From b6b9857d277c56e3d155db39d01071bcc3af05f4 Mon Sep 17 00:00:00 2001
From: megatocha <127956346+megatocha@users.noreply.github.com>
Date: Fri, 14 Aug 2026 15:48:59 +0200
Subject: [PATCH] fix(deepgram): use REST transcription for audio uploads

Set the default content type to audio/ogg and route Transcribe directly through the REST Listen endpoint, removing the streaming fallback for file-based audio.
---
 Dockerfile                  | 12 +++++++++++-
 internal/deepgram/client.go | 16 ++--------------
 2 files changed, 13 insertions(+), 15 deletions(-)

diff --git a/Dockerfile b/Dockerfile
index 8fce10b..073bdce 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -2,22 +2,32 @@
 FROM golang:1.26.5-alpine AS build
 
 WORKDIR /src
+
 COPY go.mod go.sum* ./
 RUN go mod download
+
 COPY . .
+
 ARG VERSION=dev
 ARG COMMIT=none
 ARG DATE=unknown
+
 RUN CGO_ENABLED=0 go build -trimpath \
     -ldflags "-s -w -X main.version=${VERSION} -X main.commit=${COMMIT} -X main.date=${DATE}" \
     -o /out/voicy ./cmd/voicy
 
 FROM alpine:3.22
+
 RUN apk add --no-cache ca-certificates \
     && adduser -D -H -u 10001 voicy
+
 WORKDIR /app
+
 COPY --from=build /out/voicy /app/voicy
 COPY migrations /app/migrations
+
 USER voicy
+
 EXPOSE 8080
-ENTRYPOINT ["/app/voicy"]
+
+ENTRYPOINT ["/app/voicy"]
\ No newline at end of file
diff --git a/internal/deepgram/client.go b/internal/deepgram/client.go
index 79a93e4..f4cdc2f 100644
--- a/internal/deepgram/client.go
+++ b/internal/deepgram/client.go
@@ -203,7 +203,7 @@ func isNormalClose(err error) bool {
 
 func (c *Client) ListenFile(ctx context.Context, audio []byte, contentType string) (Result, error) {
 	if contentType == "" {
-		contentType = "application/octet-stream"
+		contentType = "audio/ogg"
 	}
 	endpoint := c.rest + ListenPath + "?" + restQuery()
 	req, err := http.NewRequestWithContext(ctx, http.MethodPost, endpoint, bytes.NewReader(audio))
@@ -227,20 +227,8 @@ func (c *Client) ListenFile(ctx context.Context, audio []byte, contentType strin
 	return ExtractPrerecorded(body)
 }
 
-// Transcribe streams first and falls back to the REST Listen endpoint.
 func (c *Client) Transcribe(ctx context.Context, audio []byte, contentType string, onPartial func(string)) (Result, error) {
-	res, err := c.StreamFile(ctx, audio, contentType, onPartial)
-	if err == nil && strings.TrimSpace(res.Text) != "" {
-		return res, nil
-	}
-	rest, restErr := c.ListenFile(ctx, audio, contentType)
-	if restErr != nil {
-		if err != nil {
-			return Result{}, fmt.Errorf("stream: %v; rest: %w", err, restErr)
-		}
-		return Result{}, restErr
-	}
-	return rest, nil
+	return c.ListenFile(ctx, audio, contentType)
 }
 
 func redact(key string, err error) error {