* fuzzystrmatch.c
*
* Functions for "fuzzy" comparison of strings
*
* Joe Conway <mail@joeconway.com>
*
* contrib/fuzzystrmatch/fuzzystrmatch.c
* Copyright (c) 2001-2012, PostgreSQL Global Development Group
* ALL RIGHTS RESERVED;
*
* metaphone()
* -----------
* Modified for PostgreSQL by Joe Conway.
* Based on CPAN's "Text-Metaphone-1.96" by Michael G Schwern <schwern@pobox.com>
* Code slightly modified for use as openGauss function (palloc, elog, etc).
* Metaphone was originally created by Lawrence Philips and presented in article
* in "Computer Language" December 1990 issue.
*
* Permission to use, copy, modify, and distribute this software and its
* documentation for any purpose, without fee, and without a written agreement
* is hereby granted, provided that the above copyright notice and this
* paragraph and the following two paragraphs appear in all copies.
*
* IN NO EVENT SHALL THE AUTHORS OR DISTRIBUTORS BE LIABLE TO ANY PARTY FOR
* DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING
* LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS
* DOCUMENTATION, EVEN IF THE AUTHOR OR DISTRIBUTORS HAVE BEEN ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*
* THE AUTHORS AND DISTRIBUTORS SPECIFICALLY DISCLAIM ANY WARRANTIES,
* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
* AND FITNESS FOR A PARTICULAR PURPOSE. THE SOFTWARE PROVIDED HEREUNDER IS
* ON AN "AS IS" BASIS, AND THE AUTHOR AND DISTRIBUTORS HAS NO OBLIGATIONS TO
* PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
*
*/
#include "postgres.h"
#include "knl/knl_variable.h"
#include <ctype.h>
#include "mb/pg_wchar.h"
#include "utils/builtins.h"
PG_MODULE_MAGIC;
* External declarations for exported functions
*/
extern Datum levenshtein_with_costs(PG_FUNCTION_ARGS);
extern Datum levenshtein(PG_FUNCTION_ARGS);
extern Datum levenshtein_less_equal_with_costs(PG_FUNCTION_ARGS);
extern Datum levenshtein_less_equal(PG_FUNCTION_ARGS);
extern Datum metaphone(PG_FUNCTION_ARGS);
extern Datum soundex(PG_FUNCTION_ARGS);
extern Datum difference(PG_FUNCTION_ARGS);
* Soundex
*/
static void _soundex(const char* instr, char* outstr);
#define SOUNDEX_LEN 4
static const char* soundex_table = "01230120022455012623010202";
static char soundex_code(char letter)
{
letter = toupper((unsigned char)letter);
if (letter >= 'A' && letter <= 'Z')
return soundex_table[letter - 'A'];
return letter;
}
* Metaphone
*/
#define MAX_METAPHONE_STRLEN 255
* Original code by Michael G Schwern starts here.
* Code slightly modified for use as openGauss function.
*/
metaphone -- Breaks english phrases down into their phonemes.
Input
word -- An english word to be phonized
max_phonemes -- How many phonemes to calculate. If 0, then it
will phonize the entire phrase.
phoned_word -- The final phonized word. (We'll allocate the
memory.)
Output
error -- A simple error flag, returns TRUE or FALSE
NOTES: ALL non-alpha characters are ignored, this includes whitespace,
although non-alpha characters will break up phonemes.
****************************************************************************/
my constants -- constants I like
Probably redundant.
***************************************************************************/
#define META_ERROR FALSE
#define META_SUCCESS TRUE
#define META_FAILURE FALSE
might find in books. Define this if you want metaphone to behave
traditionally */
#undef USE_TRADITIONAL_METAPHONE
#define SH 'X'
#define TH '0'
static char Lookahead(char* word, int how_far);
static int _metaphone(char* word, int max_phonemes, char** phoned_word);
static const char _codes[26] = {
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
};
static int getcode(char c)
{
if (isalpha((unsigned char)c)) {
c = toupper((unsigned char)c);
if (c >= 'A' && c <= 'Z')
return _codes[c - 'A'];
}
return 0;
}
#define isvowel(c) (getcode(c) & 1)
#define NOCHANGE(c) (getcode(c) & 2)
#define AFFECTH(c) (getcode(c) & 4)
#define MAKESOFT(c) (getcode(c) & 8)
#define NOGHTOF(c) (getcode(c) & 16)
static inline bool rest_of_char_same(const char* s1, const char* s2, int len)
{
while (len > 0) {
len--;
if (s1[len] != s2[len])
return false;
}
return true;
}
#include "levenshtein.cpp"
#define LEVENSHTEIN_LESS_EQUAL
#include "levenshtein.cpp"
PG_FUNCTION_INFO_V1(levenshtein_with_costs);
Datum levenshtein_with_costs(PG_FUNCTION_ARGS)
{
text* src = PG_GETARG_TEXT_PP(0);
text* dst = PG_GETARG_TEXT_PP(1);
int ins_c = PG_GETARG_INT32(2);
int del_c = PG_GETARG_INT32(3);
int sub_c = PG_GETARG_INT32(4);
PG_RETURN_INT32(levenshtein_internal(src, dst, ins_c, del_c, sub_c));
}
PG_FUNCTION_INFO_V1(levenshtein);
Datum levenshtein(PG_FUNCTION_ARGS)
{
text* src = PG_GETARG_TEXT_PP(0);
text* dst = PG_GETARG_TEXT_PP(1);
PG_RETURN_INT32(levenshtein_internal(src, dst, 1, 1, 1));
}
PG_FUNCTION_INFO_V1(levenshtein_less_equal_with_costs);
Datum levenshtein_less_equal_with_costs(PG_FUNCTION_ARGS)
{
text* src = PG_GETARG_TEXT_PP(0);
text* dst = PG_GETARG_TEXT_PP(1);
int ins_c = PG_GETARG_INT32(2);
int del_c = PG_GETARG_INT32(3);
int sub_c = PG_GETARG_INT32(4);
int max_d = PG_GETARG_INT32(5);
PG_RETURN_INT32(levenshtein_less_equal_internal(src, dst, ins_c, del_c, sub_c, max_d));
}
PG_FUNCTION_INFO_V1(levenshtein_less_equal);
Datum levenshtein_less_equal(PG_FUNCTION_ARGS)
{
text* src = PG_GETARG_TEXT_PP(0);
text* dst = PG_GETARG_TEXT_PP(1);
int max_d = PG_GETARG_INT32(2);
PG_RETURN_INT32(levenshtein_less_equal_internal(src, dst, 1, 1, 1, max_d));
}
* Calculates the metaphone of an input string.
* Returns number of characters requested
* (suggested value is 4)
*/
PG_FUNCTION_INFO_V1(metaphone);
Datum metaphone(PG_FUNCTION_ARGS)
{
char* str_i = TextDatumGetCString(PG_GETARG_DATUM(0));
size_t str_i_len = strlen(str_i);
int reqlen;
char* metaph = NULL;
int retval;
if (!(str_i_len > 0))
PG_RETURN_TEXT_P(cstring_to_text(""));
if (str_i_len > MAX_METAPHONE_STRLEN)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("argument exceeds the maximum length of %d bytes", MAX_METAPHONE_STRLEN)));
if (!(str_i_len > 0))
ereport(ERROR, (errcode(ERRCODE_ZERO_LENGTH_CHARACTER_STRING), errmsg("argument is empty string")));
reqlen = PG_GETARG_INT32(1);
if (reqlen > MAX_METAPHONE_STRLEN)
ereport(ERROR,
(errcode(ERRCODE_INVALID_PARAMETER_VALUE),
errmsg("output exceeds the maximum length of %d bytes", MAX_METAPHONE_STRLEN)));
if (!(reqlen > 0))
ereport(ERROR, (errcode(ERRCODE_ZERO_LENGTH_CHARACTER_STRING), errmsg("output cannot be empty string")));
retval = _metaphone(str_i, reqlen, &metaph);
if (retval == META_SUCCESS)
PG_RETURN_TEXT_P(cstring_to_text(metaph));
else {
elog(ERROR, "metaphone: failure");
PG_RETURN_NULL();
}
}
* Original code by Michael G Schwern starts here.
* Code slightly modified for use as openGauss
* function (palloc, etc).
*/
* accessing the array directly... */
#define Next_Letter (toupper((unsigned char)word[w_idx + 1]))
#define Curr_Letter (toupper((unsigned char)word[w_idx]))
#define Look_Back_Letter(n) (w_idx >= (n) ? toupper((unsigned char)word[w_idx - (n)]) : '\0')
#define Prev_Letter (Look_Back_Letter(1))
#define After_Next_Letter (Next_Letter != '\0' ? toupper((unsigned char)word[w_idx + 2]) : '\0')
#define Look_Ahead_Letter(n) toupper((unsigned char)Lookahead(word + w_idx, n))
static char Lookahead(char* word, int how_far)
{
char letter_ahead = '\0';
int idx;
for (idx = 0; idx < how_far && word[idx] != '\0'; idx++)
;
letter_ahead = word[idx];
* end of the string */
return letter_ahead;
}
#define Phonize(c) \
do { \
(*phoned_word)[p_idx++] = c; \
} while (0)
#define End_Phoned_Word \
do { \
(*phoned_word)[p_idx] = '\0'; \
} while (0)
#define Phone_Len (p_idx)
#define Isbreak(c) (!isalpha((unsigned char)(c)))
static int _metaphone(char* word,
int max_phonemes, char** phoned_word)
{
int w_idx = 0;
int p_idx = 0;
* Shouldn't be necessary, but left these here anyway jec Aug 3, 2001
*/
if (!(max_phonemes > 0))
elog(ERROR, "metaphone: Requested output length must be > 0");
if ((word == NULL) || !(strlen(word) > 0))
elog(ERROR, "metaphone: Input string length must be > 0");
if (max_phonemes == 0) {
*phoned_word = (char*)palloc(sizeof(char) * strlen(word) + 1);
} else {
*phoned_word = (char*)palloc(sizeof(char) * max_phonemes + 1);
}
for (; !isalpha((unsigned char)(Curr_Letter)); w_idx++) {
if (Curr_Letter == '\0') {
End_Phoned_Word;
return META_SUCCESS;
}
}
switch (Curr_Letter) {
case 'A':
if (Next_Letter == 'E') {
Phonize('E');
w_idx += 2;
}
else {
Phonize('A');
w_idx++;
}
break;
case 'G':
case 'K':
case 'P':
if (Next_Letter == 'N') {
Phonize('N');
w_idx += 2;
}
break;
* WH becomes H, WR becomes R W if followed by a vowel
*/
case 'W':
if (Next_Letter == 'H' || Next_Letter == 'R') {
Phonize(Next_Letter);
w_idx += 2;
} else if (isvowel(Next_Letter)) {
Phonize('W');
w_idx += 2;
}
break;
case 'X':
Phonize('S');
w_idx++;
break;
* We did A already case 'A': case 'a':
*/
case 'E':
case 'I':
case 'O':
case 'U':
Phonize(Curr_Letter);
w_idx++;
break;
default:
break;
}
for (; Curr_Letter != '\0' && (max_phonemes == 0 || Phone_Len < max_phonemes); w_idx++) {
* How many letters to skip because an earlier encoding handled
* multiple letters
*/
unsigned short int skip_letter = 0;
* THOUGHT: It would be nice if, rather than having things like...
* well, SCI. For SCI you encode the S, then have to remember to skip
* the C. So the phonome SCI invades both S and C. It would be
* better, IMHO, to skip the C from the S part of the encoding. Hell,
* I'm trying it.
*/
if (!isalpha((unsigned char)(Curr_Letter)))
continue;
if (Curr_Letter == Prev_Letter && Curr_Letter != 'C')
continue;
switch (Curr_Letter) {
case 'B':
if (Prev_Letter != 'M')
Phonize('B');
break;
* 'sh' if -CIA- or -CH, but not SCH, except SCHW. (SCHW is
* handled in S) S if -CI-, -CE- or -CY- dropped if -SCI-,
* SCE-, -SCY- (handed in S) else K
*/
case 'C':
if (MAKESOFT(Next_Letter)) {
if (After_Next_Letter == 'A' && Next_Letter == 'I') {
Phonize(SH);
}
else if (Prev_Letter == 'S') {
} else
Phonize('S');
} else if (Next_Letter == 'H') {
#ifndef USE_TRADITIONAL_METAPHONE
if (After_Next_Letter == 'R' || Prev_Letter == 'S') {
Phonize('K');
} else
Phonize(SH);
#else
Phonize(SH);
#endif
skip_letter++;
} else
Phonize('K');
break;
* J if in -DGE-, -DGI- or -DGY- else T
*/
case 'D':
if (Next_Letter == 'G' && MAKESOFT(After_Next_Letter)) {
Phonize('J');
skip_letter++;
} else
Phonize('T');
break;
* F if in -GH and not B--GH, D--GH, -H--GH, -H---GH else
* dropped if -GNED, -GN, else dropped if -DGE-, -DGI- or
* -DGY- (handled in D) else J if in -GE-, -GI, -GY and not GG
* else K
*/
case 'G':
if (Next_Letter == 'H') {
if (!(NOGHTOF(Look_Back_Letter(3)) || Look_Back_Letter(4) == 'H')) {
Phonize('F');
skip_letter++;
} else {
}
} else if (Next_Letter == 'N') {
if (Isbreak(After_Next_Letter) || (After_Next_Letter == 'E' && Look_Ahead_Letter(3) == 'D')) {
} else
Phonize('K');
} else if (MAKESOFT(Next_Letter) && Prev_Letter != 'G')
Phonize('J');
else
Phonize('K');
break;
case 'H':
if (isvowel(Next_Letter) && !AFFECTH(Prev_Letter))
Phonize('H');
break;
* dropped if after C else K
*/
case 'K':
if (Prev_Letter != 'C')
Phonize('K');
break;
* F if before H else P
*/
case 'P':
if (Next_Letter == 'H')
Phonize('F');
else
Phonize('P');
break;
* K
*/
case 'Q':
Phonize('K');
break;
* 'sh' in -SH-, -SIO- or -SIA- or -SCHW- else S
*/
case 'S':
if (Next_Letter == 'I' && (After_Next_Letter == 'O' || After_Next_Letter == 'A'))
Phonize(SH);
else if (Next_Letter == 'H') {
Phonize(SH);
skip_letter++;
}
#ifndef USE_TRADITIONAL_METAPHONE
else if (Next_Letter == 'C' && Look_Ahead_Letter(2) == 'H' && Look_Ahead_Letter(3) == 'W') {
Phonize(SH);
skip_letter += 2;
}
#endif
else
Phonize('S');
break;
* 'sh' in -TIA- or -TIO- else 'th' before H else T
*/
case 'T':
if (Next_Letter == 'I' && (After_Next_Letter == 'O' || After_Next_Letter == 'A'))
Phonize(SH);
else if (Next_Letter == 'H') {
Phonize(TH);
skip_letter++;
} else
Phonize('T');
break;
case 'V':
Phonize('F');
break;
case 'W':
if (isvowel(Next_Letter))
Phonize('W');
break;
case 'X':
Phonize('K');
if (max_phonemes == 0 || Phone_Len < max_phonemes)
Phonize('S');
break;
case 'Y':
if (isvowel(Next_Letter))
Phonize('Y');
break;
case 'Z':
Phonize('S');
break;
case 'F':
case 'J':
case 'L':
case 'M':
case 'N':
case 'R':
Phonize(Curr_Letter);
break;
default:
break;
}
w_idx += skip_letter;
}
End_Phoned_Word;
return (META_SUCCESS);
}
* SQL function: soundex(text) returns text
*/
PG_FUNCTION_INFO_V1(soundex);
Datum soundex(PG_FUNCTION_ARGS)
{
char outstr[SOUNDEX_LEN + 1];
char* arg = NULL;
arg = text_to_cstring(PG_GETARG_TEXT_P(0));
_soundex(arg, outstr);
PG_RETURN_TEXT_P(cstring_to_text(outstr));
}
static void _soundex(const char* instr, char* outstr)
{
int count;
AssertArg(instr);
AssertArg(outstr);
outstr[SOUNDEX_LEN] = '\0';
while (!isalpha((unsigned char)instr[0]) && instr[0])
++instr;
if (!instr[0]) {
outstr[0] = (char)0;
return;
}
*outstr++ = (char)toupper((unsigned char)*instr++);
count = 1;
while (*instr && count < SOUNDEX_LEN) {
if (isalpha((unsigned char)*instr) && soundex_code(*instr) != soundex_code(*(instr - 1))) {
*outstr = soundex_code(instr[0]);
if (*outstr != '0') {
++outstr;
++count;
}
}
++instr;
}
while (count < SOUNDEX_LEN) {
*outstr = '0';
++outstr;
++count;
}
}
PG_FUNCTION_INFO_V1(difference);
Datum difference(PG_FUNCTION_ARGS)
{
char sndx1[SOUNDEX_LEN + 1], sndx2[SOUNDEX_LEN + 1];
int i, result;
_soundex(text_to_cstring(PG_GETARG_TEXT_P(0)), sndx1);
_soundex(text_to_cstring(PG_GETARG_TEXT_P(1)), sndx2);
result = 0;
for (i = 0; i < SOUNDEX_LEN; i++) {
if (sndx1[i] == sndx2[i])
result++;
}
PG_RETURN_INT32(result);
}