1 /*
2  * "streamable kanji code filter and converter"
3  * Copyright (c) 1998-2002 HappySize, Inc. All rights reserved.
4  *
5  * LICENSE NOTICES
6  *
7  * This file is part of "streamable kanji code filter and converter",
8  * which is distributed under the terms of GNU Lesser General Public
9  * License (version 2) as published by the Free Software Foundation.
10  *
11  * This software is distributed in the hope that it will be useful,
12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
14  * GNU Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with "streamable kanji code filter and converter";
18  * if not, write to the Free Software Foundation, Inc., 59 Temple Place,
19  * Suite 330, Boston, MA  02111-1307  USA
20  *
21  * The author of this file:
22  *
23  */
24 /*
25  * the source code included in this files was separated from mbfilter_cp936.c
26  * by rui hirokawa <hirokawa@php.net> on 11 Aug 2011.
27  *
28  */
29 
30 #ifdef HAVE_CONFIG_H
31 #include "config.h"
32 #endif
33 
34 #include "mbfilter.h"
35 #include "mbfilter_gb18030.h"
36 
37 #include "unicode_table_cp936.h"
38 #include "unicode_table_gb18030.h"
39 
40 static int mbfl_filt_ident_gb18030(int c, mbfl_identify_filter *filter);
41 
42 static const char *mbfl_encoding_gb18030_aliases[] = {"gb-18030", "gb-18030-2000", NULL};
43 
44 const mbfl_encoding mbfl_encoding_gb18030 = {
45 	mbfl_no_encoding_gb18030,
46 	"GB18030",
47 	"GB18030",
48 	(const char *(*)[])&mbfl_encoding_gb18030_aliases,
49 	NULL,
50 	MBFL_ENCTYPE_MBCS | MBFL_ENCTYPE_GL_UNSAFE
51 };
52 
53 const struct mbfl_identify_vtbl vtbl_identify_gb18030 = {
54 	mbfl_no_encoding_gb18030,
55 	mbfl_filt_ident_common_ctor,
56 	mbfl_filt_ident_common_dtor,
57 	mbfl_filt_ident_gb18030
58 };
59 
60 const struct mbfl_convert_vtbl vtbl_gb18030_wchar = {
61 	mbfl_no_encoding_gb18030,
62 	mbfl_no_encoding_wchar,
63 	mbfl_filt_conv_common_ctor,
64 	mbfl_filt_conv_common_dtor,
65 	mbfl_filt_conv_gb18030_wchar,
66 	mbfl_filt_conv_common_flush
67 };
68 
69 const struct mbfl_convert_vtbl vtbl_wchar_gb18030 = {
70 	mbfl_no_encoding_wchar,
71 	mbfl_no_encoding_gb18030,
72 	mbfl_filt_conv_common_ctor,
73 	mbfl_filt_conv_common_dtor,
74 	mbfl_filt_conv_wchar_gb18030,
75 	mbfl_filt_conv_common_flush
76 };
77 
78 #define CK(statement)	do { if ((statement) < 0) return (-1); } while (0)
79 
80 
81 int
mbfl_bisec_srch(int w,const unsigned short * tbl,int n)82 mbfl_bisec_srch(int w, const unsigned short *tbl, int n)
83 {
84 	int k, k1 = 0, k2 = n-1;
85 
86 	while (k1 < k2) {
87 		k = (k1+k2) >> 1;
88 		if (w <= tbl[2*k+1]) {
89 			k2 = k;
90 		} else if (w >= tbl[2*k+2]) {
91 			k1 = k + 1;
92 		} else {
93 			return -1;
94 		}
95 	}
96 	return k1;
97 }
98 
99 int
mbfl_bisec_srch2(int w,const unsigned short tbl[],int n)100 mbfl_bisec_srch2(int w, const unsigned short tbl[], int n)
101 {
102 	int k, k1 = 0, k2 = n;
103 
104 	if (w == tbl[0]) {
105 		return 0;
106 	}
107 
108 	while (k2 - k1 > 1) {
109 		k = (k1 + k2) >> 1;
110 		if (w < tbl[k]) {
111 			k2 = k;
112 		} else if (w > tbl[k]) {
113 			k1 = k;
114 		} else {
115 			return k;
116 		}
117 	}
118 	return -1;
119 }
120 
121 /*
122  * GB18030 => wchar
123  */
124 int
mbfl_filt_conv_gb18030_wchar(int c,mbfl_convert_filter * filter)125 mbfl_filt_conv_gb18030_wchar(int c, mbfl_convert_filter *filter)
126 {
127 	int k;
128 	int c1, c2, c3, w = -1;
129 
130 	switch (filter->status) {
131 	case 0:
132 		if (c >= 0 && c < 0x80) {	/* latin */
133 			CK((*filter->output_function)(c, filter->data));
134 		} else if (c == 0x80) {	/* euro sign */
135 			CK((*filter->output_function)(0x20ac, filter->data));
136 		} else if (c == 0xff) {
137 			CK((*filter->output_function)(0x00ff, filter->data));
138 		} else if (c > 0x80 && c < 0xff) {	/* dbcs/qbcs lead byte */
139 			filter->status = 1;
140 			filter->cache = c;
141 		} else {
142 			w = c & MBFL_WCSGROUP_MASK;
143 			w |= MBFL_WCSGROUP_THROUGH;
144 			CK((*filter->output_function)(w, filter->data));
145 		}
146 		break;
147 
148 	case 1:		/* dbcs/qbcs second byte */
149 		c1 = filter->cache;
150 		filter->status = 0;
151 
152 		if (c1 >= 0x81 && c1 <= 0x84 && c >= 0x30 && c <= 0x39) { /* 4 byte range: Unicode BMP */
153 			filter->status = 2;
154 			filter->cache = (c1 << 8) | c;
155 			return c;
156 		} else if (c1 >= 0x90 && c1 <= 0xe3 && c >= 0x30 && c <= 0x39) {
157 			/* 4 byte range: Unicode 16 planes */
158 			filter->status = 2;
159 			filter->cache = (c1 << 8) | c;
160 			return c;
161 		} else if (((c1 >= 0xaa && c1 <= 0xaf) || (c1 >= 0xf8 && c1 <= 0xfe)) &&
162 				   (c >= 0xa1 && c <= 0xfe)) { /* UDA part1,2: U+E000-U+E4C5 */
163 			w = 94*(c1 >= 0xf8 ? c1 - 0xf2 : c1 - 0xaa) + (c - 0xa1) + 0xe000;
164 			CK((*filter->output_function)(w, filter->data));
165 		} else if (c1 >= 0xa1 && c1 <= 0xa7 && c >= 0x40 && c < 0xa1 && c != 0x7f) {
166 			/* UDA part3 : U+E4C6-U+E765*/
167 			w = 96*(c1 - 0xa1) + c - (c >= 0x80 ? 0x41 : 0x40) + 0xe4c6;
168 			CK((*filter->output_function)(w, filter->data));
169 		}
170 
171 		c2 = (c1 << 8) | c;
172 
173 		if (w <= 0 &&
174 			((c2 >= 0xa2ab && c2 <= 0xa9f0 + (0xe80f-0xe801)) ||
175 			 (c2 >= 0xd7fa && c2 <= 0xd7fa + (0xe814-0xe810)) ||
176 			 (c2 >= 0xfe50 && c2 <= 0xfe80 + (0xe864-0xe844)))) {
177 			for (k = 0; k < mbfl_gb18030_pua_tbl_max; k++) {
178 				if (c2 >= mbfl_gb18030_pua_tbl[k][2] &&
179 					c2 <= mbfl_gb18030_pua_tbl[k][2] +  mbfl_gb18030_pua_tbl[k][1]
180 					-  mbfl_gb18030_pua_tbl[k][0]) {
181 					w = c2 -  mbfl_gb18030_pua_tbl[k][2] + mbfl_gb18030_pua_tbl[k][0];
182 					CK((*filter->output_function)(w, filter->data));
183 					break;
184 				}
185 			}
186 		}
187 
188 		if (w <= 0) {
189 			if ((c1 >= 0xa1 && c1 <= 0xa9 && c >= 0xa1 && c <= 0xfe) ||
190 				(c1 >= 0xb0 && c1 <= 0xf7 && c >= 0xa1 && c <= 0xfe) ||
191 				(c1 >= 0x81 && c1 <= 0xa0 && c >= 0x40 && c <= 0xfe && c != 0x7f) ||
192 				(c1 >= 0xaa && c1 <= 0xfe && c >= 0x40 && c <= 0xa0 && c != 0x7f) ||
193 				(c1 >= 0xa8 && c1 <= 0xa9 && c >= 0x40 && c <= 0xa0 && c != 0x7f)) {
194 				w = (c1 - 0x81)*192 + (c - 0x40);
195 				if (w >= 0 && w < cp936_ucs_table_size) {
196 					w = cp936_ucs_table[w];
197 				} else {
198 					w = 0;
199 				}
200 				if (w <= 0) {
201 					w = (c1 << 8) | c;
202 					w &= MBFL_WCSPLANE_MASK;
203 					w |= MBFL_WCSPLANE_GB18030;
204 				}
205 				CK((*filter->output_function)(w, filter->data));
206 			} else if ((c >= 0 && c < 0x21) || c == 0x7f) {		/* CTLs */
207 				CK((*filter->output_function)(c, filter->data));
208 			} else {
209 				w = (c1 << 8) | c;
210 				w &= MBFL_WCSGROUP_MASK;
211 				w |= MBFL_WCSGROUP_THROUGH;
212 				CK((*filter->output_function)(w, filter->data));
213 			}
214 		}
215 		break;
216 	case 2: /* qbcs third byte */
217 		c1 = (filter->cache >> 8) & 0xff;
218 		c2 = filter->cache & 0xff;
219 		filter->status = 0;
220 		filter->cache = 0;
221 		if (((c1 >= 0x81 && c1 <= 0x84) || (c1 >= 0x90 && c1 <= 0xe3)) &&
222 			c2 >= 0x30 && c2 <= 0x39 && c >= 0x81 && c <= 0xfe) {
223 			filter->cache = (c1 << 16) | (c2 << 8) | c;
224 			filter->status = 3;
225 		} else {
226 			w = (c1 << 16) | (c2 << 8) | c;
227 			w &= MBFL_WCSGROUP_MASK;
228 			w |= MBFL_WCSGROUP_THROUGH;
229 			CK((*filter->output_function)(w, filter->data));
230 		}
231  		break;
232 
233 	case 3: /* qbcs fourth byte */
234 		c1 = (filter->cache >> 16) & 0xff;
235 		c2 = (filter->cache >> 8) & 0xff;
236 		c3 = filter->cache & 0xff;
237 		filter->status = 0;
238 		filter->cache = 0;
239 		if (((c1 >= 0x81 && c1 <= 0x84) || (c1 >= 0x90 && c1 <= 0xe3)) &&
240 			c2 >= 0x30 && c2 <= 0x39 && c3 >= 0x81 && c3 <= 0xfe && c >= 0x30 && c <= 0x39) {
241 			if (c1 >= 0x90 && c1 <= 0xe3) {
242 				w = ((((c1 - 0x90)*10 + (c2 - 0x30))*126 + (c3 - 0x81)))*10 + (c - 0x30) + 0x10000;
243 			} else { /* Unicode BMP */
244 				w = (((c1 - 0x81)*10 + (c2 - 0x30))*126 + (c3 - 0x81))*10 + (c - 0x30);
245 				if (w >= 0 && w <= 39419) {
246 					k = mbfl_bisec_srch(w, mbfl_gb2uni_tbl, mbfl_gb_uni_max);
247 					if (k<0) {
248 						/* error */
249 						w = (c1 << 24) | (c2 << 16) | (c3 << 8) | c;
250 						w &= MBFL_WCSGROUP_MASK;
251 						w |= MBFL_WCSGROUP_THROUGH;
252 						CK((*filter->output_function)(w, filter->data));
253 						return c;
254 					}
255 					w += mbfl_gb_uni_ofst[k];
256 				} else {
257 					w = (c1 << 24) | (c2 << 16) | (c3 << 8) | c;
258 					w &= MBFL_WCSGROUP_MASK;
259 					w |= MBFL_WCSGROUP_THROUGH;
260 					CK((*filter->output_function)(w, filter->data));
261 					return c;
262 				}
263 			}
264 			CK((*filter->output_function)(w, filter->data));
265 		} else {
266 			w = (c1 << 24) | (c2 << 16) | (c3 << 8) | c;
267 			w &= MBFL_WCSGROUP_MASK;
268 			w |= MBFL_WCSGROUP_THROUGH;
269 			CK((*filter->output_function)(w, filter->data));
270 		}
271  		break;
272 
273 	default:
274 		filter->status = 0;
275 		break;
276 	}
277 
278 	return c;
279 }
280 
281 /*
282  * wchar => GB18030
283  */
284 int
mbfl_filt_conv_wchar_gb18030(int c,mbfl_convert_filter * filter)285 mbfl_filt_conv_wchar_gb18030(int c, mbfl_convert_filter *filter)
286 {
287 	int k, k1, k2;
288 	int c1, s = 0, s1 = 0;
289 
290 	if (c >= ucs_a1_cp936_table_min && c < ucs_a1_cp936_table_max) {
291 		s = ucs_a1_cp936_table[c - ucs_a1_cp936_table_min];
292 	} else if (c >= ucs_a2_cp936_table_min && c < ucs_a2_cp936_table_max) {
293 		s = ucs_a2_cp936_table[c - ucs_a2_cp936_table_min];
294 	} else if (c >= ucs_a3_cp936_table_min && c < ucs_a3_cp936_table_max) {
295 		s = ucs_a3_cp936_table[c - ucs_a3_cp936_table_min];
296 	} else if (c >= ucs_i_cp936_table_min && c < ucs_i_cp936_table_max) {
297 		s = ucs_i_cp936_table[c - ucs_i_cp936_table_min];
298 	} else if (c >= ucs_ci_cp936_table_min && c < ucs_ci_cp936_table_max) {
299 		/* U+F900-FA2F CJK Compatibility Ideographs */
300 		if (c == 0xf92c) {
301 			s = 0xfd9c;
302 		} else if (c == 0xf979) {
303 			s = 0xfd9d;
304 		} else if (c == 0xf995) {
305 			s = 0xfd9e;
306 		} else if (c == 0xf9e7) {
307 			s = 0xfd9f;
308 		} else if (c == 0xf9f1) {
309 			s = 0xfda0;
310 		} else if (c >= 0xfa0c && c <= 0xfa29) {
311 			s = ucs_ci_s_cp936_table[c - 0xfa0c];
312 		}
313 	} else if (c >= ucs_cf_cp936_table_min && c < ucs_cf_cp936_table_max) {
314 		/* FE30h CJK Compatibility Forms  */
315 		s = ucs_cf_cp936_table[c - ucs_cf_cp936_table_min];
316 	} else if (c >= ucs_sfv_cp936_table_min && c < ucs_sfv_cp936_table_max) {
317 		/* U+FE50-FE6F Small Form Variants */
318 		s = ucs_sfv_cp936_table[c - ucs_sfv_cp936_table_min];
319 	} else if (c >= ucs_hff_cp936_table_min && c < ucs_hff_cp936_table_max) {
320 		/* U+FF00-FFFF HW/FW Forms */
321 		if (c == 0xff04) {
322 			s = 0xa1e7;
323 		} else if (c == 0xff5e) {
324 			s = 0xa1ab;
325 		} else if (c >= 0xff01 && c <= 0xff5d) {
326 			s = c - 0xff01 + 0xa3a1;
327 		} else if (c >= 0xffe0 && c <= 0xffe5) {
328 			s = ucs_hff_s_cp936_table[c-0xffe0];
329 		}
330 	}
331 
332 	if (c == 0x20ac) { /* euro-sign */
333 		s = 0xa2e3;
334 	}
335 
336 	if (s <= 0 && c >= mbfl_gb18030_c_tbl_key[0] &&
337 		c <= mbfl_gb18030_c_tbl_key[mbfl_gb18030_c_tbl_max-1]) {
338 		k1 = mbfl_bisec_srch2(c, mbfl_gb18030_c_tbl_key, mbfl_gb18030_c_tbl_max);
339 		if (k1 >= 0) {
340 			s = mbfl_gb18030_c_tbl_val[k1];
341 		}
342 	}
343 
344 	if (c >= 0xe000 && c <= 0xe864) { /* PUA */
345 		if (c < 0xe766) {
346 			if (c < 0xe4c6) {
347 				c1 = c - 0xe000;
348 				s = (c1 % 94) + 0xa1; c1 /= 94;
349 				s |= (c1 < 0x06 ? c1 + 0xaa : c1 + 0xf2) << 8;
350 			} else {
351 				c1 = c - 0xe4c6;
352 				s = ((c1 / 96) + 0xa1) << 8; c1 %= 96;
353 				s |= c1 + (c1 >= 0x3f ? 0x41 : 0x40);
354 			}
355 		} else {
356 			/* U+E766..U+E864 */
357 			k1 = 0; k2 = mbfl_gb18030_pua_tbl_max;
358 			while (k1 < k2) {
359 				k = (k1 + k2) >> 1;
360 				if (c < mbfl_gb18030_pua_tbl[k][0]) {
361 					k2 = k;
362 				} else if (c > mbfl_gb18030_pua_tbl[k][1]) {
363 					k1 = k + 1;
364 				} else {
365 					s = c - mbfl_gb18030_pua_tbl[k][0] + mbfl_gb18030_pua_tbl[k][2];
366 					break;
367 				}
368 			}
369 		}
370 	}
371 
372 	if (s <= 0 && c >= 0x0080 && c <= 0xffff) { /* BMP */
373 		s = mbfl_bisec_srch(c, mbfl_uni2gb_tbl, mbfl_gb_uni_max);
374 		if (s >= 0) {
375 			c1 = c - mbfl_gb_uni_ofst[s];
376 			s = (c1 % 10) + 0x30; c1 /= 10;
377 			s |= ((c1 % 126) + 0x81) << 8; c1 /= 126;
378 			s |= ((c1 % 10) + 0x30) << 16; c1 /= 10;
379 			s1 = c1 + 0x81;
380 		}
381 	} else if (c >= 0x10000 && c <= 0x10ffff) { /* Code set 3: Unicode U+10000..U+10FFFF */
382 		c1 = c - 0x10000;
383 		s = (c1 % 10) + 0x30; c1 /= 10;
384 		s |= ((c1 % 126) + 0x81) << 8; c1 /= 126;
385 		s |= ((c1 % 10) + 0x30) << 16; c1 /= 10;
386 		s1 = c1 + 0x90;
387 	}
388 
389 	if (s <= 0) {
390 		c1 = c & ~MBFL_WCSPLANE_MASK;
391 		if (c1 == MBFL_WCSPLANE_WINCP936) {
392 			s = c & MBFL_WCSPLANE_MASK;
393 		}
394 		if (c == 0) {
395 			s = 0;
396 		} else if (s <= 0) {
397 			s = -1;
398 		}
399 	}
400 	if (s >= 0) {
401 		if (s <= 0x80) {	/* latin */
402 			CK((*filter->output_function)(s, filter->data));
403 		} else if (s1 > 0) { /* qbcs */
404 			CK((*filter->output_function)(s1 & 0xff, filter->data));
405 			CK((*filter->output_function)((s >> 16) & 0xff, filter->data));
406 			CK((*filter->output_function)((s >> 8) & 0xff, filter->data));
407 			CK((*filter->output_function)(s & 0xff, filter->data));
408 		} else { /* dbcs */
409 			CK((*filter->output_function)((s >> 8) & 0xff, filter->data));
410 			CK((*filter->output_function)(s & 0xff, filter->data));
411 		}
412 	} else {
413 		if (filter->illegal_mode != MBFL_OUTPUTFILTER_ILLEGAL_MODE_NONE) {
414 			CK(mbfl_filt_conv_illegal_output(c, filter));
415 		}
416 	}
417 
418 	return c;
419 }
420 
mbfl_filt_ident_gb18030(int c,mbfl_identify_filter * filter)421 static int mbfl_filt_ident_gb18030(int c, mbfl_identify_filter *filter)
422 {
423 	int c1;
424 
425 	c1 = (filter->status >> 8) & 0xff;
426 	filter->status &= 0xff;
427 
428 	if (filter->status == 0) {
429 		if (c <= 0x80 || c == 0xff) {
430 			filter->status = 0;
431 		} else {
432 			filter->status = 1;
433 			filter->status |= (c << 8);
434 		}
435 	} else if (filter->status == 1) { /* dbcs/qbcs 2nd byte */
436 		if (((c1 >= 0x81 && c1 <= 0x84) || (c1 >= 0x90 && c1 <= 0xe3)) && c >= 0x30 && c <= 0x39) { /* qbcs */
437 			filter->status = 2;
438 		} else if (((c1 >= 0xaa && c1 <= 0xaf) || (c1 >= 0xf8 && c1 <= 0xfe)) && (c >= 0xa1 && c <= 0xfe)) {
439 			filter->status = 0; /* UDA part 1,2 */
440 		} else if (c1 >= 0xa1 && c1 <= 0xa7 && c >= 0x40 && c < 0xa1 && c != 0x7f) {
441 			filter->status = 0; /* UDA part 3 */
442 		} else if ((c1 >= 0xa1 && c1 <= 0xa9 && c >= 0xa1 && c <= 0xfe) ||
443 				   (c1 >= 0xb0 && c1 <= 0xf7 && c >= 0xa1 && c <= 0xfe) ||
444 				   (c1 >= 0x81 && c1 <= 0xa0 && c >= 0x40 && c <= 0xfe && c != 0x7f) ||
445 				   (c1 >= 0xaa && c1 <= 0xfe && c >= 0x40 && c <= 0xa0 && c != 0x7f) ||
446 				   (c1 >= 0xa8 && c1 <= 0xa9 && c >= 0x40 && c <= 0xa0 && c != 0x7f)) {
447 			filter->status = 0; /* DBCS */
448 		} else {
449 			filter->flag = 1; /* bad */
450 			filter->status = 0;
451 		}
452 	} else if (filter->status == 2) { /* qbcs 3rd byte */
453 		if (c > 0x80 && c < 0xff) {
454 			filter->status = 3;
455 		} else {
456 			filter->flag = 1; /* bad */
457 			filter->status = 0;
458 		}
459 	} else if (filter->status == 3) { /* qbcs 4th byte */
460 		if (c >= 0x30 && c < 0x40) {
461 			filter->status = 0;
462 		} else {
463 			filter->flag = 1; /* bad */
464 			filter->status = 0;
465 		}
466 	} else {							/* bad */
467 		filter->flag = 1;
468 	}
469 
470 	return c;
471 }
472