1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
|
/*
* Jeffrey Friedl
* Omron Corporation ʳ
* Nagaokakyoshi, Japan 617Ĺ
*
* jfriedl@nff.ncl.omron.co.jp
*
* This work is placed under the terms of the GNU General Purpose License
* (the "GNU Copyleft").
*
* Oct 1993
*
* Feb 1995: Added fuzziness.
* See "fuzzhkana.h" for overall comments.
*/
#include "config.h"
#include "assert.h"
#include "output.h"
#include "fuzzkana.h"
#include "kanaid.h"
/*
* Given
*
* IN -- some regex pattern string whose kana is to be fuzzified.
* OUT -- where to stick the new pattern string
* OUT_SIZE -- size of area pointed to by OUT.
*
* FLAGS --
* If FUZZ_LONG_VOWELS is set, the pattern will be written such
* that longness of vowels doesn't matter.
*
* If FUZZ_SMALL_TSU is set, the pattern will be written such
* that small TSUs won't matter.
*
* The number of bytes written to the output is returned, or zero if the
* output buffer was overflowed (or, I suppose, if the input is empty)
*
* If OUT is zero, nothing is written (obviously)... the number of bytes
* that would be required is returned (OUT_SIZE is ignored in this case)
*
* An example with both flags set
* in ֤Ȥ
* out ֤[]*?[]*
*
*/
#define LEAST_KANJI_HI_BYTE 0260
#define GREATEST_KANJI_HI_BYTE 0364
#define IS_EUC(HighByte) ((HighByte) & 0x80)
unsigned fuzzkana(const unsigned char *in,
unsigned char *out,
unsigned out_size,
unsigned flags)
{
const unsigned char *orig_out = out;
const unsigned char *out_end = out + out_size;
unsigned char hi, lo;
int just_want_size = (out == 0);
int do_voiced = (flags & FUZZ_VOICED);
int do_vowels = (flags & FUZZ_LONG_VOWELS);
int do_tsu = (flags & FUZZ_SMALL_TSU);
#ifdef FUZZ_REPEATER
int do_repeat = (flags & FUZZ_REPEATER);
#endif
if (in == 0 || *in == 0)
return 0;
/*
* Shove the given character to the 'out' buffer,
* aborting the function if the buffer is overflowed.
* However, if just requesting how much buffer is required
* (buy supplying no out buffer), just bump up the counter.
*/
#define SENDOUT(c) \
macro_start { \
unsigned char value = (c); \
if (just_want_size) \
out++; \
else \
{ \
if (out >= out_end) \
return 0; \
*out++ = value; \
} \
} macro_end
/* to tack on the given string to the output */
#define add(STR) \
macro_start { \
const unsigned char *str = (const unsigned char *)(STR); \
while (*str) \
SENDOUT(*str++); \
} macro_end
/* run through the line */
while (hi = *in++, hi != 0)
{
unsigned next_id;
unsigned vsound;
unsigned id;
/* Just pass through ASCII characters */
if (!IS_EUC(hi))
{
SENDOUT(hi);
continue;
}
lo = *in++; /* get the next byte of the EUC char. */
/* if this char is not kana, so just pass through and continue */
if (!IS_KANA(hi,lo))
{
SENDOUT(hi);
SENDOUT(lo);
#ifdef FUZZ_REPEATER_no_no_no /* now done in jregex.c */
#define REPEATER_HI_BYTE 0241 /* high byte of */
#define REPEATER_LO_BYTE 0271 /* low byte of */
if (do_repeat &&
hi >= LEAST_KANJI_HI_BYTE && hi <= GREATEST_KANJI_HI_BYTE &&
((in[0] == hi && in[1] == lo) ||
(in[0] == REPEATER_HI_BYTE && in[1] == REPEATER_LO_BYTE)))
{
SENDOUT('[');
SENDOUT(REPEATER_HI_BYTE);
SENDOUT(REPEATER_LO_BYTE);
SENDOUT(hi);
SENDOUT(lo);
SENDOUT(']');
in += 2;
}
#endif /* FUZZ_REPEATER */
continue;
}
id = KANA_ID(hi,lo); /* the the id flags for this kana */
/* if we're doing voiced fuzz, fuzz dual-chracter'd sounds */
if (do_voiced && (id & KID_DUAL))
{
switch(id)
{
default:
die("oops, %02x %02x -> id is %x\n", hi, lo, id);
break;
case KID_E | KID_VOWEL | KID_DUAL: /* */
case KID_ARCHAIC | KID_DUAL: /* */
add(IS_HIRAGANA(hi,lo) ? "[]" : "[]");
break;
case KID_O | KID_VOWEL | KID_DUAL: /* */
case KID_o | KID_DUAL: /* */
add(IS_HIRAGANA(hi,lo) ? "[]" : "[]");
break;
case KID_I | KID_Z | KID_DUAL: /* */
case KID_I | KID_D | KID_DUAL: /* */
add(IS_HIRAGANA(hi,lo) ? "[]" : "[]");
break;
case KID_U | KID_D | KID_DUAL: /* */
case KID_U | KID_Z | KID_DUAL: /* */
add(IS_HIRAGANA(hi,lo) ? "[Ť]" : "[ť]");
break;
}
} else {
/* otherwise, just pass through */
SENDOUT(hi);
SENDOUT(lo);
}
vsound = id & KID_VSOUND;
#define GET_NEXT_ID(ptr) \
macro_start { \
if (!IS_EUC((ptr)[0])) \
next_id = 0; \
else if (IS_DASH((ptr)[0], (ptr)[1])) \
next_id = vsound | KID_VOWEL; \
else if (!IS_KANA((ptr)[0], (ptr)[1])) \
next_id = 0; \
else \
next_id = KANA_ID((ptr)[0], (ptr)[1]); \
} macro_end
GET_NEXT_ID(in);
#if 0
/* consider the O and U sound to be the same */
if (vsound & (KID_O|KID_U))
vsound = KID_O|KID_U;
#endif
if (do_vowels)
{
/*
* If current character has a vowel sound and is not followed
* by a small y-consonant sound, allow to be doubled.
*/
if (vsound && (next_id & (KID_SMALL|KID_Y)) != (KID_SMALL|KID_Y))
{
/*
* Unless they have case folding off, it won't matter which
* of these two we use, but since it would matter if they
* had it off, we'll separate them....
*/
if (IS_HIRAGANA(hi,lo))
switch (vsound) {
default: assert(0); break;
case KID_A: add("[]*"); break;
case KID_I: add("[]*"); break;
case KID_U: add("[]*"); break;
case KID_O: add("[]*"); vsound |= KID_U; break;
case KID_E: add("[]*"); break;
}
else
switch (vsound) {
default: assert(0); break;
case KID_A: add("[]*"); break;
case KID_I: add("[]*"); break;
case KID_U: add("[]*"); break;
case KID_O: add("[]*"); vsound |= KID_U; break;
case KID_E: add("[]*"); break;
}
/*
* If the next char is the simple vowel we've just
* allowed to be doubled, skip it.
*
* If the input string is something like֤
* This IF will cause each֤pair to combine into
* a֤[]*pattern. However, if you make the IF
* a WHILE, it would collapse them all down into one.
* Either method has their benefits.
*/
while ((next_id & KID_VOWEL) && (next_id & vsound))
{
in += 2; /* skip the vowel we've just replaced */
GET_NEXT_ID(in);
}
}
}
if (do_tsu)
{
unsigned next_is_small_tsu = (next_id & (KID_T|KID_U|KID_SMALL)) ==
(KID_T|KID_U|KID_SMALL);
/*
* if the next thing is a hard-coded small-TSU, or all of
* + currently have a vowel sound, and
* + the next character has an "appropriate" consonant sound,
* + and the next character isn't small.
* then we'll add a possible small tsu.
*/
#ifndef SMALL_TSU_OK
#define SMALL_TSU_OK (KID_K|KID_S|KID_T|KID_D|KID_P|KID_W|KID_M)
#endif
if (next_is_small_tsu ||
(vsound && (next_id & SMALL_TSU_OK) && !(next_id & KID_SMALL)))
{
/* add a possability for a small TSU */
add("?");
if (next_is_small_tsu)
in += 2; /* skip small tsu that's there */
}
}
}
SENDOUT(0); /* tack on a null */
return out - orig_out;
}
|
