Merge pull request #59 from mutablelogic/v1

djthorpe · web-flow · commit eacdcfba4fae · 2024-08-10T11:17:02.000+02:00
Reorg of pkg folder
diff --git a/Makefile b/Makefile
@@ -58,7 +58,9 @@ test: generate libwhisper libggml
 	@echo "Running tests (sys)"
 	@PKG_CONFIG_PATH=${ROOT_PATH}/${BUILD_DIR} ${GO} test -v ./sys/whisper/...
 	@echo "Running tests (pkg)"
-	@PKG_CONFIG_PATH=${ROOT_PATH}/${BUILD_DIR} ${GO} test -v ./pkg/whisper/...
+	@PKG_CONFIG_PATH=${ROOT_PATH}/${BUILD_DIR} ${GO} test -v ./pkg/...
+	@echo "Running tests (whisper)"
+	@PKG_CONFIG_PATH=${ROOT_PATH}/${BUILD_DIR} ${GO} test -v ./
 
 # Build whisper-static-library
 libwhisper: submodule
diff --git a/README.md b/README.md
@@ -96,8 +96,7 @@ See all the other targets in the `Makefile` for more information.
 
 ## Developing
 
-The `cmd/examples` directory contains a simple example of how to use the `whisper` package
-in your own code.
+TODO
 
 ## Status
 
diff --git a/cmd/whisper/server.go b/cmd/whisper/server.go
@@ -5,7 +5,7 @@ import (
 
 	// Packages
 	"github.com/mutablelogic/go-server/pkg/httpserver"
-	"github.com/mutablelogic/go-whisper/pkg/whisper/api"
+	"github.com/mutablelogic/go-whisper/pkg/api"
 )
 
 type ServerCmd struct {
diff --git a/cmd/whisper/transcribe.go b/cmd/whisper/transcribe.go
@@ -6,9 +6,9 @@ import (
 
 	// Packages
 	whisper "github.com/mutablelogic/go-whisper"
-	schema "github.com/mutablelogic/go-whisper/pkg/whisper/schema"
-	segmenter "github.com/mutablelogic/go-whisper/pkg/whisper/segmenter"
-	task "github.com/mutablelogic/go-whisper/pkg/whisper/task"
+	schema "github.com/mutablelogic/go-whisper/pkg/schema"
+	segmenter "github.com/mutablelogic/go-whisper/pkg/segmenter"
+	task "github.com/mutablelogic/go-whisper/pkg/task"
 
 	// Namespace imports
 	. "github.com/djthorpe/go-errors"
diff --git a/doc/notes.md b/doc/notes.md
diff --git a/etc/Dockerfile b/etc/Dockerfile
@@ -14,7 +14,7 @@ ARG ARCH
 ARG OS
 
 RUN apt-get -y update \
- && apt-get -y install software-properties-common curl \
+ && apt-get -y install software-properties-common curl libgomp1 \
  && add-apt-repository -y ppa:ubuntuhandbook1/ffmpeg6 \
  && apt-get -y update \
  && apt-get -y install libavformat-dev libavcodec-dev libavdevice-dev libavfilter-dev libavutil-dev libswscale-dev libswresample-dev
@@ -35,7 +35,10 @@ RUN make -j$(nproc)
 # Setup runtime container
 FROM ${BASE_RUN_CONTAINER} AS runtime
 RUN apt-get -y update \
- && apt-get -y install libavformat58 libavcodec58 libavdevice58 libavfilter7 libavutil56 libswscale5 libswresample3
+ && apt-get -y install software-properties-common libgomp1 \
+ && add-apt-repository -y ppa:ubuntuhandbook1/ffmpeg6 \
+ && apt-get -y update \
+ && apt-get -y install libavformat60 libavcodec60 libavdevice60 libavfilter9 libavutil58 libswscale7 libswresample4 
 COPY --from=build --chmod=755 /app/build/whisper /usr/local/bin/whisper
 COPY --from=build /app/build/whisper /usr/local/bin/whisper
 COPY --chmod=755 etc/entrypoint.sh .
diff --git a/pkg/api/logging.go b/pkg/api/logging.go
diff --git a/pkg/api/models.go b/pkg/api/models.go
@@ -12,7 +12,7 @@ import (
 	"github.com/mutablelogic/go-server/pkg/httprequest"
 	"github.com/mutablelogic/go-server/pkg/httpresponse"
 	"github.com/mutablelogic/go-whisper"
-	"github.com/mutablelogic/go-whisper/pkg/whisper/schema"
+	"github.com/mutablelogic/go-whisper/pkg/schema"
 )
 
 ///////////////////////////////////////////////////////////////////////////////
diff --git a/pkg/api/register.go b/pkg/api/register.go
diff --git a/pkg/api/transcribe.go b/pkg/api/transcribe.go
@@ -13,9 +13,9 @@ import (
 	"github.com/mutablelogic/go-server/pkg/httprequest"
 	"github.com/mutablelogic/go-server/pkg/httpresponse"
 	"github.com/mutablelogic/go-whisper"
-	"github.com/mutablelogic/go-whisper/pkg/whisper/schema"
-	"github.com/mutablelogic/go-whisper/pkg/whisper/segmenter"
-	"github.com/mutablelogic/go-whisper/pkg/whisper/task"
+	"github.com/mutablelogic/go-whisper/pkg/schema"
+	"github.com/mutablelogic/go-whisper/pkg/segmenter"
+	"github.com/mutablelogic/go-whisper/pkg/task"
 
 	// Namespace imports
 	. "github.com/djthorpe/go-errors"
diff --git a/pkg/client/client.go b/pkg/client/client.go
@@ -11,7 +11,7 @@ import (
 	"github.com/mutablelogic/go-client"
 	"github.com/mutablelogic/go-client/pkg/multipart"
 	"github.com/mutablelogic/go-server/pkg/httprequest"
-	"github.com/mutablelogic/go-whisper/pkg/whisper/schema"
+	"github.com/mutablelogic/go-whisper/pkg/schema"
 )
 
 ///////////////////////////////////////////////////////////////////////////////
diff --git a/pkg/client/opts.go b/pkg/client/opts.go
diff --git a/pkg/pool/contextpool.go b/pkg/pool/contextpool.go
@@ -5,8 +5,8 @@ import (
 	"fmt"
 
 	// Packages
-	schema "github.com/mutablelogic/go-whisper/pkg/whisper/schema"
-	task "github.com/mutablelogic/go-whisper/pkg/whisper/task"
+	schema "github.com/mutablelogic/go-whisper/pkg/schema"
+	task "github.com/mutablelogic/go-whisper/pkg/task"
 
 	// Namespace imports
 	. "github.com/djthorpe/go-errors"
diff --git a/pkg/pool/contextpool_test.go b/pkg/pool/contextpool_test.go
@@ -4,8 +4,8 @@ import (
 	"testing"
 
 	// Packages
-	pool "github.com/mutablelogic/go-whisper/pkg/whisper/pool"
-	schema "github.com/mutablelogic/go-whisper/pkg/whisper/schema"
+	pool "github.com/mutablelogic/go-whisper/pkg/pool"
+	schema "github.com/mutablelogic/go-whisper/pkg/schema"
 )
 
 func Test_contextpool_001(t *testing.T) {
diff --git a/pkg/pool/pool.go b/pkg/pool/pool.go
diff --git a/pkg/pool/pool_test.go b/pkg/pool/pool_test.go
@@ -5,7 +5,7 @@ import (
 	"testing"
 
 	// Packages
-	"github.com/mutablelogic/go-whisper/pkg/whisper/pool"
+	"github.com/mutablelogic/go-whisper/pkg/pool"
 )
 
 type Item struct {
diff --git a/pkg/schema/model.go b/pkg/schema/model.go
diff --git a/pkg/schema/segment.go b/pkg/schema/segment.go
diff --git a/pkg/schema/transcription.go b/pkg/schema/transcription.go
diff --git a/pkg/segmenter/doc.go b/pkg/segmenter/doc.go
@@ -0,0 +1,2 @@
+/* segmenter package provides a segmenter for audio files and streams */
+package segmenter
diff --git a/pkg/segmenter/segmenter.go b/pkg/segmenter/segmenter.go
@@ -14,6 +14,8 @@ import (
 	. "github.com/djthorpe/go-errors"
 )
 
+// A segmenter reads audio samples from a reader and segments them into
+// fixed-size chunks. The segmenter can be used to process audio samples
 type Segmenter struct {
 	ts          time.Duration
 	sample_rate int
@@ -29,8 +31,12 @@ type SegmentFunc func(time.Duration, []float32) error
 //////////////////////////////////////////////////////////////////////////////
 // LIFECYCLE
 
-// Create a new segmenter for "NumSamples" with a reader r
-// If NumSamples is zero then no segmenting is performed
+// Create a new segmenter for a specific "dur" duration of samples with
+// a reader r. If dur is zero then no segmenting is performed, the whole
+// audio file is read, which could cause some memory issues.
+// The sample rate is the number of samples per second.
+// At the moment, the audio format is auto-detected, but there should be
+// a way to specify the audio format.
 func New(r io.Reader, dur time.Duration, sample_rate int) (*Segmenter, error) {
 	segmenter := new(Segmenter)
 
@@ -75,9 +81,11 @@ func (s *Segmenter) Close() error {
 //////////////////////////////////////////////////////////////////////////////
 // PUBLIC METHODS
 
-// TODO: segments are output through a callback, with the samples and a timestamp
+// Segments are output through a callback, with the samples and a timestamp
 // TODO: we could do some basic silence and voice detection to segment to ensure
 // we don't overtax the CPU/GPU with silence and non-speech
+// TODO: We whould be able to select the audio stream to use. At the moment
+// the "best" audio stream is used, based on ffmpeg heuristic.
 func (s *Segmenter) Decode(ctx context.Context, fn SegmentFunc) error {
 	// Check input parameters
 	if fn == nil {
@@ -129,7 +137,7 @@ func (s *Segmenter) Decode(ctx context.Context, fn SegmentFunc) error {
 	return nil
 }
 
-// Return the duration from the file or timestamp
+// Return the file duration from the file or timestamp
 func (s *Segmenter) Duration() time.Duration {
 	if s.reader != nil {
 		return s.reader.Duration()
diff --git a/pkg/segmenter/segmenter_test.go b/pkg/segmenter/segmenter_test.go
@@ -7,11 +7,11 @@ import (
 	"time"
 
 	// Packages
-	segmenter "github.com/mutablelogic/go-whisper/pkg/whisper/segmenter"
+	segmenter "github.com/mutablelogic/go-whisper/pkg/segmenter"
 	assert "github.com/stretchr/testify/assert"
 )
 
-const SAMPLE = "../../../samples/OlivierL.wav"
+const SAMPLE = "../../samples/OlivierL.wav"
 
 func Test_segmenter_001(t *testing.T) {
 	assert := assert.New(t)
diff --git a/pkg/store/doc.go b/pkg/store/doc.go
@@ -0,0 +1,2 @@
+/* store implements a model store which allows downloading models from a remote server */
+package store
diff --git a/pkg/store/store.go b/pkg/store/store.go
@@ -1,4 +1,4 @@
-package model
+package store
 
 import (
 	"context"
@@ -12,7 +12,7 @@ import (
 	"sync"
 
 	// Packages
-	schema "github.com/mutablelogic/go-whisper/pkg/whisper/schema"
+	schema "github.com/mutablelogic/go-whisper/pkg/schema"
 	whisper "github.com/mutablelogic/go-whisper/sys/whisper"
 
 	// Namespace imports
diff --git a/pkg/store/writer.go b/pkg/store/writer.go
@@ -1,4 +1,4 @@
-package model
+package store
 
 import (
 	"io"
diff --git a/pkg/task/context.go b/pkg/task/context.go
@@ -9,7 +9,7 @@ import (
 	"time"
 
 	// Packages
-	schema "github.com/mutablelogic/go-whisper/pkg/whisper/schema"
+	schema "github.com/mutablelogic/go-whisper/pkg/schema"
 	whisper "github.com/mutablelogic/go-whisper/sys/whisper"
 
 	// Namespace imports
diff --git a/pkg/task/transcription.go b/pkg/task/transcription.go
@@ -8,7 +8,7 @@ import (
 	"time"
 
 	// Packages
-	"github.com/mutablelogic/go-whisper/pkg/whisper/schema"
+	"github.com/mutablelogic/go-whisper/pkg/schema"
 	"github.com/mutablelogic/go-whisper/sys/whisper"
 )
 
diff --git a/whisper.go b/whisper.go
@@ -10,10 +10,10 @@ import (
 
 	// Packages
 	ffmpeg "github.com/mutablelogic/go-media/pkg/ffmpeg"
-	model "github.com/mutablelogic/go-whisper/pkg/whisper/model"
-	pool "github.com/mutablelogic/go-whisper/pkg/whisper/pool"
-	schema "github.com/mutablelogic/go-whisper/pkg/whisper/schema"
-	task "github.com/mutablelogic/go-whisper/pkg/whisper/task"
+	pool "github.com/mutablelogic/go-whisper/pkg/pool"
+	schema "github.com/mutablelogic/go-whisper/pkg/schema"
+	store "github.com/mutablelogic/go-whisper/pkg/store"
+	task "github.com/mutablelogic/go-whisper/pkg/task"
 	whisper "github.com/mutablelogic/go-whisper/sys/whisper"
 
 	// Namespace imports
@@ -26,7 +26,7 @@ import (
 // Whisper represents a whisper service for running transcription and translation
 type Whisper struct {
 	pool  *pool.ContextPool
-	store *model.Store
+	store *store.Store
 }
 
 //////////////////////////////////////////////////////////////////////////////
@@ -61,7 +61,7 @@ func New(path string, opt ...Opt) (*Whisper, error) {
 
 	// Create a new whisper service
 	w := new(Whisper)
-	if store, err := model.NewStore(path, extModel, defaultModelUrl); err != nil {
+	if store, err := store.NewStore(path, extModel, defaultModelUrl); err != nil {
 		return nil, err
 	} else {
 		w.store = store
@@ -112,7 +112,7 @@ func (w *Whisper) Close() error {
 
 func (w *Whisper) MarshalJSON() ([]byte, error) {
 	return json.Marshal(struct {
-		Store *model.Store      `json:"store"`
+		Store *store.Store      `json:"store"`
 		Pool  *pool.ContextPool `json:"pool"`
 	}{
 		Store: w.store,
diff --git a/whisper_test.go b/whisper_test.go
@@ -9,7 +9,7 @@ import (
 	// Packages
 	wav "github.com/go-audio/wav"
 	whisper "github.com/mutablelogic/go-whisper"
-	task "github.com/mutablelogic/go-whisper/pkg/whisper/task"
+	task "github.com/mutablelogic/go-whisper/pkg/task"
 	assert "github.com/stretchr/testify/assert"
 
 	// Namespace imports

Original file line number	Diff line number	Diff line change
`@@ -5,7 +5,7 @@ import (`
`5`	`5`
`6`	`6`	`// Packages`
`7`	`7`	`"github.com/mutablelogic/go-server/pkg/httpserver"`
`8`		`- "github.com/mutablelogic/go-whisper/pkg/whisper/api"`
	`8`	`+ "github.com/mutablelogic/go-whisper/pkg/api"`
`9`	`9`	`)`
`10`	`10`
`11`	`11`	`type ServerCmd struct {`
Original file line number	Diff line number	Diff line change
`@@ -12,7 +12,7 @@ import (`
`12`	`12`	`"github.com/mutablelogic/go-server/pkg/httprequest"`
`13`	`13`	`"github.com/mutablelogic/go-server/pkg/httpresponse"`
`14`	`14`	`"github.com/mutablelogic/go-whisper"`
`15`		`- "github.com/mutablelogic/go-whisper/pkg/whisper/schema"`
	`15`	`+ "github.com/mutablelogic/go-whisper/pkg/schema"`
`16`	`16`	`)`
`17`	`17`
`18`	`18`	`///////////////////////////////////////////////////////////////////////////////`