cc	82.62% <0.00%> (-0.31%)	⬇️
hybridse/include/codec/list_iterator_codec.h	80.85% <0.00%> (-0.21%)	⬇️
src/tablet/tablet_impl.cc	59.37% <0.00%> (-0.20%)	⬇️
...4paradigm/openmldb/batch/nodes/WindowAggPlan.scala	60.98% <0.00%> (-0.17%)	⬇️
hybridse/src/vm/runner.cc	67.77% <0.00%> (-0.04%)	⬇️
... and 27 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update b78fb44...68a7b7e. Read the comment docs.

[github-actions]

SDK Test Report

  75 files    75 suites   6m 7s ⏱️
163 tests 161 ✔️ 2 💤 0 ❌
204 runs  202 ✔️ 2 💤 0 ❌

Results for commit 68a7b7e.

♻️ This comment has been updated with latest results.

[github-actions]

Linux Test Report

       54 files       196 suites   49m 12s ⏱️
  8 993 tests   8 989 ✔️ 4 💤 0 ❌
13 239 runs  13 235 ✔️ 4 💤 0 ❌

Results for commit 68a7b7e.

♻️ This comment has been updated with latest results.

[mangoGoForward]

Hi @aceforeverd , the step of generate python coverage report failed, seems zookeeper can't connect. should we add a zookeeper container?

[aceforeverd]

Hi @aceforeverd , the step of generate python coverage report failed, seems zookeeper can't connect. should we add a zookeeper container?

Yes it dumps, should use steps/ut_zookeeper.sh to start zookeeper first, you can have a reference from the test sqlalchemy step, it have a script file

OpenMLDB/steps/test_python.sh

Lines 24 to 25 in af0272c

	test -d /rambuild/ut_zookeeper && rm -rf /rambuild/ut_zookeeper/*
	cp steps/zoo.cfg "$THIRDSRC/zookeeper-3.4.14/conf"

[mangoGoForward]

should use steps/ut_zookeeper.sh to start zookeeper first, you can have a reference from the test sqlalchemy step, it have a script file

OpenMLDB/steps/test_python.sh

Lines 24 to 25 in af0272c

	test -d /rambuild/ut_zookeeper && rm -rf /rambuild/ut_zookeeper/*
	cp steps/zoo.cfg "$THIRDSRC/zookeeper-3.4.14/conf"

Thanks so much.

[aceforeverd]

should use steps/ut_zookeeper.sh to start zookeeper first, you can have a reference from the test sqlalchemy step, it have a script file

OpenMLDB/steps/test_python.sh

Lines 24 to 25 in af0272c

	test -d /rambuild/ut_zookeeper && rm -rf /rambuild/ut_zookeeper/*
	cp steps/zoo.cfg "$THIRDSRC/zookeeper-3.4.14/conf"

Thanks so much.

also test_python requires you to start the openmldb cluster, so I think you can just add a few coverage lines into test_python.sh for coverage

[mangoGoForward]

Hi @aceforeverd , the CI of python-sdk has passed

[aceforeverd]

Hi @aceforeverd , the CI of python-sdk has passed

nice work. I see the coverage report but saidly it only show files in python/test, I guess that's because it is tested with the packaged openmdlb wheel, not from souce. we may further fix it

[tobegit3hub]

LGTM