shaka-packager/packager/media/formats/webvtt/webvtt_parser_unittest.cc

468 lines
13 KiB
C++

// 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/formats/webvtt/webvtt_parser.h"
#include <gtest/gtest.h>
#include "packager/base/bind.h"
#include "packager/media/base/stream_info.h"
#include "packager/media/base/text_sample.h"
namespace shaka {
namespace media {
namespace {
const uint32_t kStreamId = 0;
const uint32_t kTimeScale = 1000;
const char* kNoId = "";
void ExpectNoStyle(const TextFragmentStyle& style) {
EXPECT_FALSE(style.underline);
EXPECT_FALSE(style.bold);
EXPECT_FALSE(style.italic);
}
void ExpectPlainCueWithBody(const TextFragment& fragment,
const std::string& expected) {
ExpectNoStyle(fragment.style);
ASSERT_TRUE(fragment.body.empty());
ASSERT_FALSE(fragment.newline);
if (expected.empty()) {
EXPECT_TRUE(fragment.sub_fragments.empty());
} else {
ASSERT_EQ(fragment.sub_fragments.size(), 1u);
ExpectNoStyle(fragment.sub_fragments[0].style);
EXPECT_EQ(fragment.sub_fragments[0].body, expected);
}
}
} // namespace
class WebVttParserTest : public testing::Test {
protected:
void SetUpAndInitialize() {
parser_ = std::make_shared<WebVttParser>();
parser_->Init(
base::Bind(&WebVttParserTest::InitCB, base::Unretained(this)),
base::Bind(&WebVttParserTest::NewMediaSampleCB, base::Unretained(this)),
base::Bind(&WebVttParserTest::NewTextSampleCB, base::Unretained(this)),
nullptr);
}
void InitCB(const std::vector<std::shared_ptr<StreamInfo>>& streams) {
streams_ = streams;
}
bool NewMediaSampleCB(uint32_t stream_id,
std::shared_ptr<MediaSample> sample) {
ADD_FAILURE() << "Should not get media samples";
return false;
}
bool NewTextSampleCB(uint32_t stream_id, std::shared_ptr<TextSample> sample) {
EXPECT_EQ(stream_id, kStreamId);
samples_.emplace_back(std::move(sample));
return true;
}
std::shared_ptr<WebVttParser> parser_;
std::vector<std::shared_ptr<StreamInfo>> streams_;
std::vector<std::shared_ptr<TextSample>> samples_;
};
TEST_F(WebVttParserTest, FailToParseEmptyFile) {
const uint8_t text[] = "";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_TRUE(streams_.empty());
ASSERT_TRUE(samples_.empty());
}
TEST_F(WebVttParserTest, ParseOnlyHeader) {
const uint8_t text[] =
"WEBVTT\n"
"\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_TRUE(streams_.empty());
ASSERT_TRUE(samples_.empty());
}
TEST_F(WebVttParserTest, ParseHeaderWithBOM) {
const uint8_t text[] =
"\xEF\xBB\xBFWEBVTT\n"
"\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_TRUE(streams_.empty());
ASSERT_TRUE(samples_.empty());
}
TEST_F(WebVttParserTest, FailToParseHeaderWrongWord) {
const uint8_t text[] =
"NOT WEBVTT\n"
"\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_FALSE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(streams_.empty());
ASSERT_TRUE(samples_.empty());
}
TEST_F(WebVttParserTest, FailToParseHeaderNotOneLine) {
const uint8_t text[] =
"WEBVTT\n"
"WEBVTT\n"
"\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_FALSE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(streams_.empty());
ASSERT_TRUE(samples_.empty());
}
TEST_F(WebVttParserTest, SendsStreamInfo) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:00:00.000 --> 00:01:00.000\n"
"Testing\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
EXPECT_EQ(streams_[0]->time_scale(), kTimeScale);
EXPECT_EQ(streams_[0]->is_encrypted(), false);
EXPECT_EQ(streams_[0]->codec(), kCodecWebVtt);
EXPECT_EQ(streams_[0]->codec_string(), "wvtt");
}
TEST_F(WebVttParserTest, IgnoresZeroDurationCues) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:01:00.000 --> 00:01:00.000\n"
"This subtitle would never show\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_TRUE(samples_.empty());
}
TEST_F(WebVttParserTest, ParseOneCue) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:01:00.000 --> 01:00:00.000\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
EXPECT_EQ(samples_[0]->id(), kNoId);
EXPECT_EQ(samples_[0]->start_time(), 60000u);
EXPECT_EQ(samples_[0]->duration(), 3540000u);
ExpectPlainCueWithBody(samples_[0]->body(), "subtitle");
// No settings
const auto& settings = samples_[0]->settings();
EXPECT_FALSE(settings.line);
EXPECT_FALSE(settings.position);
EXPECT_FALSE(settings.size);
EXPECT_EQ(settings.region, "");
EXPECT_EQ(settings.writing_direction, WritingDirection::kHorizontal);
EXPECT_EQ(settings.text_alignment, TextAlignment::kCenter);
}
TEST_F(WebVttParserTest, ParseOneCueWithStyle) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"STYLE\n"
"::cue { color:lime }\n"
"\n"
"REGION\n"
"id:scroll\n"
"scroll:up\n"
"\n"
"00:01:00.000 --> 01:00:00.000\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
auto* stream = static_cast<const TextStreamInfo*>(streams_[0].get());
EXPECT_EQ(stream->css_styles(), "::cue { color:lime }");
EXPECT_EQ(samples_[0]->id(), kNoId);
EXPECT_EQ(samples_[0]->start_time(), 60000u);
EXPECT_EQ(samples_[0]->duration(), 3540000u);
ExpectPlainCueWithBody(samples_[0]->body(), "subtitle");
}
TEST_F(WebVttParserTest, ParseOneEmptyCue) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:01:00.000 --> 01:00:00.000\n"
"\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
ExpectPlainCueWithBody(samples_[0]->body(), "");
}
TEST_F(WebVttParserTest, FailToParseCueWithArrowInId) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"-->\n"
"00:01:00.000 --> 01:00:00.000\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_FALSE(parser_->Flush());
}
TEST_F(WebVttParserTest, ParseOneCueWithId) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"id\n"
"00:01:00.000 --> 01:00:00.000\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
EXPECT_EQ(samples_[0]->id(), "id");
ExpectPlainCueWithBody(samples_[0]->body(), "subtitle");
}
TEST_F(WebVttParserTest, ParseOneEmptyCueWithId) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"id\n"
"00:01:00.000 --> 01:00:00.000\n"
"\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
EXPECT_EQ(samples_[0]->id(), "id");
ExpectPlainCueWithBody(samples_[0]->body(), "");
}
TEST_F(WebVttParserTest, ParseSettingSize) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:01:00.000 --> 01:00:00.000 size:50%\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
ASSERT_TRUE(samples_[0]->settings().size);
EXPECT_EQ(samples_[0]->settings().size->type, TextUnitType::kPercent);
EXPECT_EQ(samples_[0]->settings().size->value, 50.0f);
}
TEST_F(WebVttParserTest, ParseOneCueWithManySettings) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:01:00.000 --> 01:00:00.000 line:5 vertical:lr region:foo"
" align:right position:20%\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
EXPECT_EQ(samples_[0]->settings().writing_direction,
WritingDirection::kVerticalGrowingRight);
EXPECT_EQ(samples_[0]->settings().text_alignment, TextAlignment::kRight);
EXPECT_FALSE(samples_[0]->settings().size);
ASSERT_TRUE(samples_[0]->settings().position);
EXPECT_EQ(samples_[0]->settings().position->type, TextUnitType::kPercent);
EXPECT_EQ(samples_[0]->settings().position->value, 20.0f);
ASSERT_TRUE(samples_[0]->settings().line);
EXPECT_EQ(samples_[0]->settings().line->type, TextUnitType::kLines);
EXPECT_EQ(samples_[0]->settings().line->value, 5.0f);
}
TEST_F(WebVttParserTest, ParseRegions) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"REGION\n"
"id:foo\n"
"width:20%\n"
"lines:6\n"
"viewportanchor:25%,75%\n"
"scroll:up\n"
"\n"
"00:01:00.000 --> 01:00:00.000 region:foo\n"
"subtitle\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 1u);
auto* stream = static_cast<const TextStreamInfo*>(streams_[0].get());
const auto& regions = stream->regions();
ASSERT_EQ(regions.size(), 1u);
ASSERT_EQ(regions.count("foo"), 1u);
EXPECT_EQ(samples_[0]->settings().region, "foo");
const auto& region = regions.at("foo");
EXPECT_EQ(region.width.value, 20.0f);
EXPECT_EQ(region.width.type, TextUnitType::kPercent);
EXPECT_EQ(region.height.value, 6.0f);
EXPECT_EQ(region.height.type, TextUnitType::kLines);
EXPECT_EQ(region.window_anchor_x.value, 25.0f);
EXPECT_EQ(region.window_anchor_x.type, TextUnitType::kPercent);
EXPECT_EQ(region.window_anchor_y.value, 75.0f);
EXPECT_EQ(region.window_anchor_y.type, TextUnitType::kPercent);
EXPECT_TRUE(region.scroll);
}
// Verify that a typical case with mulitple cues work.
TEST_F(WebVttParserTest, ParseMultipleCues) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"00:00:01.000 --> 00:00:05.200\n"
"subtitle A\n"
"\n"
"00:00:02.321 --> 00:00:07.000\n"
"subtitle B\n"
"\n"
"00:00:05.800 --> 00:00:08.000\n"
"subtitle C\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 3u);
EXPECT_EQ(samples_[0]->start_time(), 1000u);
EXPECT_EQ(samples_[0]->duration(), 4200u);
ExpectPlainCueWithBody(samples_[0]->body(), "subtitle A");
EXPECT_EQ(samples_[1]->start_time(), 2321u);
EXPECT_EQ(samples_[1]->duration(), 4679u);
ExpectPlainCueWithBody(samples_[1]->body(), "subtitle B");
EXPECT_EQ(samples_[2]->start_time(), 5800u);
EXPECT_EQ(samples_[2]->duration(), 2200u);
ExpectPlainCueWithBody(samples_[2]->body(), "subtitle C");
}
// Verify that a typical case with mulitple cues work even when comments are
// present.
TEST_F(WebVttParserTest, ParseWithComments) {
const uint8_t text[] =
"WEBVTT\n"
"\n"
"NOTE This is a one line comment\n"
"\n"
"00:00:01.000 --> 00:00:05.200\n"
"subtitle A\n"
"\n"
"NOTE\n"
"This is a multi-line comment\n"
"\n"
"00:00:02.321 --> 00:00:07.000\n"
"subtitle B\n"
"\n"
"NOTE This is a single line comment that\n"
"spans two lines\n"
"\n"
"NOTE\tThis is a comment that using a tab\n"
"\n"
"00:00:05.800 --> 00:00:08.000\n"
"subtitle C\n";
ASSERT_NO_FATAL_FAILURE(SetUpAndInitialize());
ASSERT_TRUE(parser_->Parse(text, sizeof(text) - 1));
ASSERT_TRUE(parser_->Flush());
ASSERT_EQ(streams_.size(), 1u);
ASSERT_EQ(samples_.size(), 3u);
ExpectPlainCueWithBody(samples_[0]->body(), "subtitle A");
ExpectPlainCueWithBody(samples_[1]->body(), "subtitle B");
ExpectPlainCueWithBody(samples_[2]->body(), "subtitle C");
}
} // namespace media
} // namespace shaka