Skip to content

CHARACTER_ENCODING_AND_DECODING_RULES.md

Latest commit

 

History

History
114 lines (91 loc) · 5.04 KB

CHARACTER_ENCODING_AND_DECODING_RULES.md

File metadata and controls

114 lines (91 loc) · 5.04 KB

Character Encoding and Decoding Rules

This document establishes the rules of encoding and decoding entities. XML specifies five predefined entities which are &, <, > ,' and ".

Parsing XML

While parsing the string representing of an XML document:

  • For the value of CDATA, comment and processing instruction nodes; no special processing is needed and all characters from the input string is passed to the node value as-is.
  • For the value of text and attribute nodes; all five predefined entitites, numeric character references (&#nnnn;) and hexadecimal numeric character references (&#xhhhh;) are decoded (note that the x must be lowercase). The two examples given below for the Yen character applies to all numeric character references.
Value Decoded Value
& &
&lt; <
&gt; >
&apos; '
&quot; "
&#9; \t
&#10; \n
&#xA; \n
&#13; \r
&#xD; \r
&#165; ¥ (e.g.)
&#xA5; ¥ (e.g.)

Serializing XML

While serializing an XML document to its string representation:

  • For the value of CDATA, comment and processing instruction nodes; no special processing is needed.
  • For the value of text nodes; &, < and > are encoded into predefined entitites, other characters are left as-is.
Value Encoded Value
& &amp;
< &lt;
> &gt;
  • For the value of attribute nodes, &, <, > and " are encoded (apostrophe is not encoded since attribute values are always serialized with double quotes). In additition, \t, \nand \r are also encoded into numeric character references.
Value Encoded Value
& &amp;
< &lt;
> &gt;
" &quot;
\t &#9;
\n &#10;
\r &#13;

Round Trip: String -> XML -> String

A round trip will potentially result in different string representations as described below.

  • CDATA, comment and processing instruction nodes can safely round trip as their values are not processed at all.
  • Value of text nodes may potentially change due to differences in parsing and serialization.
Value Decoded Value Encoded Value Same?
&amp; & &amp; ✓
&lt; < &lt; ✓
&gt; > &gt; ✓
&apos; ' ' ✗
&quot; " " ✗
&#9; \t \t ✗
&#10; \n \n ✗
&#xA; \n \n ✗
&#13; \r \r ✗
&#xD; \r \r ✗
&#165; ¥ ¥ ✗
&#xA5; ¥ ¥ ✗
  • Value of attribute nodes may also change as below.
Value Decoded Value Encoded Value Same?
&amp; & &amp; ✓
&lt; < &lt; ✓
&gt; > &gt; ✓
&apos; ' ' ✗
&quot; " &quot; ✓
&#9; \t &#9; ✓
&#10; \n &#10; ✓
&#xA; \n &#10; ✗
&#13; \r &#13; ✓
&#xD; \r &#13; ✗
&#165; ¥ ¥ ✗
&#xA5; ¥ ¥ ✗

Round Trip: XML -> String -> XML

A round from an XML document to its string representation and back to an XML document is safe.

  • CDATA, comment and processing instruction can safely round trip as their values are not processed at all.
  • Text nodes are safe because all encoded values are decoded back to their original character representations.
Value Encoded Value Decoded Value Same?
& &amp; & ✓
< &lt; < ✓
> &gt; > ✓
  • Attribute nodes are also safe.
Value Encoded Value Decoded Value Same?
& &amp; & ✓
< &lt; < ✓
> &gt; > ✓
" &quot; " ✓
\t &#9; \t ✓
\n &#10; \n ✓
\r &#13; \r ✓