Merge "Add AAC extractor"

10 files changed, 442 insertions, 52 deletions

diff --git a/media/libstagefright/AACExtractor.cpp b/media/libstagefright/AACExtractor.cpp
new file mode 100644
index 0000000..4203b6e
--- /dev/null
+++ b/media/libstagefright/AACExtractor.cpp
@@ -0,0 +1,323 @@
+/*
+ * Copyright (C) 2011 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+//#define LOG_NDEBUG 0
+#define LOG_TAG "AACExtractor"
+#include <utils/Log.h>
+
+#include "include/AACExtractor.h"
+#include "include/avc_utils.h"
+
+#include <media/stagefright/foundation/ABuffer.h>
+#include <media/stagefright/DataSource.h>
+#include <media/stagefright/MediaBufferGroup.h>
+#include <media/stagefright/MediaDebug.h>
+#include <media/stagefright/MediaDefs.h>
+#include <media/stagefright/MediaErrors.h>
+#include <media/stagefright/MediaSource.h>
+#include <media/stagefright/MetaData.h>
+#include <utils/String8.h>
+
+namespace android {
+
+#define ADTS_HEADER_LENGTH 7
+
+class AACSource : public MediaSource {
+public:
+    AACSource(const sp<DataSource> &source,
+              const sp<MetaData> &meta,
+              const Vector<uint64_t> &offset_vector,
+              int64_t frame_duration_us);
+
+    virtual status_t start(MetaData *params = NULL);
+    virtual status_t stop();
+
+    virtual sp<MetaData> getFormat();
+
+    virtual status_t read(
+            MediaBuffer **buffer, const ReadOptions *options = NULL);
+
+protected:
+    virtual ~AACSource();
+
+private:
+    static const size_t kMaxFrameSize;
+    sp<DataSource> mDataSource;
+    sp<MetaData> mMeta;
+
+    off64_t mOffset;
+    int64_t mCurrentTimeUs;
+    bool mStarted;
+    MediaBufferGroup *mGroup;
+
+    Vector<uint64_t> mOffsetVector;
+    int64_t mFrameDurationUs;
+
+    AACSource(const AACSource &);
+    AACSource &operator=(const AACSource &);
+};
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Returns the sample rate based on the sampling frequency index
+uint32_t get_sample_rate(const uint8_t sf_index)
+{
+    static const uint32_t sample_rates[] =
+    {
+        96000, 88200, 64000, 48000, 44100, 32000,
+        24000, 22050, 16000, 12000, 11025, 8000
+    };
+
+    if (sf_index < sizeof(sample_rates) / sizeof(sample_rates[0])) {
+        return sample_rates[sf_index];
+    }
+
+    return 0;
+}
+
+static size_t getFrameSize(const sp<DataSource> &source, off64_t offset) {
+    size_t frameSize = 0;
+
+    uint8_t syncword[2];
+    if (source->readAt(0, &syncword, 2) != 2) {
+        return 0;
+    }
+    if ((syncword[0] != 0xff) || ((syncword[1] & 0xf6) != 0xf0)) {
+        return 0;
+    }
+
+    uint8_t protectionAbsent;
+    if (source->readAt(offset + 1, &protectionAbsent, 1) < 1) {
+        return 0;
+    }
+    protectionAbsent &= 0x1;
+
+    uint8_t header[3];
+    if (source->readAt(offset + 3, &header, 3) < 3) {
+        return 0;
+    }
+
+    frameSize = (header[0] & 0x3) << 11 | header[1] << 3 | header[2] >> 5;
+    frameSize += ADTS_HEADER_LENGTH + protectionAbsent ? 0 : 2;
+
+    return frameSize;
+}
+
+AACExtractor::AACExtractor(const sp<DataSource> &source)
+    : mDataSource(source),
+      mInitCheck(NO_INIT),
+      mFrameDurationUs(0) {
+    String8 mimeType;
+    float confidence;
+    if (!SniffAAC(mDataSource, &mimeType, &confidence, NULL)) {
+        return;
+    }
+
+    uint8_t profile, sf_index, channel, header[2];
+    if (mDataSource->readAt(2, &header, 2) < 2) {
+        return;
+    }
+
+    profile = (header[0] >> 6) & 0x3;
+    sf_index = (header[0] >> 2) & 0xf;
+    uint32_t sr = get_sample_rate(sf_index);
+    if (sr == 0) {
+        return;
+    }
+    channel = (header[0] & 0x1) << 2 | (header[1] >> 6);
+
+    mMeta = MakeAACCodecSpecificData(profile, sf_index, channel);
+
+    off64_t offset = 0;
+    off64_t streamSize, numFrames = 0;
+    size_t frameSize = 0;
+    int64_t duration = 0;
+
+    if (mDataSource->getSize(&streamSize) == OK) {
+         while (offset < streamSize) {
+            if ((frameSize = getFrameSize(source, offset)) == 0) {
+                return;
+            }
+
+            mOffsetVector.push(offset);
+
+            offset += frameSize;
+            numFrames ++;
+        }
+
+        // Round up and get the duration
+        mFrameDurationUs = (1024 * 1000000ll + (sr - 1)) / sr;
+        duration = numFrames * mFrameDurationUs;
+        mMeta->setInt64(kKeyDuration, duration);
+    }
+
+    mInitCheck = OK;
+}
+
+AACExtractor::~AACExtractor() {
+}
+
+sp<MetaData> AACExtractor::getMetaData() {
+    sp<MetaData> meta = new MetaData;
+
+    if (mInitCheck != OK) {
+        return meta;
+    }
+
+    meta->setCString(kKeyMIMEType, MEDIA_MIMETYPE_AUDIO_AAC_ADTS);
+
+    return meta;
+}
+
+size_t AACExtractor::countTracks() {
+    return mInitCheck == OK ? 1 : 0;
+}
+
+sp<MediaSource> AACExtractor::getTrack(size_t index) {
+    if (mInitCheck != OK || index != 0) {
+        return NULL;
+    }
+
+    return new AACSource(mDataSource, mMeta, mOffsetVector, mFrameDurationUs);
+}
+
+sp<MetaData> AACExtractor::getTrackMetaData(size_t index, uint32_t flags) {
+    if (mInitCheck != OK || index != 0) {
+        return NULL;
+    }
+
+    return mMeta;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// 8192 = 2^13, 13bit AAC frame size (in bytes)
+const size_t AACSource::kMaxFrameSize = 8192;
+
+AACSource::AACSource(
+        const sp<DataSource> &source, const sp<MetaData> &meta,
+        const Vector<uint64_t> &offset_vector,
+        int64_t frame_duration_us)
+    : mDataSource(source),
+      mMeta(meta),
+      mOffset(0),
+      mCurrentTimeUs(0),
+      mStarted(false),
+      mGroup(NULL),
+      mOffsetVector(offset_vector),
+      mFrameDurationUs(frame_duration_us) {
+}
+
+AACSource::~AACSource() {
+    if (mStarted) {
+        stop();
+    }
+}
+
+status_t AACSource::start(MetaData *params) {
+    CHECK(!mStarted);
+
+    mOffset = 0;
+    mCurrentTimeUs = 0;
+    mGroup = new MediaBufferGroup;
+    mGroup->add_buffer(new MediaBuffer(kMaxFrameSize));
+    mStarted = true;
+
+    return OK;
+}
+
+status_t AACSource::stop() {
+    CHECK(mStarted);
+
+    delete mGroup;
+    mGroup = NULL;
+
+    mStarted = false;
+    return OK;
+}
+
+sp<MetaData> AACSource::getFormat() {
+    return mMeta;
+}
+
+status_t AACSource::read(
+        MediaBuffer **out, const ReadOptions *options) {
+    *out = NULL;
+
+    int64_t seekTimeUs;
+    ReadOptions::SeekMode mode;
+    if (options && options->getSeekTo(&seekTimeUs, &mode)) {
+        if (mFrameDurationUs > 0) {
+            int64_t seekFrame = seekTimeUs / mFrameDurationUs;
+            mCurrentTimeUs = seekFrame * mFrameDurationUs;
+
+            mOffset = mOffsetVector.itemAt(seekFrame);
+        }
+    }
+
+    size_t frameSize, frameSizeWithoutHeader;
+    if ((frameSize = getFrameSize(mDataSource, mOffset)) == 0) {
+        return ERROR_END_OF_STREAM;
+    }
+
+    MediaBuffer *buffer;
+    status_t err = mGroup->acquire_buffer(&buffer);
+    if (err != OK) {
+        return err;
+    }
+
+    frameSizeWithoutHeader = frameSize - ADTS_HEADER_LENGTH;
+    if (mDataSource->readAt(mOffset + ADTS_HEADER_LENGTH, buffer->data(),
+                frameSizeWithoutHeader) != (ssize_t)frameSizeWithoutHeader) {
+        buffer->release();
+        buffer = NULL;
+
+        return ERROR_IO;
+    }
+
+    buffer->set_range(0, frameSizeWithoutHeader);
+    buffer->meta_data()->setInt64(kKeyTime, mCurrentTimeUs);
+    buffer->meta_data()->setInt32(kKeyIsSyncFrame, 1);
+
+    mOffset += frameSize;
+    mCurrentTimeUs += mFrameDurationUs;
+
+    *out = buffer;
+    return OK;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+bool SniffAAC(
+        const sp<DataSource> &source, String8 *mimeType, float *confidence,
+        sp<AMessage> *) {
+    uint8_t header[2];
+
+    if (source->readAt(0, &header, 2) != 2) {
+        return false;
+    }
+
+    // ADTS syncword
+    if ((header[0] == 0xff) && ((header[1] & 0xf6) == 0xf0)) {
+        *mimeType = MEDIA_MIMETYPE_AUDIO_AAC_ADTS;
+        *confidence = 0.2;
+        return true;
+    }
+
+    return false;
+}
+
+}  // namespace android
diff --git a/media/libstagefright/Android.mk b/media/libstagefright/Android.mk
index 029b238..53435f8 100644
--- a/media/libstagefright/Android.mk
+++ b/media/libstagefright/Android.mk
@@ -5,6 +5,7 @@ include frameworks/base/media/libstagefright/codecs/common/Config.mk
 
 LOCAL_SRC_FILES:=                         \
         ACodec.cpp                        \
+        AACExtractor.cpp                  \
         AMRExtractor.cpp                  \
         AMRWriter.cpp                     \
         AudioPlayer.cpp                   \
diff --git a/media/libstagefright/DataSource.cpp b/media/libstagefright/DataSource.cpp
index e06fa81..8f9c150 100644
--- a/media/libstagefright/DataSource.cpp
+++ b/media/libstagefright/DataSource.cpp
@@ -24,6 +24,7 @@
 #include "include/NuHTTPDataSource.h"
 #include "include/DRMExtractor.h"
 #include "include/FLACExtractor.h"
+#include "include/AACExtractor.h"
 
 #include "matroska/MatroskaExtractor.h"
 
@@ -109,6 +110,7 @@ void DataSource::RegisterDefaultSniffers() {
     RegisterSniffer(SniffAMR);
     RegisterSniffer(SniffMPEG2TS);
     RegisterSniffer(SniffMP3);
+    RegisterSniffer(SniffAAC);
 
     char value[PROPERTY_VALUE_MAX];
     if (property_get("drm.service.enabled", value, NULL)
diff --git a/media/libstagefright/MediaDefs.cpp b/media/libstagefright/MediaDefs.cpp
index b50af89..0be7261 100644
--- a/media/libstagefright/MediaDefs.cpp
+++ b/media/libstagefright/MediaDefs.cpp
@@ -36,6 +36,7 @@ const char *MEDIA_MIMETYPE_AUDIO_G711_ALAW = "audio/g711-alaw";
 const char *MEDIA_MIMETYPE_AUDIO_G711_MLAW = "audio/g711-mlaw";
 const char *MEDIA_MIMETYPE_AUDIO_RAW = "audio/raw";
 const char *MEDIA_MIMETYPE_AUDIO_FLAC = "audio/flac";
+const char *MEDIA_MIMETYPE_AUDIO_AAC_ADTS = "audio/aac-adts";
 
 const char *MEDIA_MIMETYPE_CONTAINER_MPEG4 = "video/mpeg4";
 const char *MEDIA_MIMETYPE_CONTAINER_WAV = "audio/wav";
diff --git a/media/libstagefright/MediaExtractor.cpp b/media/libstagefright/MediaExtractor.cpp
index 08ed206..dbd0829 100644
--- a/media/libstagefright/MediaExtractor.cpp
+++ b/media/libstagefright/MediaExtractor.cpp
@@ -27,6 +27,7 @@
 #include "include/DRMExtractor.h"
 #include "include/WVMExtractor.h"
 #include "include/FLACExtractor.h"
+#include "include/AACExtractor.h"
 
 #include "matroska/MatroskaExtractor.h"
 
@@ -98,6 +99,8 @@ sp<MediaExtractor> MediaExtractor::Create(
         return new MPEG2TSExtractor(source);
     } else if (!strcasecmp(mime, MEDIA_MIMETYPE_CONTAINER_WVM)) {
         return new WVMExtractor(source);
+    } else if (!strcasecmp(mime, MEDIA_MIMETYPE_AUDIO_AAC_ADTS)) {
+        return new AACExtractor(source);
     }
 
     return NULL;
diff --git a/media/libstagefright/avc_utils.cpp b/media/libstagefright/avc_utils.cpp
index fa12cf0..95cf2d3 100644
--- a/media/libstagefright/avc_utils.cpp
+++ b/media/libstagefright/avc_utils.cpp
@@ -329,5 +329,52 @@ bool IsIDR(const sp<ABuffer> &buffer) {
     return foundIDR;
 }
 
+sp<MetaData> MakeAACCodecSpecificData(
+        unsigned profile, unsigned sampling_freq_index,
+        unsigned channel_configuration) {
+    sp<MetaData> meta = new MetaData;
+    meta->setCString(kKeyMIMEType, MEDIA_MIMETYPE_AUDIO_AAC);
+
+    CHECK_LE(sampling_freq_index, 11u);
+    static const int32_t kSamplingFreq[] = {
+        96000, 88200, 64000, 48000, 44100, 32000, 24000, 22050,
+        16000, 12000, 11025, 8000
+    };
+    meta->setInt32(kKeySampleRate, kSamplingFreq[sampling_freq_index]);
+    meta->setInt32(kKeyChannelCount, channel_configuration);
+
+    static const uint8_t kStaticESDS[] = {
+        0x03, 22,
+        0x00, 0x00,     // ES_ID
+        0x00,           // streamDependenceFlag, URL_Flag, OCRstreamFlag
+
+        0x04, 17,
+        0x40,                       // Audio ISO/IEC 14496-3
+        0x00, 0x00, 0x00, 0x00,
+        0x00, 0x00, 0x00, 0x00,
+        0x00, 0x00, 0x00, 0x00,
+
+        0x05, 2,
+        // AudioSpecificInfo follows
+
+        // oooo offf fccc c000
+        // o - audioObjectType
+        // f - samplingFreqIndex
+        // c - channelConfig
+    };
+    sp<ABuffer> csd = new ABuffer(sizeof(kStaticESDS) + 2);
+    memcpy(csd->data(), kStaticESDS, sizeof(kStaticESDS));
+
+    csd->data()[sizeof(kStaticESDS)] =
+        ((profile + 1) << 3) | (sampling_freq_index >> 1);
+
+    csd->data()[sizeof(kStaticESDS) + 1] =
+        ((sampling_freq_index << 7) & 0x80) | (channel_configuration << 3);
+
+    meta->setData(kKeyESDS, 0, csd->data(), csd->size());
+
+    return meta;
+}
+
 }  // namespace android
 
diff --git a/media/libstagefright/include/AACExtractor.h b/media/libstagefright/include/AACExtractor.h
new file mode 100644
index 0000000..8e5657b
--- /dev/null
+++ b/media/libstagefright/include/AACExtractor.h
@@ -0,0 +1,61 @@
+/*
+ * Copyright (C) 2011 The Android Open Source Project
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#ifndef AAC_EXTRACTOR_H_
+
+#define AAC_EXTRACTOR_H_
+
+#include <media/stagefright/MediaExtractor.h>
+
+#include <utils/Vector.h>
+
+namespace android {
+
+struct AMessage;
+class String8;
+
+class AACExtractor : public MediaExtractor {
+public:
+    AACExtractor(const sp<DataSource> &source);
+
+    virtual size_t countTracks();
+    virtual sp<MediaSource> getTrack(size_t index);
+    virtual sp<MetaData> getTrackMetaData(size_t index, uint32_t flags);
+
+    virtual sp<MetaData> getMetaData();
+
+protected:
+    virtual ~AACExtractor();
+
+private:
+    sp<DataSource> mDataSource;
+    sp<MetaData> mMeta;
+    status_t mInitCheck;
+
+    Vector<uint64_t> mOffsetVector;
+    int64_t mFrameDurationUs;
+
+    AACExtractor(const AACExtractor &);
+    AACExtractor &operator=(const AACExtractor &);
+};
+
+bool SniffAAC(
+        const sp<DataSource> &source, String8 *mimeType, float *confidence,
+        sp<AMessage> *);
+
+}  // namespace android
+
+#endif  // AAC_EXTRACTOR_H_
diff --git a/media/libstagefright/include/avc_utils.h b/media/libstagefright/include/avc_utils.h
index 3aeb07f..0218755 100644
--- a/media/libstagefright/include/avc_utils.h
+++ b/media/libstagefright/include/avc_utils.h
@@ -52,6 +52,10 @@ bool IsIDR(const sp<ABuffer> &accessUnit);
 
 const char *AVCProfileToString(uint8_t profile);
 
+sp<MetaData> MakeAACCodecSpecificData(
+        unsigned profile, unsigned sampling_freq_index,
+        unsigned channel_configuration);
+
 }  // namespace android
 
 #endif  // AVC_UTILS_H_
diff --git a/media/libstagefright/mpeg2ts/ESQueue.cpp b/media/libstagefright/mpeg2ts/ESQueue.cpp
index 73efdfe..dcaf9f7 100644
--- a/media/libstagefright/mpeg2ts/ESQueue.cpp
+++ b/media/libstagefright/mpeg2ts/ESQueue.cpp
@@ -341,54 +341,6 @@ int64_t ElementaryStreamQueue::fetchTimestamp(size_t size) {
     return timeUs;
 }
 
-// static
-sp<MetaData> ElementaryStreamQueue::MakeAACCodecSpecificData(
-        unsigned profile, unsigned sampling_freq_index,
-        unsigned channel_configuration) {
-    sp<MetaData> meta = new MetaData;
-    meta->setCString(kKeyMIMEType, MEDIA_MIMETYPE_AUDIO_AAC);
-
-    CHECK_LE(sampling_freq_index, 11u);
-    static const int32_t kSamplingFreq[] = {
-        96000, 88200, 64000, 48000, 44100, 32000, 24000, 22050,
-        16000, 12000, 11025, 8000
-    };
-    meta->setInt32(kKeySampleRate, kSamplingFreq[sampling_freq_index]);
-    meta->setInt32(kKeyChannelCount, channel_configuration);
-
-    static const uint8_t kStaticESDS[] = {
-        0x03, 22,
-        0x00, 0x00,     // ES_ID
-        0x00,           // streamDependenceFlag, URL_Flag, OCRstreamFlag
-
-        0x04, 17,
-        0x40,                       // Audio ISO/IEC 14496-3
-        0x00, 0x00, 0x00, 0x00,
-        0x00, 0x00, 0x00, 0x00,
-        0x00, 0x00, 0x00, 0x00,
-
-        0x05, 2,
-        // AudioSpecificInfo follows
-
-        // oooo offf fccc c000
-        // o - audioObjectType
-        // f - samplingFreqIndex
-        // c - channelConfig
-    };
-    sp<ABuffer> csd = new ABuffer(sizeof(kStaticESDS) + 2);
-    memcpy(csd->data(), kStaticESDS, sizeof(kStaticESDS));
-
-    csd->data()[sizeof(kStaticESDS)] =
-        ((profile + 1) << 3) | (sampling_freq_index >> 1);
-
-    csd->data()[sizeof(kStaticESDS) + 1] =
-        ((sampling_freq_index << 7) & 0x80) | (channel_configuration << 3);
-
-    meta->setData(kKeyESDS, 0, csd->data(), csd->size());
-
-    return meta;
-}
-
 struct NALPosition {
     size_t nalOffset;
     size_t nalSize;
diff --git a/media/libstagefright/mpeg2ts/ESQueue.h b/media/libstagefright/mpeg2ts/ESQueue.h
index 5b7957e..d081995 100644
--- a/media/libstagefright/mpeg2ts/ESQueue.h
+++ b/media/libstagefright/mpeg2ts/ESQueue.h
@@ -61,10 +61,6 @@ private:
     // returns its timestamp in us (or -1 if no time information).
     int64_t fetchTimestamp(size_t size);
 
-    static sp<MetaData> MakeAACCodecSpecificData(
-            unsigned profile, unsigned sampling_freq_index,
-            unsigned channel_configuration);
-
     DISALLOW_EVIL_CONSTRUCTORS(ElementaryStreamQueue);
 };