This is an automated email from the ASF dual-hosted git repository.
mobuchowski pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/airflow.git
The following commit(s) were added to refs/heads/main by this push:
new 57dab1303c typo: wrong OpenLineage facet key in spec (#39782)
57dab1303c is described below
commit 57dab1303c5850b9b2ff81e64cbbe5bd4d1032f9
Author: Kacper Muda <[email protected]>
AuthorDate: Thu May 23 15:53:07 2024 +0200
typo: wrong OpenLineage facet key in spec (#39782)
Signed-off-by: Kacper Muda <[email protected]>
---
.../providers/google/cloud/openlineage/BigQueryErrorRunFacet.json | 2 +-
airflow/providers/google/cloud/openlineage/utils.py | 6 +-----
2 files changed, 2 insertions(+), 6 deletions(-)
diff --git
a/airflow/providers/google/cloud/openlineage/BigQueryErrorRunFacet.json
b/airflow/providers/google/cloud/openlineage/BigQueryErrorRunFacet.json
index f85756ee92..3213f9b8b2 100644
--- a/airflow/providers/google/cloud/openlineage/BigQueryErrorRunFacet.json
+++ b/airflow/providers/google/cloud/openlineage/BigQueryErrorRunFacet.json
@@ -23,7 +23,7 @@
},
"type": "object",
"properties": {
- "bigQueryJob": {
+ "bigQuery_error": {
"$ref": "#/$defs/BigQueryErrorRunFacet"
}
}
diff --git a/airflow/providers/google/cloud/openlineage/utils.py
b/airflow/providers/google/cloud/openlineage/utils.py
index 9f18aac9f4..fb0d4c663b 100644
--- a/airflow/providers/google/cloud/openlineage/utils.py
+++ b/airflow/providers/google/cloud/openlineage/utils.py
@@ -269,7 +269,7 @@ class _BigQueryOpenLineageMixin:
if hasattr(self, "log"):
self.log.warning("Cannot retrieve job details from
BigQuery.Client. %s", e, exc_info=True)
exception_msg = traceback.format_exc()
- # TODO: remove ErrorMessageRunFacet in next release
+ # TODO: remove BigQueryErrorRunFacet in next release
run_facets.update(
{
"errorMessage": ErrorMessageRunFacet(
@@ -282,10 +282,6 @@ class _BigQueryOpenLineageMixin:
}
)
deduplicated_outputs = self._deduplicate_outputs(outputs)
- # For complex scripts there can be multiple outputs - in that case
keep them all in `outputs` and
- # leave the `output` empty to avoid providing misleading information.
When the script has a single
- # output (f.e. a single statement with some variable declarations),
treat it as a regular non-script
- # job and put the output in `output` as an addition to new `outputs`.
`output` is deprecated.
return inputs, deduplicated_outputs, run_facets
def _deduplicate_outputs(self, outputs: list[Dataset | None]) ->
list[Dataset]:
