1 /***************************************************************************
3 * Open \______ \ ____ ____ | | _\_ |__ _______ ___
4 * Source | _// _ \_/ ___\| |/ /| __ \ / _ \ \/ /
5 * Jukebox | | ( <_> ) \___| < | \_\ ( <_> > < <
6 * Firmware |____|_ /\____/ \___ >__|_ \|___ /\____/__/\_ \
10 * Copyright (C) 2005 by Gadi Cohen
12 * Largely based on php_hebrev by Zeev Suraski <zeev@php.net>
13 * Heavily modified by Gadi Cohen aka Kinslayer <dragon@wastelands.net>
15 * All files in this archive are subject to the GNU General Public License.
16 * See the file COPYING in the source tree root for full license agreement.
18 * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
19 * KIND, either express or implied.
21 ****************************************************************************/
27 #include "rbunicode.h"
29 #include "scroll_engine.h"
31 //#define _HEB_BUFFER_LENGTH (MAX_PATH + LCD_WIDTH/2 + 3 + 2 + 2) * 2
32 #define _HEB_BLOCK_TYPE_ENG 1
33 #define _HEB_BLOCK_TYPE_HEB 0
34 #define _HEB_ORIENTATION_LTR 1
35 #define _HEB_ORIENTATION_RTL 0
37 #define ischar(c) ((c > 0x0589 && c < 0x0700) || (c >= 0xfb50 && c <= 0xfefc) ? 1 : 0)
38 #define _isblank(c) ((c==' ' || c=='\t') ? 1 : 0)
39 #define _isnewline(c) ((c=='\n' || c=='\r') ? 1 : 0)
40 #define XOR(a,b) ((a||b) && !(a&&b))
42 const arab_t
* arab_lookup(unsigned short uchar
)
44 if (uchar
>= 0x621 && uchar
<= 0x63a)
45 return &(jointable
[uchar
- 0x621]);
46 if (uchar
>= 0x640 && uchar
<= 0x64a)
47 return &(jointable
[uchar
- 0x621 - 5]);
48 if (uchar
>= 0x671 && uchar
<= 0x6d5)
49 return &(jointable
[uchar
- 0x621 - 5 - 38]);
50 if (uchar
== 0x200D) /* Support for the zero-width joiner */
55 void arabjoin(unsigned short * stringprt
, int length
){
57 bool connected
= false;
58 unsigned short * writeprt
= stringprt
;
60 const arab_t
* prev
= 0;
62 const arab_t
* ligature
= 0;
66 for (i
= 0; i
<= length
; i
++) {
67 cur
= arab_lookup(uchar
= *stringprt
++);
69 /* Skip non-arabic chars */
72 /* Finish the last char */
74 *writeprt
++ = prev
->final
;
77 *writeprt
++ = prev
->isolated
;
86 /* nothing to do for arabic char if the previous was non-arabic */
92 /* if it's LAM, check for LAM+ALEPH ligatures */
93 if (prev
->isolated
== 0xfedd) {
94 switch (cur
->isolated
) {
96 ligature
= &(lamaleph
[0]);
99 ligature
= &(lamaleph
[1]);
102 ligature
= &(lamaleph
[2]);
105 ligature
= &(lamaleph
[3]);
109 if (ligature
) { /* replace the 2 glyphs by their ligature */
113 if (connected
) { /* previous char has something connected to it */
114 if (prev
->medial
&& cur
->final
) /* Can we connect to it? */
115 *writeprt
++ = prev
->medial
;
117 *writeprt
++ = prev
->final
;
121 if (prev
->initial
&& cur
->final
) { /* Can we connect to it? */
122 *writeprt
++ = prev
->initial
;
125 *writeprt
++ = prev
->isolated
;
132 unsigned short *bidi_l2v(const unsigned char *str
, int orientation
)
134 int length
= utf8length(str
);
135 static unsigned short utf16_buf
[SCROLL_LINE_SIZE
];
136 static unsigned short bidi_buf
[SCROLL_LINE_SIZE
];
137 unsigned short *heb_str
, *target
, *tmp
; // *broken_str
138 int block_start
, block_end
, block_type
, block_length
, i
;
140 //int begin, end, char_count, orig_begin;
143 target
= tmp
= utf16_buf
;
145 str
= utf8decode(str
, target
++);
148 if (target
== utf16_buf
) /* empty string */
151 /* properly join any arabic chars */
152 arabjoin(utf16_buf
, length
);
154 block_start
=block_end
=block_length
=0;
160 target
= heb_str
+ length
;
166 block_type
= _HEB_BLOCK_TYPE_HEB
;
168 block_type
= _HEB_BLOCK_TYPE_ENG
;
171 while((XOR(ischar(*(tmp
+1)),block_type
)
172 || _isblank(*(tmp
+1)) || ispunct((int)*(tmp
+1))
174 && block_end
< length
-1) {
180 if (block_type
!= orientation
) {
181 while ((_isblank(*tmp
) || ispunct((int)*tmp
))
182 && *tmp
!='/' && *tmp
!='-' && block_end
>block_start
) {
188 for (i
=block_start
; i
<=block_end
; i
++) {
189 *target
= (block_type
== orientation
) ? *(utf16_buf
+i
) : *(utf16_buf
+block_end
-i
+block_start
);
190 if (block_type
!=orientation
) {
202 target
+= orientation
? 1 : -1;
204 block_type
= !block_type
;
205 block_start
=block_end
+1;
206 } while(block_end
<length
-1);
210 #if 0 /* Is this code really necessary? */
211 broken_str
= utf16_buf
;
217 while ((!max_chars
|| char_count
<max_chars
) && begin
>0) {
220 if (begin
<=0 || _isnewline(heb_str
[begin
])) {
221 while(begin
>0 && _isnewline(heb_str
[begin
-1])) {
228 if (char_count
==max_chars
) { /* try to avoid breaking words */
229 int new_char_count
= char_count
;
230 int new_begin
= begin
;
232 while (new_char_count
>0) {
233 if (_isblank(heb_str
[new_begin
]) ||
234 _isnewline(heb_str
[new_begin
])) {
240 if (new_char_count
>0) {
241 char_count
=new_char_count
;
247 /* if (_isblank(heb_str[begin])) {
251 /* skip leading newlines */
252 while (begin
<=end
&& _isnewline(heb_str
[begin
])) {
257 for (i
=begin
; i
<=end
; i
++) {
258 *target
= heb_str
[i
];
262 for (i
=orig_begin
; i
<=end
&& _isnewline(heb_str
[i
]); i
++) {
263 *target
= heb_str
[i
];