libgo/go/html/entity.go

   1 // Copyright 2010 The Go Authors. All rights reserved.
   2 // Use of this source code is governed by a BSD-style
   3 // license that can be found in the LICENSE file.
   4
   5 package html
   6
   7 import (
   8         "utf8"
   9 )
  10
  11 // entity is a map from HTML entity names to their values. The semicolon matters:
  12 // http://www.whatwg.org/specs/web-apps/current-work/multipage/named-character-references.html
  13 // lists both "amp" and "amp;" as two separate entries.
  14 //
  15 // TODO(nigeltao): Take the complete map from the HTML5 spec section 10.5 "Named character references".
  16 // http://www.whatwg.org/specs/web-apps/current-work/multipage/named-character-references.html
  17 // Note that the HTML5 list is larger than the HTML4 list at
  18 // http://www.w3.org/TR/html4/sgml/entities.html
  19 var entity = map[string]int{
  20         "aacute":  '\U000000E1',
  21         "aacute;": '\U000000E1',
  22         "amp;":    '\U00000026',
  23         "apos;":   '\U00000027',
  24         "gt;":     '\U0000003E',
  25         "lt;":     '\U0000003C',
  26         "quot;":   '\U00000022',
  27 }
  28
  29 func init() {
  30         // We verify that the length of UTF-8 encoding of each value is <= 1 + len(key).
  31         // The +1 comes from the leading "&". This property implies that the length of
  32         // unescaped text is <= the length of escaped text.
  33         for k, v := range entity {
  34                 if 1+len(k) < utf8.RuneLen(v) {
  35                         panic("escaped entity &" + k + " is shorter than its UTF-8 encoding " + string(v))
  36                 }
  37         }
  38 }