Fixed a path-duplication bug and added regression coverage

This commit is contained in:
2026-05-03 22:38:22 -05:00
parent d1802f99f0
commit 4252de4b99
5 changed files with 201 additions and 3 deletions

View File

@@ -190,9 +190,7 @@ func discoverRawTranscripts(m *manifest.Manifest, paths artifacts.SessionPaths)
if p == "" {
continue
}
if !filepath.IsAbs(p) {
p = filepath.Join(paths.Root, p)
}
p = artifacts.ResolveSessionLocalPathForRead(paths, p)
fromManifest = append(fromManifest, filepath.Clean(p))
}
}

View File

@@ -2,6 +2,7 @@ package stage
import (
"context"
"os"
"path/filepath"
"strings"
"testing"
@@ -153,9 +154,90 @@ func TestMergeStageFallsBackToRawDirectoryWhenTranscribeOutputsMissing(t *testin
}
}
func TestMergeStageResolvesWorkspaceQualifiedManifestOutputsWithoutDuplication(t *testing.T) {
env, m := setupMergeEnvWithRelativeWorkspaceRoot(t)
paths := env.ArtifactStore.SessionPaths(m.SessionID)
rawPath := filepath.Join(paths.TranscriptsRawDir, "alice.json")
writeFile(t, rawPath, `{"segments":[]}`)
writeFile(t, filepath.Join(paths.InputsDir, "speakers.yml"), "match: []\n")
writeFile(t, filepath.Join(paths.InputsDir, "autocorrect.yml"), "rules: []\n")
m.MarkStageSucceeded("transcribe", time.Now().UTC(), []manifest.ArtifactRecord{
{Kind: "transcript_raw", LocalPath: filepath.Join(env.Config.Pipeline.Workspace.Root, "work", m.SessionID, "transcripts", "raw", "alice.json")},
})
fake := &seriatim.FakeRunner{}
env.Seriatim = fake
if _, err := (mergeStage{}).Run(context.Background(), env, m); err != nil {
t.Fatalf("merge.Run() error = %v", err)
}
if len(fake.Requests) != 1 {
t.Fatalf("fake requests = %d, want 1", len(fake.Requests))
}
got := fake.Requests[0].InputTranscriptPaths
if len(got) != 1 {
t.Fatalf("input transcript paths = %#v, want len 1", got)
}
if got[0] != filepath.Clean(rawPath) {
t.Fatalf("resolved transcript path = %q, want %q", got[0], filepath.Clean(rawPath))
}
}
func TestMergeStageResolvesSessionRelativeManifestOutputs(t *testing.T) {
env, m := setupMergeEnv(t)
paths := env.ArtifactStore.SessionPaths(m.SessionID)
rawPath := filepath.Join(paths.TranscriptsRawDir, "alice.json")
writeFile(t, rawPath, `{"segments":[]}`)
writeFile(t, filepath.Join(paths.InputsDir, "speakers.yml"), "match: []\n")
writeFile(t, filepath.Join(paths.InputsDir, "autocorrect.yml"), "rules: []\n")
m.MarkStageSucceeded("transcribe", time.Now().UTC(), []manifest.ArtifactRecord{
{Kind: "transcript_raw", LocalPath: filepath.Join("transcripts", "raw", "alice.json")},
})
fake := &seriatim.FakeRunner{}
env.Seriatim = fake
if _, err := (mergeStage{}).Run(context.Background(), env, m); err != nil {
t.Fatalf("merge.Run() error = %v", err)
}
if len(fake.Requests) != 1 {
t.Fatalf("fake requests = %d, want 1", len(fake.Requests))
}
got := fake.Requests[0].InputTranscriptPaths
if len(got) != 1 {
t.Fatalf("input transcript paths = %#v, want len 1", got)
}
if got[0] != filepath.Clean(rawPath) {
t.Fatalf("resolved transcript path = %q, want %q", got[0], filepath.Clean(rawPath))
}
}
func setupMergeEnv(t *testing.T) (*Env, *manifest.Manifest) {
t.Helper()
workspace := t.TempDir()
return setupMergeEnvWithWorkspace(t, workspace)
}
func setupMergeEnvWithRelativeWorkspaceRoot(t *testing.T) (*Env, *manifest.Manifest) {
t.Helper()
cwd, err := os.Getwd()
if err != nil {
t.Fatalf("Getwd() error = %v", err)
}
workspaceAbs := t.TempDir()
workspaceRel, err := filepath.Rel(cwd, workspaceAbs)
if err != nil {
t.Fatalf("Rel() error = %v", err)
}
return setupMergeEnvWithWorkspace(t, workspaceRel)
}
func setupMergeEnvWithWorkspace(t *testing.T, workspace string) (*Env, *manifest.Manifest) {
t.Helper()
cfgDir := t.TempDir()
sessionPath := filepath.Join(cfgDir, "session.yml")
pipelinePath := filepath.Join(cfgDir, "pipeline.yml")