Merge "Implements TtmlRenderer."

1 files changed, 751 insertions, 0 deletions

diff --git a/media/java/android/media/TtmlRenderer.java b/media/java/android/media/TtmlRenderer.java
new file mode 100644
index 0000000..0309334
--- /dev/null
+++ b/media/java/android/media/TtmlRenderer.java
@@ -0,0 +1,751 @@
+/*
+ * Copyright (C) 2014 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.
+ */
+
+package android.media;
+
+import android.content.Context;
+import android.graphics.Color;
+import android.media.SubtitleTrack.RenderingWidget.OnChangedListener;
+import android.text.Layout.Alignment;
+import android.text.SpannableStringBuilder;
+import android.text.TextUtils;
+import android.util.ArrayMap;
+import android.util.AttributeSet;
+import android.util.Log;
+import android.view.Gravity;
+import android.view.View;
+import android.view.ViewGroup;
+import android.view.View.MeasureSpec;
+import android.view.ViewGroup.LayoutParams;
+import android.view.accessibility.CaptioningManager;
+import android.view.accessibility.CaptioningManager.CaptionStyle;
+import android.view.accessibility.CaptioningManager.CaptioningChangeListener;
+import android.widget.LinearLayout;
+import android.widget.TextView;
+
+import com.android.internal.widget.SubtitleView;
+
+import java.io.IOException;
+import java.io.StringReader;
+import java.util.ArrayList;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.TreeSet;
+import java.util.Vector;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+import org.xmlpull.v1.XmlPullParser;
+import org.xmlpull.v1.XmlPullParserException;
+import org.xmlpull.v1.XmlPullParserFactory;
+
+/** @hide */
+public class TtmlRenderer extends SubtitleController.Renderer {
+    private final Context mContext;
+
+    private static final String MEDIA_MIMETYPE_TEXT_TTML = "application/ttml+xml";
+
+    private TtmlRenderingWidget mRenderingWidget;
+
+    public TtmlRenderer(Context context) {
+        mContext = context;
+    }
+
+    @Override
+    public boolean supports(MediaFormat format) {
+        if (format.containsKey(MediaFormat.KEY_MIME)) {
+            return format.getString(MediaFormat.KEY_MIME).equals(MEDIA_MIMETYPE_TEXT_TTML);
+        }
+        return false;
+    }
+
+    @Override
+    public SubtitleTrack createTrack(MediaFormat format) {
+        if (mRenderingWidget == null) {
+            mRenderingWidget = new TtmlRenderingWidget(mContext);
+        }
+        return new TtmlTrack(mRenderingWidget, format);
+    }
+}
+
+/**
+ * A class which provides utillity methods for TTML parsing.
+ *
+ * @hide
+ */
+final class TtmlUtils {
+    public static final String TAG_TT = "tt";
+    public static final String TAG_HEAD = "head";
+    public static final String TAG_BODY = "body";
+    public static final String TAG_DIV = "div";
+    public static final String TAG_P = "p";
+    public static final String TAG_SPAN = "span";
+    public static final String TAG_BR = "br";
+    public static final String TAG_STYLE = "style";
+    public static final String TAG_STYLING = "styling";
+    public static final String TAG_LAYOUT = "layout";
+    public static final String TAG_REGION = "region";
+    public static final String TAG_METADATA = "metadata";
+    public static final String TAG_SMPTE_IMAGE = "smpte:image";
+    public static final String TAG_SMPTE_DATA = "smpte:data";
+    public static final String TAG_SMPTE_INFORMATION = "smpte:information";
+    public static final String PCDATA = "#pcdata";
+    public static final String ATTR_BEGIN = "begin";
+    public static final String ATTR_DURATION = "dur";
+    public static final String ATTR_END = "end";
+    public static final long INVALID_TIMESTAMP = Long.MAX_VALUE;
+
+    /**
+     * Time expression RE according to the spec:
+     * http://www.w3.org/TR/ttaf1-dfxp/#timing-value-timeExpression
+     */
+    private static final Pattern CLOCK_TIME = Pattern.compile(
+            "^([0-9][0-9]+):([0-9][0-9]):([0-9][0-9])"
+            + "(?:(\\.[0-9]+)|:([0-9][0-9])(?:\\.([0-9]+))?)?$");
+
+    private static final Pattern OFFSET_TIME = Pattern.compile(
+            "^([0-9]+(?:\\.[0-9]+)?)(h|m|s|ms|f|t)$");
+
+    private TtmlUtils() {
+    }
+
+    /**
+     * Parses the given time expression and returns a timestamp in millisecond.
+     * <p>
+     * For the format of the time expression, please refer <a href=
+     * "http://www.w3.org/TR/ttaf1-dfxp/#timing-value-timeExpression">timeExpression</a>
+     *
+     * @param time A string which includes time expression.
+     * @param frameRate the framerate of the stream.
+     * @param subframeRate the sub-framerate of the stream
+     * @param tickRate the tick rate of the stream.
+     * @return the parsed timestamp in micro-second.
+     * @throws NumberFormatException if the given string does not match to the
+     *             format.
+     */
+    public static long parseTimeExpression(String time, int frameRate, int subframeRate,
+            int tickRate) throws NumberFormatException {
+        Matcher matcher = CLOCK_TIME.matcher(time);
+        if (matcher.matches()) {
+            String hours = matcher.group(1);
+            double durationSeconds = Long.parseLong(hours) * 3600;
+            String minutes = matcher.group(2);
+            durationSeconds += Long.parseLong(minutes) * 60;
+            String seconds = matcher.group(3);
+            durationSeconds += Long.parseLong(seconds);
+            String fraction = matcher.group(4);
+            durationSeconds += (fraction != null) ? Double.parseDouble(fraction) : 0;
+            String frames = matcher.group(5);
+            durationSeconds += (frames != null) ? ((double)Long.parseLong(frames)) / frameRate : 0;
+            String subframes = matcher.group(6);
+            durationSeconds += (subframes != null) ? ((double)Long.parseLong(subframes))
+                    / subframeRate / frameRate
+                    : 0;
+            return (long)(durationSeconds * 1000);
+        }
+        matcher = OFFSET_TIME.matcher(time);
+        if (matcher.matches()) {
+            String timeValue = matcher.group(1);
+            double value = Double.parseDouble(timeValue);
+            String unit = matcher.group(2);
+            if (unit.equals("h")) {
+                value *= 3600L * 1000000L;
+            } else if (unit.equals("m")) {
+                value *= 60 * 1000000;
+            } else if (unit.equals("s")) {
+                value *= 1000000;
+            } else if (unit.equals("ms")) {
+                value *= 1000;
+            } else if (unit.equals("f")) {
+                value = value / frameRate * 1000000;
+            } else if (unit.equals("t")) {
+                value = value / tickRate * 1000000;
+            }
+            return (long)value;
+        }
+        throw new NumberFormatException("Malformed time expression : " + time);
+    }
+
+    /**
+     * Applies <a href
+     * src="http://www.w3.org/TR/ttaf1-dfxp/#content-attribute-space">the
+     * default space policy</a> to the given string.
+     *
+     * @param in A string to apply the policy.
+     */
+    public static String applyDefaultSpacePolicy(String in) {
+        return applySpacePolicy(in, true);
+    }
+
+    /**
+     * Applies the space policy to the given string. This applies <a href
+     * src="http://www.w3.org/TR/ttaf1-dfxp/#content-attribute-space">the
+     * default space policy</a> with linefeed-treatment as treat-as-space
+     * or preserve.
+     *
+     * @param in A string to apply the policy.
+     * @param treatLfAsSpace Whether convert line feeds to spaces or not.
+     */
+    public static String applySpacePolicy(String in, boolean treatLfAsSpace) {
+        // Removes CR followed by LF. ref:
+        // http://www.w3.org/TR/xml/#sec-line-ends
+        String crRemoved = in.replaceAll("\r\n", "\n");
+        // Apply suppress-at-line-break="auto" and
+        // white-space-treatment="ignore-if-surrounding-linefeed"
+        String spacesNeighboringLfRemoved = crRemoved.replaceAll(" *\n *", "\n");
+        // Apply linefeed-treatment="treat-as-space"
+        String lfToSpace = treatLfAsSpace ? spacesNeighboringLfRemoved.replaceAll("\n", " ")
+                : spacesNeighboringLfRemoved;
+        // Apply white-space-collapse="true"
+        String spacesCollapsed = lfToSpace.replaceAll("[ \t\\x0B\f\r]+", " ");
+        return spacesCollapsed;
+    }
+
+    /**
+     * Returns the timed text for the given time period.
+     *
+     * @param root The root node of the TTML document.
+     * @param startUs The start time of the time period in microsecond.
+     * @param endUs The end time of the time period in microsecond.
+     */
+    public static String extractText(TtmlNode root, long startUs, long endUs) {
+        StringBuilder text = new StringBuilder();
+        extractText(root, startUs, endUs, text, false);
+        return text.toString().replaceAll("\n$", "");
+    }
+
+    private static void extractText(TtmlNode node, long startUs, long endUs, StringBuilder out,
+            boolean inPTag) {
+        if (node.mName.equals(TtmlUtils.PCDATA) && inPTag) {
+            out.append(node.mText);
+        } else if (node.mName.equals(TtmlUtils.TAG_BR) && inPTag) {
+            out.append("\n");
+        } else if (node.mName.equals(TtmlUtils.TAG_METADATA)) {
+            // do nothing.
+        } else if (node.isActive(startUs, endUs)) {
+            boolean pTag = node.mName.equals(TtmlUtils.TAG_P);
+            int length = out.length();
+            for (int i = 0; i < node.mChildren.size(); ++i) {
+                extractText(node.mChildren.get(i), startUs, endUs, out, pTag || inPTag);
+            }
+            if (pTag && length != out.length()) {
+                out.append("\n");
+            }
+        }
+    }
+
+    /**
+     * Returns a TTML fragment string for the given time period.
+     *
+     * @param root The root node of the TTML document.
+     * @param startUs The start time of the time period in microsecond.
+     * @param endUs The end time of the time period in microsecond.
+     */
+    public static String extractTtmlFragment(TtmlNode root, long startUs, long endUs) {
+        StringBuilder fragment = new StringBuilder();
+        extractTtmlFragment(root, startUs, endUs, fragment);
+        return fragment.toString();
+    }
+
+    private static void extractTtmlFragment(TtmlNode node, long startUs, long endUs,
+            StringBuilder out) {
+        if (node.mName.equals(TtmlUtils.PCDATA)) {
+            out.append(node.mText);
+        } else if (node.mName.equals(TtmlUtils.TAG_BR)) {
+            out.append("<br/>");
+        } else if (node.isActive(startUs, endUs)) {
+            out.append("<");
+            out.append(node.mName);
+            out.append(node.mAttributes);
+            out.append(">");
+            for (int i = 0; i < node.mChildren.size(); ++i) {
+                extractTtmlFragment(node.mChildren.get(i), startUs, endUs, out);
+            }
+            out.append("</");
+            out.append(node.mName);
+            out.append(">");
+        }
+    }
+}
+
+/**
+ * A container class which represents a cue in TTML.
+ * @hide
+ */
+class TtmlCue extends SubtitleTrack.Cue {
+    public String mText;
+    public String mTtmlFragment;
+
+    public TtmlCue(long startTimeMs, long endTimeMs, String text, String ttmlFragment) {
+        this.mStartTimeMs = startTimeMs;
+        this.mEndTimeMs = endTimeMs;
+        this.mText = text;
+        this.mTtmlFragment = ttmlFragment;
+    }
+}
+
+/**
+ * A container class which represents a node in TTML.
+ *
+ * @hide
+ */
+class TtmlNode {
+    public final String mName;
+    public final String mAttributes;
+    public final TtmlNode mParent;
+    public final String mText;
+    public final List<TtmlNode> mChildren = new ArrayList<TtmlNode>();
+    public final long mRunId;
+    public final long mStartTimeMs;
+    public final long mEndTimeMs;
+
+    public TtmlNode(String name, String attributes, String text, long startTimeMs, long endTimeMs,
+            TtmlNode parent, long runId) {
+        this.mName = name;
+        this.mAttributes = attributes;
+        this.mText = text;
+        this.mStartTimeMs = startTimeMs;
+        this.mEndTimeMs = endTimeMs;
+        this.mParent = parent;
+        this.mRunId = runId;
+    }
+
+    /**
+     * Check if this node is active in the given time range.
+     *
+     * @param startTimeMs The start time of the range to check in microsecond.
+     * @param endTimeMs The end time of the range to check in microsecond.
+     * @return return true if the given range overlaps the time range of this
+     *         node.
+     */
+    public boolean isActive(long startTimeMs, long endTimeMs) {
+        return this.mEndTimeMs > startTimeMs && this.mStartTimeMs < endTimeMs;
+    }
+}
+
+/**
+ * A simple TTML parser (http://www.w3.org/TR/ttaf1-dfxp/) which supports DFXP
+ * presentation profile.
+ * <p>
+ * Supported features in this parser are:
+ * <ul>
+ * <li>content
+ * <li>core
+ * <li>presentation
+ * <li>profile
+ * <li>structure
+ * <li>time-offset
+ * <li>timing
+ * <li>tickRate
+ * <li>time-clock-with-frames
+ * <li>time-clock
+ * <li>time-offset-with-frames
+ * <li>time-offset-with-ticks
+ * </ul>
+ * </p>
+ *
+ * @hide
+ */
+class TtmlParser {
+    static final String TAG = "TtmlParser";
+
+    // TODO: read and apply the following attributes if specified.
+    private static final int DEFAULT_FRAMERATE = 30;
+    private static final int DEFAULT_SUBFRAMERATE = 1;
+    private static final int DEFAULT_TICKRATE = 1;
+
+    private XmlPullParser mParser;
+    private final TtmlNodeListener mListener;
+    private long mCurrentRunId;
+
+    public TtmlParser(TtmlNodeListener listener) {
+        mListener = listener;
+    }
+
+    /**
+     * Parse TTML data. Once this is called, all the previous data are
+     * reset and it starts parsing for the given text.
+     *
+     * @param ttmlText TTML text to parse.
+     * @throws XmlPullParserException
+     * @throws IOException
+     */
+    public void parse(String ttmlText, long runId) throws XmlPullParserException, IOException {
+        mParser = null;
+        mCurrentRunId = runId;
+        loadParser(ttmlText);
+        parseTtml();
+    }
+
+    private void loadParser(String ttmlFragment) throws XmlPullParserException {
+        XmlPullParserFactory factory = XmlPullParserFactory.newInstance();
+        factory.setNamespaceAware(false);
+        mParser = factory.newPullParser();
+        StringReader in = new StringReader(ttmlFragment);
+        mParser.setInput(in);
+    }
+
+    private void extractAttribute(XmlPullParser parser, int i, StringBuilder out) {
+        out.append(" ");
+        out.append(parser.getAttributeName(i));
+        out.append("=\"");
+        out.append(parser.getAttributeValue(i));
+        out.append("\"");
+    }
+
+    private void parseTtml() throws XmlPullParserException, IOException {
+        LinkedList<TtmlNode> nodeStack = new LinkedList<TtmlNode>();
+        int depthInUnsupportedTag = 0;
+        boolean active = true;
+        while (!isEndOfDoc()) {
+            int eventType = mParser.getEventType();
+            TtmlNode parent = nodeStack.peekLast();
+            if (active) {
+                if (eventType == XmlPullParser.START_TAG) {
+                    if (!isSupportedTag(mParser.getName())) {
+                        Log.w(TAG, "Unsupported tag " + mParser.getName() + " is ignored.");
+                        depthInUnsupportedTag++;
+                        active = false;
+                    } else {
+                        TtmlNode node = parseNode(parent);
+                        nodeStack.addLast(node);
+                        if (parent != null) {
+                            parent.mChildren.add(node);
+                        }
+                    }
+                } else if (eventType == XmlPullParser.TEXT) {
+                    String text = TtmlUtils.applyDefaultSpacePolicy(mParser.getText());
+                    if (!TextUtils.isEmpty(text)) {
+                        parent.mChildren.add(new TtmlNode(
+                                TtmlUtils.PCDATA, "", text, 0, TtmlUtils.INVALID_TIMESTAMP,
+                                parent, mCurrentRunId));
+
+                    }
+                } else if (eventType == XmlPullParser.END_TAG) {
+                    if (mParser.getName().equals(TtmlUtils.TAG_P)) {
+                        mListener.onTtmlNodeParsed(nodeStack.getLast());
+                    } else if (mParser.getName().equals(TtmlUtils.TAG_TT)) {
+                        mListener.onRootNodeParsed(nodeStack.getLast());
+                    }
+                    nodeStack.removeLast();
+                }
+            } else {
+                if (eventType == XmlPullParser.START_TAG) {
+                    depthInUnsupportedTag++;
+                } else if (eventType == XmlPullParser.END_TAG) {
+                    depthInUnsupportedTag--;
+                    if (depthInUnsupportedTag == 0) {
+                        active = true;
+                    }
+                }
+            }
+            mParser.next();
+        }
+    }
+
+    private TtmlNode parseNode(TtmlNode parent) throws XmlPullParserException, IOException {
+        int eventType = mParser.getEventType();
+        if (!(eventType == XmlPullParser.START_TAG)) {
+            return null;
+        }
+        StringBuilder attrStr = new StringBuilder();
+        long start = 0;
+        long end = TtmlUtils.INVALID_TIMESTAMP;
+        long dur = 0;
+        for (int i = 0; i < mParser.getAttributeCount(); ++i) {
+            String attr = mParser.getAttributeName(i);
+            String value = mParser.getAttributeValue(i);
+            // TODO: check if it's safe to ignore the namespace of attributes as follows.
+            attr = attr.replaceFirst("^.*:", "");
+            if (attr.equals(TtmlUtils.ATTR_BEGIN)) {
+                start = TtmlUtils.parseTimeExpression(value, DEFAULT_FRAMERATE,
+                        DEFAULT_SUBFRAMERATE, DEFAULT_TICKRATE);
+            } else if (attr.equals(TtmlUtils.ATTR_END)) {
+                end = TtmlUtils.parseTimeExpression(value, DEFAULT_FRAMERATE, DEFAULT_SUBFRAMERATE,
+                        DEFAULT_TICKRATE);
+            } else if (attr.equals(TtmlUtils.ATTR_DURATION)) {
+                dur = TtmlUtils.parseTimeExpression(value, DEFAULT_FRAMERATE, DEFAULT_SUBFRAMERATE,
+                        DEFAULT_TICKRATE);
+            } else {
+                extractAttribute(mParser, i, attrStr);
+            }
+        }
+        if (parent != null) {
+            start += parent.mStartTimeMs;
+            if (end != TtmlUtils.INVALID_TIMESTAMP) {
+                end += parent.mStartTimeMs;
+            }
+        }
+        if (dur > 0) {
+            if (end != TtmlUtils.INVALID_TIMESTAMP) {
+                Log.e(TAG, "'dur' and 'end' attributes are defined at the same time." +
+                        "'end' value is ignored.");
+            }
+            end = start + dur;
+        }
+        if (parent != null) {
+            // If the end time remains unspecified, then the end point is
+            // interpreted as the end point of the external time interval.
+            if (end == TtmlUtils.INVALID_TIMESTAMP &&
+                    parent.mEndTimeMs != TtmlUtils.INVALID_TIMESTAMP &&
+                    end > parent.mEndTimeMs) {
+                end = parent.mEndTimeMs;
+            }
+        }
+        TtmlNode node = new TtmlNode(mParser.getName(), attrStr.toString(), null, start, end,
+                parent, mCurrentRunId);
+        return node;
+    }
+
+    private boolean isEndOfDoc() throws XmlPullParserException {
+        return (mParser.getEventType() == XmlPullParser.END_DOCUMENT);
+    }
+
+    private static boolean isSupportedTag(String tag) {
+        if (tag.equals(TtmlUtils.TAG_TT) || tag.equals(TtmlUtils.TAG_HEAD) ||
+                tag.equals(TtmlUtils.TAG_BODY) || tag.equals(TtmlUtils.TAG_DIV) ||
+                tag.equals(TtmlUtils.TAG_P) || tag.equals(TtmlUtils.TAG_SPAN) ||
+                tag.equals(TtmlUtils.TAG_BR) || tag.equals(TtmlUtils.TAG_STYLE) ||
+                tag.equals(TtmlUtils.TAG_STYLING) || tag.equals(TtmlUtils.TAG_LAYOUT) ||
+                tag.equals(TtmlUtils.TAG_REGION) || tag.equals(TtmlUtils.TAG_METADATA) ||
+                tag.equals(TtmlUtils.TAG_SMPTE_IMAGE) || tag.equals(TtmlUtils.TAG_SMPTE_DATA) ||
+                tag.equals(TtmlUtils.TAG_SMPTE_INFORMATION)) {
+            return true;
+        }
+        return false;
+    }
+}
+
+/** @hide */
+interface TtmlNodeListener {
+    void onTtmlNodeParsed(TtmlNode node);
+    void onRootNodeParsed(TtmlNode node);
+}
+
+/** @hide */
+class TtmlTrack extends SubtitleTrack implements TtmlNodeListener {
+    private static final String TAG = "TtmlTrack";
+
+    private final TtmlParser mParser = new TtmlParser(this);
+    private final TtmlRenderingWidget mRenderingWidget;
+    private String mParsingData;
+    private Long mCurrentRunID;
+
+    private final LinkedList<TtmlNode> mTtmlNodes;
+    private final TreeSet<Long> mTimeEvents;
+    private TtmlNode mRootNode;
+
+    TtmlTrack(TtmlRenderingWidget renderingWidget, MediaFormat format) {
+        super(format);
+
+        mTtmlNodes = new LinkedList<TtmlNode>();
+        mTimeEvents = new TreeSet<Long>();
+        mRenderingWidget = renderingWidget;
+        mParsingData = "";
+    }
+
+    @Override
+    public TtmlRenderingWidget getRenderingWidget() {
+        return mRenderingWidget;
+    }
+
+    @Override
+    public void onData(String data, boolean eos, long runID) {
+        // implement intermixing restriction for TTML.
+        synchronized(mParser) {
+            if (mCurrentRunID != null && runID != mCurrentRunID) {
+                throw new IllegalStateException(
+                        "Run #" + mCurrentRunID +
+                        " in progress.  Cannot process run #" + runID);
+            }
+            mCurrentRunID = runID;
+            mParsingData += data;
+            if (eos) {
+                try {
+                    mParser.parse(mParsingData, mCurrentRunID);
+                } catch (XmlPullParserException e) {
+                    e.printStackTrace();
+                } catch (IOException e) {
+                    e.printStackTrace();
+                }
+                finishedRun(runID);
+                mParsingData = "";
+                mCurrentRunID = null;
+            }
+        }
+    }
+
+    @Override
+    public void onTtmlNodeParsed(TtmlNode node) {
+        mTtmlNodes.addLast(node);
+        addTimeEvents(node);
+    }
+
+    @Override
+    public void onRootNodeParsed(TtmlNode node) {
+        mRootNode = node;
+        TtmlCue cue = null;
+        while ((cue = getNextResult()) != null) {
+            addCue(cue);
+        }
+        mRootNode = null;
+        mTtmlNodes.clear();
+        mTimeEvents.clear();
+    }
+
+    @Override
+    public void updateView(Vector<SubtitleTrack.Cue> activeCues) {
+        if (!mVisible) {
+            // don't keep the state if we are not visible
+            return;
+        }
+
+        if (DEBUG && mTimeProvider != null) {
+            try {
+                Log.d(TAG, "at " +
+                        (mTimeProvider.getCurrentTimeUs(false, true) / 1000) +
+                        " ms the active cues are:");
+            } catch (IllegalStateException e) {
+                Log.d(TAG, "at (illegal state) the active cues are:");
+            }
+        }
+
+        mRenderingWidget.setActiveCues(activeCues);
+    }
+
+    /**
+     * Returns a {@link TtmlCue} in the presentation time order.
+     * {@code null} is returned if there is no more timed text to show.
+     */
+    public TtmlCue getNextResult() {
+        while (mTimeEvents.size() >= 2) {
+            long start = mTimeEvents.pollFirst();
+            long end = mTimeEvents.first();
+            List<TtmlNode> activeCues = getActiveNodes(start, end);
+            if (!activeCues.isEmpty()) {
+                return new TtmlCue(start, end,
+                        TtmlUtils.applySpacePolicy(TtmlUtils.extractText(
+                                mRootNode, start, end), false),
+                        TtmlUtils.extractTtmlFragment(mRootNode, start, end));
+            }
+        }
+        return null;
+    }
+
+    private void addTimeEvents(TtmlNode node) {
+        mTimeEvents.add(node.mStartTimeMs);
+        mTimeEvents.add(node.mEndTimeMs);
+        for (int i = 0; i < node.mChildren.size(); ++i) {
+            addTimeEvents(node.mChildren.get(i));
+        }
+    }
+
+    private List<TtmlNode> getActiveNodes(long startTimeUs, long endTimeUs) {
+        List<TtmlNode> activeNodes = new ArrayList<TtmlNode>();
+        for (int i = 0; i < mTtmlNodes.size(); ++i) {
+            TtmlNode node = mTtmlNodes.get(i);
+            if (node.isActive(startTimeUs, endTimeUs)) {
+                activeNodes.add(node);
+            }
+        }
+        return activeNodes;
+    }
+}
+
+/**
+ * Widget capable of rendering TTML captions.
+ *
+ * @hide
+ */
+class TtmlRenderingWidget extends LinearLayout implements SubtitleTrack.RenderingWidget {
+
+    /** Callback for rendering changes. */
+    private OnChangedListener mListener;
+    private final TextView mTextView;
+
+    public TtmlRenderingWidget(Context context) {
+        this(context, null);
+    }
+
+    public TtmlRenderingWidget(Context context, AttributeSet attrs) {
+        this(context, attrs, 0);
+    }
+
+    public TtmlRenderingWidget(Context context, AttributeSet attrs, int defStyleAttr) {
+        this(context, attrs, defStyleAttr, 0);
+    }
+
+    public TtmlRenderingWidget(Context context, AttributeSet attrs, int defStyleAttr,
+            int defStyleRes) {
+        super(context, attrs, defStyleAttr, defStyleRes);
+        // Cannot render text over video when layer type is hardware.
+        setLayerType(View.LAYER_TYPE_SOFTWARE, null);
+
+        CaptioningManager captionManager = (CaptioningManager) context.getSystemService(
+                Context.CAPTIONING_SERVICE);
+        mTextView = new TextView(context);
+        mTextView.setTextColor(captionManager.getUserStyle().foregroundColor);
+        addView(mTextView, LayoutParams.MATCH_PARENT, LayoutParams.MATCH_PARENT);
+        mTextView.setGravity(Gravity.BOTTOM | Gravity.CENTER_HORIZONTAL);
+    }
+
+    @Override
+    public void setOnChangedListener(OnChangedListener listener) {
+        mListener = listener;
+    }
+
+    @Override
+    public void setSize(int width, int height) {
+        final int widthSpec = MeasureSpec.makeMeasureSpec(width, MeasureSpec.EXACTLY);
+        final int heightSpec = MeasureSpec.makeMeasureSpec(height, MeasureSpec.EXACTLY);
+
+        measure(widthSpec, heightSpec);
+        layout(0, 0, width, height);
+    }
+
+    @Override
+    public void setVisible(boolean visible) {
+        if (visible) {
+            setVisibility(View.VISIBLE);
+        } else {
+            setVisibility(View.GONE);
+        }
+    }
+
+    @Override
+    public void onAttachedToWindow() {
+        super.onAttachedToWindow();
+    }
+
+    @Override
+    public void onDetachedFromWindow() {
+        super.onDetachedFromWindow();
+    }
+
+    public void setActiveCues(Vector<SubtitleTrack.Cue> activeCues) {
+        final int count = activeCues.size();
+        String subtitleText = "";
+        for (int i = 0; i < count; i++) {
+            TtmlCue cue = (TtmlCue) activeCues.get(i);
+            subtitleText += cue.mText + "\n";
+        }
+        mTextView.setText(subtitleText);
+
+        if (mListener != null) {
+            mListener.onChanged(this);
+        }
+    }
+}