使用jsoup解析XML - 防止jsoup“清理”<link>标记

Question

In most case, I have no problem with using jsoup to parse XML. However, if there are <link> tags in the XML document, jsoup will change <link>some text here</link> to <link />some text here . This makes it impossible to extract text inside the <link> tag using CSS selector.

So how to prevent jsoup from "cleaning" <link> tags?

Answer 1

In jsoup 1.6.2 I have added an XML parser mode, which parses the input as-is, without applying the HTML5 parse rules (contents of element, document structure, etc). This mode will keep text in a <link> tag, and allow multiples of it, etc.

Here's an example:

String xml = "<link>One</link><link>Two</link>";
Document xmlDoc = Jsoup.parse(xml, "", Parser.xmlParser());

Elements links = xmlDoc.select("link");
System.out.println("Link text 1: " + links.get(0).text());
System.out.println("Link text 2: " + links.get(1).text());

Returns:

Link text 1: One
Link text 2: Two

Answer 2

Do not store any text inside <link> element - it's invalid. If you need extra information, keep it inside HTML5 data-* attributes. I'm sure jsoup won't touch it.

<link rel="..." data-city="Warsaw" />

Answer 3

There can be a workaround for this. Before passing XML to jsoup. Transform XML file to replace all with some dummy tag say and do what you want to do.