hashtags/lmms
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 1 Wiki Activity

Fix regressions from #6610 (#7040)

Browse Source 
Fixes crashes when using decodeSampleOggVorbis, as well as missing supported audio formats
This commit is contained in:
saker
2023-12-31 18:05:46 -05:00
committed by GitHub
parent 1868fa170a
commit f79695e4f8
1 changed files with 44 additions and 16 deletions
Download Patch File Download Diff File Expand all files Collapse all files

60
src/core/SampleDecoder.cpp
View File

@@ -43,7 +43,7 @@ namespace lmms {
namespace {
using Decoder = std::optional<SampleDecoder::Result>(*)(const QString&);
using Decoder = std::optional<SampleDecoder::Result> (*)(const QString&);
auto decodeSampleSF(const QString& audioFile) -> std::optional<SampleDecoder::Result>;
auto decodeSampleDS(const QString& audioFile) -> std::optional<SampleDecoder::Result>;
@@ -62,7 +62,7 @@ auto decodeSampleSF(const QString& audioFile) -> std::optional<SampleDecoder::Re
SNDFILE* sndFile = nullptr;
auto sfInfo = SF_INFO{};
// Use QFile to handle unicode file names on Windows
// TODO: Remove use of QFile
auto file = QFile{audioFile};
if (!file.open(QIODevice::ReadOnly)) { return std::nullopt; }
@@ -116,15 +116,47 @@ auto decodeSampleDS(const QString& audioFile) -> std::optional<SampleDecoder::Re
#ifdef LMMS_HAVE_OGGVORBIS
auto decodeSampleOggVorbis(const QString& audioFile) -> std::optional<SampleDecoder::Result>
{
auto vorbisFile = OggVorbis_File{};
const auto openError = ov_fopen(audioFile.toLocal8Bit(), &vorbisFile);
static auto s_read = [](void* buffer, size_t size, size_t count, void* stream) -> size_t {
auto file = static_cast<QFile*>(stream);
return file->read(static_cast<char*>(buffer), size * count);
};
if (openError != 0) { return std::nullopt; }
static auto s_seek = [](void* stream, ogg_int64_t offset, int whence) -> int {
auto file = static_cast<QFile*>(stream);
if (whence == SEEK_SET) { file->seek(offset); }
else if (whence == SEEK_CUR) { file->seek(file->pos() + offset); }
else if (whence == SEEK_END) { file->seek(file->size() + offset); }
else { return -1; }
return 0;
};
static auto s_close = [](void* stream) -> int {
auto file = static_cast<QFile*>(stream);
file->close();
return 0;
};
static auto s_tell = [](void* stream) -> long {
auto file = static_cast<QFile*>(stream);
return file->pos();
};
static ov_callbacks s_callbacks = {s_read, s_seek, s_close, s_tell};
// TODO: Remove use of QFile
auto file = QFile{audioFile};
if (!file.open(QIODevice::ReadOnly)) { return std::nullopt; }
auto vorbisFile = OggVorbis_File{};
if (ov_open_callbacks(&file, &vorbisFile, nullptr, 0, s_callbacks) < 0) { return std::nullopt; }
const auto vorbisInfo = ov_info(&vorbisFile, -1);
if (vorbisInfo == nullptr) { return std::nullopt; }
const auto numChannels = vorbisInfo->channels;
const auto sampleRate = vorbisInfo->rate;
const auto numSamples = ov_pcm_total(&vorbisFile, -1);
if (numSamples < 0) { return std::nullopt; }
auto buffer = std::vector<float>(numSamples);
auto output = static_cast<float**>(nullptr);
@@ -141,14 +173,14 @@ auto decodeSampleOggVorbis(const QString& audioFile) -> std::optional<SampleDeco
totalSamplesRead += samplesRead;
}
ov_clear(&vorbisFile);
auto result = std::vector<sampleFrame>(numSamples / numChannels);
for (int i = 0; i < buffer.size(); ++i)
auto result = std::vector<sampleFrame>(totalSamplesRead / numChannels);
for (int i = 0; i < result.size(); ++i)
{
if (numChannels == 1) { result[i] = {buffer[i], buffer[i]}; }
else if (numChannels > 1) { result[i] = {buffer[i * numChannels], buffer[i * numChannels + 1]}; }
}
ov_clear(&vorbisFile);
return SampleDecoder::Result{std::move(result), static_cast<int>(sampleRate)};
}
#endif // LMMS_HAVE_OGGVORBIS
@@ -156,8 +188,7 @@ auto decodeSampleOggVorbis(const QString& audioFile) -> std::optional<SampleDeco
auto SampleDecoder::supportedAudioTypes() -> const std::vector<AudioType>&
{
static const auto s_audioTypes = []
{
static const auto s_audioTypes = [] {
auto types = std::vector<AudioType>();
// Add DrumSynth by default since that support comes from us
@@ -167,8 +198,8 @@ auto SampleDecoder::supportedAudioTypes() -> const std::vector<AudioType>&
auto simpleTypeCount = 0;
sf_command(nullptr, SFC_GET_SIMPLE_FORMAT_COUNT, &simpleTypeCount, sizeof(int));
// TODO: Ideally, this code should be iterating over the major formats, but some important extensions such as *.ogg
// are not included. This is planned for future versions of sndfile.
// TODO: Ideally, this code should be iterating over the major formats, but some important extensions such as
// *.ogg are not included. This is planned for future versions of sndfile.
for (int simple = 0; simple < simpleTypeCount; ++simple)
{
sfFormatInfo.format = simple;
@@ -182,12 +213,9 @@ auto SampleDecoder::supportedAudioTypes() -> const std::vector<AudioType>&
std::transform(name.begin(), name.end(), name.begin(), [](unsigned char ch) { return std::toupper(ch); });
types.push_back(AudioType{std::move(name), sfFormatInfo.extension});
return types;
}
std::sort(types.begin(), types.end(),
[&](const AudioType& a, const AudioType& b) { return a.name < b.name; });
std::sort(types.begin(), types.end(), [&](const AudioType& a, const AudioType& b) { return a.name < b.name; });
return types;
}();
return s_audioTypes;