Files
seriatim/internal/builtin/postprocess_test.go

144 lines
4.1 KiB
Go

package builtin
import (
"context"
"strings"
"testing"
"gitea.maximumdirect.net/eric/seriatim/internal/config"
"gitea.maximumdirect.net/eric/seriatim/internal/model"
)
func TestValidateOutputSucceedsAfterAssignIDs(t *testing.T) {
merged := model.MergedTranscript{
Segments: []model.Segment{
{Source: "input.json", Speaker: "Alice", Start: 1, End: 2, Text: "hello"},
{Source: "input.json", Speaker: "Alice", Start: 3, End: 4, Text: "again"},
},
}
withIDs, _, err := assignIDs{}.Process(context.Background(), merged, testConfig())
if err != nil {
t.Fatalf("assign IDs: %v", err)
}
got, events, err := validateOutput{}.Process(context.Background(), withIDs, testConfig())
if err != nil {
t.Fatalf("validate output: %v", err)
}
if len(got.Segments) != 2 {
t.Fatalf("segment count = %d, want 2", len(got.Segments))
}
if len(events) != 1 || !strings.Contains(events[0].Message, "validated 2 output segment(s)") {
t.Fatalf("events = %#v", events)
}
}
func TestValidateOutputFailsBeforeAssignIDs(t *testing.T) {
merged := model.MergedTranscript{
Segments: []model.Segment{
{Source: "input.json", Speaker: "Alice", Start: 1, End: 2, Text: "hello"},
},
}
_, _, err := validateOutput{}.Process(context.Background(), merged, testConfig())
if err == nil {
t.Fatal("expected validation error")
}
if !strings.Contains(err.Error(), "segment 0 has id 0; want 1") {
t.Fatalf("unexpected error: %v", err)
}
}
func TestValidateOutputUsesMinimalSchemaWhenConfigured(t *testing.T) {
merged := model.MergedTranscript{
Segments: []model.Segment{
{
ID: 1,
Source: "input.json",
SourceRef: "word-run:1:1:1",
DerivedFrom: []string{"input.json#0"},
Speaker: "Alice",
Start: 1,
End: 2,
Text: "hello",
Categories: []string{"backchannel"},
OverlapGroupID: 1,
},
},
}
cfg := testConfig()
cfg.OutputSchema = config.OutputSchemaMinimal
got, events, err := validateOutput{}.Process(context.Background(), merged, cfg)
if err != nil {
t.Fatalf("validate output: %v", err)
}
if len(got.Segments) != 1 {
t.Fatalf("segment count = %d, want 1", len(got.Segments))
}
if len(events) != 1 || !strings.Contains(events[0].Message, "validated 1 output segment(s)") {
t.Fatalf("events = %#v", events)
}
}
func TestValidateOutputUsesSeriatimSchemaWhenConfigured(t *testing.T) {
merged := model.MergedTranscript{
Segments: []model.Segment{
{
ID: 1,
Source: "input.json",
SourceRef: "word-run:1:1:1",
DerivedFrom: []string{"input.json#0"},
Speaker: "Alice",
Start: 1,
End: 2,
Text: "hello",
Categories: []string{"backchannel"},
OverlapGroupID: 1,
},
},
}
cfg := testConfig()
cfg.OutputSchema = config.OutputSchemaSeriatim
got, events, err := validateOutput{}.Process(context.Background(), merged, cfg)
if err != nil {
t.Fatalf("validate output: %v", err)
}
if len(got.Segments) != 1 {
t.Fatalf("segment count = %d, want 1", len(got.Segments))
}
if len(events) != 1 || !strings.Contains(events[0].Message, "validated 1 output segment(s)") {
t.Fatalf("events = %#v", events)
}
}
func TestValidateOutputMinimalFailsBeforeAssignIDs(t *testing.T) {
merged := model.MergedTranscript{
Segments: []model.Segment{
{Source: "input.json", Speaker: "Alice", Start: 1, End: 2, Text: "hello"},
},
}
cfg := testConfig()
cfg.OutputSchema = config.OutputSchemaMinimal
_, _, err := validateOutput{}.Process(context.Background(), merged, cfg)
if err == nil {
t.Fatal("expected validation error")
}
if !strings.Contains(err.Error(), "segment 0 has id 0; want 1") {
t.Fatalf("unexpected error: %v", err)
}
}
func testConfig() config.Config {
return config.Config{
InputReader: config.DefaultInputReader,
InputFiles: []string{"input.json"},
OutputSchema: config.DefaultOutputSchema,
PreprocessingModules: []string{"validate-raw", "normalize-speakers", "trim-text"},
PostprocessingModules: []string{"assign-ids", "validate-output"},
OutputModules: []string{"json"},
}
}