Home

Awesome

Xsoup

Build Status

XPath selector based on Jsoup.

Get started:

    @Test
    public void testSelect() {

        String html = "<html><div><a href='https://github.com'>github.com</a></div>" +
                "<table><tr><td>a</td><td>b</td></tr></table></html>";

        Document document = Jsoup.parse(html);

        String result = Xsoup.compile("//a/@href").evaluate(document).get();
        Assert.assertEquals("https://github.com", result);

        List<String> list = Xsoup.compile("//tr/td/text()").evaluate(document).list();
        Assert.assertEquals("a", list.get(0));
        Assert.assertEquals("b", list.get(1));
    }

Performance:

Xsoup use Jsoup as HTML parser.

Compare with another most used XPath selector for HTML - HtmlCleaner, Xsoup is much faster:

Normal HTML, size 44KB
XPath: "//a"	
Run for 2000 times

Environment：Mac Air MD231CH/A 
CPU: 1.8Ghz Intel Core i5
<table> <tr> <td width="100">Operation</td> <td width="100">Xsoup</td> <td>HtmlCleaner</td> </tr> <tr> <td>parse</td> <td>3,207(ms)</td> <td>7,999(ms)</td> </tr> <tr> <td>select</td> <td>95(ms)</td> <td>380(ms)</td> </tr> </table>

Syntax supported:

XPath1.0:

<table> <tr> <td width="100">Name</td> <td width="100">Expression</td> <td>Support</td> </tr> <tr> <td>nodename</td> <td>nodename</td> <td>yes</td> </tr> <tr> <td>immediate parent</td> <td>/</td> <td>yes</td> </tr> <tr> <td>parent</td> <td>//</td> <td>yes</td> </tr> <tr> <td>attribute</td> <td>[@key=value]</td> <td>yes</td> </tr> <tr> <td>nth child</td> <td>tag[n]</td> <td>yes</td> </tr> <tr> <td>attribute</td> <td>/@key</td> <td>yes</td> </tr> <tr> <td>wildcard in tagname</td> <td>/*</td> <td>yes</td> </tr> <tr> <td>wildcard in attribute</td> <td>/[@*]</td> <td>yes</td> </tr> <tr> <td>function</td> <td>function()</td> <td>part</td> </tr> <tr> <td>or</td> <td>a | b</td> <td>yes since 0.2.0</td> </tr> <tr> <td>parent in path</td> <td>. or ..</td> <td>no</td> </tr> <tr> <td>predicates</td> <td>price>35</td> <td>no</td> </tr> <tr> <td>predicates logic</td> <td>@class=a or @class=b</td> <td>yes since 0.2.0</td> </tr> </table>

Function supported:

In Xsoup, we use some function (maybe not in Standard XPath 1.0):

<table> <tr> <td width="100">Expression</td> <td width="100">Description</td> <td>Standard XPath</td> </tr> <tr> <td width="100">text(n)</td> <td width="100">nth text content of element(0 for all)</td> <td>text() only</td> </tr> <tr> <td width="100">allText()</td> <td width="100">text including children</td> <td>not support</td> </tr> </tr> <tr> <td width="100">tidyText()</td> <td width="100">text including children, well formatted</td> <td>not support</td> </tr> <tr> <td width="100">html()</td> <td width="100">innerhtml of element</td> <td>not support</td> </tr> <tr> <td width="100">outerHtml()</td> <td width="100">outerHtml of element</td> <td>not support</td> </tr> <tr> <td width="100">regex(@attr,expr,group)</td> <td width="100">use regex to extract content</td> <td>not support</td> </tr> </table>

Extended syntax supported:

These XPath syntax are extended only in Xsoup (for convenience in extracting HTML, refer to Jsoup CSS Selector):

<table> <tr> <td width="100">Name</td> <td width="100">Expression</td> <td>Support</td> </tr> <tr> <td>attribute value not equals</td> <td>[@key!=value]</td> <td>yes</td> </tr> <tr> <td>attribute value start with</td> <td>[@key~=value]</td> <td>yes</td> </tr> <tr> <td>attribute value end with</td> <td>[@key$=value]</td> <td>yes</td> </tr> <tr> <td>attribute value contains</td> <td>[@key*=value]</td> <td>yes</td> </tr> <tr> <td>attribute value match regex</td> <td>[@key~=value]</td> <td>yes</td> </tr> </table>

License

MIT License, see file LICENSE

Bitdeli Badge