Mirror-Tools
/
shaka-packager
mirror of https://github.com/shaka-project/shaka-packager.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
shaka-packager/packager/media/chunking/chunking_handler_unittest.cc

262 lines
11 KiB
C++
Raw Normal View History

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/chunking/chunking_handler.h"
#include <gmock/gmock.h>
#include <gtest/gtest.h>
#include "packager/media/base/media_handler_test_base.h"
#include "packager/media/base/test/status_test_util.h"
using ::testing::ElementsAre;
using ::testing::IsEmpty;
namespace shaka {
namespace media {
namespace {
const int kStreamIndex0 = 0;
const int kStreamIndex1 = 1;
const uint32_t kTimeScale0 = 800;
const uint32_t kTimeScale1 = 1000;
const int64_t kDuration0 = 200;
const int64_t kDuration1 = 300;
const bool kKeyFrame = true;
const bool kIsSubsegment = true;
const bool kEncrypted = true;
} // namespace
class ChunkingHandlerTest : public MediaHandlerTestBase {
public:
void SetUpChunkingHandler(int num_inputs,
const ChunkingOptions& chunking_options) {
chunking_handler_.reset(new ChunkingHandler(chunking_options));
SetUpGraph(num_inputs, num_inputs, chunking_handler_);
ASSERT_OK(chunking_handler_->Initialize());
}
Status Process(std::unique_ptr<StreamData> stream_data) {
return chunking_handler_->Process(std::move(stream_data));
}
Rename FlushStream to OnFlushRequest; add FlushOutput Change-Id: I2b37030b0c35aa34e9c72e43825f157924034227
2017-02-22 20:14:26 +00:00
Status OnFlushRequest(int stream_index) {
return chunking_handler_->OnFlushRequest(stream_index);
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
}
protected:
std::shared_ptr<ChunkingHandler> chunking_handler_;
};
TEST_F(ChunkingHandlerTest, AudioNoSubsegmentsThenFlush) {
ChunkingOptions chunking_options;
chunking_options.segment_duration_in_seconds = 1;
SetUpChunkingHandler(1, chunking_options);
ASSERT_OK(Process(GetAudioStreamInfoStreamData(kStreamIndex0, kTimeScale0)));
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(IsStreamInfo(kStreamIndex0, kTimeScale0, !kEncrypted)));
for (int i = 0; i < 5; ++i) {
ClearOutputStreamDataVector();
ASSERT_OK(Process(GetMediaSampleStreamData(kStreamIndex0, i * kDuration1,
kDuration1, kKeyFrame)));
// One output stream_data except when i == 3, which also has SegmentInfo.
if (i == 3) {
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(
IsSegmentInfo(kStreamIndex0, 0, kDuration1 * 3, !kIsSubsegment),
IsMediaSample(kStreamIndex0, i * kDuration1, kDuration1)));
} else {
EXPECT_THAT(GetOutputStreamDataVector(),
ElementsAre(IsMediaSample(kStreamIndex0, i * kDuration1,
kDuration1)));
}
}
ClearOutputStreamDataVector();
Rename FlushStream to OnFlushRequest; add FlushOutput Change-Id: I2b37030b0c35aa34e9c72e43825f157924034227
2017-02-22 20:14:26 +00:00
ASSERT_OK(OnFlushRequest(kStreamIndex0));
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
EXPECT_THAT(GetOutputStreamDataVector(),
ElementsAre(IsSegmentInfo(kStreamIndex0, kDuration1 * 3,
kDuration1 * 2, !kIsSubsegment)));
Integrate ChunkingHandler Also moved MediaHandler output validation to Initialize instead. This CL also addresses #122 with consistent chunking. Change-Id: I60c0da6d1b33421d7828bcb827d18899e71884ce
2017-02-24 01:17:47 +00:00
}
TEST_F(ChunkingHandlerTest, AudioWithSubsegments) {
ChunkingOptions chunking_options;
chunking_options.segment_duration_in_seconds = 1;
chunking_options.subsegment_duration_in_seconds = 0.5;
SetUpChunkingHandler(1, chunking_options);
ASSERT_OK(Process(GetAudioStreamInfoStreamData(kStreamIndex0, kTimeScale0)));
for (int i = 0; i < 5; ++i) {
ASSERT_OK(Process(GetMediaSampleStreamData(kStreamIndex0, i * kDuration1,
kDuration1, kKeyFrame)));
}
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(
IsStreamInfo(kStreamIndex0, kTimeScale0, !kEncrypted),
IsMediaSample(kStreamIndex0, 0, kDuration1),
IsMediaSample(kStreamIndex0, kDuration1, kDuration1),
IsSegmentInfo(kStreamIndex0, 0, kDuration1 * 2, kIsSubsegment),
IsMediaSample(kStreamIndex0, 2 * kDuration1, kDuration1),
IsSegmentInfo(kStreamIndex0, 0, kDuration1 * 3, !kIsSubsegment),
IsMediaSample(kStreamIndex0, 3 * kDuration1, kDuration1),
IsMediaSample(kStreamIndex0, 4 * kDuration1, kDuration1)));
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
}
TEST_F(ChunkingHandlerTest, VideoAndSubsegmentAndNonzeroStart) {
ChunkingOptions chunking_options;
chunking_options.segment_duration_in_seconds = 1;
chunking_options.subsegment_duration_in_seconds = 0.3;
SetUpChunkingHandler(1, chunking_options);
ASSERT_OK(Process(GetVideoStreamInfoStreamData(kStreamIndex0, kTimeScale1)));
const int64_t kVideoStartTimestamp = 12345;
for (int i = 0; i < 6; ++i) {
// Alternate key frame.
const bool is_key_frame = (i % 2) == 1;
ASSERT_OK(Process(GetMediaSampleStreamData(
kStreamIndex0, kVideoStartTimestamp + i * kDuration1, kDuration1,
is_key_frame)));
}
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(
IsStreamInfo(kStreamIndex0, kTimeScale1, !kEncrypted),
// The first samples @ kStartTimestamp is discarded - not key frame.
IsMediaSample(kStreamIndex0, kVideoStartTimestamp + kDuration1,
kDuration1),
IsMediaSample(kStreamIndex0, kVideoStartTimestamp + kDuration1 * 2,
kDuration1),
// The next segment boundary 13245 / 1000 != 12645 / 1000.
IsSegmentInfo(kStreamIndex0, kVideoStartTimestamp + kDuration1,
kDuration1 * 2, !kIsSubsegment),
IsMediaSample(kStreamIndex0, kVideoStartTimestamp + kDuration1 * 3,
kDuration1),
IsMediaSample(kStreamIndex0, kVideoStartTimestamp + kDuration1 * 4,
kDuration1),
// The subsegment has duration kDuration1 * 2 since it can only
// terminate before key frame.
IsSegmentInfo(kStreamIndex0, kVideoStartTimestamp + kDuration1 * 3,
kDuration1 * 2, kIsSubsegment),
IsMediaSample(kStreamIndex0, kVideoStartTimestamp + kDuration1 * 5,
kDuration1)));
}
TEST_F(ChunkingHandlerTest, AudioAndVideo) {
ChunkingOptions chunking_options;
chunking_options.segment_duration_in_seconds = 1;
chunking_options.subsegment_duration_in_seconds = 0.3;
SetUpChunkingHandler(2, chunking_options);
ASSERT_OK(Process(GetAudioStreamInfoStreamData(kStreamIndex0, kTimeScale0)));
ASSERT_OK(Process(GetVideoStreamInfoStreamData(kStreamIndex1, kTimeScale1)));
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(IsStreamInfo(kStreamIndex0, kTimeScale0, !kEncrypted),
IsStreamInfo(kStreamIndex1, kTimeScale1, !kEncrypted)));
ClearOutputStreamDataVector();
// Equivalent to 12345 in video timescale.
const int64_t kAudioStartTimestamp = 9876;
const int64_t kVideoStartTimestamp = 12345;
for (int i = 0; i < 5; ++i) {
ASSERT_OK(Process(GetMediaSampleStreamData(
kStreamIndex0, kAudioStartTimestamp + kDuration0 * i, kDuration0,
true)));
// Alternate key frame.
const bool is_key_frame = (i % 2) == 1;
ASSERT_OK(Process(GetMediaSampleStreamData(
kStreamIndex1, kVideoStartTimestamp + kDuration1 * i, kDuration1,
is_key_frame)));
}
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(
// The first samples @ kStartTimestamp is discarded - not key frame.
IsMediaSample(kStreamIndex0, kAudioStartTimestamp + kDuration0,
kDuration0),
IsMediaSample(kStreamIndex1, kVideoStartTimestamp + kDuration1,
kDuration1),
IsMediaSample(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 2,
kDuration0),
IsMediaSample(kStreamIndex1, kVideoStartTimestamp + kDuration1 * 2,
kDuration1),
IsMediaSample(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 3,
kDuration0),
// The audio segment is terminated together with video stream.
IsSegmentInfo(kStreamIndex0, kAudioStartTimestamp + kDuration0,
kDuration0 * 3, !kIsSubsegment),
// The next segment boundary 13245 / 1000 != 12645 / 1000.
IsSegmentInfo(kStreamIndex1, kVideoStartTimestamp + kDuration1,
kDuration1 * 2, !kIsSubsegment),
IsMediaSample(kStreamIndex1, kVideoStartTimestamp + kDuration1 * 3,
kDuration1),
IsMediaSample(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 4,
kDuration0),
IsMediaSample(kStreamIndex1, kVideoStartTimestamp + kDuration1 * 4,
kDuration1)));
ClearOutputStreamDataVector();
// The side comments below show the equivalent timestamp in video timescale.
// The audio and video are made ~aligned.
ASSERT_OK(Process(GetMediaSampleStreamData(kStreamIndex0,
kAudioStartTimestamp + kDuration0 * 5,
kDuration0, true))); // 13595
ASSERT_OK(Process(GetMediaSampleStreamData(kStreamIndex1,
kVideoStartTimestamp + kDuration1 * 5,
kDuration1, true))); // 13845
ASSERT_OK(Process(GetMediaSampleStreamData(kStreamIndex0,
kAudioStartTimestamp + kDuration0 * 6,
kDuration0, true))); // 13845
// This expectation are separated from the expectation above because
// ElementsAre supports at most 10 elements.
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(
IsMediaSample(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 5,
kDuration0),
// Audio is terminated along with video below.
IsSegmentInfo(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 4,
kDuration0 * 2, kIsSubsegment),
// The subsegment has duration kDuration1 * 2 since it can only
// terminate before key frame.
IsSegmentInfo(kStreamIndex1, kVideoStartTimestamp + kDuration1 * 3,
kDuration1 * 2, kIsSubsegment),
IsMediaSample(kStreamIndex1, kVideoStartTimestamp + kDuration1 * 5,
kDuration1)));
ClearOutputStreamDataVector();
Rename FlushStream to OnFlushRequest; add FlushOutput Change-Id: I2b37030b0c35aa34e9c72e43825f157924034227
2017-02-22 20:14:26 +00:00
ASSERT_OK(OnFlushRequest(kStreamIndex0));
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
EXPECT_THAT(
GetOutputStreamDataVector(),
ElementsAre(
IsMediaSample(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 6,
kDuration0),
IsSegmentInfo(kStreamIndex0, kAudioStartTimestamp + kDuration0 * 4,
kDuration0 * 3, !kIsSubsegment)));
ClearOutputStreamDataVector();
Rename FlushStream to OnFlushRequest; add FlushOutput Change-Id: I2b37030b0c35aa34e9c72e43825f157924034227
2017-02-22 20:14:26 +00:00
ASSERT_OK(OnFlushRequest(kStreamIndex1));
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
EXPECT_THAT(GetOutputStreamDataVector(),
ElementsAre(IsSegmentInfo(kStreamIndex1,
kVideoStartTimestamp + kDuration1 * 3,
kDuration1 * 3, !kIsSubsegment)));
// Flush again will do nothing.
ClearOutputStreamDataVector();
Rename FlushStream to OnFlushRequest; add FlushOutput Change-Id: I2b37030b0c35aa34e9c72e43825f157924034227
2017-02-22 20:14:26 +00:00
ASSERT_OK(OnFlushRequest(kStreamIndex0));
ASSERT_OK(OnFlushRequest(kStreamIndex1));
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
EXPECT_THAT(GetOutputStreamDataVector(), IsEmpty());
}
} // namespace media
} // namespace shaka