Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
// Copyright 2017 Google Inc. All rights reserved.
|
|
|
|
//
|
|
|
|
// Use of this source code is governed by a BSD-style
|
|
|
|
// license that can be found in the LICENSE file or at
|
|
|
|
// https://developers.google.com/open-source/licenses/bsd
|
|
|
|
|
|
|
|
#include "packager/media/base/media_handler_test_base.h"
|
|
|
|
|
|
|
|
#include "packager/media/base/audio_stream_info.h"
|
|
|
|
#include "packager/media/base/video_stream_info.h"
|
2017-06-29 22:23:53 +00:00
|
|
|
#include "packager/status_test_util.h"
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
|
|
|
|
namespace {
|
|
|
|
|
|
|
|
const int kTrackId = 1;
|
|
|
|
const uint64_t kDuration = 10000;
|
|
|
|
const char kCodecString[] = "codec string";
|
|
|
|
const uint8_t kSampleBits = 1;
|
|
|
|
const uint8_t kNumChannels = 2;
|
|
|
|
const uint32_t kSamplingFrequency = 48000;
|
|
|
|
const uint64_t kSeekPrerollNs = 12345;
|
|
|
|
const uint64_t kCodecDelayNs = 56789;
|
|
|
|
const uint32_t kMaxBitrate = 13579;
|
|
|
|
const uint32_t kAvgBitrate = 13000;
|
|
|
|
const char kLanguage[] = "eng";
|
|
|
|
const uint16_t kWidth = 10u;
|
|
|
|
const uint16_t kHeight = 20u;
|
|
|
|
const uint32_t kPixelWidth = 2u;
|
|
|
|
const uint32_t kPixelHeight = 3u;
|
2017-05-15 16:22:06 +00:00
|
|
|
const int16_t kTrickPlayFactor = 0;
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
const uint8_t kNaluLengthSize = 1u;
|
|
|
|
const bool kEncrypted = true;
|
|
|
|
|
|
|
|
// Use H264 code config.
|
|
|
|
const uint8_t kCodecConfig[]{
|
|
|
|
// Header
|
|
|
|
0x01, 0x64, 0x00, 0x1e, 0xff,
|
|
|
|
// SPS count (ignore top three bits)
|
|
|
|
0xe1,
|
|
|
|
// SPS
|
|
|
|
0x00, 0x19, // Size
|
|
|
|
0x67, 0x64, 0x00, 0x1e, 0xac, 0xd9, 0x40, 0xa0, 0x2f, 0xf9, 0x70, 0x11,
|
|
|
|
0x00, 0x00, 0x03, 0x03, 0xe9, 0x00, 0x00, 0xea, 0x60, 0x0f, 0x16, 0x2d,
|
|
|
|
0x96,
|
|
|
|
// PPS count
|
|
|
|
0x01,
|
|
|
|
// PPS
|
|
|
|
0x00, 0x06, // Size
|
|
|
|
0x68, 0xeb, 0xe3, 0xcb, 0x22, 0xc0,
|
|
|
|
};
|
|
|
|
|
|
|
|
// Mock data, we don't really care about what is inside.
|
|
|
|
const uint8_t kData[]{
|
|
|
|
0x01, 0x02, 0x03, 0x04, 0x05, 0x06,
|
|
|
|
};
|
|
|
|
|
|
|
|
} // namespace
|
|
|
|
|
|
|
|
namespace shaka {
|
|
|
|
namespace media {
|
|
|
|
|
2017-09-14 16:23:29 +00:00
|
|
|
bool FakeInputMediaHandler::ValidateOutputStreamIndex(size_t index) const {
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
Status FakeInputMediaHandler::InitializeInternal() {
|
|
|
|
return Status::OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
Status FakeInputMediaHandler::Process(std::unique_ptr<StreamData> stream_data) {
|
|
|
|
return Status::OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
Status MockOutputMediaHandler::InitializeInternal() {
|
|
|
|
return Status::OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
Status MockOutputMediaHandler::Process(
|
|
|
|
std::unique_ptr<StreamData> stream_data) {
|
|
|
|
OnProcess(stream_data.get());
|
|
|
|
return Status::OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
Status MockOutputMediaHandler::OnFlushRequest(size_t index) {
|
|
|
|
OnFlush(index);
|
|
|
|
return Status::OK;
|
|
|
|
}
|
|
|
|
|
2017-03-21 23:14:46 +00:00
|
|
|
Status FakeMediaHandler::InitializeInternal() {
|
|
|
|
return Status::OK;
|
|
|
|
}
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
|
2017-03-21 23:14:46 +00:00
|
|
|
Status FakeMediaHandler::Process(std::unique_ptr<StreamData> stream_data) {
|
|
|
|
stream_data_vector_.push_back(std::move(stream_data));
|
|
|
|
return Status::OK;
|
|
|
|
}
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
|
2017-03-21 23:14:46 +00:00
|
|
|
Status FakeMediaHandler::OnFlushRequest(size_t input_stream_index) {
|
|
|
|
return Status::OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool FakeMediaHandler::ValidateOutputStreamIndex(size_t stream_index) const {
|
|
|
|
return true;
|
|
|
|
}
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
|
|
|
|
MediaHandlerTestBase::MediaHandlerTestBase()
|
|
|
|
: next_handler_(new FakeMediaHandler),
|
|
|
|
some_handler_(new FakeMediaHandler) {}
|
|
|
|
|
2017-09-12 17:24:24 +00:00
|
|
|
bool MediaHandlerTestBase::IsVideoCodec(Codec codec) const {
|
|
|
|
return codec >= kCodecVideo && codec < kCodecVideoMaxPlusOne;
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<StreamInfo> MediaHandlerTestBase::GetVideoStreamInfo(
|
|
|
|
uint32_t time_scale) const {
|
|
|
|
return GetVideoStreamInfo(time_scale, kCodecVP9, kWidth, kHeight);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<StreamInfo> MediaHandlerTestBase::GetVideoStreamInfo(
|
|
|
|
uint32_t time_scale,
|
|
|
|
uint32_t width,
|
|
|
|
uint64_t height) const {
|
|
|
|
return GetVideoStreamInfo(time_scale, kCodecVP9, width, height);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<StreamInfo> MediaHandlerTestBase::GetVideoStreamInfo(
|
|
|
|
uint32_t time_scale,
|
|
|
|
Codec codec) const {
|
|
|
|
return GetVideoStreamInfo(time_scale, codec, kWidth, kHeight);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<StreamInfo> MediaHandlerTestBase::GetVideoStreamInfo(
|
|
|
|
uint32_t time_scale,
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
Codec codec,
|
2017-09-12 17:24:24 +00:00
|
|
|
uint32_t width,
|
|
|
|
uint64_t height) const {
|
|
|
|
return std::unique_ptr<VideoStreamInfo>(new VideoStreamInfo(
|
|
|
|
kTrackId, time_scale, kDuration, codec, H26xStreamFormat::kUnSpecified,
|
|
|
|
kCodecString, kCodecConfig, sizeof(kCodecConfig), width, height,
|
|
|
|
kPixelWidth, kPixelHeight, kTrickPlayFactor, kNaluLengthSize, kLanguage,
|
|
|
|
!kEncrypted));
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<StreamInfo> MediaHandlerTestBase::GetAudioStreamInfo(
|
|
|
|
uint32_t time_scale) const {
|
|
|
|
return GetAudioStreamInfo(time_scale, kCodecAAC);
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<StreamInfo> MediaHandlerTestBase::GetAudioStreamInfo(
|
|
|
|
uint32_t time_scale,
|
|
|
|
Codec codec) const {
|
|
|
|
return std::unique_ptr<AudioStreamInfo>(new AudioStreamInfo(
|
|
|
|
kTrackId, time_scale, kDuration, codec, kCodecString, kCodecConfig,
|
|
|
|
sizeof(kCodecConfig), kSampleBits, kNumChannels, kSamplingFrequency,
|
|
|
|
kSeekPrerollNs, kCodecDelayNs, kMaxBitrate, kAvgBitrate, kLanguage,
|
|
|
|
!kEncrypted));
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<MediaSample> MediaHandlerTestBase::GetMediaSample(
|
|
|
|
int64_t timestamp,
|
|
|
|
int64_t duration,
|
|
|
|
bool is_keyframe) const {
|
|
|
|
return GetMediaSample(timestamp, duration, is_keyframe, kData, sizeof(kData));
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
}
|
|
|
|
|
2017-09-12 17:24:24 +00:00
|
|
|
std::unique_ptr<MediaSample> MediaHandlerTestBase::GetMediaSample(
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
int64_t timestamp,
|
|
|
|
int64_t duration,
|
2017-09-12 17:24:24 +00:00
|
|
|
bool is_keyframe,
|
|
|
|
const uint8_t* data,
|
|
|
|
size_t data_length) const {
|
|
|
|
std::unique_ptr<MediaSample> sample(
|
|
|
|
new MediaSample(data, data_length, nullptr, 0, is_keyframe));
|
|
|
|
sample->set_dts(timestamp);
|
|
|
|
sample->set_duration(duration);
|
|
|
|
|
|
|
|
return sample;
|
|
|
|
}
|
|
|
|
|
|
|
|
std::unique_ptr<SegmentInfo> MediaHandlerTestBase::GetSegmentInfo(
|
2017-03-11 02:48:04 +00:00
|
|
|
int64_t start_timestamp,
|
|
|
|
int64_t duration,
|
2017-09-12 17:24:24 +00:00
|
|
|
bool is_subsegment) const {
|
|
|
|
std::unique_ptr<SegmentInfo> info(new SegmentInfo);
|
|
|
|
info->start_timestamp = start_timestamp;
|
|
|
|
info->duration = duration;
|
|
|
|
info->is_subsegment = is_subsegment;
|
|
|
|
|
|
|
|
return info;
|
2017-03-11 02:48:04 +00:00
|
|
|
}
|
|
|
|
|
2017-03-28 06:04:29 +00:00
|
|
|
void MediaHandlerTestBase::SetUpGraph(size_t num_inputs,
|
|
|
|
size_t num_outputs,
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
std::shared_ptr<MediaHandler> handler) {
|
|
|
|
// Input handler is not really used anywhere but just to satisfy one input
|
|
|
|
// one output restriction for the encryption handler.
|
|
|
|
auto input_handler = std::make_shared<FakeMediaHandler>();
|
2017-03-28 06:04:29 +00:00
|
|
|
for (size_t i = 0; i < num_inputs; ++i)
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
ASSERT_OK(input_handler->SetHandler(i, handler));
|
|
|
|
// All outputs are routed to |next_handler_|.
|
2017-03-28 06:04:29 +00:00
|
|
|
for (size_t i = 0; i < num_outputs; ++i)
|
Implement ChunkingHandler
This handler is a multi-in multi-out handler. If more than one input is
provided, there should be one and only one video stream; also, all inputs
should come from the same thread and are synchronized.
There can be multiple chunking handler running in different threads or even
different processes, we use the "consistent chunking algorithm" to make sure
the chunks in different streams are aligned without explicit communcating
with each other - which is not efficient and often difficult.
Consistent Chunking Algorithm:
1. Find the consistent chunkable boundary
Let the timestamps for video frames be (t1, t2, t3, ...). Then a
consistent chunkable boundary is simply the first chunkable boundary after
(tk / N) != (tk-1 / N), where '/' denotes integer division, and N is the
intended chunk duration.
2. Chunk only at the consistent chunkable boundary
This algorithm will make sure the chunks from different video streams are
aligned if they have aligned GoPs. However, this algorithm will only work
for video streams. To be able to chunk non video streams at similar
positions as video streams, ChunkingHandler is designed to accept one video
input and multiple non video inputs, the non video inputs are chunked when
the video input is chunked. If the inputs are synchronized - which is true
if the inputs come from the same demuxer, the video and non video chunks
are aligned.
Change-Id: Id3bad51ab14f311efdb8713b6cd36d36cf9e4639
2017-02-07 18:58:47 +00:00
|
|
|
ASSERT_OK(handler->SetHandler(i, next_handler_));
|
|
|
|
}
|
|
|
|
|
|
|
|
const std::vector<std::unique_ptr<StreamData>>&
|
|
|
|
MediaHandlerTestBase::GetOutputStreamDataVector() const {
|
|
|
|
return next_handler_->stream_data_vector();
|
|
|
|
}
|
|
|
|
|
|
|
|
void MediaHandlerTestBase::ClearOutputStreamDataVector() {
|
|
|
|
next_handler_->clear_stream_data_vector();
|
|
|
|
}
|
|
|
|
|
|
|
|
} // namespace media
|
|
|
|
} // namespace shaka
|