mirror of
https://github.com/coder/coder.git
synced 2026-06-03 21:18:24 +00:00
Thomas Kosiewski
517 lines
16 KiB
Go
517 lines
16 KiB
Go
package chatdebug_test
|
|
|
|
import (
|
|
"encoding/json"
|
|
"testing"
|
|
"time"
|
|
"unicode/utf8"
|
|
|
|
"charm.land/fantasy"
|
|
"github.com/google/uuid"
|
|
"github.com/sqlc-dev/pqtype"
|
|
"github.com/stretchr/testify/require"
|
|
|
|
"github.com/coder/coder/v2/coderd/database"
|
|
"github.com/coder/coder/v2/coderd/x/chatd/chatdebug"
|
|
)
|
|
|
|
func TestTruncateLabel(t *testing.T) {
|
|
t.Parallel()
|
|
|
|
tests := []struct {
|
|
name string
|
|
input string
|
|
maxLen int
|
|
want string
|
|
}{
|
|
{name: "Empty", input: "", maxLen: 10, want: ""},
|
|
{name: "WhitespaceOnly", input: " \t\n ", maxLen: 10, want: ""},
|
|
{name: "ShortText", input: "hello world", maxLen: 20, want: "hello world"},
|
|
{name: "ExactLength", input: "abcde", maxLen: 5, want: "abcde"},
|
|
{name: "LongTextTruncated", input: "abcdefghij", maxLen: 5, want: "abcd…"},
|
|
{name: "NegativeMaxLen", input: "hello", maxLen: -1, want: ""},
|
|
{name: "ZeroMaxLen", input: "hello", maxLen: 0, want: ""},
|
|
{name: "SingleRuneLimit", input: "hello", maxLen: 1, want: "…"},
|
|
{name: "MultipleWhitespaceRuns", input: " hello world \t again ", maxLen: 100, want: "hello world again"},
|
|
{name: "UnicodeRunes", input: "こんにちは世界", maxLen: 3, want: "こん…"},
|
|
}
|
|
|
|
for _, tc := range tests {
|
|
t.Run(tc.name, func(t *testing.T) {
|
|
t.Parallel()
|
|
got := chatdebug.TruncateLabel(tc.input, tc.maxLen)
|
|
require.Equal(t, tc.want, got)
|
|
require.LessOrEqual(t, utf8.RuneCountInString(got), max(tc.maxLen, 0))
|
|
})
|
|
}
|
|
}
|
|
|
|
func TestSeedSummary(t *testing.T) {
|
|
t.Parallel()
|
|
|
|
t.Run("NonEmptyLabel", func(t *testing.T) {
|
|
t.Parallel()
|
|
got := chatdebug.SeedSummary("hello world")
|
|
require.Equal(t, map[string]any{"first_message": "hello world"}, got)
|
|
})
|
|
|
|
t.Run("EmptyLabel", func(t *testing.T) {
|
|
t.Parallel()
|
|
got := chatdebug.SeedSummary("")
|
|
require.Nil(t, got)
|
|
})
|
|
}
|
|
|
|
func TestExtractFirstUserText(t *testing.T) {
|
|
t.Parallel()
|
|
|
|
t.Run("EmptyPrompt", func(t *testing.T) {
|
|
t.Parallel()
|
|
got := chatdebug.ExtractFirstUserText(fantasy.Prompt{})
|
|
require.Equal(t, "", got)
|
|
})
|
|
|
|
t.Run("NoUserMessages", func(t *testing.T) {
|
|
t.Parallel()
|
|
prompt := fantasy.Prompt{
|
|
{
|
|
Role: fantasy.MessageRoleSystem,
|
|
Content: []fantasy.MessagePart{fantasy.TextPart{Text: "system"}},
|
|
},
|
|
{
|
|
Role: fantasy.MessageRoleAssistant,
|
|
Content: []fantasy.MessagePart{fantasy.TextPart{Text: "assistant"}},
|
|
},
|
|
}
|
|
got := chatdebug.ExtractFirstUserText(prompt)
|
|
require.Equal(t, "", got)
|
|
})
|
|
|
|
t.Run("FirstUserMessageMixedParts", func(t *testing.T) {
|
|
t.Parallel()
|
|
prompt := fantasy.Prompt{
|
|
{
|
|
Role: fantasy.MessageRoleUser,
|
|
Content: []fantasy.MessagePart{
|
|
fantasy.TextPart{Text: "hello "},
|
|
fantasy.FilePart{Filename: "test.png"},
|
|
fantasy.TextPart{Text: "world"},
|
|
},
|
|
},
|
|
}
|
|
got := chatdebug.ExtractFirstUserText(prompt)
|
|
require.Equal(t, "hello world", got)
|
|
})
|
|
|
|
t.Run("MultipleUserMessagesReturnsFirst", func(t *testing.T) {
|
|
t.Parallel()
|
|
prompt := fantasy.Prompt{
|
|
{
|
|
Role: fantasy.MessageRoleSystem,
|
|
Content: []fantasy.MessagePart{fantasy.TextPart{Text: "system"}},
|
|
},
|
|
{
|
|
Role: fantasy.MessageRoleUser,
|
|
Content: []fantasy.MessagePart{fantasy.TextPart{Text: "first"}},
|
|
},
|
|
{
|
|
Role: fantasy.MessageRoleUser,
|
|
Content: []fantasy.MessagePart{fantasy.TextPart{Text: "second"}},
|
|
},
|
|
}
|
|
got := chatdebug.ExtractFirstUserText(prompt)
|
|
require.Equal(t, "first", got)
|
|
})
|
|
}
|
|
|
|
func TestService_AggregateRunSummary(t *testing.T) {
|
|
t.Parallel()
|
|
|
|
t.Run("NilRunID", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, uuid.Nil, nil)
|
|
require.NoError(t, err)
|
|
require.Nil(t, got)
|
|
})
|
|
|
|
t.Run("ZeroSteps", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
// No steps created. Call with a base summary containing
|
|
// first_message so we can verify it is preserved.
|
|
base := map[string]any{"first_message": "hello world"}
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, base)
|
|
require.NoError(t, err)
|
|
require.Equal(t, "hello world", got["first_message"])
|
|
require.EqualValues(t, 0, got["step_count"])
|
|
require.EqualValues(t, int64(0), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(0), got["total_output_tokens"])
|
|
require.NotContains(t, got, "total_reasoning_tokens")
|
|
require.NotContains(t, got, "total_cache_creation_tokens")
|
|
require.NotContains(t, got, "total_cache_read_tokens")
|
|
require.NotContains(t, got, "has_error")
|
|
require.NotContains(t, got, "endpoint_label")
|
|
})
|
|
|
|
t.Run("NilBaseSummary", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
// Create a step with usage.
|
|
step := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step.ID, 10, 5, 0, 0)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
require.NotNil(t, got)
|
|
require.EqualValues(t, 1, got["step_count"])
|
|
require.EqualValues(t, int64(10), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(5), got["total_output_tokens"])
|
|
})
|
|
|
|
t.Run("PreservesFirstMessage", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step.ID, 20, 10, 0, 0)
|
|
|
|
base := map[string]any{"first_message": "hello world"}
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, base)
|
|
require.NoError(t, err)
|
|
require.Equal(t, "hello world", got["first_message"])
|
|
require.EqualValues(t, 1, got["step_count"])
|
|
require.EqualValues(t, int64(20), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(10), got["total_output_tokens"])
|
|
})
|
|
|
|
t.Run("ClearsStaleDerivedFields", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step.ID, 10, 5, 0, 0)
|
|
|
|
base := map[string]any{
|
|
"first_message": "hello world",
|
|
"step_count": 9,
|
|
"total_input_tokens": 999,
|
|
"total_output_tokens": 888,
|
|
"total_reasoning_tokens": 777,
|
|
"total_cache_creation_tokens": 100,
|
|
"total_cache_read_tokens": 200,
|
|
"has_error": true,
|
|
"endpoint_label": "POST /stale",
|
|
}
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, base)
|
|
require.NoError(t, err)
|
|
require.Equal(t, "hello world", got["first_message"])
|
|
require.EqualValues(t, 1, got["step_count"])
|
|
require.EqualValues(t, int64(10), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(5), got["total_output_tokens"])
|
|
// Stale reasoning tokens must be cleared because the step
|
|
// has zero reasoning tokens.
|
|
require.NotContains(t, got, "total_reasoning_tokens")
|
|
require.NotContains(t, got, "total_cache_creation_tokens")
|
|
require.NotContains(t, got, "total_cache_read_tokens")
|
|
// has_error must be cleared because the step is not in error
|
|
// status and has no error payload.
|
|
require.NotContains(t, got, "has_error")
|
|
require.NotContains(t, got, "endpoint_label")
|
|
})
|
|
|
|
t.Run("RecomputesHasErrorAndCompletedEndpointLabel", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step1 := createTestStep(t, fixture, run.ID)
|
|
_, err := fixture.svc.UpdateStep(fixture.ctx, chatdebug.UpdateStepParams{
|
|
ID: step1.ID,
|
|
ChatID: fixture.chat.ID,
|
|
Status: chatdebug.StatusError,
|
|
Attempts: []chatdebug.Attempt{{
|
|
Number: 1,
|
|
Status: "failed",
|
|
Method: "POST",
|
|
Path: "/failed",
|
|
}},
|
|
})
|
|
require.NoError(t, err)
|
|
|
|
step2 := createTestStepN(t, fixture, run.ID, 2)
|
|
_, err = fixture.svc.UpdateStep(fixture.ctx, chatdebug.UpdateStepParams{
|
|
ID: step2.ID,
|
|
ChatID: fixture.chat.ID,
|
|
Status: chatdebug.StatusCompleted,
|
|
Attempts: []chatdebug.Attempt{{
|
|
Number: 1,
|
|
Status: "completed",
|
|
Method: "POST",
|
|
Path: "/v1/messages",
|
|
}},
|
|
})
|
|
require.NoError(t, err)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
require.Equal(t, true, got["has_error"])
|
|
require.Equal(t, "POST /v1/messages", got["endpoint_label"])
|
|
})
|
|
|
|
t.Run("EndpointLabelPathOnlyWhenMethodEmpty", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step := createTestStep(t, fixture, run.ID)
|
|
_, err := fixture.svc.UpdateStep(fixture.ctx, chatdebug.UpdateStepParams{
|
|
ID: step.ID,
|
|
ChatID: fixture.chat.ID,
|
|
Status: chatdebug.StatusCompleted,
|
|
Attempts: []chatdebug.Attempt{{
|
|
Number: 1,
|
|
Status: "completed",
|
|
Method: "",
|
|
Path: "/v1/messages",
|
|
}},
|
|
})
|
|
require.NoError(t, err)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
require.Equal(t, "/v1/messages", got["endpoint_label"],
|
|
"endpoint_label should be path-only when method is empty")
|
|
})
|
|
|
|
t.Run("InterruptedStepWithErrorExcludedFromHasError", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
// An interrupted step with a real error payload should NOT
|
|
// trigger has_error. Interrupted means user-initiated
|
|
// cancellation (e.g. clicking Stop).
|
|
step := createTestStep(t, fixture, run.ID)
|
|
_, err := fixture.svc.UpdateStep(fixture.ctx, chatdebug.UpdateStepParams{
|
|
ID: step.ID,
|
|
ChatID: fixture.chat.ID,
|
|
Status: chatdebug.StatusInterrupted,
|
|
Error: map[string]any{"message": "user canceled"},
|
|
})
|
|
require.NoError(t, err)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
require.NotContains(t, got, "has_error",
|
|
"interrupted steps should not trigger has_error even with error payload")
|
|
})
|
|
|
|
t.Run("MultipleStepsSumTokens", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step1 := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step1.ID, 10, 5, 2, 3)
|
|
|
|
step2 := createTestStepN(t, fixture, run.ID, 2)
|
|
updateTestStepWithUsage(t, fixture, step2.ID, 15, 7, 1, 4)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
require.EqualValues(t, 2, got["step_count"])
|
|
require.EqualValues(t, int64(25), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(12), got["total_output_tokens"])
|
|
require.EqualValues(t, int64(3), got["total_cache_creation_tokens"])
|
|
require.EqualValues(t, int64(7), got["total_cache_read_tokens"])
|
|
})
|
|
|
|
t.Run("StepWithNilUsageContributesZeroTokens", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
// Step with usage.
|
|
step1 := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step1.ID, 10, 5, 0, 0)
|
|
|
|
// Step without usage (just complete it, no usage).
|
|
step2 := createTestStepN(t, fixture, run.ID, 2)
|
|
_, err := fixture.svc.UpdateStep(fixture.ctx, chatdebug.UpdateStepParams{
|
|
ID: step2.ID,
|
|
ChatID: fixture.chat.ID,
|
|
Status: chatdebug.StatusCompleted,
|
|
})
|
|
require.NoError(t, err)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
// Both steps are counted even though one has no usage.
|
|
require.EqualValues(t, 2, got["step_count"])
|
|
require.EqualValues(t, int64(10), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(5), got["total_output_tokens"])
|
|
})
|
|
|
|
t.Run("ZeroCacheTotalsOmitCacheFields", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step.ID, 10, 5, 0, 0)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
_, hasCacheCreation := got["total_cache_creation_tokens"]
|
|
_, hasCacheRead := got["total_cache_read_tokens"]
|
|
require.False(t, hasCacheCreation,
|
|
"cache creation tokens should be omitted when zero")
|
|
require.False(t, hasCacheRead,
|
|
"cache read tokens should be omitted when zero")
|
|
})
|
|
|
|
t.Run("ReasoningTokensSummedAcrossSteps", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step1 := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithFullUsage(t, fixture, step1.ID, 10, 5, 20, 0, 0)
|
|
|
|
step2 := createTestStepN(t, fixture, run.ID, 2)
|
|
updateTestStepWithFullUsage(t, fixture, step2.ID, 15, 7, 30, 0, 0)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
require.EqualValues(t, 2, got["step_count"])
|
|
require.EqualValues(t, int64(25), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(12), got["total_output_tokens"])
|
|
require.EqualValues(t, int64(50), got["total_reasoning_tokens"],
|
|
"reasoning tokens should be summed across steps")
|
|
})
|
|
|
|
t.Run("ZeroReasoningTokensOmitsField", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
step := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithFullUsage(t, fixture, step.ID, 10, 5, 0, 0, 0)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err)
|
|
_, hasReasoning := got["total_reasoning_tokens"]
|
|
require.False(t, hasReasoning,
|
|
"reasoning tokens should be omitted when zero")
|
|
})
|
|
|
|
t.Run("MalformedUsageJSONSkipped", func(t *testing.T) {
|
|
t.Parallel()
|
|
fixture := newFixture(t)
|
|
run := createRun(t, fixture)
|
|
|
|
// Step 1 has valid usage and should contribute to totals.
|
|
step1 := createTestStep(t, fixture, run.ID)
|
|
updateTestStepWithUsage(t, fixture, step1.ID, 10, 5, 0, 0)
|
|
|
|
// Step 2 is stamped with structurally-valid JSONB that cannot
|
|
// unmarshal into fantasy.Usage (string where int64 is
|
|
// expected). Write directly through the store so the jsonb
|
|
// cast succeeds while the Go unmarshal fails, exercising the
|
|
// "skipping malformed step usage JSON" log-and-continue path.
|
|
step2 := createTestStepN(t, fixture, run.ID, 2)
|
|
_, err := fixture.db.UpdateChatDebugStep(fixture.ctx, database.UpdateChatDebugStepParams{
|
|
ID: step2.ID,
|
|
ChatID: fixture.chat.ID,
|
|
Usage: pqtype.NullRawMessage{
|
|
RawMessage: json.RawMessage(`{"input_tokens":"not-a-number"}`),
|
|
Valid: true,
|
|
},
|
|
Now: time.Now(),
|
|
})
|
|
require.NoError(t, err)
|
|
|
|
got, err := fixture.svc.AggregateRunSummary(fixture.ctx, run.ID, nil)
|
|
require.NoError(t, err,
|
|
"malformed usage JSON must be skipped, not surfaced as an error")
|
|
|
|
// Both steps are counted, but only step1's tokens contribute.
|
|
require.EqualValues(t, 2, got["step_count"])
|
|
require.EqualValues(t, int64(10), got["total_input_tokens"])
|
|
require.EqualValues(t, int64(5), got["total_output_tokens"])
|
|
})
|
|
}
|
|
|
|
// createTestStep is a thin helper that creates a debug step with
|
|
// step number 1 for the given run.
|
|
func createTestStep(
|
|
t *testing.T,
|
|
fixture testFixture,
|
|
runID uuid.UUID,
|
|
) database.ChatDebugStep {
|
|
t.Helper()
|
|
return createTestStepN(t, fixture, runID, 1)
|
|
}
|
|
|
|
// createTestStepN creates a debug step with the given step number.
|
|
func createTestStepN(
|
|
t *testing.T,
|
|
fixture testFixture,
|
|
runID uuid.UUID,
|
|
stepNumber int32,
|
|
) database.ChatDebugStep {
|
|
t.Helper()
|
|
step, err := fixture.svc.CreateStep(fixture.ctx, chatdebug.CreateStepParams{
|
|
RunID: runID,
|
|
ChatID: fixture.chat.ID,
|
|
StepNumber: stepNumber,
|
|
Operation: chatdebug.OperationGenerate,
|
|
Status: chatdebug.StatusInProgress,
|
|
})
|
|
require.NoError(t, err)
|
|
return step
|
|
}
|
|
|
|
// updateTestStepWithUsage completes a step and sets token usage fields.
|
|
func updateTestStepWithUsage(
|
|
t *testing.T,
|
|
fixture testFixture,
|
|
stepID uuid.UUID,
|
|
input, output, cacheCreation, cacheRead int64,
|
|
) {
|
|
t.Helper()
|
|
updateTestStepWithFullUsage(t, fixture, stepID, input, output, 0, cacheCreation, cacheRead)
|
|
}
|
|
|
|
// updateTestStepWithFullUsage completes a step with all token usage
|
|
// fields, including reasoning tokens.
|
|
func updateTestStepWithFullUsage(
|
|
t *testing.T,
|
|
fixture testFixture,
|
|
stepID uuid.UUID,
|
|
input, output, reasoning, cacheCreation, cacheRead int64,
|
|
) {
|
|
t.Helper()
|
|
_, err := fixture.svc.UpdateStep(fixture.ctx, chatdebug.UpdateStepParams{
|
|
ID: stepID,
|
|
ChatID: fixture.chat.ID,
|
|
Status: chatdebug.StatusCompleted,
|
|
Usage: map[string]any{
|
|
"input_tokens": input,
|
|
"output_tokens": output,
|
|
"reasoning_tokens": reasoning,
|
|
"cache_creation_tokens": cacheCreation,
|
|
"cache_read_tokens": cacheRead,
|
|
},
|
|
})
|
|
require.NoError(t, err)
|
|
}
|