Support upload outputs and use needs context (#133)

See [Example usage of the needs context](https://docs.github.com/en/actions/learn-github-actions/contexts#example-usage-of-the-needs-context).

Related to:
- [actions-proto-def #5](https://gitea.com/gitea/actions-proto-def/pulls/5)
- [gitea #24230](https://github.com/go-gitea/gitea/pull/24230)

Reviewed-on: https://gitea.com/gitea/act_runner/pulls/133
Reviewed-by: Lunny Xiao <xiaolunwen@gmail.com>
Reviewed-by: Zettat123 <zettat123@noreply.gitea.io>
Co-authored-by: Jason Song <i@wolfogre.com>
Co-committed-by: Jason Song <i@wolfogre.com>

internal/app/run/runner.go

@@ -4,7 +4,6 @@
 package run
 
 import (
-	"bytes"
 	"context"
 	"encoding/json"
 	"fmt"
@@ -112,16 +111,11 @@ func (r *Runner) run(ctx context.Context, task *runnerv1.Task, reporter *report.
 
 	reporter.Logf("%s(version:%s) received task %v of job %v, be triggered by event: %s", r.name, ver.Version(), task.Id, task.Context.Fields["job"].GetStringValue(), task.Context.Fields["event_name"].GetStringValue())
 
-	workflow, err := model.ReadWorkflow(bytes.NewReader(task.WorkflowPayload))
+	workflow, jobID, err := generateWorkflow(task)
 	if err != nil {
 		return err
 	}
 
-	jobIDs := workflow.GetJobIDs()
-	if len(jobIDs) != 1 {
-		return fmt.Errorf("multiple jobs found: %v", jobIDs)
-	}
-	jobID := jobIDs[0]
 	plan, err := model.CombineWorkflowPlanner(workflow).PlanJob(jobID)
 	if err != nil {
 		return err
@@ -209,5 +203,7 @@ func (r *Runner) run(ctx context.Context, task *runnerv1.Task, reporter *report.
 	// add logger recorders
 	ctx = common.WithLoggerHook(ctx, reporter)
 
-	return executor(ctx)
+	execErr := executor(ctx)
+	reporter.SetOutputs(job.Outputs)
+	return execErr
 }

internal/app/run/workflow.go

@@ -0,0 +1,54 @@
+// Copyright 2023 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package run
+
+import (
+	"bytes"
+	"fmt"
+	"sort"
+	"strings"
+
+	runnerv1 "code.gitea.io/actions-proto-go/runner/v1"
+	"github.com/nektos/act/pkg/model"
+	"gopkg.in/yaml.v3"
+)
+
+func generateWorkflow(task *runnerv1.Task) (*model.Workflow, string, error) {
+	workflow, err := model.ReadWorkflow(bytes.NewReader(task.WorkflowPayload))
+	if err != nil {
+		return nil, "", err
+	}
+
+	jobIDs := workflow.GetJobIDs()
+	if len(jobIDs) != 1 {
+		return nil, "", fmt.Errorf("multiple jobs found: %v", jobIDs)
+	}
+	jobID := jobIDs[0]
+
+	needJobIDs := make([]string, 0, len(task.Needs))
+	for id, need := range task.Needs {
+		needJobIDs = append(needJobIDs, id)
+		needJob := &model.Job{
+			Outputs: need.Outputs,
+			Result:  strings.ToLower(strings.TrimPrefix(need.Result.String(), "RESULT_")),
+		}
+		workflow.Jobs[id] = needJob
+	}
+	sort.Strings(needJobIDs)
+
+	rawNeeds := yaml.Node{
+		Kind:    yaml.SequenceNode,
+		Content: make([]*yaml.Node, 0, len(needJobIDs)),
+	}
+	for _, id := range needJobIDs {
+		rawNeeds.Content = append(rawNeeds.Content, &yaml.Node{
+			Kind:  yaml.ScalarNode,
+			Value: id,
+		})
+	}
+
+	workflow.Jobs[jobID].RawNeeds = rawNeeds
+
+	return workflow, jobID, nil
+}

internal/app/run/workflow_test.go

@@ -0,0 +1,74 @@
+// Copyright 2023 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package run
+
+import (
+	"testing"
+
+	runnerv1 "code.gitea.io/actions-proto-go/runner/v1"
+	"github.com/nektos/act/pkg/model"
+	"github.com/stretchr/testify/require"
+	"gotest.tools/v3/assert"
+)
+
+func Test_generateWorkflow(t *testing.T) {
+	type args struct {
+		task *runnerv1.Task
+	}
+	tests := []struct {
+		name    string
+		args    args
+		assert  func(t *testing.T, wf *model.Workflow)
+		want1   string
+		wantErr bool
+	}{
+		{
+			name: "has needs",
+			args: args{
+				task: &runnerv1.Task{
+					WorkflowPayload: []byte(`
+name: Build and deploy
+on: push
+
+jobs:
+  job9:
+    needs: build
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v3
+      - run: ./deploy --build ${{ needs.job1.outputs.output1 }}
+      - run: ./deploy --build ${{ needs.job2.outputs.output2 }}
+`),
+					Needs: map[string]*runnerv1.TaskNeed{
+						"job1": {
+							Outputs: map[string]string{
+								"output1": "output1 value",
+							},
+							Result: runnerv1.Result_RESULT_SUCCESS,
+						},
+						"job2": {
+							Outputs: map[string]string{
+								"output2": "output2 value",
+							},
+							Result: runnerv1.Result_RESULT_SUCCESS,
+						},
+					},
+				},
+			},
+			assert: func(t *testing.T, wf *model.Workflow) {
+				assert.DeepEqual(t, wf.GetJob("job9").Needs(), []string{"job1", "job2"})
+			},
+			want1:   "job9",
+			wantErr: false,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, got1, err := generateWorkflow(tt.args.task)
+			require.NoError(t, err)
+			tt.assert(t, got)
+			assert.Equal(t, got1, tt.want1)
+		})
+	}
+}

internal/pkg/report/reporter.go

@@ -31,8 +31,10 @@ type Reporter struct {
 	logOffset   int
 	logRows     []*runnerv1.LogRow
 	logReplacer *strings.Replacer
+
 	state   *runnerv1.TaskState
-	stateM      sync.RWMutex
+	stateMu sync.RWMutex
+	outputs sync.Map
 }
 
 func NewReporter(ctx context.Context, cancel context.CancelFunc, client client.Client, task *runnerv1.Task) *Reporter {
@@ -56,8 +58,8 @@ func NewReporter(ctx context.Context, cancel context.CancelFunc, client client.C
 }
 
 func (r *Reporter) ResetSteps(l int) {
-	r.stateM.Lock()
+	r.stateMu.Lock()
-	defer r.stateM.Unlock()
+	defer r.stateMu.Unlock()
 	for i := 0; i < l; i++ {
 		r.state.Steps = append(r.state.Steps, &runnerv1.StepState{
 			Id: int64(i),
@@ -70,8 +72,8 @@ func (r *Reporter) Levels() []log.Level {
 }
 
 func (r *Reporter) Fire(entry *log.Entry) error {
-	r.stateM.Lock()
+	r.stateMu.Lock()
-	defer r.stateM.Unlock()
+	defer r.stateMu.Unlock()
 
 	log.WithFields(entry.Data).Trace(entry.Message)
 
@@ -155,9 +157,13 @@ func (r *Reporter) RunDaemon() {
 }
 
 func (r *Reporter) Logf(format string, a ...interface{}) {
-	r.stateM.Lock()
+	r.stateMu.Lock()
-	defer r.stateM.Unlock()
+	defer r.stateMu.Unlock()
 
+	r.logf(format, a...)
+}
+
+func (r *Reporter) logf(format string, a ...interface{}) {
 	if !r.duringSteps() {
 		r.logRows = append(r.logRows, &runnerv1.LogRow{
 			Time:    timestamppb.Now(),
@@ -166,10 +172,30 @@ func (r *Reporter) Logf(format string, a ...interface{}) {
 	}
 }
 
+func (r *Reporter) SetOutputs(outputs map[string]string) {
+	r.stateMu.Lock()
+	defer r.stateMu.Unlock()
+
+	for k, v := range outputs {
+		if len(k) > 255 {
+			r.logf("ignore output because the key is too long: %q", k)
+			continue
+		}
+		if l := len(v); l > 1024*1024 {
+			log.Println("ignore output because the value is too long:", k, l)
+			r.logf("ignore output because the value %q is too long: %d", k, l)
+		}
+		if _, ok := r.outputs.Load(k); ok {
+			continue
+		}
+		r.outputs.Store(k, v)
+	}
+}
+
 func (r *Reporter) Close(lastWords string) error {
 	r.closed = true
 
-	r.stateM.Lock()
+	r.stateMu.Lock()
 	if r.state.Result == runnerv1.Result_RESULT_UNSPECIFIED {
 		if lastWords == "" {
 			lastWords = "Early termination"
@@ -191,7 +217,7 @@ func (r *Reporter) Close(lastWords string) error {
 			Content: lastWords,
 		})
 	}
-	r.stateM.Unlock()
+	r.stateMu.Unlock()
 
 	return retry.Do(func() error {
 		if err := r.ReportLog(true); err != nil {
@@ -205,9 +231,9 @@ func (r *Reporter) ReportLog(noMore bool) error {
 	r.clientM.Lock()
 	defer r.clientM.Unlock()
 
-	r.stateM.RLock()
+	r.stateMu.RLock()
 	rows := r.logRows
-	r.stateM.RUnlock()
+	r.stateMu.RUnlock()
 
 	resp, err := r.client.UpdateLog(r.ctx, connect.NewRequest(&runnerv1.UpdateLogRequest{
 		TaskId: r.state.Id,
@@ -224,10 +250,10 @@ func (r *Reporter) ReportLog(noMore bool) error {
 		return fmt.Errorf("submitted logs are lost")
 	}
 
-	r.stateM.Lock()
+	r.stateMu.Lock()
 	r.logRows = r.logRows[ack-r.logOffset:]
 	r.logOffset = ack
-	r.stateM.Unlock()
+	r.stateMu.Unlock()
 
 	if noMore && ack < r.logOffset+len(rows) {
 		return fmt.Errorf("not all logs are submitted")
@@ -240,21 +266,45 @@ func (r *Reporter) ReportState() error {
 	r.clientM.Lock()
 	defer r.clientM.Unlock()
 
-	r.stateM.RLock()
+	r.stateMu.RLock()
 	state := proto.Clone(r.state).(*runnerv1.TaskState)
-	r.stateM.RUnlock()
+	r.stateMu.RUnlock()
+
+	outputs := make(map[string]string)
+	r.outputs.Range(func(k, v interface{}) bool {
+		if val, ok := v.(string); ok {
+			outputs[k.(string)] = val
+		}
+		return true
+	})
 
 	resp, err := r.client.UpdateTask(r.ctx, connect.NewRequest(&runnerv1.UpdateTaskRequest{
 		State:   state,
+		Outputs: outputs,
 	}))
 	if err != nil {
 		return err
 	}
 
+	for _, k := range resp.Msg.SentOutputs {
+		r.outputs.Store(k, struct{}{})
+	}
+
 	if resp.Msg.State != nil && resp.Msg.State.Result == runnerv1.Result_RESULT_CANCELLED {
 		r.cancel()
 	}
 
+	var noSent []string
+	r.outputs.Range(func(k, v interface{}) bool {
+		if _, ok := v.(string); ok {
+			noSent = append(noSent, k.(string))
+		}
+		return true
+	})
+	if len(noSent) > 0 {
+		return fmt.Errorf("there are still outputs that have not been sent: %v", noSent)
+	}
+
 	return nil
 }