chamikaramj commented on a change in pull request #11185: [BEAM-8019] Updates
Python SDK to handle remote SDK coders and preserve tags added by remote SDKs
and propagate restriction coders.
URL: https://github.com/apache/beam/pull/11185#discussion_r403351916
##########
File path: sdks/python/apache_beam/pipeline.py
##########
@@ -1133,29 +1141,67 @@ def from_runner_api(proto, # type:
beam_runner_api_pb2.PTransform
context # type: PipelineContext
):
# type: (...) -> AppliedPTransform
- def is_side_input(tag):
- # type: (str) -> bool
- # As per named_inputs() above.
- return tag.startswith('side')
+
+ if common_urns.primitives.PAR_DO.urn == proto.spec.urn:
+ # Preserving side input tags.
+ from apache_beam.portability.api import beam_runner_api_pb2
+ payload = (
+ proto_utils.parse_Bytes(
+ proto.spec.payload, beam_runner_api_pb2.ParDoPayload))
+ side_input_tags = list(payload.side_inputs.keys())
+ else:
+ side_input_tags = []
main_inputs = [
context.pcollections.get_by_id(id) for tag,
- id in proto.inputs.items() if not is_side_input(tag)
+ id in proto.inputs.items() if tag not in side_input_tags
]
- # Ordering is important here.
- indexed_side_inputs = [
- (get_sideinput_index(tag), context.pcollections.get_by_id(id)) for tag,
- id in proto.inputs.items() if is_side_input(tag)
- ]
- side_inputs = [si for _, si in sorted(indexed_side_inputs)]
+ def is_python_side_input(tag):
+ # type: (str) -> bool
+ # As per named_inputs() above.
+ return re.match(SIDE_INPUT_REGEX, tag)
+
+ uses_python_sideinput_tags = (
+ is_python_side_input(side_input_tags[0]) if side_input_tags else False)
+
+ if uses_python_sideinput_tags:
+ # Ordering is important here.
+ # TODO(BEAM-9635): use key, value pairs instead of depending on tags with
+ # index as a suffix.
+ indexed_side_inputs = [
+ (get_sideinput_index(tag), context.pcollections.get_by_id(id))
+ for tag,
+ id in proto.inputs.items() if tag in side_input_tags
+ ]
+ side_inputs = [si for _, si in sorted(indexed_side_inputs)]
+ else:
+ side_inputs = [
+ context.pcollections.get_by_id(id) for tag,
+ id in proto.inputs.items() if tag in side_input_tags
+ ]
+
transform = ptransform.PTransform.from_runner_api(proto, context)
+ if isinstance(transform, RunnerAPIPTransformHolder):
+ # For external transforms that are ParDos, we have to set side-inputs
+ # manually and preserve input tags.
+ transform.side_inputs = [pvalue.AsMultiMap(pc) for pc in side_inputs]
+ input_tags_to_preserve = {
+ context.pcollections.get_by_id(id): tag
+ for tag,
+ id in proto.inputs.items()
+ }
+ else:
Review comment:
That's done in respective from_runner_api methods. For example,
https://github.com/apache/beam/blob/master/sdks/python/apache_beam/transforms/core.py#L1361
But RunnerAPIPTransformHolder is a holder type that is constructed directly
so setting it here seemed like the best option (we may be able to move this to
constructor but that would require passing in additional parameters to get
access to inputs etc. since using Python only
pvalue.SideInputData.from_runner_api is not an option).
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
