Keep scenario tests with their implementation

[onechat] / turn_test.go

diff --git a/turn_test.go b/turn_test.go
index c378469574dfa8d163e78d8c6ab3010a93fd866d..811487e8771fa54bcf33640606c7945eb6a4d59b 100644 (file)
--- a/turn_test.go
+++ b/turn_test.go
@@ -3,7 +3,6 @@ package main
 import (
        "context"
        "encoding/json"
 import (
        "context"
        "encoding/json"

-       "errors"

        "net/http"
        "path/filepath"
        "strings"
        "net/http"
        "path/filepath"
        "strings"


@@ -29,7 +28,7 @@ func TestTurnRunsToolsUntilTheFinalAnswer(t *testing.T) {
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("It said hi.")}, 120, 100, 5))
        }
 
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("It said hi.")}, 120, 100, 5))
        }
 

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        _, err := test.loop.Submit(terminalSource, "Run echo hi.")
        require.NoError(t, err)
        test.events.WaitFor(t, "idle", 1)
        _, err := test.loop.Submit(terminalSource, "Run echo hi.")
        require.NoError(t, err)
        test.events.WaitFor(t, "idle", 1)


@@ -61,6 +60,7 @@ func TestTurnRunsToolsUntilTheFinalAnswer(t *testing.T) {
        require.Equal(t, []string{"user", "busy", "tool", "echo", "reply", "idle"}, test.events.Kinds())
 
        var turns []string
        require.Equal(t, []string{"user", "busy", "tool", "echo", "reply", "idle"}, test.events.Kinds())
 
        var turns []string

+

        for _, record := range modelReadUsage(t, test.usagePath) {
                turns = append(turns, record.Turn)
        }
        for _, record := range modelReadUsage(t, test.usagePath) {
                turns = append(turns, record.Turn)
        }


@@ -70,7 +70,7 @@ func TestTurnRunsToolsUntilTheFinalAnswer(t *testing.T) {
 
 func TestFirstRequestHoldsTheViewTheStateAndTheMessage(t *testing.T) {
        server := newModelFake(t)
 
 func TestFirstRequestHoldsTheViewTheStateAndTheMessage(t *testing.T) {
        server := newModelFake(t)

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        _, err := test.memory.Append("user", "an earlier message")
        require.NoError(t, err)
 
        _, err := test.memory.Append("user", "an earlier message")
        require.NoError(t, err)
 


@@ -82,7 +82,7 @@ func TestFirstRequestHoldsTheViewTheStateAndTheMessage(t *testing.T) {
                "<chat>\n0+1|user: an earlier message\n",
                "</chat>\n\nNow: Thursday 2026-01-01 12:00 +00:00",
                "The new message.",
                "<chat>\n0+1|user: an earlier message\n",
                "</chat>\n\nNow: Thursday 2026-01-01 12:00 +00:00",
                "The new message.",

-       }, inputTexts(t, server.Call(0)))
+       }, turnInputTexts(t, server.Call(0)))

 }
 
 func TestMessageDuringATurnJoinsTheNextRequest(t *testing.T) {
 }
 
 func TestMessageDuringATurnJoinsTheNextRequest(t *testing.T) {


@@ -101,16 +101,18 @@ func TestMessageDuringATurnJoinsTheNextRequest(t *testing.T) {
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("done")}, 100, 0, 5))
        }
 
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("done")}, 100, 0, 5))
        }
 

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        _, err := test.loop.Submit(terminalSource, "first")
        require.NoError(t, err)
 
        <-arrived
        _, err := test.loop.Submit(terminalSource, "first")
        require.NoError(t, err)
 
        <-arrived

+

        _, err = test.loop.Submit(terminalSource, "second, during the turn")
        require.NoError(t, err)
        close(release)
 
        test.events.WaitFor(t, "idle", 1)
        _, err = test.loop.Submit(terminalSource, "second, during the turn")
        require.NoError(t, err)
        close(release)
 
        test.events.WaitFor(t, "idle", 1)

+

        input := server.Call(1).body["input"].([]any)
        require.Equal(t, map[string]any{
                "role":    "user",
        input := server.Call(1).body["input"].([]any)
        require.Equal(t, map[string]any{
                "role":    "user",


@@ -135,11 +137,12 @@ func TestMessageDuringTheFinalRequestStartsANewTurn(t *testing.T) {
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("answer")}, 100, 0, 5))
        }
 
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("answer")}, 100, 0, 5))
        }
 

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        _, err := test.loop.Submit(terminalSource, "first")
        require.NoError(t, err)
 
        <-arrived
        _, err := test.loop.Submit(terminalSource, "first")
        require.NoError(t, err)
 
        <-arrived

+

        _, err = test.loop.Submit(terminalSource, "second")
        require.NoError(t, err)
        close(release)
        _, err = test.loop.Submit(terminalSource, "second")
        require.NoError(t, err)
        close(release)


@@ -147,7 +150,7 @@ func TestMessageDuringTheFinalRequestStartsANewTurn(t *testing.T) {
        test.events.WaitFor(t, "idle", 2)
        require.Equal(t, 2, server.Calls())
 
        test.events.WaitFor(t, "idle", 2)
        require.Equal(t, 2, server.Calls())
 

-       texts := inputTexts(t, server.Call(1))
+       texts := turnInputTexts(t, server.Call(1))

        require.Equal(t, "second", texts[len(texts)-1])
        require.Contains(t, strings.Join(texts, ""), "|reply: answer\n")
 }
        require.Equal(t, "second", texts[len(texts)-1])
        require.Contains(t, strings.Join(texts, ""), "|reply: answer\n")
 }


@@ -156,7 +159,7 @@ func TestTurnWaitsForTheSummaries(t *testing.T) {
        server := newModelFake(t)
        summarizer := newFakeSummarizer()
        summarizer.gate = make(chan struct{})
        server := newModelFake(t)
        summarizer := newFakeSummarizer()
        summarizer.gate = make(chan struct{})

-       test := newTestLoopWith(t, server, summarizer)
+       test := newTestTurnLoopWith(t, server, summarizer)

 
        _, err := test.memory.Append("echo", strings.Repeat("long output ", 100))
        require.NoError(t, err)
 
        _, err := test.memory.Append("echo", strings.Repeat("long output ", 100))
        require.NoError(t, err)


@@ -170,7 +173,7 @@ func TestTurnWaitsForTheSummaries(t *testing.T) {
        close(summarizer.gate)
        test.events.WaitFor(t, "idle", 1)
        require.Equal(t, 1, server.Calls())
        close(summarizer.gate)
        test.events.WaitFor(t, "idle", 1)
        require.Equal(t, 1, server.Calls())

-       require.NotContains(t, inputTexts(t, server.Call(0))[0], pending)
+       require.NotContains(t, turnInputTexts(t, server.Call(0))[0], viewPending)

 }
 
 func TestInterruptEndsTheTurn(t *testing.T) {
 }
 
 func TestInterruptEndsTheTurn(t *testing.T) {


@@ -186,7 +189,7 @@ func TestInterruptEndsTheTurn(t *testing.T) {
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("answer")}, 100, 0, 5))
        }
 
                modelWriteEvents(w, modelCompletedEvent([]any{modelMessageItem("answer")}, 100, 0, 5))
        }
 

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        _, err := test.loop.Submit(terminalSource, "a long task")
        require.NoError(t, err)
 
        _, err := test.loop.Submit(terminalSource, "a long task")
        require.NoError(t, err)
 


@@ -203,15 +206,16 @@ func TestInterruptEndsTheTurn(t *testing.T) {
        require.NoError(t, err)
        test.events.WaitFor(t, "idle", 2)
        require.Equal(t, "reply: answer", test.Log(t)[2])
        require.NoError(t, err)
        test.events.WaitFor(t, "idle", 2)
        require.Equal(t, "reply: answer", test.Log(t)[2])

+       require.Empty(t, test.logs.String())

 }
 
 func TestFailedCallEndsTheTurnWithoutARetry(t *testing.T) {
        server := newModelFake(t)
 }
 
 func TestFailedCallEndsTheTurnWithoutARetry(t *testing.T) {
        server := newModelFake(t)

-       server.reply = func(w http.ResponseWriter, call ModelFakeCall) {
+       server.reply = func(w http.ResponseWriter, _ ModelFakeCall) {

                http.Error(w, "busy", http.StatusServiceUnavailable)
        }
 
                http.Error(w, "busy", http.StatusServiceUnavailable)
        }
 

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        _, err := test.loop.Submit(terminalSource, "hello")
        require.NoError(t, err)
        test.events.WaitFor(t, "idle", 1)
        _, err := test.loop.Submit(terminalSource, "hello")
        require.NoError(t, err)
        test.events.WaitFor(t, "idle", 1)


@@ -219,11 +223,15 @@ func TestFailedCallEndsTheTurnWithoutARetry(t *testing.T) {
        time.Sleep(50 * time.Millisecond)
        require.Equal(t, 1, server.Calls())
        require.Contains(t, test.events.Last("error"), "HTTP 503")
        time.Sleep(50 * time.Millisecond)
        require.Equal(t, 1, server.Calls())
        require.Contains(t, test.events.Last("error"), "HTTP 503")

+       require.Equal(t, 1, strings.Count(test.logs.String(), "\n"))
+       require.Contains(t, test.logs.String(), "<3>turn failed: ")
+       require.Contains(t, test.logs.String(), "HTTP 503")
+       require.NotContains(t, test.logs.String(), "hello")

 }
 
 func TestViewMessagesOnlyGrowBetweenTurns(t *testing.T) {
        server := newModelFake(t)
 }
 
 func TestViewMessagesOnlyGrowBetweenTurns(t *testing.T) {
        server := newModelFake(t)

-       test := newTestLoop(t, server)
+       test := newTestTurnLoop(t, server)

        for n, text := range []string{"one", "two", "three"} {
                _, err := test.loop.Submit(terminalSource, text)
                require.NoError(t, err)
        for n, text := range []string{"one", "two", "three"} {
                _, err := test.loop.Submit(terminalSource, text)
                require.NoError(t, err)


@@ -232,6 +240,7 @@ func TestViewMessagesOnlyGrowBetweenTurns(t *testing.T) {
 
        first := server.Call(1).body["input"].([]any)
        second := server.Call(2).body["input"].([]any)
 
        first := server.Call(1).body["input"].([]any)
        second := server.Call(2).body["input"].([]any)

+

        require.Len(t, first, 4)
        require.Len(t, second, 5)
        require.Equal(t, first[:2], second[:2])
        require.Len(t, first, 4)
        require.Len(t, second, 5)
        require.Equal(t, first[:2], second[:2])


@@ -241,23 +250,25 @@ func TestViewMessagesOnlyGrowBetweenTurns(t *testing.T) {
                "2+1|user: two\n3+1|reply: a line\n",
                "</chat>\n\nNow: Thursday 2026-01-01 12:00 +00:00",
                "three",
                "2+1|user: two\n3+1|reply: a line\n",
                "</chat>\n\nNow: Thursday 2026-01-01 12:00 +00:00",
                "three",

-       }, inputTexts(t, server.Call(2)))
+       }, turnInputTexts(t, server.Call(2)))

 }
 
 }
 

-// testLoop is a loop that runs in the background of a test, with its parts.
-type testLoop struct {
+// TurnTestLoop is a loop that runs in the background of a test, with its parts.
+type TurnTestLoop struct {

        agent     *Agent
        memory    *Memory
        agent     *Agent
        memory    *Memory

-       loop      *loop
-       events    *eventRecorder
+       loop      *TurnLoop
+       events    *TurnEventRecorder
+       logs      *SyncBuffer

        usagePath string
 }
 
 // Log returns the logged messages as "kind: text".
        usagePath string
 }
 
 // Log returns the logged messages as "kind: text".

-func (l *testLoop) Log(t *testing.T) []string {
+func (l *TurnTestLoop) Log(t *testing.T) []string {

        t.Helper()
 
        var lines []string
        t.Helper()
 
        var lines []string

+

        for id := range l.memory.Len() {
                message, err := l.memory.Message(id)
                require.NoError(t, err)
        for id := range l.memory.Len() {
                message, err := l.memory.Message(id)
                require.NoError(t, err)


@@ -268,20 +279,22 @@ func (l *testLoop) Log(t *testing.T) []string {
        return lines
 }
 
        return lines
 }
 

-// eventRecorder keeps the events of a loop.
-type eventRecorder struct {
+// TurnEventRecorder keeps the events of a loop.
+type TurnEventRecorder struct {

        mutex  sync.Mutex
        events []Event
 }
 
        mutex  sync.Mutex
        events []Event
 }
 

-func (r *eventRecorder) Add(event Event) {
-       r.mutex.Lock()
-       defer r.mutex.Unlock()
+// WaitFor waits until count events of kind came.
+func (r *TurnEventRecorder) WaitFor(t *testing.T, kind string, count int) {
+       t.Helper()

 
 

-       r.events = append(r.events, event)
+       require.Eventually(t, func() bool {
+               return r.Count(kind) >= count
+       }, 10*time.Second, time.Millisecond)

 }
 
 }
 

-func (r *eventRecorder) Kinds() []string {
+func (r *TurnEventRecorder) Kinds() []string {

        r.mutex.Lock()
        defer r.mutex.Unlock()
 
        r.mutex.Lock()
        defer r.mutex.Unlock()
 


@@ -293,7 +306,7 @@ func (r *eventRecorder) Kinds() []string {
        return kinds
 }
 
        return kinds
 }
 

-func (r *eventRecorder) Count(kind string) int {
+func (r *TurnEventRecorder) Count(kind string) int {

        count := 0
        for _, each := range r.Kinds() {
                if each == kind {
        count := 0
        for _, each := range r.Kinds() {
                if each == kind {


@@ -305,57 +318,97 @@ func (r *eventRecorder) Count(kind string) int {
 }
 
 // Last returns the text of the last event of kind.
 }
 
 // Last returns the text of the last event of kind.

-func (r *eventRecorder) Last(kind string) string {
+func (r *TurnEventRecorder) Last(kind string) string {
+       return r.LastEvent(kind).Text
+}
+
+// LastEvent returns the last event of kind.
+func (r *TurnEventRecorder) LastEvent(kind string) Event {

        r.mutex.Lock()
        defer r.mutex.Unlock()
 
        for n := len(r.events) - 1; n >= 0; n-- {
                if r.events[n].Kind == kind {
        r.mutex.Lock()
        defer r.mutex.Unlock()
 
        for n := len(r.events) - 1; n >= 0; n-- {
                if r.events[n].Kind == kind {

-                       return r.events[n].Text
+                       return r.events[n]

                }
        }
 
                }
        }
 

-       return ""
+       return Event{}

 }
 
 }
 

-// WaitFor waits until count events of kind came.
-func (r *eventRecorder) WaitFor(t *testing.T, kind string, count int) {
+func (r *TurnEventRecorder) Add(event Event) {
+       r.mutex.Lock()
+       defer r.mutex.Unlock()
+
+       r.events = append(r.events, event)
+}
+
+func newTestTurnLoop(t *testing.T, server *ModelFake) *TurnTestLoop {

        t.Helper()
 
        t.Helper()
 

-       require.Eventually(t, func() bool {
-               return r.Count(kind) >= count
-       }, 10*time.Second, time.Millisecond)
+       return newTestTurnLoopWith(t, server, newFakeSummarizer())

 }
 
 }
 

-func newTestLoop(t *testing.T, server *ModelFake) *testLoop {
+// turnInputTexts returns the text of each user message of the input of call.
+func turnInputTexts(t *testing.T, call ModelFakeCall) []string {

        t.Helper()
 
        t.Helper()
 

-       return newTestLoopWith(t, server, newFakeSummarizer())
+       var body struct {
+               Input []struct {
+                       Role    string `json:"role"`
+                       Content []struct {
+                               Text string `json:"text"`
+                       } `json:"content"`
+               } `json:"input"`
+       }
+
+       require.NoError(t, json.Unmarshal(call.raw, &body))
+
+       var texts []string
+
+       for _, item := range body.Input {
+               if item.Role != "user" {
+                       continue
+               }
+
+               var text strings.Builder
+
+               for _, part := range item.Content {
+                       text.WriteString(part.Text)
+               }
+
+               texts = append(texts, text.String())
+       }
+
+       return texts

 }
 
 }
 

-// newTestLoopWith returns a running loop on a new agent whose builds use
+// newTestTurnLoopWith returns a running loop on a new agent whose builds use

 // summarizer and whose commands run in a temporary directory.
 // summarizer and whose commands run in a temporary directory.

-func newTestLoopWith(t *testing.T, server *ModelFake, summarizer Summarizer) *testLoop {
+func newTestTurnLoopWith(t *testing.T, server *ModelFake, summarizer Summarizer) *TurnTestLoop {

        t.Helper()
 
        t.Helper()
 

-       agent, err := openAgent(t.TempDir())
+       agent, err := openAgent(filepath.Join(t.TempDir(), "agent"))

        require.NoError(t, err)
 
        agent.config.Workdir = t.TempDir()
        clock := newTestClock()
        require.NoError(t, err)
 
        agent.config.Workdir = t.TempDir()
        clock := newTestClock()

-       memory, err := openMemory(t.Context(), t.TempDir(), clock.Now, summarizer)
+       memory, err := openMemory(t.Context(), t.TempDir(), clock.Now, summarizer, discardLog())

        require.NoError(t, err)
 
        require.NoError(t, err)
 

-       test := &testLoop{
+       test := &TurnTestLoop{

                agent:     agent,
                memory:    memory,
                agent:     agent,
                memory:    memory,

-               events:    &eventRecorder{},
+               events:    &TurnEventRecorder{},
+               logs:      &SyncBuffer{},

                usagePath: filepath.Join(t.TempDir(), "usage.jsonl"),
        }
 
        client := newModelClient(server.URL, ModelAPIKey(modelTestKey), agent.CacheKey(), test.usagePath)
                usagePath: filepath.Join(t.TempDir(), "usage.jsonl"),
        }
 
        client := newModelClient(server.URL, ModelAPIKey(modelTestKey), agent.CacheKey(), test.usagePath)

-       test.loop, err = newLoop(agent, memory, client, clock.Now, test.events.Add)
+       whatsapp, _, _ := whatsappStartTest(t, t.TempDir(), clock)
+       test.loop, err = newTurnLoop(agent, memory, newTestSubagents(t), whatsapp, newTestSchedules(t, clock.Now), client, clock.Now, daemonTestLogger(test.logs), test.events.Add)

        require.NoError(t, err)
        require.NoError(t, err)

+

        ctx, cancel := context.WithCancel(t.Context())
        done := make(chan error)
        go func() {
        ctx, cancel := context.WithCancel(t.Context())
        done := make(chan error)
        go func() {


@@ -364,41 +417,18 @@ func newTestLoopWith(t *testing.T, server *ModelFake, summarizer Summarizer) *te
 
        t.Cleanup(func() {
                cancel()
 
        t.Cleanup(func() {
                cancel()

-               require.True(t, errors.Is(<-done, context.Canceled))
+               require.ErrorIs(t, <-done, context.Canceled)

                require.NoError(t, memory.Close())
        })
 
        return test
 }
 
                require.NoError(t, memory.Close())
        })
 
        return test
 }
 

-// inputTexts returns the text of each user message of the input of call.
-func inputTexts(t *testing.T, call ModelFakeCall) []string {
-       t.Helper()
-
-       var body struct {
-               Input []struct {
-                       Role    string `json:"role"`
-                       Content []struct {
-                               Text string `json:"text"`
-                       } `json:"content"`
-               } `json:"input"`
-       }
-
-       require.NoError(t, json.Unmarshal(call.raw, &body))
-
-       var texts []string
-       for _, item := range body.Input {
-               if item.Role != "user" {
-                       continue
-               }
-
-               var text strings.Builder
-               for _, part := range item.Content {
-                       text.WriteString(part.Text)
-               }
-
-               texts = append(texts, text.String())
+func modelReasoningItem(id, encrypted string) map[string]any {
+       return map[string]any{
+               "type":              "reasoning",
+               "id":                id,
+               "encrypted_content": encrypted,
+               "summary":           []any{},

        }
        }

-
-       return texts

 }
 }