Index: media/filters/source_buffer_stream_unittest.cc |
diff --git a/media/filters/source_buffer_stream_unittest.cc b/media/filters/source_buffer_stream_unittest.cc |
index b0e9316e72abdcdffb475c3a4dc99f6304bec30c..2fdcce24d26da104a596b008cf0d06e8f4ecc2f9 100644 |
--- a/media/filters/source_buffer_stream_unittest.cc |
+++ b/media/filters/source_buffer_stream_unittest.cc |
@@ -306,7 +306,7 @@ class SourceBufferStreamTest : public testing::Test { |
} |
// Handle preroll buffers. |
- if (base::EndsWith(timestamps[i], "P", true)) { |
+ if (base::EndsWith(timestamps[i], "P", base::CompareCase::SENSITIVE)) { |
ASSERT_TRUE(buffer->is_key_frame()); |
scoped_refptr<StreamParserBuffer> preroll_buffer; |
preroll_buffer.swap(buffer); |
@@ -500,40 +500,42 @@ class SourceBufferStreamTest : public testing::Test { |
bool is_duration_estimated = false; |
// Handle splice frame starts. |
- if (base::StartsWithASCII(timestamps[i], "S(", true)) { |
+ if (base::StartsWith(timestamps[i], "S(", base::CompareCase::SENSITIVE)) { |
CHECK(!splice_frame); |
splice_frame = true; |
// Remove the "S(" off of the token. |
timestamps[i] = timestamps[i].substr(2, timestamps[i].length()); |
} |
- if (splice_frame && base::EndsWith(timestamps[i], ")", true)) { |
+ if (splice_frame && |
+ base::EndsWith(timestamps[i], ")", base::CompareCase::SENSITIVE)) { |
splice_frame = false; |
last_splice_frame = true; |
// Remove the ")" off of the token. |
timestamps[i] = timestamps[i].substr(0, timestamps[i].length() - 1); |
} |
// Handle config changes within the splice frame. |
- if (splice_frame && base::EndsWith(timestamps[i], "C", true)) { |
+ if (splice_frame && |
+ base::EndsWith(timestamps[i], "C", base::CompareCase::SENSITIVE)) { |
splice_config_id++; |
CHECK(splice_config_id < stream_->audio_configs_.size() || |
splice_config_id < stream_->video_configs_.size()); |
// Remove the "C" off of the token. |
timestamps[i] = timestamps[i].substr(0, timestamps[i].length() - 1); |
} |
- if (base::EndsWith(timestamps[i], "K", true)) { |
+ if (base::EndsWith(timestamps[i], "K", base::CompareCase::SENSITIVE)) { |
is_keyframe = true; |
// Remove the "K" off of the token. |
timestamps[i] = timestamps[i].substr(0, timestamps[i].length() - 1); |
} |
// Handle preroll buffers. |
- if (base::EndsWith(timestamps[i], "P", true)) { |
+ if (base::EndsWith(timestamps[i], "P", base::CompareCase::SENSITIVE)) { |
is_keyframe = true; |
has_preroll = true; |
// Remove the "P" off of the token. |
timestamps[i] = timestamps[i].substr(0, timestamps[i].length() - 1); |
} |
- if (base::EndsWith(timestamps[i], "E", true)) { |
+ if (base::EndsWith(timestamps[i], "E", base::CompareCase::SENSITIVE)) { |
is_duration_estimated = true; |
// Remove the "E" off of the token. |
timestamps[i] = timestamps[i].substr(0, timestamps[i].length() - 1); |