| commit | author | age | ||
| 2207d6 | 1 | Core.LegacyEntityDecoder |
| W | 2 | TYPE: bool |
| 3 | VERSION: 4.9.0 | |
| 4 | DEFAULT: false | |
| 5 | --DESCRIPTION-- | |
| 6 | <p> | |
| 7 | Prior to HTML Purifier 4.9.0, entities were decoded by performing | |
| 8 | a global search replace for all entities whose decoded versions | |
| 9 | did not have special meanings under HTML, and replaced them with | |
| 10 | their decoded versions. We would match all entities, even if they did | |
| 11 | not have a trailing semicolon, but only if there weren't any trailing | |
| 12 | alphanumeric characters. | |
| 13 | </p> | |
| 14 | <table> | |
| 15 | <tr><th>Original</th><th>Text</th><th>Attribute</th></tr> | |
| 16 | <tr><td>&yen;</td><td>¥</td><td>¥</td></tr> | |
| 17 | <tr><td>&yen</td><td>¥</td><td>¥</td></tr> | |
| 18 | <tr><td>&yena</td><td>&yena</td><td>&yena</td></tr> | |
| 19 | <tr><td>&yen=</td><td>¥=</td><td>¥=</td></tr> | |
| 20 | </table> | |
| 21 | <p> | |
| 22 | In HTML Purifier 4.9.0, we changed the behavior of entity parsing | |
| 23 | to match entities that had missing trailing semicolons in less | |
| 24 | cases, to more closely match HTML5 parsing behavior: | |
| 25 | </p> | |
| 26 | <table> | |
| 27 | <tr><th>Original</th><th>Text</th><th>Attribute</th></tr> | |
| 28 | <tr><td>&yen;</td><td>¥</td><td>¥</td></tr> | |
| 29 | <tr><td>&yen</td><td>¥</td><td>¥</td></tr> | |
| 30 | <tr><td>&yena</td><td>¥a</td><td>&yena</td></tr> | |
| 31 | <tr><td>&yen=</td><td>¥=</td><td>&yen=</td></tr> | |
| 32 | </table> | |
| 33 | <p> | |
| 34 | This flag reverts back to pre-HTML Purifier 4.9.0 behavior. | |
| 35 | </p> | |
| 36 | --# vim: et sw=4 sts=4 | |
