TPG 图片加速 SDK 接入-文档中心-腾讯云

本文档以 Android 客户端接入 SDK 为例，介绍如何使用 SDK。
具体实现可以参考 Android 客户端 SDK Java 目录下的 TPGDemo 范例。
Android SDK 接入步骤一：加载解码库将解码库放在工程的lib目录下，工程代码里需加载 TPG 解码库：
System.loadLibrary（“TPGDecoder”）
该解码库提供了以下JNI函数：
// interface using bitstream
private native int CreateDecoder(byte[] pStream);
﻿
private native int DecodeImage(long hObj, byte[] pStream, int index,
        TPGOutFrame tpgOutFrame);
﻿
private native int DecodeImageToBitmap(long hObj, byte[] pStream, int index,
        Bitmap bitmap, Integer delayTime);
﻿
private native void CloseDecoder(long hObj);
﻿
private native int ParseHeader(byte[] pStream, TPGFeature info);
﻿
private native int GetDelayTime(long hObj, byte[] pStream, int index);
﻿
private native byte[] GetAdditionalInfo(long hObj, byte[] pStream, int id);
﻿
// interface using file path as input
private native long CreateDecoder2(String tpg_path);
﻿
private native int DecodeImage2(long hObj, int index, TPGOutFrame tpgOutFrame);
﻿
private native int DecodeImageToBitmap2(long hObj, int index, Bitmap bitmap,
        Integer delayTime);
﻿
private native void CloseDecoder2(long hObj);
﻿
private native int ParseHeader2(String tpgpath, TPGFeature info);
﻿
private native int GetDelayTime2(long hObj, int index);
﻿
private native byte[] GetAdditionalInfo2(long hObj, int id);
﻿
private native int GetVersion(TPGVersionNum hObj);
步骤二：调用 TPGDecoder 接口
详细接口信息，见 API 接口。
接口使用方法
1. 创建 TPGDecoder 对象，调用如下接口。确认输入图片是否为 tpg 格式。
private native int ParseHeader(byte[] pStream, TPGFeature info)
﻿
从传入的 TPGFeature对象可以判断当前图片的宽度、高度、帧数和图片类型：
public class TPGFeature {
     int width;
     int height;
     int headerSize;
     int imageMode;
     int frameCount;
     int version;
}
图片类型可分为4类：
public static final int IMAGE_MODE_Normal = 0;
public static final int IMAGE_MODE_EncodeAlpha = 1;
public static final int IMAGE_MODE_BlendAlpha = 2;
public static final int IMAGE_MODE_Animation = 3;
public static final int IMAGE_MODE_AnimationWithAlpha = 4;
2. 图片为动图和非动图时实现方式如下：
如果图片为动图IMAGE_MODE_Animation或IMAGE_MODE_AnimationWithAlpha，可参考 TPGDemo 里 TPGDecoder.java 中的decodeOneFrame2 ()函数的实现。
public int decodeOneFrame2(int index, int[] outData, Bitmap bm,
        int[] delayTime) {
﻿
	int res = 0;
TPGOutFrame tpgOutFrame = new TPGOutFrame();
tpgOutFrame.pOutBuf = outData;
tpgOutFrame.dstWidth = mFeatureInfo.width;
tpgOutFrame.dstHeight = mFeatureInfo.height;
tpgOutFrame.fmt = Utils.FORMAT_BGRA;
if (DecodeImage2(mhDec, index, tpgOutFrame) > 0) {
    System.out.println("decode error: ");
}
delayTime[0] = tpgOutFrame.delayTime;
﻿
	bm.setPixels(outData, 0, mFeatureInfo.width, 0, 0, mFeatureInfo.width,
        mFeatureInfo.height);
﻿
	return res;
}
如果图片为非动图，可参考 TPGDemo 里 TPGDecoder.java 中的decodeTPG2()函数的实现。
public Bitmap decodeTPG2(String tpgPath, int format, int dstWidth) {
    Bitmap bm = null;
﻿
	int dstHeight = 0;
﻿
	TPGFeature info = new TPGFeature();
﻿
	int res = ParseHeader2(tpgPath, info);
﻿
	if (res != Utils.TPG_STATUS_OK) {
    return null;
}
mhDec = CreateDecoder2(tpgPath);
﻿
	if (mhDec == 0) {
    return null;
}
﻿
	imageWidth = info.width;
imageHeight = info.height;
﻿
	dstHeight = (int) ((double) imageHeight / (double) imageWidth * dstWidth);
﻿
	if (dstWidth > imageWidth || dstHeight > imageHeight) {
    dstWidth = imageWidth;
    dstHeight = imageHeight;
}
// System.out.println("createBitmap start!" + imageWidth*imageHeight*4);
bm = Bitmap.createBitmap(dstWidth, dstHeight, Bitmap.Config.ARGB_8888);
if(info.imageMode==Utils.IMAGE_MODE_Normal)
{
    bm.setHasAlpha(false);
}
﻿
	// System.out.println("createBitmap end!");
if (null == bm) {
    System.out.println("no memory!");
}
int delayTime = 0;
if (DecodeImageToBitmap2(mhDec, 0, bm, delayTime) > 0) {
    System.out.println("decode image error!");	
}
﻿
	//TPGVersionNum foo = new TPGVersionNum();
﻿
	//GetVersion(foo);
//System.out.println("version num:"+foo.num+"version str:"+foo.str);
CloseDecoder2(mhDec);
mhDec = 0;
﻿
	return bm;
}