Commit4c35ec5

committed

Allow 4 bytes UTF-8 (UCS-4 range 00010000-001FFFFF)

This is necessary to support JIS X 0213 <--> UTF-8 conversion.

1 parent6b77e3a commit4c35ec5Copy full SHA for 4c35ec5

File tree

+11

-4

lines changed

+11

-4

lines changed

Lines changed: 11 additions & 4 deletions

Original file line number	Diff line number	Diff line change
`@@ -1,6 +1,8 @@`
`1`	`1`	`#`
`2`		`-# $PostgreSQL: pgsql/src/backend/utils/mb/Unicode/ucs2utf.pl,v 1.2 2003/11/29 22:40:01 pgsql Exp $`
`3`		`-# convert UCS-2 to UTF-8`
	`2`	`+# Copyright (c) 2001-2007, PostgreSQL Global Development Group`
	`3`	`+#`
	`4`	`+# $PostgreSQL: pgsql/src/backend/utils/mb/Unicode/ucs2utf.pl,v 1.3 2007/03/23 13:51:30 ishii Exp $`
	`5`	`+# convert UCS-4 to UTF-8`
`4`	`6`	`#`
`5`	`7`	`subucs2utf {`
`6`	`8`	`local($ucs)=@_;`
`@@ -10,11 +12,16 @@ sub ucs2utf {`
`10`	`12`	`$utf=$ucs;`
`11`	`13`	`}elsif ($ucs>0x007f&&$ucs<=0x07ff) {`
`12`	`14`	`$utf= (($ucs &0x003f)\|0x80)\| ((($ucs>>6)\|0xc0)<<8);`
`13`		`-}else {`
	`15`	`+}elsif ($ucs>0x07ff&&$ucs<=0xffff) {`
`14`	`16`	`$utf= ((($ucs>>12)\|0xe0)<<16)\|`
`15`	`17`	`(((($ucs &0x0fc0)>>6)\|0x80)<<8)\|`
`16`	`18`	`(($ucs &0x003f)\|0x80);`
`17`		`-}`
	`19`	`+}else {`
	`20`	`+$utf= ((($ucs>>18)\|0xf0)<<24)\|`
	`21`	`+ (((($ucs &0x3ffff)>>12)\|0x80)<<16)\|`
	`22`	`+ (((($ucs &0x0fc0)>>6)\|0x80)<<8)\|`
	`23`	`+ (($ucs &0x003f)\|0x80);`
	`24`	`+ }`
`18`	`25`	`return($utf);`
`19`	`26`	`}`
`20`	`27`	`1;`

Comments

(0)