Commit bfadac75 authored by yihua.huang's avatar yihua.huang

fix an attribute bug

parent 14562855
...@@ -8,6 +8,7 @@ import org.htmlcleaner.DomSerializer; ...@@ -8,6 +8,7 @@ import org.htmlcleaner.DomSerializer;
import org.htmlcleaner.HtmlCleaner; import org.htmlcleaner.HtmlCleaner;
import org.htmlcleaner.TagNode; import org.htmlcleaner.TagNode;
import org.w3c.dom.Document; import org.w3c.dom.Document;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList; import org.w3c.dom.NodeList;
import javax.xml.namespace.NamespaceContext; import javax.xml.namespace.NamespaceContext;
...@@ -70,7 +71,7 @@ public class Xpath2Selector implements Selector { ...@@ -70,7 +71,7 @@ public class Xpath2Selector implements Selector {
private XPath2NamespaceContext() { private XPath2NamespaceContext() {
put("fn", NamespaceConstant.FN); put("fn", NamespaceConstant.FN);
put("xslt",NamespaceConstant.XSLT); put("xslt", NamespaceConstant.XSLT);
} }
@Override @Override
...@@ -116,15 +117,20 @@ public class Xpath2Selector implements Selector { ...@@ -116,15 +117,20 @@ public class Xpath2Selector implements Selector {
result = xPathExpression.evaluate(document, XPathConstants.STRING); result = xPathExpression.evaluate(document, XPathConstants.STRING);
} }
if (result instanceof NodeList) { if (result instanceof NodeList) {
StreamResult xmlOutput = new StreamResult(new StringWriter());
Transformer transformer = TransformerFactory.newInstance().newTransformer();
transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
NodeList nodeList = (NodeList) result; NodeList nodeList = (NodeList) result;
if (nodeList.getLength() == 0) { if (nodeList.getLength() == 0) {
return null; return null;
} }
transformer.transform(new DOMSource(nodeList.item(0)), xmlOutput); Node item = nodeList.item(0);
return xmlOutput.getWriter().toString(); if (item.getNodeType() == Node.ATTRIBUTE_NODE || item.getNodeType() == Node.TEXT_NODE) {
return item.getTextContent();
} else {
StreamResult xmlOutput = new StreamResult(new StringWriter());
Transformer transformer = TransformerFactory.newInstance().newTransformer();
transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
transformer.transform(new DOMSource(item), xmlOutput);
return xmlOutput.getWriter().toString();
}
} }
return result.toString(); return result.toString();
} catch (Exception e) { } catch (Exception e) {
...@@ -152,9 +158,14 @@ public class Xpath2Selector implements Selector { ...@@ -152,9 +158,14 @@ public class Xpath2Selector implements Selector {
StreamResult xmlOutput = new StreamResult(); StreamResult xmlOutput = new StreamResult();
transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes"); transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "yes");
for (int i = 0; i < nodeList.getLength(); i++) { for (int i = 0; i < nodeList.getLength(); i++) {
xmlOutput.setWriter(new StringWriter()); Node item = nodeList.item(i);
transformer.transform(new DOMSource(nodeList.item(i)), xmlOutput); if (item.getNodeType() == Node.ATTRIBUTE_NODE || item.getNodeType() == Node.TEXT_NODE) {
results.add(xmlOutput.getWriter().toString()); results.add(item.getTextContent());
} else {
xmlOutput.setWriter(new StringWriter());
transformer.transform(new DOMSource(item), xmlOutput);
results.add(xmlOutput.getWriter().toString());
}
} }
} else { } else {
results.add(result.toString()); results.add(result.toString());
......
package us.codecraft.webmagic.oo; package us.codecraft.webmagic.oo;
import us.codecraft.webmagic.Page; import us.codecraft.webmagic.Page;
import us.codecraft.webmagic.Site;
import java.util.List; import java.util.List;
...@@ -9,7 +10,7 @@ import java.util.List; ...@@ -9,7 +10,7 @@ import java.util.List;
* @date: 13-8-1 <br> * @date: 13-8-1 <br>
* Time: 下午10:18 <br> * Time: 下午10:18 <br>
*/ */
@TargetUrl(value="http://my.oschina.net/flashsword/blog/*",sourceRegion = "//div[@class='BlogLinks']") @TargetUrl(value="http://my.oschina.net/flashsword/blog/*",sourceRegion = "//div[@class='BlogLinks']//a/@href")
public class OschinaBlog implements AfterExtractor { public class OschinaBlog implements AfterExtractor {
@ExtractBy("//title") @ExtractBy("//title")
...@@ -23,6 +24,13 @@ public class OschinaBlog implements AfterExtractor { ...@@ -23,6 +24,13 @@ public class OschinaBlog implements AfterExtractor {
@Override @Override
public void afterProcess(Page page) { public void afterProcess(Page page) {
content = null; System.out.println("title:\t"+title);
System.out.println("content:\t"+content);
System.out.println("tags:\t" + tags);
} }
}
public static void main(String[] args) {
OOSpider.create(Site.me().addStartUrl("http://my.oschina.net/flashsword/blog/145796"), OschinaBlog.class)
.run();
}
}
\ No newline at end of file
...@@ -1418,7 +1418,7 @@ public class XpathSelectorTest { ...@@ -1418,7 +1418,7 @@ public class XpathSelectorTest {
@Test @Test
public void testXpath2Selector() { public void testXpath2Selector() {
Xpath2Selector xpath2Selector = new Xpath2Selector("//a"); Xpath2Selector xpath2Selector = new Xpath2Selector("//a/@href");
String select = xpath2Selector.select(html); String select = xpath2Selector.select(html);
Assert.assertNotNull(select); Assert.assertNotNull(select);
} }
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment