Merge pull request #1 from code4craft/master

merge

webmagic-scripts/pom.xml

@@ -16,6 +16,11 @@
             <artifactId>jruby</artifactId>
             <version>1.7.6</version>
         </dependency>
+        <dependency>
+            <groupId>org.codehaus.groovy</groupId>
+            <artifactId>groovy-all</artifactId>
+            <version>2.1.6</version>
+        </dependency>
         <dependency><groupId>org.python</groupId>
             <artifactId>jython</artifactId>
             <version>2.5.3</version>

webmagic-scripts/src/main/groovy/Github.groovy

+Github {
+    Site {
+        sleepTime 0
+        timeOut 100
+        retryTimes 3
+        userAgent ['a','b','c'].random
+    }
+    match "https://github.com/\\w+/\\w+" {
+        addUrl(url.regex("https://github.com/\\w+/\\w+"))
+        return  {
+            name: html.xpath("//h1[@class='entry-title public']/strong/a/text()")
+            author: html.xpath "https://github\\.com/(\\w+)/.*"
+            readme: html.xpath "//div[@id='readme']/tidyText()"
+            star : toInt(html.xpath("//div[@id='readme']/tidyText()"))
+        }
+    }
+
+}