Necessary to encode characters in HTML links?

Should I be encoding characters contained within a url?

Example:

<a href="http://google.com?topicId=1&pageId=1">Some link 开发者_如何学编程using &</a>

or

<a href="http://google.com?topicId=1&amp;pageId=1">Some link using &amp;</a>

Yes.

In HTML (including XHTML and HTML5, as far as I know), all attribute values and tag content should be encoded:

Authors should also use "&" in attribute values since character references are allowed within CDATA attribute values.

Depends how your files are being served up and identified.

For XHTML, yes and it's required.

For HTML, no and it's incorrect to do it.

There are two different kinds of encoding which are needed for different purposes in web programming, and it is easy to get confused.

Special characters in text which is to be displayed as HTML need to be encoded as HTML entities. This is particularly characters such as '<' which are part of HTML markup, but it may also be useful for other special characters if there is any doubt about the character encoding to be used.

Special characters in a URL need to be URL-encoded (replaced by %nn codes).

There is no harm in putting an HTML entity into a URL if it is going to be treated as HTML text by whatever receives it; but if it is part of an instruction to a program (such as the & used to separate arguments in a CGI query string) you should not encode it.