1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
|
package executor
import (
"context"
"fmt"
"log/slog"
"sync"
"time"
"github.com/thepeterstone/claudomator/internal/storage"
"github.com/thepeterstone/claudomator/internal/task"
"github.com/google/uuid"
)
// Runner executes a single task and returns the result.
type Runner interface {
Run(ctx context.Context, t *task.Task, exec *storage.Execution) error
}
// Pool manages a bounded set of concurrent task workers.
type Pool struct {
maxConcurrent int
runner Runner
store *storage.DB
logger *slog.Logger
mu sync.Mutex
active int
resultCh chan *Result
Questions *QuestionRegistry
}
// Result is emitted when a task execution completes.
type Result struct {
TaskID string
Execution *storage.Execution
Err error
}
func NewPool(maxConcurrent int, runner Runner, store *storage.DB, logger *slog.Logger) *Pool {
if maxConcurrent < 1 {
maxConcurrent = 1
}
return &Pool{
maxConcurrent: maxConcurrent,
runner: runner,
store: store,
logger: logger,
resultCh: make(chan *Result, maxConcurrent*2),
Questions: NewQuestionRegistry(),
}
}
// Submit dispatches a task for execution. Blocks if pool is at capacity.
func (p *Pool) Submit(ctx context.Context, t *task.Task) error {
p.mu.Lock()
if p.active >= p.maxConcurrent {
active := p.active
max := p.maxConcurrent
p.mu.Unlock()
return fmt.Errorf("executor pool at capacity (%d/%d)", active, max)
}
p.active++
p.mu.Unlock()
go p.execute(ctx, t)
return nil
}
// Results returns the channel for reading execution results.
func (p *Pool) Results() <-chan *Result {
return p.resultCh
}
// ActiveCount returns the number of currently running tasks.
func (p *Pool) ActiveCount() int {
p.mu.Lock()
defer p.mu.Unlock()
return p.active
}
func (p *Pool) execute(ctx context.Context, t *task.Task) {
defer func() {
p.mu.Lock()
p.active--
p.mu.Unlock()
}()
// Wait for all dependencies to complete before starting execution.
if len(t.DependsOn) > 0 {
if err := p.waitForDependencies(ctx, t); err != nil {
now := time.Now().UTC()
exec := &storage.Execution{
ID: uuid.New().String(),
TaskID: t.ID,
StartTime: now,
EndTime: now,
Status: "FAILED",
ErrorMsg: err.Error(),
}
if createErr := p.store.CreateExecution(exec); createErr != nil {
p.logger.Error("failed to create execution record", "error", createErr)
}
p.store.UpdateTaskState(t.ID, task.StateFailed)
p.resultCh <- &Result{TaskID: t.ID, Execution: exec, Err: err}
return
}
}
execID := uuid.New().String()
exec := &storage.Execution{
ID: execID,
TaskID: t.ID,
StartTime: time.Now().UTC(),
Status: "RUNNING",
}
// Record execution start.
if err := p.store.CreateExecution(exec); err != nil {
p.logger.Error("failed to create execution record", "error", err)
}
if err := p.store.UpdateTaskState(t.ID, task.StateRunning); err != nil {
p.logger.Error("failed to update task state", "error", err)
}
// Apply task timeout.
var cancel context.CancelFunc
if t.Timeout.Duration > 0 {
ctx, cancel = context.WithTimeout(ctx, t.Timeout.Duration)
} else {
ctx, cancel = context.WithCancel(ctx)
}
defer cancel()
// Run the task.
err := p.runner.Run(ctx, t, exec)
exec.EndTime = time.Now().UTC()
if err != nil {
if ctx.Err() == context.DeadlineExceeded {
exec.Status = "TIMED_OUT"
exec.ErrorMsg = "execution timed out"
p.store.UpdateTaskState(t.ID, task.StateTimedOut)
} else if ctx.Err() == context.Canceled {
exec.Status = "CANCELLED"
exec.ErrorMsg = "execution cancelled"
p.store.UpdateTaskState(t.ID, task.StateCancelled)
} else {
exec.Status = "FAILED"
exec.ErrorMsg = err.Error()
p.store.UpdateTaskState(t.ID, task.StateFailed)
}
} else {
if t.ParentTaskID == "" {
exec.Status = "READY"
p.store.UpdateTaskState(t.ID, task.StateReady)
} else {
exec.Status = "COMPLETED"
p.store.UpdateTaskState(t.ID, task.StateCompleted)
}
}
if updateErr := p.store.UpdateExecution(exec); updateErr != nil {
p.logger.Error("failed to update execution", "error", updateErr)
}
p.resultCh <- &Result{TaskID: t.ID, Execution: exec, Err: err}
}
// terminalFailureStates are dependency states that cause the waiting task to fail immediately.
var terminalFailureStates = map[task.State]bool{
task.StateFailed: true,
task.StateTimedOut: true,
task.StateCancelled: true,
task.StateBudgetExceeded: true,
}
// waitForDependencies polls storage until all tasks in t.DependsOn reach COMPLETED,
// or until a dependency enters a terminal failure state or the context is cancelled.
func (p *Pool) waitForDependencies(ctx context.Context, t *task.Task) error {
for {
allDone := true
for _, depID := range t.DependsOn {
dep, err := p.store.GetTask(depID)
if err != nil {
return fmt.Errorf("dependency %q not found: %w", depID, err)
}
if dep.State == task.StateCompleted {
continue
}
if terminalFailureStates[dep.State] {
return fmt.Errorf("dependency %q ended in state %s", depID, dep.State)
}
allDone = false
}
if allDone {
return nil
}
select {
case <-ctx.Done():
return ctx.Err()
case <-time.After(5 * time.Second):
}
}
}
|
