7 #define LEN(a) (sizeof(a) / sizeof((a)[0]))
9 /* return the length of a utf-8 character */
12 int c
= (unsigned char) s
[0];
28 /* the number of utf-8 characters in s */
37 /* the unicode codepoint of the given utf-8 character */
40 int c
= (unsigned char) s
[0];
45 return ((c
& 0x1f) << 6) | (s
[1] & 0x3f);
47 return ((c
& 0x0f) << 12) | ((s
[1] & 0x3f) << 6) | (s
[2] & 0x3f);
49 c
= (0x3f >> --l
) & (unsigned char) *s
++;
51 c
= (c
<< 6) | ((unsigned char) *s
++ & 0x3f);
55 /* find the beginning of the character at s[i] */
56 char *uc_beg(char *beg
, char *s
)
58 while (s
> beg
&& (((unsigned char) *s
) & 0xc0) == 0x80)
63 /* find the end of the character at s[i] */
66 if (!*s
|| !((unsigned char) *s
& 0x80))
68 if (((unsigned char) *s
& 0xc0) == 0xc0)
70 while (((unsigned char) *s
& 0xc0) == 0x80)
75 /* return a pointer to the character following s */
76 char *uc_next(char *s
)
79 return *s
? s
+ 1 : s
;
82 /* return a pointer to the character preceding s */
83 char *uc_prev(char *beg
, char *s
)
85 return s
== beg
? beg
: uc_beg(beg
, s
- 1);
93 char *uc_lastline(char *s
)
95 char *r
= strrchr(s
, '\n');
99 /* allocate and return an array for the characters in s */
100 char **uc_chop(char *s
, int *n
)
105 chrs
= malloc((*n
+ 1) * sizeof(chrs
[0]));
106 for (i
= 0; i
< *n
+ 1; i
++) {
113 char *uc_chr(char *s
, int off
)
121 return s
&& (off
< 0 || i
== off
) ? s
: "";
124 /* the number of characters between s and s + off */
125 int uc_off(char *s
, int off
)
129 for (i
= 0; s
< e
&& *s
; i
++)
134 char *uc_sub(char *s
, int beg
, int end
)
136 char *sbeg
= uc_chr(s
, beg
);
137 char *send
= uc_chr(s
, end
);
138 int len
= sbeg
&& send
&& sbeg
<= send
? send
- sbeg
: 0;
139 char *r
= malloc(len
+ 1);
140 memcpy(r
, sbeg
, len
);
145 char *uc_dup(char *s
)
147 char *r
= malloc(strlen(s
) + 1);
148 return r
? strcpy(r
, s
) : NULL
;
151 int uc_isspace(char *s
)
153 int c
= s
? (unsigned char) *s
: 0;
154 return c
<= 0x7f && isspace(c
);
157 int uc_isprint(char *s
)
159 int c
= s
? (unsigned char) *s
: 0;
160 return c
> 0x7f || isprint(c
);
163 int uc_isalpha(char *s
)
165 int c
= s
? (unsigned char) *s
: 0;
166 return c
> 0x7f || isalpha(c
);
169 int uc_isdigit(char *s
)
171 int c
= s
? (unsigned char) *s
: 0;
172 return c
<= 0x7f && isdigit(c
);
179 if (uc_isalpha(c
) || uc_isdigit(c
) || c
[0] == '_')
184 #define UC_R2L(ch) (((ch) & 0xff00) == 0x0600 || \
185 ((ch) & 0xfffc) == 0x200c || \
186 ((ch) & 0xff00) == 0xfb00 || \
187 ((ch) & 0xff00) == 0xfc00 || \
188 ((ch) & 0xff00) == 0xfe00)
190 /* sorted list of characters that can be shaped */
191 static struct achar
{
192 unsigned c
; /* utf-8 code */
193 unsigned s
; /* single form */
194 unsigned i
; /* initial form */
195 unsigned m
; /* medial form */
196 unsigned f
; /* final form */
198 {0x0621, 0xfe80}, /* hamza */
199 {0x0622, 0xfe81, 0, 0, 0xfe82}, /* alef madda */
200 {0x0623, 0xfe83, 0, 0, 0xfe84}, /* alef hamza above */
201 {0x0624, 0xfe85, 0, 0, 0xfe86}, /* waw hamza */
202 {0x0625, 0xfe87, 0, 0, 0xfe88}, /* alef hamza below */
203 {0x0626, 0xfe89, 0xfe8b, 0xfe8c, 0xfe8a}, /* yeh hamza */
204 {0x0627, 0xfe8d, 0, 0, 0xfe8e}, /* alef */
205 {0x0628, 0xfe8f, 0xfe91, 0xfe92, 0xfe90}, /* beh */
206 {0x0629, 0xfe93, 0, 0, 0xfe94}, /* teh marbuta */
207 {0x062a, 0xfe95, 0xfe97, 0xfe98, 0xfe96}, /* teh */
208 {0x062b, 0xfe99, 0xfe9b, 0xfe9c, 0xfe9a}, /* theh */
209 {0x062c, 0xfe9d, 0xfe9f, 0xfea0, 0xfe9e}, /* jeem */
210 {0x062d, 0xfea1, 0xfea3, 0xfea4, 0xfea2}, /* hah */
211 {0x062e, 0xfea5, 0xfea7, 0xfea8, 0xfea6}, /* khah */
212 {0x062f, 0xfea9, 0, 0, 0xfeaa}, /* dal */
213 {0x0630, 0xfeab, 0, 0, 0xfeac}, /* thal */
214 {0x0631, 0xfead, 0, 0, 0xfeae}, /* reh */
215 {0x0632, 0xfeaf, 0, 0, 0xfeb0}, /* zain */
216 {0x0633, 0xfeb1, 0xfeb3, 0xfeb4, 0xfeb2}, /* seen */
217 {0x0634, 0xfeb5, 0xfeb7, 0xfeb8, 0xfeb6}, /* sheen */
218 {0x0635, 0xfeb9, 0xfebb, 0xfebc, 0xfeba}, /* sad */
219 {0x0636, 0xfebd, 0xfebf, 0xfec0, 0xfebe}, /* dad */
220 {0x0637, 0xfec1, 0xfec3, 0xfec4, 0xfec2}, /* tah */
221 {0x0638, 0xfec5, 0xfec7, 0xfec8, 0xfec6}, /* zah */
222 {0x0639, 0xfec9, 0xfecb, 0xfecc, 0xfeca}, /* ain */
223 {0x063a, 0xfecd, 0xfecf, 0xfed0, 0xfece}, /* ghain */
224 {0x0640, 0x640, 0x640, 0x640}, /* tatweel */
225 {0x0641, 0xfed1, 0xfed3, 0xfed4, 0xfed2}, /* feh */
226 {0x0642, 0xfed5, 0xfed7, 0xfed8, 0xfed6}, /* qaf */
227 {0x0643, 0xfed9, 0xfedb, 0xfedc, 0xfeda}, /* kaf */
228 {0x0644, 0xfedd, 0xfedf, 0xfee0, 0xfede}, /* lam */
229 {0x0645, 0xfee1, 0xfee3, 0xfee4, 0xfee2}, /* meem */
230 {0x0646, 0xfee5, 0xfee7, 0xfee8, 0xfee6}, /* noon */
231 {0x0647, 0xfee9, 0xfeeb, 0xfeec, 0xfeea}, /* heh */
232 {0x0648, 0xfeed, 0, 0, 0xfeee}, /* waw */
233 {0x0649, 0xfeef, 0, 0, 0xfef0}, /* alef maksura */
234 {0x064a, 0xfef1, 0xfef3, 0xfef4, 0xfef2}, /* yeh */
235 {0x067e, 0xfb56, 0xfb58, 0xfb59, 0xfb57}, /* peh */
236 {0x0686, 0xfb7a, 0xfb7c, 0xfb7d, 0xfb7b}, /* tcheh */
237 {0x0698, 0xfb8a, 0, 0, 0xfb8b}, /* jeh */
238 {0x06a9, 0xfb8e, 0xfb90, 0xfb91, 0xfb8f}, /* fkaf */
239 {0x06af, 0xfb92, 0xfb94, 0xfb95, 0xfb93}, /* gaf */
240 {0x06cc, 0xfbfc, 0xfbfe, 0xfbff, 0xfbfd}, /* fyeh */
242 {0x200d, 0, 0x200d, 0x200d}, /* ZWJ */
245 static struct achar
*find_achar(int c
)
250 /* using binary search to find c */
253 if (achars
[m
].c
== c
)
263 static int can_join(int c1
, int c2
)
265 struct achar
*a1
= find_achar(c1
);
266 struct achar
*a2
= find_achar(c2
);
267 return a1
&& a2
&& (a1
->i
|| a1
->m
) && (a2
->f
|| a2
->m
);
270 static int uc_cshape(int cur
, int prev
, int next
)
273 int join_prev
, join_next
;
274 struct achar
*ac
= find_achar(c
);
275 if (!ac
) /* ignore non-Arabic characters */
277 join_prev
= can_join(prev
, c
);
278 join_next
= can_join(c
, next
);
279 if (join_prev
&& join_next
)
281 if (join_prev
&& !join_next
)
283 if (!join_prev
&& join_next
)
285 if (!join_prev
&& !join_next
)
286 c
= ac
->c
; /* some fonts do not have a glyph for ac->s */
291 * return nonzero for Arabic combining characters
293 * The standard Arabic diacritics:
302 * + 0x0653: madda above
303 * + 0x0654: hamza above
304 * + 0x0655: hamza below
305 * + 0x0670: superscript alef
307 static int uc_comb(int c
)
309 return (c
>= 0x064b && c
<= 0x0655) || /* the standard diacritics */
310 (c
>= 0xfc5e && c
<= 0xfc63) || /* shadda ligatures */
311 c
== 0x0670; /* superscript alef */
314 static void uc_cput(char *d
, int c
)
318 *d
++ = 0xf0 | (c
>> 18);
320 } else if (c
> 0x7ff) {
321 *d
++ = 0xe0 | (c
>> 12);
323 } else if (c
> 0x7f) {
324 *d
++ = 0xc0 | (c
>> 6);
330 *d
++ = 0x80 | ((c
>> (l
* 6)) & 0x3f);
334 /* shape the given arabic character; returns a static buffer */
335 char *uc_shape(char *beg
, char *s
)
341 int curr
= uc_code(s
);
342 if (!curr
|| !UC_R2L(curr
))
346 r
= uc_beg(beg
, r
- 1);
347 if (!uc_comb(uc_code(r
))) {
355 if (!uc_comb(uc_code(r
))) {
360 uc_cput(out
, uc_cshape(curr
, prev
, next
));