URI Encoding in Attributes

URI Encoding (or Percent Encoding) applies to attributes of type URI and URIs.

Important: After any required URI-encoding (see below), any remaining ampersand (&) characters must then be escaped as & (or &) before inserting the URI into the attribute. Also, if using a pair of single quote (') characters to delimit the attribute, any single quotes within the attribute value itself must be escaped as ' (or ' if compatibility with HTML isn't necessary - see Character Entities Predefined by XML 1.0). These requirements are necessary to comply with the constraints of the *XML 1.0 Specification.

The characters A-Z, a-z, 0-9, hyphen/dash (-), underscore (_), full stop (.) and tilde (~) should be used without encoding in a URI. Others may have to be encoded, depending upon their purpose in the URI.

Characters within a URI may be grouped into the following categories:

Unreserved characters

These are characters which may be used as they are, without any encoding, and comprise:
abcdefghijklmnopqrstuvwxyz
ABCDEFGHIJKLMNOPQRSTUVWXYZ
0123456789-_.~

Reserved characters

These 19 characters have a reserved purpose and should usually be *URI-encoded unless they are being used for that purpose. Reserved characters comprise:
:/?#[]@!$&'()*+,;=%
Examples:

• The plus sign (+) is often used to represent a space character and should not be encoded if used for that purpose.
• The percent sign (%) is used to begin an encoded character and, if not being used for that purpose, it should be encoded as %25.

Other characters

All other characters (not listed above), including all whitespace, must be *URI-encoded.

See *RFC 3986: URI Generic Syntax (Section 2) for more information on the structure of URIs.

The table below shows how common characters should be URI-encoded.