Update: 提供测试用例。

webmagic-saxon/src/test/java/us/codecraft/webmagic/selector/XpathSelectorTest.java

@@ -1393,12 +1393,13 @@ public class XpathSelectorTest {
     public void htmlCleanerParseTest() {
         Spider.create(new RuoxiaPageProcessor()).addUrl("http://www.ruoxia.com/top/dianji/month").thread(1).run();
     }
+
     class RuoxiaPageProcessor implements PageProcessor {
         @Override
         public void process(Page page) {
-            List<Selectable> nodes = page.getHtml().xpath("//div[@class=\"bd\"]//tbody/tr").nodes();
-            for (Selectable node:nodes) {
-                String name = node.xpath("//td[3]/div/a[1]/text()").get();
+            List<String> items = new Xpath2Selector("//div[@class=\"bd\"]//tbody/tr").selectList(page.getRawText());
+            for (String item : items) {
+                String name = new Xpath2Selector("//td[3]/div/a[1]/text()").select(item);
                 System.out.println(name);
             }
         }
@@ -1408,31 +1409,31 @@ public class XpathSelectorTest {
     @Test
     public void performanceTest() {
         Xpath2Selector xpath2Selector = new Xpath2Selector("//a");
-        long time =System.currentTimeMillis();
+        long time = System.currentTimeMillis();
         for (int i = 0; i < 1000; i++) {
             xpath2Selector.selectList(html);
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
         XpathSelector xpathSelector = new XpathSelector("//a");
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 1000; i++) {
             xpathSelector.selectList(html);
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 1000; i++) {
             xpath2Selector.selectList(html);
         }
         System.out.println(System.currentTimeMillis() - time);
 
         CssSelector cssSelector = new CssSelector("a");
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 1000; i++) {
             cssSelector.selectList(html);
         }
-        System.out.println("css "+(System.currentTimeMillis()-time));
+        System.out.println("css " + (System.currentTimeMillis() - time));
     }
 
     @Ignore("take long time")
@@ -1444,54 +1445,54 @@ public class XpathSelectorTest {
         TagNode tagNode = htmlCleaner.clean(html);
         Document document = Jsoup.parse(html);
 
-        long time =System.currentTimeMillis();
+        long time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             htmlCleaner.clean(html);
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             tagNode.evaluateXPath("//a");
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
         System.out.println("=============");
 
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             Jsoup.parse(html);
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             document.select("a");
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
         System.out.println("=============");
 
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             htmlCleaner.clean(html);
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             tagNode.evaluateXPath("//a");
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
         System.out.println("=============");
 
         XPathEvaluator compile = Xsoup.compile("//a");
-        time =System.currentTimeMillis();
+        time = System.currentTimeMillis();
         for (int i = 0; i < 2000; i++) {
             compile.evaluate(document);
         }
-        System.out.println(System.currentTimeMillis()-time);
+        System.out.println(System.currentTimeMillis() - time);
 
     }