.HTML File Format – Character references

The three forms

How references are read

• Longest match. After & the parser takes the longest name in the table that the text starts with. Names are case sensitive: É is É and é is é.
• No semicolon. A hundred and six names from old HTML also work without their semicolon (listed below), which is a parse error but still decodes. &notit; reads as ¬ followed by it;, because not is such a name and the text does not continue into another name.
• In attribute values a legacy name without a semicolon is left alone when the next character is = or a letter or digit, so href="?a=1&copy=2" keeps its &copy=2.
• Unknown names stay as literal text; if they end with a semicolon it is the parse error unknown-named-character-reference.
• Numeric values. Zero, a value above U+10FFFF, or a surrogate reads as U+FFFD. A value in the C1 control range U+0080 to U+009F is remapped as if the code were Windows-1252: € is € (U+20AC), ’ is ’ (U+2019), and so on for 0x80, 82 to 8C, 8E, 91 to 9C, 9E and 9F. A noncharacter or other control character decodes but is a parse error. A reference with no digits is left as text, with the parse error absence-of-digits-in-numeric-character-reference.
• Where they count. References are read in text, in attribute values and in RCDATA (title, textarea), but not in raw text such as script and style, and not in comments.

Names that work without a semicolon

AElig AMP Aacute Acirc Agrave Aring Atilde Auml COPY Ccedil ETH Eacute Ecirc Egrave Euml GT Iacute Icirc Igrave Iuml LT Ntilde Oacute Ocirc Ograve Oslash Otilde Ouml QUOT REG THORN Uacute Ucirc Ugrave Uuml Yacute aacute acirc acute aelig agrave amp aring atilde auml brvbar ccedil cedil cent copy curren deg divide eacute ecirc egrave eth euml frac12 frac14 frac34 gt iacute icirc iexcl igrave iquest iuml laquo lt macr micro middot nbsp not ntilde oacute ocirc ograve ordf ordm oslash otilde ouml para plusmn pound quot raquo reg sect shy sup1 sup2 sup3 szlig thorn times uacute ucirc ugrave uml uuml yacute yen yuml

All 2,125 names

Each entry is the name, with its semicolon, its code point or points in hexadecimal, and the character it stands for. Names are grouped by first letter.