[
https://issues.apache.org/jira/browse/BEAM-14347?focusedWorklogId=767706&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-767706
]
ASF GitHub Bot logged work on BEAM-14347:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 08/May/22 19:05
Start Date: 08/May/22 19:05
Worklog Time Spent: 10m
Work Description: lostluck commented on code in PR #17574:
URL: https://github.com/apache/beam/pull/17574#discussion_r867529520
##########
sdks/go/pkg/beam/registration/emitter.go:
##########
@@ -0,0 +1,173 @@
+// Licensed to the Apache Software Foundation (ASF) under one or more
+// contributor license agreements. See the NOTICE file distributed with
+// this work for additional information regarding copyright ownership.
+// The ASF licenses this file to You under the Apache License, Version 2.0
+// (the "License"); you may not use this file except in compliance with
+// the License. You may obtain a copy of the License at
+//
+// http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+package registration
+
+import (
+ "context"
+ "reflect"
+
+ "github.com/apache/beam/sdks/v2/go/pkg/beam/core/runtime/exec"
+ "github.com/apache/beam/sdks/v2/go/pkg/beam/core/typex"
+)
+
+type emit1[T any] struct {
+ n exec.ElementProcessor
+
+ ctx context.Context
+ ws []typex.Window
+ et typex.EventTime
+ value exec.FullValue
Review Comment:
Another thought: We could make a simple emitImpl and iterImpl types that
cover the Init method, and then embed those types, which would deduplicate
things a little bit...
https://go101.org/article/type-embedding.html
But it might not be worthwhile since invoke, and Value() would still need to
be hight level type anyway :/. Hmm.
##########
sdks/go/pkg/beam/registration/registration_test.go:
##########
@@ -155,6 +157,48 @@ func TestRegister_RegistersTypes(t *testing.T) {
}
}
+func TestEmitter1(t *testing.T) {
+ Emitter1[int]()
+ if !exec.IsEmitterRegistered(reflect.TypeOf((*func(int))(nil)).Elem()) {
+
t.Fatalf("exec.IsEmitterRegistered(reflect.TypeOf((*func(int))(nil)).Elem()) =
false, want true")
+ }
+}
+
+func TestEmitter2(t *testing.T) {
+ Emitter2[int, string]()
+ if !exec.IsEmitterRegistered(reflect.TypeOf((*func(int,
string))(nil)).Elem()) {
+ t.Fatalf("exec.IsEmitterRegistered(reflect.TypeOf((*func(int,
string))(nil)).Elem()) = false, want true")
+ }
+}
+
+func TestEmitter2_WithTimestamp(t *testing.T) {
+ Emitter2[typex.EventTime, string]()
+ if !exec.IsEmitterRegistered(reflect.TypeOf((*func(typex.EventTime,
string))(nil)).Elem()) {
+
t.Fatalf("exec.IsEmitterRegistered(reflect.TypeOf((*func(typex.EventTime,
string))(nil)).Elem()) = false, want true")
+ }
+}
+
+func TestEmitter3(t *testing.T) {
+ Emitter3[typex.EventTime, int, string]()
Review Comment:
I just noted coverage is pretty low for the emitters and iterators (12% of
the diff), but since the `exec.make*` calls are unexported, that makes it
slightly more difficult.
I think even instantiating them manually only loses coverage of the factory
closures anyway so we're probably fine if we were to do it that way. This way
we'll be testing *everything* that is instantiated by users too.
Issue Time Tracking
-------------------
Worklog Id: (was: 767706)
Time Spent: 9h 40m (was: 9.5h)
> [Go SDK] Allow users to optimize DoFns with a single generic registration
> function
> ----------------------------------------------------------------------------------
>
> Key: BEAM-14347
> URL: https://issues.apache.org/jira/browse/BEAM-14347
> Project: Beam
> Issue Type: New Feature
> Components: sdk-go
> Reporter: Danny McCormick
> Assignee: Danny McCormick
> Priority: P2
> Time Spent: 9h 40m
> Remaining Estimate: 0h
>
> Right now, to optimize DoFn execution, users have to use the code generator.
> This updates to allow them to use generics instead.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
