Implement real transcribe stage

This commit is contained in:
2026-05-03 15:42:44 -05:00
parent 24833194db
commit 2438a0b7a5
7 changed files with 660 additions and 26 deletions

View File

@@ -1,6 +1,12 @@
package whisperx
import "context"
import (
"context"
"os"
"path/filepath"
)
var minimalTranscriptJSON = []byte(`{"schema":"speaker_transcript.v1","segments":[]}`)
// NoopClient is a deterministic no-op WhisperX adapter.
type NoopClient struct{}
@@ -10,6 +16,9 @@ func (n *NoopClient) Transcribe(ctx context.Context, req TranscribeRequest) (Tra
if err := ctx.Err(); err != nil {
return TranscribeResult{}, err
}
if err := writeMinimalJSON(req.OutputRawTranscriptPath); err != nil {
return TranscribeResult{}, err
}
return TranscribeResult{
OutputRawTranscriptPath: req.OutputRawTranscriptPath,
Attempts: 1,
@@ -22,9 +31,10 @@ func (n *NoopClient) Transcribe(ctx context.Context, req TranscribeRequest) (Tra
// FakeClient captures requests and returns deterministic responses for tests.
type FakeClient struct {
Requests []TranscribeRequest
Err error
Result TranscribeResult
Requests []TranscribeRequest
Err error
Result TranscribeResult
TranscribeFn func(ctx context.Context, req TranscribeRequest) (TranscribeResult, error)
}
// Transcribe records the request and returns either configured error or result.
@@ -33,6 +43,9 @@ func (f *FakeClient) Transcribe(ctx context.Context, req TranscribeRequest) (Tra
return TranscribeResult{}, err
}
f.Requests = append(f.Requests, req)
if f.TranscribeFn != nil {
return f.TranscribeFn(ctx, req)
}
if f.Err != nil {
return TranscribeResult{}, f.Err
}
@@ -46,5 +59,18 @@ func (f *FakeClient) Transcribe(ctx context.Context, req TranscribeRequest) (Tra
if res.Metadata == nil {
res.Metadata = map[string]any{"fake": true}
}
if err := writeMinimalJSON(res.OutputRawTranscriptPath); err != nil {
return TranscribeResult{}, err
}
return res, nil
}
func writeMinimalJSON(path string) error {
if path == "" {
return nil
}
if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
return err
}
return os.WriteFile(path, minimalTranscriptJSON, 0o644)
}