* chain nested conditions
* add missing license
* add dependency check to remove unnecessary when expression
* minor improvement to check for the exact condition rather than just the condition task
* make results to store in multi-steps if needed
* add warning messages on oversized parameter
* regenerate tesrs
* fix copy-results format
* optimize the bin packing problem
* add docs and warning test cases
eliminate the usage of ContainerOp in testcases and
update the testcase accordingly, including filenames starting
with `o` to `u`
Signed-off-by: Yihong Wang <yh.wang@ibm.com>
* add test for taskRef-vs-taskSpec params
* fix: taskSpec accepting params
* add license headers
* m: no spaces around "buzz"
* fix: test file custom_task_spec.yaml
* m: follow links on newlines
* fix: big_data_passing test result yamls
* fix: yamls of other tests
* factor out conditions to process params
* fix big data rewrite for regular tasks
* Change to use component to replace containerOp for any sequencer
* Remove context input and change results path
* 1: Update to use component rather than containerOp
2: Fix lint/license issue
* add auto apply custom task CRs before pipelineRun creation
* fix lint
* store loop resource in ordered dict
* Update kfp-admin-guide.md
* sorted pipeline params to avoid randomness
* sorted pipeline params to avoid randomness
* regenerate tests
* convert template string from yaml to json
* regenerate tests
* clean up unnecessary helper function
* resolve merge conflicts
* resolve merge conflicts
* remove unnecessary go mod package
* only opt-in to embed the recursion loop annotations
* only opt-in to embed the recursion loop annotations
* enable opt-in only via package variable
* add example for how to inline recursive loops
* resolve conflicts
* regenerate tests
* make auto generate and apply recursive pipelineloop in annotations as default
* move parameter sorting to the end
* add parameter sorting for recursion loops
* sort recursive task parameters
* add more details on how to opt-out auto apply custom resource
* TODO1: PipelineLoop inlined including nested loops and added tests.
* Supported recursive and parallel tasks inlining.
For recursive tasks, we simply skip them.
For Parallel tasks, relevant tests were added.
* Run all tests against inlined taskSpec and normal taskRefs
* Added CLI Flag disable/enable spec inlining.
* Regenerate E2E test logs for release 0.8.0
* Regerate "golden" log files
* Regenerate "golden" logs after PR #567
* Update list of ignored tests
* Add cond_recur.yaml to list of ignored tests
* Regenerate E2E logs after updating "golden" YAML files of unit tests
* Regenerate E2E logs after enabling auto-strip EOF
* Regenerate "golden" YAML for unit tests
* Rename loop-static CRDs
* Regenerate E2E logs, enable EOF stripping only for E2E test
* add test for artifacts of ops with long names
* add test on correct artifact_items
* fix: stepTemplate generated
* fix: artifact_items name collision
* refactor: mount_path to a separate variable
* add license to the new test file
* m: two free lines after imports in testcase
* m: other linting
* add cel custom task poc
* apply suggestions with minor fixes
* add initial version of plugable cel template
* fix lint
* update staging code for converting containerOp to custom task
* update staging code for converting containerOp to custom task
* remove duplicate cel conditions
* update flip-coin example to use custom task
* update flip-coin example to use custom task
* remove dsl dependencies;
* add POC comments
* fix lint
* fix self-defined CEL custom task mapping
* move custom task condition example to a separate folder
* fix tests
* make the new custom task spec and CEL condition feature as opt-in
* update instructions and generalize custom task images
* only let custom task result comparasion to skip the condition container to provide backward compatibility
* only let custom task result comparasion to skip the condition container to provide backward compatibility
* update example
* update tests
* fix comment
* KFP 1.5.0-rc.0 Rebase
* Resolve backend and API conflicts
* Resolve UI conflicts
* Apply changes from KFP 1.5.0-rc.1
* Resolve backend and API conflicts from RC1
* Resolve UI conflicts from RC1
* Apply changes from KFP 1.5.0-rc.2
* Resolve backend conflicts from RC2
* Build SDK based on kfp RC2 from Github instead of PyPI
* Regenerate unittest's Golden YAML files
* fix nested loop params issue, and ensure multi nested loop works well
* address the issue of ut failed on python3.6
* Trigger CI
* address the issue of ut failed on python3.6
* Use kfp client to run E2E tests
* Update Tekton server and tkn client versions
* Regenerate pipelinerun logs
* Give unique names to pipelineruns resembling file names
* Delete previous pipelineruns
* Mark tests as skipped instead of not generating them
* Regenerate pipeline-loop CRDs, update copyright statements
for f in sdk/python/tests/compiler/testdata/*_pipelineloop_cr*.yaml; do
echo ${f/_pipelineloop_cr*.yaml/.py};
done | sort -u | while read f; do
echo $f; dsl-compile-tekton --py $f --output ${f/.py/.yaml};
done;
make unit_test GENERATE_GOLDEN_YAML=True
* update code to address the review comments
* change the dir name
* change the dir name
* update yaml file after rebase master
* update yaml file to trigger travic
* address comments from Tommy
* Now including logs for 'finally' tasks
* Add env var for SLEEP_BETWEEN_TEST_PHASES
* Add 'retry.yaml' to ignored tests
* Make node_selector more general
* Use pipelinerun status.conditions[].reason instead of .type
* Treat status "Completed" as "Succeeded"
* Tolerate erratic log lines in E2E tests
* Add option to re-run only tests that failed previously
* Add option to exclude/include certain test_cases only
* Add volume_op.yaml to ignored tests as the created PVC
needs to be deleted before/after test run
* Regenerate "golden" log files after recent commits
changed the log output regarding artifacts
* Add option to override expected Tekton Pipeline and CLI
versions, default: 0.14 and 0.10
Closes#229
* fix bugs, update examples, and add tests
* Address comments
* Update readme to a shared prerequisites section
* update instructions to use python commands if possible
* add details on the yaml name produced by the SDK
* Update samples/README.md
Co-authored-by: Animesh Singh <singhan@us.ibm.com>
Co-authored-by: Animesh Singh <singhan@us.ibm.com>
* further big_data_passing fixes after PR #166
* further big_data_passing fixes after PR #166
* address comments from reviewers
* Address comments form Tomcli
Prashant Sharma