Create a job that embeds captions in the output video

import (
	"context"
	"fmt"
	"io"

	transcoder "cloud.google.com/go/video/transcoder/apiv1"
	"cloud.google.com/go/video/transcoder/apiv1/transcoderpb"
)

// createJobWithEmbeddedCaptions creates a job that embeds closed captions in the
// output video. See https://cloud.google.com/transcoder/docs/how-to/captions-and-subtitles
// for more information.
func createJobWithEmbeddedCaptions(w io.Writer, projectID string, location string, inputVideoURI string, inputCaptionsURI string, outputURI string) error {
	// projectID := "my-project-id"
	// location := "us-central1"
	// inputVideoURI := "gs://my-bucket/my-video-file"
	// inputCaptionsURI := "gs://my-bucket/my-captions-file"
	// outputURI := "gs://my-bucket/my-output-folder/"

	ctx := context.Background()
	client, err := transcoder.NewClient(ctx)
	if err != nil {
		return fmt.Errorf("NewClient: %w", err)
	}
	defer client.Close()

	// Set up elementary streams. The InputKey field refers to inputs in
	// the Inputs array defined the job config.
	elementaryStreams := []*transcoderpb.ElementaryStream{
		{
			Key: "video_stream0",
			ElementaryStream: &transcoderpb.ElementaryStream_VideoStream{
				VideoStream: &transcoderpb.VideoStream{
					CodecSettings: &transcoderpb.VideoStream_H264{
						H264: &transcoderpb.VideoStream_H264CodecSettings{
							BitrateBps:   550000,
							FrameRate:    60,
							HeightPixels: 360,
							WidthPixels:  640,
						},
					},
				},
			},
		},
		{
			Key: "audio_stream0",
			ElementaryStream: &transcoderpb.ElementaryStream_AudioStream{
				AudioStream: &transcoderpb.AudioStream{
					Codec:      "aac",
					BitrateBps: 64000,
				},
			},
		},
		{
			Key: "cea_stream0",
			ElementaryStream: &transcoderpb.ElementaryStream_TextStream{
				TextStream: &transcoderpb.TextStream{
					Codec: "cea608",
					Mapping: []*transcoderpb.TextStream_TextMapping{
						{
							AtomKey:    "atom0",
							InputKey:   "caption_input0",
							InputTrack: 0,
						},
					},
					LanguageCode: "en-US",
					DisplayName:  "English",
				},
			},
		},
	}

	req := &transcoderpb.CreateJobRequest{
		Parent: fmt.Sprintf("projects/%s/locations/%s", projectID, location),
		Job: &transcoderpb.Job{
			OutputUri: outputURI,
			JobConfig: &transcoderpb.Job_Config{
				Config: &transcoderpb.JobConfig{
					Inputs: []*transcoderpb.Input{
						{
							Key: "input0",
							Uri: inputVideoURI,
						},
						{
							Key: "caption_input0",
							Uri: inputCaptionsURI,
						},
					},
					EditList: []*transcoderpb.EditAtom{
						{
							Key:    "atom0",
							Inputs: []string{"input0", "caption_input0"},
						},
					},
					ElementaryStreams: elementaryStreams,
					MuxStreams: []*transcoderpb.MuxStream{
						{
							Key:               "sd-hls",
							Container:         "ts",
							ElementaryStreams: []string{"video_stream0", "audio_stream0"},
						},
						{
							Key:               "sd-dash",
							Container:         "fmp4",
							ElementaryStreams: []string{"video_stream0"},
						},
						{
							Key:               "audio-dash",
							Container:         "fmp4",
							ElementaryStreams: []string{"audio_stream0"},
						},
					},
					Manifests: []*transcoderpb.Manifest{
						{
							FileName:   "manifest.m3u8",
							Type:       transcoderpb.Manifest_HLS,
							MuxStreams: []string{"sd-hls"},
						},
						{
							FileName:   "manifest.mpd",
							Type:       transcoderpb.Manifest_DASH,
							MuxStreams: []string{"sd-dash", "audio-dash"},
						},
					},
				},
			},
		},
	}
	// Creates the job. Jobs take a variable amount of time to run.
	// You can query for the job state; see getJob() in get_job.go.
	response, err := client.CreateJob(ctx, req)
	if err != nil {
		return fmt.Errorf("CreateJob: %w", err)
	}

	fmt.Fprintf(w, "Job: %v", response.GetName())
	return nil
}

import com.google.cloud.video.transcoder.v1.AudioStream;
import com.google.cloud.video.transcoder.v1.CreateJobRequest;
import com.google.cloud.video.transcoder.v1.EditAtom;
import com.google.cloud.video.transcoder.v1.ElementaryStream;
import com.google.cloud.video.transcoder.v1.Input;
import com.google.cloud.video.transcoder.v1.Job;
import com.google.cloud.video.transcoder.v1.JobConfig;
import com.google.cloud.video.transcoder.v1.LocationName;
import com.google.cloud.video.transcoder.v1.Manifest;
import com.google.cloud.video.transcoder.v1.Manifest.ManifestType;
import com.google.cloud.video.transcoder.v1.MuxStream;
import com.google.cloud.video.transcoder.v1.Output;
import com.google.cloud.video.transcoder.v1.TextStream;
import com.google.cloud.video.transcoder.v1.TextStream.TextMapping;
import com.google.cloud.video.transcoder.v1.TranscoderServiceClient;
import com.google.cloud.video.transcoder.v1.VideoStream;
import java.io.IOException;

public class CreateJobWithEmbeddedCaptions {

  public static void main(String[] args) throws IOException {
    // TODO(developer): Replace these variables before running the sample.
    String projectId = "my-project-id";
    String location = "us-central1";
    String inputVideoUri = "gs://my-bucket/my-video-file";
    String inputCaptionsUri = "gs://my-bucket/my-captions-file";
    String outputUri = "gs://my-bucket/my-output-folder/";

    createJobWithEmbeddedCaptions(projectId, location, inputVideoUri, inputCaptionsUri, outputUri);
  }

  // Creates a job from an ad-hoc configuration that embeds captions in the output video.
  public static void createJobWithEmbeddedCaptions(
      String projectId,
      String location,
      String inputVideoUri,
      String inputCaptionsUri,
      String outputUri)
      throws IOException {
    // Initialize client that will be used to send requests. This client only needs to be created
    // once, and can be reused for multiple requests.
    try (TranscoderServiceClient transcoderServiceClient = TranscoderServiceClient.create()) {

      VideoStream videoStream0 =
          VideoStream.newBuilder()
              .setH264(
                  VideoStream.H264CodecSettings.newBuilder()
                      .setBitrateBps(550000)
                      .setFrameRate(60)
                      .setHeightPixels(360)
                      .setWidthPixels(640))
              .build();

      AudioStream audioStream0 =
          AudioStream.newBuilder().setCodec("aac").setBitrateBps(64000).build();

      TextStream textStream0 =
          TextStream.newBuilder()
              .setCodec("cea608")
              .addMapping(
                  0,
                  TextMapping.newBuilder()
                      .setAtomKey("atom0")
                      .setInputKey("caption_input0")
                      .setInputTrack(0)
                      .build())
              .build();

      JobConfig config =
          JobConfig.newBuilder()
              .addInputs(Input.newBuilder().setKey("input0").setUri(inputVideoUri))
              .addInputs(Input.newBuilder().setKey("caption_input0").setUri(inputCaptionsUri))
              .addEditList(
                  0, // Index in the edit list
                  EditAtom.newBuilder()
                      .setKey("atom0")
                      .addInputs("input0")
                      .addInputs("caption_input0")
                      .build())
              .setOutput(Output.newBuilder().setUri(outputUri))
              .addElementaryStreams(
                  ElementaryStream.newBuilder()
                      .setKey("video_stream0")
                      .setVideoStream(videoStream0))
              .addElementaryStreams(
                  ElementaryStream.newBuilder()
                      .setKey("audio_stream0")
                      .setAudioStream(audioStream0))
              .addElementaryStreams(
                  ElementaryStream.newBuilder().setKey("cea_stream0").setTextStream(textStream0))
              .addMuxStreams(
                  0,
                  MuxStream.newBuilder()
                      .setKey("sd")
                      .setContainer("mp4")
                      .addElementaryStreams("video_stream0")
                      .addElementaryStreams("audio_stream0")
                      .build())
              .addMuxStreams(
                  1,
                  MuxStream.newBuilder()
                      .setKey("sd_hls")
                      .setContainer("ts")
                      .addElementaryStreams("video_stream0")
                      .addElementaryStreams("audio_stream0")
                      .build())
              .addMuxStreams(
                  2,
                  MuxStream.newBuilder()
                      .setKey("sd_dash")
                      .setContainer("fmp4")
                      .addElementaryStreams("video_stream0")
                      .build())
              .addMuxStreams(
                  3,
                  MuxStream.newBuilder()
                      .setKey("audio_dash")
                      .setContainer("fmp4")
                      .addElementaryStreams("audio_stream0")
                      .build())
              .addManifests(
                  0,
                  Manifest.newBuilder()
                      .setFileName("manifest.m3u8")
                      .setType(ManifestType.HLS)
                      .addMuxStreams("sd_hls")
                      .build())
              .addManifests(
                  1,
                  Manifest.newBuilder()
                      .setFileName("manifest.mpd")
                      .setType(ManifestType.DASH)
                      .addMuxStreams("sd_dash")
                      .addMuxStreams("audio_dash")
                      .build())
              .build();

      CreateJobRequest createJobRequest =
          CreateJobRequest.newBuilder()
              .setJob(Job.newBuilder().setOutputUri(outputUri).setConfig(config).build())
              .setParent(LocationName.of(projectId, location).toString())
              .build();

      // Send the job creation request and process the response.
      Job job = transcoderServiceClient.createJob(createJobRequest);
      System.out.println("Job: " + job.getName());
    }
  }
}

/**
 * TODO(developer): Uncomment these variables before running the sample.
 */
// projectId = 'my-project-id';
// location = 'us-central1';
// inputVideoUri = 'gs://my-bucket/my-video-file';
// inputCaptionsUri = 'gs://my-bucket/my-captions-file';
// outputUri = 'gs://my-bucket/my-output-folder/';

// Imports the Transcoder library
const {TranscoderServiceClient} =
  require('@google-cloud/video-transcoder').v1;

// Instantiates a client
const transcoderServiceClient = new TranscoderServiceClient();

async function createJobWithEmbeddedCaptions() {
  // Construct request
  const request = {
    parent: transcoderServiceClient.locationPath(projectId, location),
    job: {
      outputUri: outputUri,
      config: {
        inputs: [
          {
            key: 'input0',
            uri: inputVideoUri,
          },
          {
            key: 'caption_input0',
            uri: inputCaptionsUri,
          },
        ],
        editList: [
          {
            key: 'atom0',
            inputs: ['input0', 'caption_input0'],
          },
        ],
        elementaryStreams: [
          {
            key: 'video-stream0',
            videoStream: {
              h264: {
                heightPixels: 360,
                widthPixels: 640,
                bitrateBps: 550000,
                frameRate: 60,
              },
            },
          },
          {
            key: 'audio-stream0',
            audioStream: {
              codec: 'aac',
              bitrateBps: 64000,
            },
          },
          {
            key: 'cea-stream0',
            textStream: {
              codec: 'cea608',
              mapping: [
                {
                  atomKey: 'atom0',
                  inputKey: 'caption_input0',
                  inputTrack: 0,
                },
              ],
              languageCode: 'en-US',
              displayName: 'English',
            },
          },
        ],
        muxStreams: [
          {
            key: 'sd-hls',
            container: 'ts',
            elementaryStreams: ['video-stream0', 'audio-stream0'],
          },
          {
            key: 'sd-dash',
            container: 'fmp4',
            elementaryStreams: ['video-stream0'],
          },
          {
            key: 'audio-dash',
            container: 'fmp4',
            elementaryStreams: ['audio-stream0'],
          },
        ],
        manifests: [
          {
            fileName: 'manifest.m3u8',
            type: 'HLS',
            muxStreams: ['sd-hls'],
          },
          {
            fileName: 'manifest.mpd',
            type: 'DASH',
            muxStreams: ['sd-dash', 'audio-dash'],
          },
        ],
      },
    },
  };

  // Run request
  const [response] = await transcoderServiceClient.createJob(request);
  console.log(`Job: ${response.name}`);
}

createJobWithEmbeddedCaptions();

import argparse

from google.cloud.video import transcoder_v1
from google.cloud.video.transcoder_v1.services.transcoder_service import (
    TranscoderServiceClient,
)


def create_job_with_embedded_captions(
    project_id: str,
    location: str,
    input_video_uri: str,
    input_captions_uri: str,
    output_uri: str,
) -> transcoder_v1.types.resources.Job:
    """Creates a job based on an ad-hoc job configuration that embeds closed captions in the output video.

    Args:
        project_id (str): The GCP project ID.
        location (str): The location to start the job in.
        input_video_uri (str): Uri of the input video in the Cloud Storage
          bucket.
        input_captions_uri (str): Uri of the input captions file in the Cloud
          Storage bucket.
        output_uri (str): Uri of the video output folder in the Cloud Storage
          bucket.

    Returns:
        The job resource.
    """

    client = TranscoderServiceClient()

    parent = f"projects/{project_id}/locations/{location}"
    job = transcoder_v1.types.Job()
    job.output_uri = output_uri
    job.config = transcoder_v1.types.JobConfig(
        inputs=[
            transcoder_v1.types.Input(
                key="input0",
                uri=input_video_uri,
            ),
            transcoder_v1.types.Input(
                key="caption-input0",
                uri=input_captions_uri,
            ),
        ],
        edit_list=[
            transcoder_v1.types.EditAtom(
                key="atom0",
                inputs=["input0", "caption-input0"],
            ),
        ],
        elementary_streams=[
            transcoder_v1.types.ElementaryStream(
                key="video-stream0",
                video_stream=transcoder_v1.types.VideoStream(
                    h264=transcoder_v1.types.VideoStream.H264CodecSettings(
                        height_pixels=360,
                        width_pixels=640,
                        bitrate_bps=550000,
                        frame_rate=60,
                    ),
                ),
            ),
            transcoder_v1.types.ElementaryStream(
                key="audio-stream0",
                audio_stream=transcoder_v1.types.AudioStream(
                    codec="aac",
                    bitrate_bps=64000,
                ),
            ),
            transcoder_v1.types.ElementaryStream(
                key="cea-stream0",
                text_stream=transcoder_v1.types.TextStream(
                    codec="cea608",
                    mapping_=[
                        transcoder_v1.types.TextStream.TextMapping(
                            atom_key="atom0",
                            input_key="caption-input0",
                            input_track=0,
                        ),
                    ],
                    language_code="en-US",
                    display_name="English",
                ),
            ),
        ],
        mux_streams=[
            transcoder_v1.types.MuxStream(
                key="sd-hls",
                container="ts",
                elementary_streams=["video-stream0", "audio-stream0"],
            ),
            transcoder_v1.types.MuxStream(
                key="sd-dash",
                container="fmp4",
                elementary_streams=["video-stream0"],
            ),
            transcoder_v1.types.MuxStream(
                key="audio-dash",
                container="fmp4",
                elementary_streams=["audio-stream0"],
            ),
        ],
        manifests=[
            transcoder_v1.types.Manifest(
                file_name="manifest.m3u8",
                type_="HLS",
                mux_streams=["sd-hls"],
            ),
            transcoder_v1.types.Manifest(
                file_name="manifest.mpd",
                type_="DASH",
                mux_streams=["sd-dash", "audio-dash"],
            ),
        ],
    )
    response = client.create_job(parent=parent, job=job)
    print(f"Job: {response.name}")
    return response