FileDocCategorySizeDatePackage
MediaDataBox.javaAPI Docmp4parser 1.0-RC-177742Wed Dec 19 20:10:38 GMT 2012com.coremedia.iso.boxes.mdat

MediaDataBox.java

/*  
 * Copyright 2008 CoreMedia AG, Hamburg
 *
 * 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. 
 */

package com.coremedia.iso.boxes.mdat;

import com.coremedia.iso.BoxParser;
import com.coremedia.iso.ChannelHelper;
import com.coremedia.iso.boxes.Box;
import com.coremedia.iso.boxes.ContainerBox;
import com.googlecode.mp4parser.AbstractBox;

import java.io.IOException;
import java.nio.ByteBuffer;
import java.nio.channels.FileChannel;
import java.nio.channels.ReadableByteChannel;
import java.nio.channels.WritableByteChannel;
import java.util.logging.Logger;

import static com.googlecode.mp4parser.util.CastUtils.l2i;

/**
 * This box contains the media data. In video tracks, this box would contain video frames. A presentation may
 * contain zero or more Media Data Boxes. The actual media data follows the type field; its structure is described
 * by the metadata (see {@link com.coremedia.iso.boxes.SampleTableBox}).<br>
 * In large presentations, it may be desirable to have more data in this box than a 32-bit size would permit. In this
 * case, the large variant of the size field is used.<br>
 * There may be any number of these boxes in the file (including zero, if all the media data is in other files). The
 * metadata refers to media data by its absolute offset within the file (see {@link com.coremedia.iso.boxes.StaticChunkOffsetBox});
 * so Media Data Box headers and free space may easily be skipped, and files without any box structure may
 * also be referenced and used.
 */
public final class MediaDataBox implements Box {
    private static Logger LOG = Logger.getLogger(MediaDataBox.class.getName());

    public static final String TYPE = "mdat";
    public static final int BUFFER_SIZE = 10 * 1024 * 1024;
    ContainerBox parent;

    ByteBuffer header;

    // These fields are for the special case of a FileChannel as input.
    private FileChannel fileChannel;
    private long startPosition;
    private long contentSize;


    /**
     * If the whole content is just in one mapped buffer keep a strong reference to it so it is
     * not evicted from the cache.
     */
    private ByteBuffer content;

    public ContainerBox getParent() {
        return parent;
    }

    public void setParent(ContainerBox parent) {
        this.parent = parent;
    }

    public String getType() {
        return TYPE;
    }

    private static void transfer(FileChannel from, long position, long count, WritableByteChannel to) throws IOException {
        long maxCount = (64 * 1024 * 1024) - (32 * 1024);
        // Transfer data in chunks a bit less than 64MB
        // People state that this is a kind of magic number on Windows.
        // I don't care. The size seems reasonable.
        long offset = 0;
        while (offset < count) {
            offset += from.transferTo(position + offset, Math.min(maxCount, count - offset), to);
        }
    }

    public void getBox(WritableByteChannel writableByteChannel) throws IOException {
        if (fileChannel != null) {
            assert checkStillOk();
            transfer(fileChannel, startPosition - header.limit(), contentSize + header.limit(), writableByteChannel);
        } else {
            header.rewind();
            writableByteChannel.write(header);
            writableByteChannel.write(content);
        }
    }

    /**
     * If someone use the same file as source and sink it could the case that
     * inserting a few bytes before the mdat results in overwriting data we still
     * need to write this mdat here. This method just makes sure that we haven't already
     * overwritten the mdat contents.
     *
     * @return true if ok
     */
    private boolean checkStillOk() {
        try {
            fileChannel.position(startPosition - header.limit());
            ByteBuffer h2 = ByteBuffer.allocate(header.limit());
            fileChannel.read(h2);
            header.rewind();
            h2.rewind();
            assert h2.equals(header) : "It seems that the content I want to read has already been overwritten.";
            return true;
        } catch (IOException e) {
            e.printStackTrace();
            return false;
        }

    }


    public long getSize() {
        long size = header.limit();
        size += contentSize;
        return size;
    }

    public long getDataStartPosition() {
        return startPosition;
    }

    public long getDataEndPosition() {
        return startPosition + contentSize;
    }

    public void parse(ReadableByteChannel readableByteChannel, ByteBuffer header, long contentSize, BoxParser boxParser) throws IOException {
        this.header = header;
        this.contentSize = contentSize;

        if (readableByteChannel instanceof FileChannel && (contentSize > AbstractBox.MEM_MAP_THRESHOLD)) {
            this.fileChannel = ((FileChannel) readableByteChannel);
            this.startPosition = ((FileChannel) readableByteChannel).position();
            ((FileChannel) readableByteChannel).position(((FileChannel) readableByteChannel).position() + contentSize);
        } else {
            content = ChannelHelper.readFully(readableByteChannel, l2i(contentSize));
            startPosition = 0;
            for (Box box : this.getParent().getBoxes()) {
                startPosition += box.getSize();
            }
            startPosition += header.limit();
            cacheSliceCurrentlyInUse = content;
            cacheSliceCurrentlyInUseStart = 0;
        }
    }

    ByteBuffer cacheSliceCurrentlyInUse = null;
    long cacheSliceCurrentlyInUseStart = Long.MAX_VALUE;

    public synchronized ByteBuffer getContent(long offset, int length) {
        // most likely the last used cache slice will be used again
        if (cacheSliceCurrentlyInUseStart <= offset && cacheSliceCurrentlyInUse != null && offset + length <= cacheSliceCurrentlyInUseStart + cacheSliceCurrentlyInUse.limit()) {
            ByteBuffer cachedSample = cacheSliceCurrentlyInUse.asReadOnlyBuffer();
            cachedSample.position((int) (offset - cacheSliceCurrentlyInUseStart));
            cachedSample.mark();
            cachedSample.limit((int) (offset - cacheSliceCurrentlyInUseStart) + length);
            return cachedSample;
        }

        // CACHE MISS
        ByteBuffer cacheEntry;
        try {
            // Just mapping 10MB at a time. Seems reasonable.
            cacheEntry = fileChannel.map(FileChannel.MapMode.READ_ONLY, startPosition + offset, Math.min(BUFFER_SIZE, contentSize - offset));
        } catch (IOException e1) {
            LOG.fine("Even mapping just 10MB of the source file into the memory failed. " + e1);
            throw new RuntimeException(
                    "Delayed reading of mdat content failed. Make sure not to close " +
                            "the FileChannel that has been used to create the IsoFile!", e1);
        }
        cacheSliceCurrentlyInUse = cacheEntry;
        cacheSliceCurrentlyInUseStart = offset;
        ByteBuffer cachedSample = cacheSliceCurrentlyInUse.asReadOnlyBuffer();
        cachedSample.position(0);
        cachedSample.mark();
        cachedSample.limit(length);
        return cachedSample;
    }


    public ByteBuffer getHeader() {
        return header;
    }

}