add scripts

pom.xml

@@ -48,6 +48,7 @@
     <modules>
         <module>webmagic-core</module>
         <module>webmagic-extension/</module>
+        <module>webmagic-scripts</module>
     </modules>
 
     <dependencyManagement>

webmagic-core/src/main/java/us/codecraft/webmagic/selector/PlainText.java

@@ -45,6 +45,16 @@ public class PlainText implements Selectable {
         throw new UnsupportedOperationException();
     }
 
+    @Override
+    public Selectable css(String selector) {
+        return $(selector);
+    }
+
+    @Override
+    public Selectable css(String selector, String attrName) {
+        return $(selector, attrName);
+    }
+
     @Override
     public Selectable smartContent() {
         throw new UnsupportedOperationException();

webmagic-core/src/main/java/us/codecraft/webmagic/selector/Selectable.java

@@ -35,6 +35,23 @@ public interface Selectable {
      */
     public Selectable $(String selector, String attrName);
 
+    /**
+     * select list with css selector
+     *
+     * @param selector css selector expression
+     * @return new Selectable after extract
+     */
+    public Selectable css(String selector);
+
+    /**
+     * select list with css selector
+     *
+     * @param selector css selector expression
+     * @param attrName attribute name of css selector
+     * @return new Selectable after extract
+     */
+    public Selectable css(String selector, String attrName);
+
     /**
      * select smart content with ReadAbility algorithm
      *

webmagic-scripts/pom.xml

+<?xml version="1.0" encoding="UTF-8"?>
+<project xmlns="http://maven.apache.org/POM/4.0.0"
+         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
+    <parent>
+        <artifactId>webmagic-parent</artifactId>
+        <groupId>us.codecraft</groupId>
+        <version>0.4.1-SNAPSHOT</version>
+    </parent>
+    <modelVersion>4.0.0</modelVersion>
+
+    <groupId>us.codecraft</groupId>
+    <artifactId>webmagic-scripts</artifactId>
+    <version>0.4.1-SNAPSHOT</version>
+
+    <dependencies>
+        <dependency>
+            <groupId>org.jruby</groupId>
+            <artifactId>jruby</artifactId>
+            <version>1.7.6</version>
+        </dependency>
+        <dependency>
+            <groupId>junit</groupId>
+            <artifactId>junit</artifactId>
+            <scope>test</scope>
+        </dependency>
+        <dependency>
+            <groupId>us.codecraft</groupId>
+            <artifactId>webmagic-core</artifactId>
+            <version>${project.version}</version>
+        </dependency>
+    </dependencies>
+
+
+</project>
\ No newline at end of file

webmagic-scripts/src/main/java/us/codecraft/webmagic/processor/ScriptProcessor.java

+package us.codecraft.webmagic.processor;
+
+import org.apache.commons.io.IOUtils;
+import us.codecraft.webmagic.Page;
+import us.codecraft.webmagic.Site;
+import us.codecraft.webmagic.Spider;
+
+import javax.script.ScriptContext;
+import javax.script.ScriptEngine;
+import javax.script.ScriptEngineManager;
+import javax.script.ScriptException;
+import java.io.IOException;
+import java.io.InputStream;
+
+/**
+ * @author code4crafter@gmail.com
+ */
+public class ScriptProcessor implements PageProcessor{
+
+    private ScriptEngine rubyEngine;
+
+    private String defines;
+
+    ScriptProcessor(){
+        ScriptEngineManager manager = new ScriptEngineManager();
+        rubyEngine = manager.getEngineByName("jruby");
+        InputStream resourceAsStream = this.getClass().getClassLoader().getResourceAsStream("ruby/defines.rb");
+        try {
+            defines = IOUtils.toString(resourceAsStream);
+        } catch (IOException e) {
+            e.printStackTrace();
+        }
+    }
+
+    @Override
+    public void process(Page page) {
+        ScriptContext context = rubyEngine.getContext();
+        context.setAttribute("page", page, ScriptContext.ENGINE_SCOPE);
+        String script;
+        try {
+            InputStream resourceAsStream = this.getClass().getClassLoader().getResourceAsStream("ruby/oschina.rb");
+            try {
+                script = IOUtils.toString(resourceAsStream);
+                rubyEngine.eval(defines+script, context);
+            } catch (IOException e) {
+                e.printStackTrace();
+            }
+        } catch (ScriptException e) {
+            e.printStackTrace();
+        }
+    }
+
+    @Override
+    public Site getSite() {
+        return Site.me();
+    }
+
+    public static void main(String[] args) {
+        Spider.create(new ScriptProcessor()).addUrl("http://my.oschina.net/flashsword/blog").run();
+    }
+}

webmagic-scripts/src/main/resources/ruby/defines.rb

+def xpath str
+  $page.getHtml().xpath(str).toString()
+end
+def css str
+  $page.getHtml().css(str).toString()
+end
+def urls str
+  links = $page.getHtml().links().regex(str).all();
+  $page.addTargetRequests(links);
+end
+

webmagic-scripts/src/main/resources/ruby/oschina.rb

+title = css "div.BlogTitle h1"
+content = css "div.BlogContent"
+urls "http://my\\.oschina\\.net/flashsword/blog/\\d+"
+puts title
+puts content
\ No newline at end of file

webmagic-scripts/src/test/java/us/codecraft/webmagic/jruby/TestJRubyCall.java

+package us.codecraft.webmagic.jruby;
+
+import org.junit.Test;
+
+import javax.script.ScriptContext;
+import javax.script.ScriptEngine;
+import javax.script.ScriptEngineManager;
+import javax.script.ScriptException;
+
+/**
+ * @author code4crafter@gmail.com
+ */
+public class TestJRubyCall {
+
+    @Test
+    public void test() throws ScriptException {
+        ScriptEngineManager manager = new ScriptEngineManager();
+        ScriptEngine rubyEngine = manager.getEngineByName("jruby");
+        ScriptContext context = rubyEngine.getContext();
+
+        context.setAttribute("a", "sad", ScriptContext.ENGINE_SCOPE);
+//        rubyEngine.eval("", context);
+        rubyEngine.eval("b=1; puts b", context);
+    }
+}