Line data Source code
1 : /*
2 : * fuzzystrmatch.c
3 : *
4 : * Functions for "fuzzy" comparison of strings
5 : *
6 : * Joe Conway <mail@joeconway.com>
7 : *
8 : * contrib/fuzzystrmatch/fuzzystrmatch.c
9 : * Copyright (c) 2001-2023, PostgreSQL Global Development Group
10 : * ALL RIGHTS RESERVED;
11 : *
12 : * metaphone()
13 : * -----------
14 : * Modified for PostgreSQL by Joe Conway.
15 : * Based on CPAN's "Text-Metaphone-1.96" by Michael G Schwern <schwern@pobox.com>
16 : * Code slightly modified for use as PostgreSQL function (palloc, elog, etc).
17 : * Metaphone was originally created by Lawrence Philips and presented in article
18 : * in "Computer Language" December 1990 issue.
19 : *
20 : * Permission to use, copy, modify, and distribute this software and its
21 : * documentation for any purpose, without fee, and without a written agreement
22 : * is hereby granted, provided that the above copyright notice and this
23 : * paragraph and the following two paragraphs appear in all copies.
24 : *
25 : * IN NO EVENT SHALL THE AUTHORS OR DISTRIBUTORS BE LIABLE TO ANY PARTY FOR
26 : * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING
27 : * LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS
28 : * DOCUMENTATION, EVEN IF THE AUTHOR OR DISTRIBUTORS HAVE BEEN ADVISED OF THE
29 : * POSSIBILITY OF SUCH DAMAGE.
30 : *
31 : * THE AUTHORS AND DISTRIBUTORS SPECIFICALLY DISCLAIM ANY WARRANTIES,
32 : * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
33 : * AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
34 : * ON AN "AS IS" BASIS, AND THE AUTHOR AND DISTRIBUTORS HAS NO OBLIGATIONS TO
35 : * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
36 : *
37 : */
38 :
39 : #include "postgres.h"
40 :
41 : #include <ctype.h>
42 :
43 : #include "mb/pg_wchar.h"
44 : #include "utils/builtins.h"
45 : #include "utils/varlena.h"
46 : #include "varatt.h"
47 :
48 2 : PG_MODULE_MAGIC;
49 :
50 : /*
51 : * Soundex
52 : */
53 : static void _soundex(const char *instr, char *outstr);
54 :
55 : #define SOUNDEX_LEN 4
56 :
57 : /* ABCDEFGHIJKLMNOPQRSTUVWXYZ */
58 : static const char *soundex_table = "01230120022455012623010202";
59 :
60 : static char
61 254 : soundex_code(char letter)
62 : {
63 254 : letter = toupper((unsigned char) letter);
64 : /* Defend against non-ASCII letters */
65 254 : if (letter >= 'A' && letter <= 'Z')
66 252 : return soundex_table[letter - 'A'];
67 2 : return letter;
68 : }
69 :
70 : /*
71 : * Metaphone
72 : */
73 : #define MAX_METAPHONE_STRLEN 255
74 :
75 : /*
76 : * Original code by Michael G Schwern starts here.
77 : * Code slightly modified for use as PostgreSQL function.
78 : */
79 :
80 :
81 : /**************************************************************************
82 : metaphone -- Breaks english phrases down into their phonemes.
83 :
84 : Input
85 : word -- An english word to be phonized
86 : max_phonemes -- How many phonemes to calculate. If 0, then it
87 : will phonize the entire phrase.
88 : phoned_word -- The final phonized word. (We'll allocate the
89 : memory.)
90 : Output
91 : error -- A simple error flag, returns true or false
92 :
93 : NOTES: ALL non-alpha characters are ignored, this includes whitespace,
94 : although non-alpha characters will break up phonemes.
95 : ****************************************************************************/
96 :
97 :
98 : /* I add modifications to the traditional metaphone algorithm that you
99 : might find in books. Define this if you want metaphone to behave
100 : traditionally */
101 : #undef USE_TRADITIONAL_METAPHONE
102 :
103 : /* Special encodings */
104 : #define SH 'X'
105 : #define TH '0'
106 :
107 : static char Lookahead(char *word, int how_far);
108 : static void _metaphone(char *word, int max_phonemes, char **phoned_word);
109 :
110 : /* Metachar.h ... little bits about characters for metaphone */
111 :
112 :
113 : /*-- Character encoding array & accessing macros --*/
114 : /* Stolen directly out of the book... */
115 : static const char _codes[26] = {
116 : 1, 16, 4, 16, 9, 2, 4, 16, 9, 2, 0, 2, 2, 2, 1, 4, 0, 2, 4, 4, 1, 0, 0, 0, 8, 0
117 : /* a b c d e f g h i j k l m n o p q r s t u v w x y z */
118 : };
119 :
120 : static int
121 2 : getcode(char c)
122 : {
123 2 : if (isalpha((unsigned char) c))
124 : {
125 2 : c = toupper((unsigned char) c);
126 : /* Defend against non-ASCII letters */
127 2 : if (c >= 'A' && c <= 'Z')
128 2 : return _codes[c - 'A'];
129 : }
130 0 : return 0;
131 : }
132 :
133 : #define isvowel(c) (getcode(c) & 1) /* AEIOU */
134 :
135 : /* These letters are passed through unchanged */
136 : #define NOCHANGE(c) (getcode(c) & 2) /* FJMNR */
137 :
138 : /* These form diphthongs when preceding H */
139 : #define AFFECTH(c) (getcode(c) & 4) /* CGPST */
140 :
141 : /* These make C and G soft */
142 : #define MAKESOFT(c) (getcode(c) & 8) /* EIY */
143 :
144 : /* These prevent GH from becoming F */
145 : #define NOGHTOF(c) (getcode(c) & 16) /* BDH */
146 :
147 4 : PG_FUNCTION_INFO_V1(levenshtein_with_costs);
148 : Datum
149 2 : levenshtein_with_costs(PG_FUNCTION_ARGS)
150 : {
151 2 : text *src = PG_GETARG_TEXT_PP(0);
152 2 : text *dst = PG_GETARG_TEXT_PP(1);
153 2 : int ins_c = PG_GETARG_INT32(2);
154 2 : int del_c = PG_GETARG_INT32(3);
155 2 : int sub_c = PG_GETARG_INT32(4);
156 : const char *s_data;
157 : const char *t_data;
158 : int s_bytes,
159 : t_bytes;
160 :
161 : /* Extract a pointer to the actual character data */
162 2 : s_data = VARDATA_ANY(src);
163 2 : t_data = VARDATA_ANY(dst);
164 : /* Determine length of each string in bytes */
165 2 : s_bytes = VARSIZE_ANY_EXHDR(src);
166 2 : t_bytes = VARSIZE_ANY_EXHDR(dst);
167 :
168 2 : PG_RETURN_INT32(varstr_levenshtein(s_data, s_bytes, t_data, t_bytes,
169 : ins_c, del_c, sub_c, false));
170 : }
171 :
172 :
173 4 : PG_FUNCTION_INFO_V1(levenshtein);
174 : Datum
175 2 : levenshtein(PG_FUNCTION_ARGS)
176 : {
177 2 : text *src = PG_GETARG_TEXT_PP(0);
178 2 : text *dst = PG_GETARG_TEXT_PP(1);
179 : const char *s_data;
180 : const char *t_data;
181 : int s_bytes,
182 : t_bytes;
183 :
184 : /* Extract a pointer to the actual character data */
185 2 : s_data = VARDATA_ANY(src);
186 2 : t_data = VARDATA_ANY(dst);
187 : /* Determine length of each string in bytes */
188 2 : s_bytes = VARSIZE_ANY_EXHDR(src);
189 2 : t_bytes = VARSIZE_ANY_EXHDR(dst);
190 :
191 2 : PG_RETURN_INT32(varstr_levenshtein(s_data, s_bytes, t_data, t_bytes,
192 : 1, 1, 1, false));
193 : }
194 :
195 :
196 2 : PG_FUNCTION_INFO_V1(levenshtein_less_equal_with_costs);
197 : Datum
198 0 : levenshtein_less_equal_with_costs(PG_FUNCTION_ARGS)
199 : {
200 0 : text *src = PG_GETARG_TEXT_PP(0);
201 0 : text *dst = PG_GETARG_TEXT_PP(1);
202 0 : int ins_c = PG_GETARG_INT32(2);
203 0 : int del_c = PG_GETARG_INT32(3);
204 0 : int sub_c = PG_GETARG_INT32(4);
205 0 : int max_d = PG_GETARG_INT32(5);
206 : const char *s_data;
207 : const char *t_data;
208 : int s_bytes,
209 : t_bytes;
210 :
211 : /* Extract a pointer to the actual character data */
212 0 : s_data = VARDATA_ANY(src);
213 0 : t_data = VARDATA_ANY(dst);
214 : /* Determine length of each string in bytes */
215 0 : s_bytes = VARSIZE_ANY_EXHDR(src);
216 0 : t_bytes = VARSIZE_ANY_EXHDR(dst);
217 :
218 0 : PG_RETURN_INT32(varstr_levenshtein_less_equal(s_data, s_bytes,
219 : t_data, t_bytes,
220 : ins_c, del_c, sub_c,
221 : max_d, false));
222 : }
223 :
224 :
225 4 : PG_FUNCTION_INFO_V1(levenshtein_less_equal);
226 : Datum
227 4 : levenshtein_less_equal(PG_FUNCTION_ARGS)
228 : {
229 4 : text *src = PG_GETARG_TEXT_PP(0);
230 4 : text *dst = PG_GETARG_TEXT_PP(1);
231 4 : int max_d = PG_GETARG_INT32(2);
232 : const char *s_data;
233 : const char *t_data;
234 : int s_bytes,
235 : t_bytes;
236 :
237 : /* Extract a pointer to the actual character data */
238 4 : s_data = VARDATA_ANY(src);
239 4 : t_data = VARDATA_ANY(dst);
240 : /* Determine length of each string in bytes */
241 4 : s_bytes = VARSIZE_ANY_EXHDR(src);
242 4 : t_bytes = VARSIZE_ANY_EXHDR(dst);
243 :
244 4 : PG_RETURN_INT32(varstr_levenshtein_less_equal(s_data, s_bytes,
245 : t_data, t_bytes,
246 : 1, 1, 1,
247 : max_d, false));
248 : }
249 :
250 :
251 : /*
252 : * Calculates the metaphone of an input string.
253 : * Returns number of characters requested
254 : * (suggested value is 4)
255 : */
256 4 : PG_FUNCTION_INFO_V1(metaphone);
257 : Datum
258 2 : metaphone(PG_FUNCTION_ARGS)
259 : {
260 2 : char *str_i = TextDatumGetCString(PG_GETARG_DATUM(0));
261 2 : size_t str_i_len = strlen(str_i);
262 : int reqlen;
263 : char *metaph;
264 :
265 : /* return an empty string if we receive one */
266 2 : if (!(str_i_len > 0))
267 0 : PG_RETURN_TEXT_P(cstring_to_text(""));
268 :
269 2 : if (str_i_len > MAX_METAPHONE_STRLEN)
270 0 : ereport(ERROR,
271 : (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
272 : errmsg("argument exceeds the maximum length of %d bytes",
273 : MAX_METAPHONE_STRLEN)));
274 :
275 2 : reqlen = PG_GETARG_INT32(1);
276 2 : if (reqlen > MAX_METAPHONE_STRLEN)
277 0 : ereport(ERROR,
278 : (errcode(ERRCODE_INVALID_PARAMETER_VALUE),
279 : errmsg("output exceeds the maximum length of %d bytes",
280 : MAX_METAPHONE_STRLEN)));
281 :
282 2 : if (!(reqlen > 0))
283 0 : ereport(ERROR,
284 : (errcode(ERRCODE_ZERO_LENGTH_CHARACTER_STRING),
285 : errmsg("output cannot be empty string")));
286 :
287 2 : _metaphone(str_i, reqlen, &metaph);
288 2 : PG_RETURN_TEXT_P(cstring_to_text(metaph));
289 : }
290 :
291 :
292 : /*
293 : * Original code by Michael G Schwern starts here.
294 : * Code slightly modified for use as PostgreSQL
295 : * function (palloc, etc).
296 : */
297 :
298 : /* I suppose I could have been using a character pointer instead of
299 : * accessing the array directly... */
300 :
301 : /* Look at the next letter in the word */
302 : #define Next_Letter (toupper((unsigned char) word[w_idx+1]))
303 : /* Look at the current letter in the word */
304 : #define Curr_Letter (toupper((unsigned char) word[w_idx]))
305 : /* Go N letters back. */
306 : #define Look_Back_Letter(n) \
307 : (w_idx >= (n) ? toupper((unsigned char) word[w_idx-(n)]) : '\0')
308 : /* Previous letter. I dunno, should this return null on failure? */
309 : #define Prev_Letter (Look_Back_Letter(1))
310 : /* Look two letters down. It makes sure you don't walk off the string. */
311 : #define After_Next_Letter \
312 : (Next_Letter != '\0' ? toupper((unsigned char) word[w_idx+2]) : '\0')
313 : #define Look_Ahead_Letter(n) toupper((unsigned char) Lookahead(word+w_idx, n))
314 :
315 :
316 : /* Allows us to safely look ahead an arbitrary # of letters */
317 : /* I probably could have just used strlen... */
318 : static char
319 0 : Lookahead(char *word, int how_far)
320 : {
321 0 : char letter_ahead = '\0'; /* null by default */
322 : int idx;
323 :
324 0 : for (idx = 0; word[idx] != '\0' && idx < how_far; idx++);
325 : /* Edge forward in the string... */
326 :
327 0 : letter_ahead = word[idx]; /* idx will be either == to how_far or at the
328 : * end of the string */
329 0 : return letter_ahead;
330 : }
331 :
332 :
333 : /* phonize one letter */
334 : #define Phonize(c) do {(*phoned_word)[p_idx++] = c;} while (0)
335 : /* Slap a null character on the end of the phoned word */
336 : #define End_Phoned_Word do {(*phoned_word)[p_idx] = '\0';} while (0)
337 : /* How long is the phoned word? */
338 : #define Phone_Len (p_idx)
339 :
340 : /* Note is a letter is a 'break' in the word */
341 : #define Isbreak(c) (!isalpha((unsigned char) (c)))
342 :
343 :
344 : static void
345 2 : _metaphone(char *word, /* IN */
346 : int max_phonemes,
347 : char **phoned_word) /* OUT */
348 : {
349 2 : int w_idx = 0; /* point in the phonization we're at. */
350 2 : int p_idx = 0; /* end of the phoned phrase */
351 :
352 : /*-- Parameter checks --*/
353 :
354 : /*
355 : * Shouldn't be necessary, but left these here anyway jec Aug 3, 2001
356 : */
357 :
358 : /* Negative phoneme length is meaningless */
359 2 : if (!(max_phonemes > 0))
360 : /* internal error */
361 0 : elog(ERROR, "metaphone: Requested output length must be > 0");
362 :
363 : /* Empty/null string is meaningless */
364 2 : if ((word == NULL) || !(strlen(word) > 0))
365 : /* internal error */
366 0 : elog(ERROR, "metaphone: Input string length must be > 0");
367 :
368 : /*-- Allocate memory for our phoned_phrase --*/
369 2 : if (max_phonemes == 0)
370 : { /* Assume largest possible */
371 0 : *phoned_word = palloc(sizeof(char) * strlen(word) + 1);
372 : }
373 : else
374 : {
375 2 : *phoned_word = palloc(sizeof(char) * max_phonemes + 1);
376 : }
377 :
378 : /*-- The first phoneme has to be processed specially. --*/
379 : /* Find our first letter */
380 2 : for (; !isalpha((unsigned char) (Curr_Letter)); w_idx++)
381 : {
382 : /* On the off chance we were given nothing but crap... */
383 0 : if (Curr_Letter == '\0')
384 : {
385 0 : End_Phoned_Word;
386 0 : return;
387 : }
388 : }
389 :
390 2 : switch (Curr_Letter)
391 : {
392 : /* AE becomes E */
393 0 : case 'A':
394 0 : if (Next_Letter == 'E')
395 : {
396 0 : Phonize('E');
397 0 : w_idx += 2;
398 : }
399 : /* Remember, preserve vowels at the beginning */
400 : else
401 : {
402 0 : Phonize('A');
403 0 : w_idx++;
404 : }
405 0 : break;
406 : /* [GKP]N becomes N */
407 2 : case 'G':
408 : case 'K':
409 : case 'P':
410 2 : if (Next_Letter == 'N')
411 : {
412 0 : Phonize('N');
413 0 : w_idx += 2;
414 : }
415 2 : break;
416 :
417 : /*
418 : * WH becomes H, WR becomes R W if followed by a vowel
419 : */
420 0 : case 'W':
421 0 : if (Next_Letter == 'H' ||
422 0 : Next_Letter == 'R')
423 : {
424 0 : Phonize(Next_Letter);
425 0 : w_idx += 2;
426 : }
427 0 : else if (isvowel(Next_Letter))
428 : {
429 0 : Phonize('W');
430 0 : w_idx += 2;
431 : }
432 : /* else ignore */
433 0 : break;
434 : /* X becomes S */
435 0 : case 'X':
436 0 : Phonize('S');
437 0 : w_idx++;
438 0 : break;
439 : /* Vowels are kept */
440 :
441 : /*
442 : * We did A already case 'A': case 'a':
443 : */
444 0 : case 'E':
445 : case 'I':
446 : case 'O':
447 : case 'U':
448 0 : Phonize(Curr_Letter);
449 0 : w_idx++;
450 0 : break;
451 0 : default:
452 : /* do nothing */
453 0 : break;
454 : }
455 :
456 :
457 :
458 : /* On to the metaphoning */
459 12 : for (; Curr_Letter != '\0' &&
460 10 : (max_phonemes == 0 || Phone_Len < max_phonemes);
461 10 : w_idx++)
462 : {
463 : /*
464 : * How many letters to skip because an earlier encoding handled
465 : * multiple letters
466 : */
467 10 : unsigned short int skip_letter = 0;
468 :
469 :
470 : /*
471 : * THOUGHT: It would be nice if, rather than having things like...
472 : * well, SCI. For SCI you encode the S, then have to remember to skip
473 : * the C. So the phonome SCI invades both S and C. It would be
474 : * better, IMHO, to skip the C from the S part of the encoding. Hell,
475 : * I'm trying it.
476 : */
477 :
478 : /* Ignore non-alphas */
479 10 : if (!isalpha((unsigned char) (Curr_Letter)))
480 0 : continue;
481 :
482 : /* Drop duplicates, except CC */
483 10 : if (Curr_Letter == Prev_Letter &&
484 0 : Curr_Letter != 'C')
485 0 : continue;
486 :
487 10 : switch (Curr_Letter)
488 : {
489 : /* B -> B unless in MB */
490 2 : case 'B':
491 2 : if (Prev_Letter != 'M')
492 0 : Phonize('B');
493 2 : break;
494 :
495 : /*
496 : * 'sh' if -CIA- or -CH, but not SCH, except SCHW. (SCHW is
497 : * handled in S) S if -CI-, -CE- or -CY- dropped if -SCI-,
498 : * SCE-, -SCY- (handed in S) else K
499 : */
500 0 : case 'C':
501 0 : if (MAKESOFT(Next_Letter))
502 : { /* C[IEY] */
503 0 : if (After_Next_Letter == 'A' &&
504 0 : Next_Letter == 'I')
505 : { /* CIA */
506 0 : Phonize(SH);
507 : }
508 : /* SC[IEY] */
509 0 : else if (Prev_Letter == 'S')
510 : {
511 : /* Dropped */
512 : }
513 : else
514 0 : Phonize('S');
515 : }
516 0 : else if (Next_Letter == 'H')
517 : {
518 : #ifndef USE_TRADITIONAL_METAPHONE
519 0 : if (After_Next_Letter == 'R' ||
520 0 : Prev_Letter == 'S')
521 : { /* Christ, School */
522 0 : Phonize('K');
523 : }
524 : else
525 0 : Phonize(SH);
526 : #else
527 : Phonize(SH);
528 : #endif
529 0 : skip_letter++;
530 : }
531 : else
532 0 : Phonize('K');
533 0 : break;
534 :
535 : /*
536 : * J if in -DGE-, -DGI- or -DGY- else T
537 : */
538 0 : case 'D':
539 0 : if (Next_Letter == 'G' &&
540 0 : MAKESOFT(After_Next_Letter))
541 : {
542 0 : Phonize('J');
543 0 : skip_letter++;
544 : }
545 : else
546 0 : Phonize('T');
547 0 : break;
548 :
549 : /*
550 : * F if in -GH and not B--GH, D--GH, -H--GH, -H---GH else
551 : * dropped if -GNED, -GN, else dropped if -DGE-, -DGI- or
552 : * -DGY- (handled in D) else J if in -GE-, -GI, -GY and not GG
553 : * else K
554 : */
555 2 : case 'G':
556 2 : if (Next_Letter == 'H')
557 : {
558 0 : if (!(NOGHTOF(Look_Back_Letter(3)) ||
559 0 : Look_Back_Letter(4) == 'H'))
560 : {
561 0 : Phonize('F');
562 0 : skip_letter++;
563 : }
564 : else
565 : {
566 : /* silent */
567 : }
568 : }
569 2 : else if (Next_Letter == 'N')
570 : {
571 0 : if (Isbreak(After_Next_Letter) ||
572 0 : (After_Next_Letter == 'E' &&
573 0 : Look_Ahead_Letter(3) == 'D'))
574 : {
575 : /* dropped */
576 : }
577 : else
578 0 : Phonize('K');
579 : }
580 2 : else if (MAKESOFT(Next_Letter) &&
581 0 : Prev_Letter != 'G')
582 0 : Phonize('J');
583 : else
584 2 : Phonize('K');
585 2 : break;
586 : /* H if before a vowel and not after C,G,P,S,T */
587 0 : case 'H':
588 0 : if (isvowel(Next_Letter) &&
589 0 : !AFFECTH(Prev_Letter))
590 0 : Phonize('H');
591 0 : break;
592 :
593 : /*
594 : * dropped if after C else K
595 : */
596 0 : case 'K':
597 0 : if (Prev_Letter != 'C')
598 0 : Phonize('K');
599 0 : break;
600 :
601 : /*
602 : * F if before H else P
603 : */
604 0 : case 'P':
605 0 : if (Next_Letter == 'H')
606 0 : Phonize('F');
607 : else
608 0 : Phonize('P');
609 0 : break;
610 :
611 : /*
612 : * K
613 : */
614 0 : case 'Q':
615 0 : Phonize('K');
616 0 : break;
617 :
618 : /*
619 : * 'sh' in -SH-, -SIO- or -SIA- or -SCHW- else S
620 : */
621 0 : case 'S':
622 0 : if (Next_Letter == 'I' &&
623 0 : (After_Next_Letter == 'O' ||
624 0 : After_Next_Letter == 'A'))
625 0 : Phonize(SH);
626 0 : else if (Next_Letter == 'H')
627 : {
628 0 : Phonize(SH);
629 0 : skip_letter++;
630 : }
631 : #ifndef USE_TRADITIONAL_METAPHONE
632 0 : else if (Next_Letter == 'C' &&
633 0 : Look_Ahead_Letter(2) == 'H' &&
634 0 : Look_Ahead_Letter(3) == 'W')
635 : {
636 0 : Phonize(SH);
637 0 : skip_letter += 2;
638 : }
639 : #endif
640 : else
641 0 : Phonize('S');
642 0 : break;
643 :
644 : /*
645 : * 'sh' in -TIA- or -TIO- else 'th' before H else T
646 : */
647 0 : case 'T':
648 0 : if (Next_Letter == 'I' &&
649 0 : (After_Next_Letter == 'O' ||
650 0 : After_Next_Letter == 'A'))
651 0 : Phonize(SH);
652 0 : else if (Next_Letter == 'H')
653 : {
654 0 : Phonize(TH);
655 0 : skip_letter++;
656 : }
657 : else
658 0 : Phonize('T');
659 0 : break;
660 : /* F */
661 0 : case 'V':
662 0 : Phonize('F');
663 0 : break;
664 : /* W before a vowel, else dropped */
665 0 : case 'W':
666 0 : if (isvowel(Next_Letter))
667 0 : Phonize('W');
668 0 : break;
669 : /* KS */
670 0 : case 'X':
671 0 : Phonize('K');
672 0 : if (max_phonemes == 0 || Phone_Len < max_phonemes)
673 0 : Phonize('S');
674 0 : break;
675 : /* Y if followed by a vowel */
676 0 : case 'Y':
677 0 : if (isvowel(Next_Letter))
678 0 : Phonize('Y');
679 0 : break;
680 : /* S */
681 0 : case 'Z':
682 0 : Phonize('S');
683 0 : break;
684 : /* No transformation */
685 2 : case 'F':
686 : case 'J':
687 : case 'L':
688 : case 'M':
689 : case 'N':
690 : case 'R':
691 2 : Phonize(Curr_Letter);
692 2 : break;
693 4 : default:
694 : /* nothing */
695 4 : break;
696 : } /* END SWITCH */
697 :
698 10 : w_idx += skip_letter;
699 : } /* END FOR */
700 :
701 2 : End_Phoned_Word;
702 : } /* END metaphone */
703 :
704 :
705 : /*
706 : * SQL function: soundex(text) returns text
707 : */
708 6 : PG_FUNCTION_INFO_V1(soundex);
709 :
710 : Datum
711 14 : soundex(PG_FUNCTION_ARGS)
712 : {
713 : char outstr[SOUNDEX_LEN + 1];
714 : char *arg;
715 :
716 14 : arg = text_to_cstring(PG_GETARG_TEXT_PP(0));
717 :
718 14 : _soundex(arg, outstr);
719 :
720 14 : PG_RETURN_TEXT_P(cstring_to_text(outstr));
721 : }
722 :
723 : static void
724 26 : _soundex(const char *instr, char *outstr)
725 : {
726 : int count;
727 :
728 : Assert(instr);
729 : Assert(outstr);
730 :
731 26 : outstr[SOUNDEX_LEN] = '\0';
732 :
733 : /* Skip leading non-alphabetic characters */
734 26 : while (!isalpha((unsigned char) instr[0]) && instr[0])
735 0 : ++instr;
736 :
737 : /* No string left */
738 26 : if (!instr[0])
739 : {
740 0 : outstr[0] = (char) 0;
741 0 : return;
742 : }
743 :
744 : /* Take the first letter as is */
745 26 : *outstr++ = (char) toupper((unsigned char) *instr++);
746 :
747 26 : count = 1;
748 120 : while (*instr && count < SOUNDEX_LEN)
749 : {
750 186 : if (isalpha((unsigned char) *instr) &&
751 92 : soundex_code(*instr) != soundex_code(*(instr - 1)))
752 : {
753 70 : *outstr = soundex_code(instr[0]);
754 70 : if (*outstr != '0')
755 : {
756 46 : ++outstr;
757 46 : ++count;
758 : }
759 : }
760 94 : ++instr;
761 : }
762 :
763 : /* Fill with 0's */
764 58 : while (count < SOUNDEX_LEN)
765 : {
766 32 : *outstr = '0';
767 32 : ++outstr;
768 32 : ++count;
769 : }
770 : }
771 :
772 4 : PG_FUNCTION_INFO_V1(difference);
773 :
774 : Datum
775 6 : difference(PG_FUNCTION_ARGS)
776 : {
777 : char sndx1[SOUNDEX_LEN + 1],
778 : sndx2[SOUNDEX_LEN + 1];
779 : int i,
780 : result;
781 :
782 6 : _soundex(text_to_cstring(PG_GETARG_TEXT_PP(0)), sndx1);
783 6 : _soundex(text_to_cstring(PG_GETARG_TEXT_PP(1)), sndx2);
784 :
785 6 : result = 0;
786 30 : for (i = 0; i < SOUNDEX_LEN; i++)
787 : {
788 24 : if (sndx1[i] == sndx2[i])
789 12 : result++;
790 : }
791 :
792 6 : PG_RETURN_INT32(result);
793 : }
|
