Return to utf-8.fs CVS log

Up to [gforth] / gforth

Annotation of gforth/utf-8.fs, revision 1.2

1.1       pazsan      1: \ UTF-8 handling                                       12dec04py
                      2: 
                      3: \ Copyright (C) 2004 Free Software Foundation, Inc.
                      4: 
                      5: \ This file is part of Gforth.
                      6: 
                      7: \ Gforth is free software; you can redistribute it and/or
                      8: \ modify it under the terms of the GNU General Public License
                      9: \ as published by the Free Software Foundation; either version 2
                     10: \ of the License, or (at your option) any later version.
                     11: 
                     12: \ This program is distributed in the hope that it will be useful,
                     13: \ but WITHOUT ANY WARRANTY; without even the implied warranty of
                     14: \ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
                     15: \ GNU General Public License for more details.
                     16: 
                     17: \ You should have received a copy of the GNU General Public License
                     18: \ along with this program; if not, write to the Free Software
                     19: \ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
                     20: 
                     21: \ short: u8 means utf-8 encoded address
                     22: 
                     23: : u8len ( u -- n )
                     24:     dup      $80 u< IF  drop 1  EXIT  THEN \ special case ASCII
                     25:     $800  2 >r
                     26:     BEGIN  2dup u>=  WHILE  5 lshift r> 1+ >r  REPEAT
                     27:     2drop r> ;
                     28: 
                     29: : u8@+ ( u8addr -- u8addr' u )
                     30:     count  dup $80 and 0= ?EXIT  \ special case ASCII
                     31:     $7F and  $40 >r
                     32:     BEGIN  dup r@ and  WHILE  r@ xor
                     33:            6 lshift r> 5 lshift >r >r count
                     34: \          dup $C0 and $80 <> abort" malformed character"
                     35:            $3F and r> or
                     36:     REPEAT  rdrop ;
                     37: 
                     38: : u8!+ ( u u8addr -- u8addr' )
                     39:     over $80 < IF  tuck c! 1+  EXIT  THEN \ special case ASCII
                     40:     >r 0 swap  $3F
                     41:     BEGIN  2dup u>  WHILE
                     42:            2/ >r  dup $3F and $80 or swap 6 rshift r>
                     43:     REPEAT  $7F xor 2* or  r>
                     44:     BEGIN   over $80 u>= WHILE  tuck c! 1+  REPEAT  nip ;
                     45: 
                     46: \ scan to next/previous character
                     47: 
                     48: : u8>> ( u8addr -- u8addr' )
                     49:     BEGIN  count $C0 and $80 <>  UNTIL ;
                     50: : u8<< ( u8addr -- u8addr' )
                     51:     BEGIN  1- dup c@ $C0 and $80 <>  UNTIL ;
                     52: 
                     53: \ utf key and emit
                     54: 
                     55: : u8key ( -- u )
                     56:     defers key dup $80 and 0= ?EXIT  \ special case ASCII
                     57:     $7F and  $40 >r
                     58:     BEGIN  dup r@ and  WHILE  r@ xor
                     59:            6 lshift r> 5 lshift >r >r defers key
                     60: \          dup $C0 and $80 <> abort" malformed character"
                     61:            $3F and r> or
                     62:     REPEAT  rdrop ;
                     63: 
                     64: : u8emit ( u -- )
                     65:     dup $80 < IF  defers emit  EXIT  THEN \ special case ASCII
                     66:     0 swap  $3F
                     67:     BEGIN  2dup u>  WHILE
                     68:            2/ >r  dup $3F and $80 or swap 6 rshift r>
                     69:     REPEAT  $7F xor 2* or
                     70:     BEGIN   dup $80 u>= WHILE  defers emit  REPEAT  drop ;
                     71: 
                     72: \ input editor
                     73: 
                     74: : save-cursor ( -- )  27 emit '7 emit ;
                     75: : restore-cursor ( -- )  27 emit '8 emit ;
                     76: : .rest ( addr pos1 -- addr pos1 )
                     77:     restore-cursor 2dup type ;
                     78: : .all ( span addr pos1 -- span addr pos1 )
                     79:     restore-cursor >r 2dup swap type r> ;
                     80: 
                     81: : (u8ins)  ( max span addr pos1 u8char -- max span addr pos2 )
                     82:     >r >string over r@ u8len + swap move 2dup chars + r@ swap u8!+ drop
                     83:     r> u8len >r  rot r@ chars + -rot r> chars + .all .rest ;
                     84: : u8back  ( max span addr pos1 -- max span addr pos2 f )
1.2     ! pazsan     85:     dup  IF  over + u8<< over -  0 max .all .rest
        !            86:     ELSE  #bell emit  THEN 0 ;
1.1       pazsan     87: : u8forw  ( max span addr pos1 -- max span addr pos2 f )
                     88:     2 pick over <> IF  over + u8@+ u8emit over -  ELSE  #bell emit  THEN 0 ;
                     89: : (u8del)  ( max span addr pos1 -- max span addr pos2 )
                     90:     over + dup u8<< tuck - >r over -
                     91:     >string over r@ + -rot move
                     92:     rot r> - -rot .all 2 spaces .rest ;
                     93: : ?u8del ( max span addr pos1 -- max span addr pos2 0 )
                     94:   dup  IF  (u8del)  THEN  0 ;
                     95: : <u8del> ( max span addr pos1 -- max span addr pos2 0 )
                     96:   2 pick over <>
                     97:        IF  u8forw drop (u8del)  ELSE  #bell emit  THEN  0 ;
                     98: : u8eof  2 pick over or 0=  IF  bye  ELSE  <u8del>  THEN ;
                     99: 
                    100: ' u8forw  ctrl F bindkey
                    101: ' u8back  ctrl B bindkey
                    102: ' ?u8del  ctrl H bindkey
                    103: ' u8eof   ctrl D bindkey
                    104: ' <u8del> ctrl X bindkey
                    105: ' (u8ins) IS insert-char
                    106: ' noop IS everychar
                    107: ' save-cursor IS everyline
                    108: ' u8key IS key
                    109: ' u8emit IS emit