android:获取富文本图片和使用Jsoup抓取腾讯新闻网页数据

先看效果：

• 获取富文本中的图片

• 抓取腾讯新闻中的图片

• 首先引入要使用的jar包

 compile 'jp.wasabeef:glide-transformations:2.0.2'
 compile 'org.jsoup:jsoup:1.9.2'

一、加载富文本图片自适应

• 关键代码如下：

package tsou.cn.webviewtext;

import android.os.Build;
import android.os.Bundle;
import android.support.v7.app.AppCompatActivity;
import android.view.View;
import android.webkit.WebChromeClient;
import android.webkit.WebView;

import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;

import tsou.cn.webviewtext.util.StringBrowserUtils;
import tsou.cn.webviewtext.util.data.Data;
import tsou.cn.webviewtext.webview.MJavascriptInterface;
import tsou.cn.webviewtext.webview.MyWebViewClient;

public class WebViewDataOneActivity extends AppCompatActivity {
    private WebView webview;
    private String[] imageUrls;

    @Override
    protected void onCreate(Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_web_view_data_one);
        webview = (WebView) findViewById(R.id.webview);

        imageUrls = StringBrowserUtils.returnImageUrlsFromHtml(Data.getData());
        //设置加载进来的页面自适应手机屏幕
        //第一个方法设置webview推荐使用的窗口，设置为true。
        webview.getSettings().setUseWideViewPort(true);
        // 第二个方法是设置webview加载的页面的模式，也设置为true。
        webview.getSettings().setLoadWithOverviewMode(true);
        //webview.getSettings().setDefaultFontSize(20);
        webview.getSettings().setTextZoom(260);
        webview.getSettings().setJavaScriptEnabled(true);
        webview.getSettings().setSupportZoom(false);
        webview.getSettings().setBuiltInZoomControls(false);
        webview.getSettings().setDisplayZoomControls(false);
        webview.setScrollBarStyle(View.SCROLLBARS_INSIDE_OVERLAY); //取消滚动条白边效果
        webview.setWebChromeClient(new WebChromeClient());
        webview.addJavascriptInterface(new MJavascriptInterface(this, imageUrls), "imagelistener");
        webview.setWebViewClient(new MyWebViewClient());
        webview.getSettings().setDefaultTextEncodingName("UTF-8");
        webview.getSettings().setBlockNetworkImage(false);
        if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.LOLLIPOP) {
            webview.getSettings().setMixedContentMode(webview.getSettings()
                    .MIXED_CONTENT_ALWAYS_ALLOW);  //注意安卓5.0以上的权限
        }
        webview.loadDataWithBaseURL(null, Data.getData(), "text/html", "UTF-8", null);
    }

}

MJavascriptInterface类和MyWebViewClient类下面提供。

二、加载富文本图片按宽度适配

• 关键代码：

package tsou.cn.webviewtext;

import android.os.Build;
import android.os.Bundle;
import android.support.v7.app.AppCompatActivity;
import android.view.View;
import android.webkit.WebChromeClient;
import android.webkit.WebView;

import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;

import tsou.cn.webviewtext.util.StringBrowserUtils;
import tsou.cn.webviewtext.util.data.Data;
import tsou.cn.webviewtext.webview.MJavascriptInterface;
import tsou.cn.webviewtext.webview.MyWebViewClient;

public class WebViewDatatwoActivity extends AppCompatActivity {
    private WebView webview;
    private String[] imageUrls;

    @Override
    protected void onCreate(Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_web_view_data_one);
        webview = (WebView) findViewById(R.id.webview);

        imageUrls = StringBrowserUtils.returnImageUrlsFromHtml(Data.getData());
        webview.getSettings().setJavaScriptEnabled(true);
        webview.getSettings().setUseWideViewPort(true);
        webview.getSettings().setLoadWithOverviewMode(true);
        webview.getSettings().setDefaultFontSize(40);
        webview.getSettings().setBuiltInZoomControls(true);
        webview.getSettings().setDisplayZoomControls(false);
        webview.setScrollBarStyle(View.SCROLLBARS_INSIDE_OVERLAY); //取消滚动条白边效果
        webview.setWebChromeClient(new WebChromeClient());
        webview.addJavascriptInterface(new MJavascriptInterface(this, imageUrls), "imagelistener");
        webview.setWebViewClient(new MyWebViewClient());
        webview.getSettings().setDefaultTextEncodingName("UTF-8");
        webview.getSettings().setBlockNetworkImage(false);
        if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.LOLLIPOP) {
            webview.getSettings().setMixedContentMode(webview.getSettings()
                    .MIXED_CONTENT_ALWAYS_ALLOW);  //注意安卓5.0以上的权限
        }
        webview.loadDataWithBaseURL(null, getNewContent(Data.getData()), "text/html", "UTF-8", null);
    }

    private String getNewContent(String htmltext) {

        Document doc = Jsoup.parse(htmltext);
        Elements elements = doc.getElementsByTag("img");
        for (Element element : elements) {
            element.attr("width", "100%").attr("height", "auto");
        }

        return doc.toString();
    }
}

使用Jsoup让图片宽度充满全屏

三、加载HTML使用Jsoup爬取数据

• 关键代码：

package tsou.cn.webviewtext;

import android.os.Bundle;
import android.support.v7.app.AppCompatActivity;
import android.util.Log;
import android.webkit.WebView;

import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.select.Elements;

import java.util.ArrayList;
import java.util.List;

import tsou.cn.webviewtext.util.data.Data;
import tsou.cn.webviewtext.webview.MJavascriptInterface;
import tsou.cn.webviewtext.webview.MyWebViewClient;

public class WebViewUrlTwoActivity extends AppCompatActivity {
    private WebView webview;
    private List imageSrcList;

    @Override
    protected void onDestroy() {
        webview.clearHistory();
        super.onDestroy();
        System.gc();
    }

    @Override
    protected void onCreate(Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_web_view_data_one);
        webview = (WebView) findViewById(R.id.webview);
        webview.loadUrl(Data.getUrl());
        new Thread(new Runnable() {
            @Override
            public void run() {
                getHtmlData();
            }
        }).start();
    }

    private void getHtmlData() {
        try {
            //从一个URL加载一个Document对象。
            Document doc = Jsoup.connect(Data.getUrl())
                    .userAgent("Mozilla")
                    .timeout(50000)
                    .get();
            Elements elements = doc.select("div.main").select("img");
            imageSrcList = new ArrayList<>();
            imageSrcList.clear();
            for (int i = 0; i < elements.size(); i++) {
                Log.e("huangxiaoguo", "elements" + elements.get(i).attr("src"));
                imageSrcList.add(elements.get(i).attr("src"));
            }

            runOnUiThread(new Runnable() {
                @Override
                public void run() {
                    setWebView(imageSrcList.toArray(new String[imageSrcList.size()]));
                }
            });
        } catch (Exception e) {
            e.printStackTrace();
        }
    }

    private void setWebView(String[] imageUrls) {
        webview.addJavascriptInterface(new MJavascriptInterface(this, imageUrls), "imagelistener");
        webview.setWebViewClient(new MyWebViewClient());

    }


}

四、其他相关代码

• 数据类Data

public class Data {
    public static String getUrl() {
        return "https://view.inews.qq.com/a/20171101A07WQO00";
    }
    public static String getData() {
        return "富文本"
    }
}

富文本太长，请到https://gitee.com/huangxiaoguo/WebViewText/blob/master/app/src/main/java/tsou/cn/webviewtext/util/data/Data.java查看。

• MJavascriptInterface类

package tsou.cn.webviewtext.webview;

import android.content.Context;
import android.content.Intent;
import android.util.Log;

import tsou.cn.webviewtext.LookBigPhotoActivity;

/**
 * Created by Administrator on 2017/2/10.
 */

public class MJavascriptInterface {
    private Context context;
    private String[] imageUrls;

    public MJavascriptInterface(Context context, String[] imageUrls) {
        this.context = context;
        this.imageUrls = imageUrls;
    }

    @android.webkit.JavascriptInterface
    public void openImage(String img) {
        Intent intent = new Intent();
        intent.putExtra("imageUrls", imageUrls);
        for (int i = 0; i < imageUrls.length; i++) {
            if (imageUrls[i].equals(img))
                intent.putExtra("position", i);
        }
        intent.setClass(context, LookBigPhotoActivity.class);
        context.startActivity(intent);

    }
}

• MyWebViewClient类

package tsou.cn.webviewtext.webview;

import android.content.Intent;
import android.graphics.Bitmap;
import android.net.Uri;
import android.util.Log;
import android.webkit.WebView;
import android.webkit.WebViewClient;

public class MyWebViewClient extends WebViewClient {
    @Override
    public void onPageFinished(WebView view, String url) {
        view.getSettings().setJavaScriptEnabled(true);
        super.onPageFinished(view, url);
        //待网页加载完全后设置图片点击的监听方法
        addImageClickListener(view);
    }

    @Override
    public boolean shouldOverrideUrlLoading(WebView view, String url) {
        // 调用系统默认浏览器处理url
        view.stopLoading();
        view.getContext().startActivity(new Intent(Intent.ACTION_VIEW, Uri.parse(url)));
        return true;
    }

    @Override
    public void onPageStarted(WebView view, String url, Bitmap favicon) {
        view.getSettings().setJavaScriptEnabled(true);
        super.onPageStarted(view, url, favicon);
    }

    private void addImageClickListener(WebView webView) {
        webView.loadUrl("javascript:(function(){" +
                "var objs = document.getElementsByTagName(\"img\"); " +
                "for(var i=0;i +
                "{"
                + " objs[i].οnclick=function() " +
                " { "
                //通过js代码找到标签为img的代码块，设置点击的监听方法与本地的openImage方法进行连接
                + "  window.imagelistener.openImage(this.src); " +
                " } " +
                "}" +
                "})()");
    }
}

• ImageLoadUtil类

package tsou.cn.webviewtext.util;

import android.content.Context;
import android.support.annotation.DrawableRes;
import android.widget.ImageView;

import com.bumptech.glide.Glide;
import com.bumptech.glide.load.engine.DiskCacheStrategy;

import java.io.File;

import jp.wasabeef.glide.transformations.CropCircleTransformation;
import jp.wasabeef.glide.transformations.CropTransformation;
import jp.wasabeef.glide.transformations.RoundedCornersTransformation;
import tsou.cn.webviewtext.R;


/**
 * 图片加载类封装
 *
 * @author RS
 */
public class ImageLoadUtil {

    public static ImageView display(Context context, ImageView img, String url) {
        Glide.with(context)
                .load(url)
                .placeholder(R.drawable.app_loading_pic) //加载中的图片
                .error(R.drawable.app_loading_pic) //加载失败的图片
                .into(img);
        return img;
    }

    public static ImageView display(Context context, ImageView img, File file) {
        Glide.with(context)
                .load(file)
                .placeholder(R.drawable.app_loading_pic) //加载中的图片
                .error(R.drawable.app_loading_pic) //加载失败的图片
                .into(img);
        return img;
    }

    public static ImageView displayCircle(Context context, ImageView img, String url) {
        Glide.with(context)
                .load(url)
                .placeholder(R.drawable.app_loading_pic_round) //加载中的图片
                .error(R.drawable.app_loading_pic_round) //加载失败的图片
                .override(150, 150)
                .bitmapTransform(new CropCircleTransformation(context))
                .into(img);
        return img;
    }

    public static ImageView displayCircle(Context context, ImageView img, File file) {
        Glide.with(context)
                .load(file)
                .override(150, 150)
                .placeholder(R.drawable.app_loading_pic_round) //加载中的图片
                .error(R.drawable.app_loading_pic_round) //加载失败的图片
                .bitmapTransform(new CropCircleTransformation(context))
                .into(img);
        return img;
    }

    public static ImageView displayCircle(Context context, ImageView img, String url, @DrawableRes int defaultPic) {
        Glide.with(context)
                .load(url)
                .override(150, 150)
                .placeholder(defaultPic) //加载中的图片
                .error(defaultPic) //加载失败的图片
                .bitmapTransform(new CropCircleTransformation(context))
                .into(img);
        return img;
    }

    public static ImageView displayRound(Context context, ImageView img, String url, int round, int width, int height) {
        Glide.with(context)
                .load(url)
                .placeholder(R.drawable.app_loading_pic) //加载中的图片
                .error(R.drawable.app_loading_pic) //加载失败的图片
                .bitmapTransform(new CropTransformation(context, width, height),
                        new RoundedCornersTransformation(context, UIUtils.dip2px(context, round), 0))
                .into(img);
        return img;
    }

    public static ImageView displaySquareRound(Context context, ImageView img, String url, int round, int length) {
        displayRound(context, img, url, round, length, length);
        return img;
    }

    public static ImageView displaySquareRound(Context context, ImageView img, String url, int length) {
        displaySquareRound(context, img, url, 4, length);
        return img;
    }

    /****************************/
    public static void display(Context context, String imgPath, ImageView imageView) {
        display(context, imgPath, imageView, R.drawable.app_loading_pic);
    }

    public static void display(Context context, String imgPath, ImageView imageView, @DrawableRes int resId) {
        Glide.with(context).load(imgPath).error(resId)
                .placeholder(resId) //加载中的图片
                .diskCacheStrategy(DiskCacheStrategy.ALL)
                .into(imageView);
    }

    public static void display(Context context, @DrawableRes int imgPath, ImageView imageView) {
        display(context, imgPath, imageView, R.drawable.app_loading_pic);
    }

    public static void display(Context context, @DrawableRes int imgPath, ImageView imageView, @DrawableRes int resId) {
        Glide.with(context).load(imgPath).error(resId)
                .placeholder(resId) //加载中的图片
                .diskCacheStrategy(DiskCacheStrategy.ALL)
                .into(imageView);
    }

    public static void display(Context context, String imgPath, ImageView imageView, int width, int height) {
        display(context, imgPath, imageView, R.drawable.app_loading_pic, width, height);
    }

    public static void display(Context context, String imgPath, ImageView imageView, @DrawableRes int resId, int width,
                               int height) {
        Glide.with(context).load(imgPath).error(resId)
                .diskCacheStrategy(DiskCacheStrategy.ALL)
                .override(width, height)
                .placeholder(resId) //加载中的图片
                .into(imageView);
    }

    public static void clearMemory(final Context mContext) {
        new Thread(new Runnable() {
            @Override
            public void run() {
                Glide.get(mContext).clearDiskCache();//清理磁盘缓存需要在子线程中执行
            }
        }).start();
        Glide.get(mContext).clearMemory();//清理内存缓存可以在UI主线程中进行
    }

}

• StringBrowserUtils类

package tsou.cn.webviewtext.util;

import android.util.Log;

import java.util.ArrayList;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

/**
 * Created by Administrator on 2017/2/9.
 */

public class StringBrowserUtils {

    public static String[] returnImageUrlsFromHtml(String content) {
        List imageSrcList = new ArrayList();
        String htmlCode = content;
        Pattern p = Pattern.compile("]*\\bsrc\\b\\s*=\\s*('|\")?([^'\"\n\r\f>]+(\\.jpg|\\.bmp|\\.eps|\\.gif|\\.mif|\\.miff|\\.png|\\.tif|\\.tiff|\\.svg|\\.wmf|\\.jpe|\\.jpeg|\\.dib|\\.ico|\\.tga|\\.cut|\\.pic|\\b)\\b)[^>]*>", Pattern.CASE_INSENSITIVE);
        Matcher m = p.matcher(htmlCode);
        String quote = null;
        String src = null;
        while (m.find()) {
            quote = m.group(1);
            src = (quote == null || quote.trim().length() == 0) ? m.group(2).split("//s+")[0] : m.group(2);
            imageSrcList.add(src);
        }
        if (imageSrcList == null || imageSrcList.size() == 0) {
            Log.e("huangxiaoguo", "资讯中未匹配到图片链接");
            return null;
        }

        return imageSrcList.toArray(new String[imageSrcList.size()]);
    }


}

• UIUtils类

package tsou.cn.webviewtext.util;

import android.content.Context;


public class UIUtils {


    /**
     * dip转换px
     */
    public static int dip2px(Context context, int dip) {
        final float scale = context.getResources().getDisplayMetrics().density;
        return (int) (dip * scale + 0.5f);
    }


}

• LookBigPhotoActivity类

package tsou.cn.webviewtext;

import android.os.Bundle;
import android.support.annotation.Nullable;
import android.support.v4.view.PagerAdapter;
import android.support.v4.view.ViewPager;
import android.support.v7.app.AppCompatActivity;
import android.view.View;
import android.view.ViewGroup;
import android.widget.ImageView;
import android.widget.TextView;

import tsou.cn.webviewtext.util.ImageLoadUtil;


public class LookBigPhotoActivity extends AppCompatActivity {

    ViewPager viewpager;

    TextView mCircleindicator;

    private int position;
    private String[] imageUrlses;

    @Override
    protected void onDestroy() {
        ImageLoadUtil.clearMemory(this);
        super.onDestroy();
    }

    @Override
    protected void onCreate(@Nullable Bundle savedInstanceState) {
        super.onCreate(savedInstanceState);
        setContentView(R.layout.activity_look_big_photo);
        initView();
        initData();
        initListener();
    }

    protected void initView() {
        viewpager = (ViewPager) findViewById(R.id.viewpager);
        mCircleindicator= (TextView) findViewById(R.id.circleindicator);
    }

    protected void initData() {
        position = getIntent().getIntExtra("position", 0);
        imageUrlses = getIntent().getStringArrayExtra("imageUrls");
    }

    protected void initListener() {
        viewpager.setAdapter(new PagerAdapter() {
            @Override
            public int getCount() {
                return imageUrlses.length;
            }

            @Override
            public Object instantiateItem(ViewGroup container, int position) {
                ImageView imageView = new ImageView(LookBigPhotoActivity.this);
                ImageLoadUtil.display(LookBigPhotoActivity.this, imageView, imageUrlses[position]);
                container.addView(imageView);
                imageView.setOnClickListener(new View.OnClickListener() {
                    @Override
                    public void onClick(View v) {
                        finish();
                    }
                });
                return imageView;
            }

            @Override
            public void destroyItem(ViewGroup container, int position, Object object) {
                container.removeView((View) object);
            }

            @Override
            public boolean isViewFromObject(View view, Object object) {
                return view == object;
            }
        });
        viewpager.setCurrentItem(position);
        mCircleindicator.setText(position+1+"/"+imageUrlses.length);

        viewpager.addOnPageChangeListener(onMyPageChangeListener);
    }

    private ViewPager.OnPageChangeListener onMyPageChangeListener = new ViewPager.OnPageChangeListener() {
        @Override
        public void onPageScrolled(int position, float positionOffset, int positionOffsetPixels) {

        }

        @Override
        public void onPageSelected(int position) {
            mCircleindicator.setText(position+1+"/"+imageUrlses.length);
        }

        @Override
        public void onPageScrollStateChanged(int state) {

        }
    };

}

• LookBigPhotoActivity布局

<RelativeLayout xmlns:android="http://schemas.android.com/apk/res/android"
    android:layout_width="match_parent"
    android:layout_height="match_parent"
    android:background="@color/black">


    <TextView
        android:id="@+id/circleindicator"
        android:layout_width="wrap_content"
        android:layout_height="wrap_content"
        android:layout_alignParentBottom="true"
        android:textSize="20sp"
        android:textColor="@color/white"
        android:text="0/0"
        android:layout_marginBottom="50px"
        android:layout_marginLeft="30dp" />

    <android.support.v4.view.ViewPager
        android:id="@+id/viewpager"
        android:layout_width="match_parent"
        android:layout_height="match_parent"
        android:layout_above="@id/circleindicator"
        android:layout_marginBottom="150px"
        android:layout_marginTop="150px" />
RelativeLayout>

DEMO地址：https://gitee.com/huangxiaoguo/WebViewText

这里有个问题，在Jsoup怎么获取网页中JS动态解析出来的标签呢？
有知道的，烦请赐教，谢谢哈…………….