static FUZZY_WORD *Stem_snowball( FUZZY_OBJECT *fi, const char *inword)
{
char *out_word;
struct SN_env *snowball = fi->snowball_options;
FUZZY_WORD *fw = create_fuzzy_word( inword, 1 ); /* create place to store stemmed word */
SN_set_current(snowball,strlen(inword),(const symbol *)inword); /* Set Word to Stem */
fi->stemmer->lang_stem(snowball); /* Stem the word */
if ( 0 == snowball->l )
{
fw->error = STEM_TO_NOTHING;
return fw;
}
fw->free_strings = 1; /* flag that malloc is used */
out_word = emalloc(snowball->l + 1);
memcpy(out_word, snowball->p, snowball->l);
out_word[snowball->l] = '\0';
fw->string_list[0] = out_word;
return fw;
}
static FUZZY_WORD *double_metaphone( FUZZY_OBJECT *fi, const char *inword)
{
FUZZY_WORD *fw = create_fuzzy_word( inword, 2 ); /* create place to store stemmed word */
char *codes[2];
DoubleMetaphone( inword, codes );
if ( !(*codes[0]) ) /* was there at least one conversion? */
{
efree( codes[0] );
efree( codes[1] );
return fw;
}
fw->free_strings = 1;
fw->string_list[0] = codes[0];
/* Is double metaphone enabled? */
if ( FUZZY_DOUBLE_METAPHONE != fi->stemmer->fuzzy_mode )
return fw;
/* Is there a second metaphone that is different from the first? */
if ( *codes[1] && strcmp(codes[0], codes[1]) )
{
fw->list_size++;
fw->string_list[1] = codes[1];
}
else
{
efree( codes[1] );
}
return fw;
}
static FUZZY_WORD *Stem( FUZZY_OBJECT *fi, const char *inword)
{
char *end; /* pointer to the end of the word */
char word[MAXWORDLEN+1];
int length;
int rule_result; /* which rule is fired in replacing an end */
int i;
FUZZY_WORD *fw = create_fuzzy_word( inword, 1 );
/* Make sure the word is not too large from the start. */
if ( strlen( inword ) >= MAXWORDLEN )
{
fw->error = STEM_WORD_TOO_BIG;
return fw;
}
/* make working copy */
strcpy( word, inword );
/* Part 1: Check to ensure the word is all alphabetic */
/* no longer converts to lower case -- word should be lower before calling */
for ( end = word; *end; end++ )
if ( !isalpha( (unsigned int) *end ) )
{
fw->error = STEM_NOT_ALPHA;
return fw;
}
/* Part 2: Run through the Porter algorithm */
for (i = 0; i < (int)(sizeof(all_steps)/sizeof(all_steps[0])); i++)
{
rule_result = ReplaceEnd(word, all_steps[i]);
if ((rule_result == 106) || (rule_result == 107))
rule_result = ReplaceEnd(word, step1b1_rules);
if ( rule_result == STEM_WORD_TOO_BIG )
{
fw->error = rule_result;
return fw;
}
}
length = strlen( word );
/* Stem must be two chars or more in length */
if ( length <= 1 )
{
fw->error = STEM_TO_NOTHING;
return fw;
}
if ( length >= MAXWORDLEN )
{
fw->error = STEM_WORD_TOO_BIG;
return fw;
}
fw->free_strings = 1;
fw->string_list[0] = estrdup( word );
return fw;
}
static FUZZY_WORD *no_stem( FUZZY_OBJECT *fi, const char *inword)
{
return create_fuzzy_word( inword, 1 );
}
FUZZY_WORD *soundex( FUZZY_OBJECT *fi, const char *inword)
{
FUZZY_WORD *fw = create_fuzzy_word( inword, 1 ); /* create place to store stemmed word */
char word[MAXWORDLEN+1];
/* Misc Stuff */
char u, l ;
int i, j, n;
/* Resultant Sound Code */
char soundCode[5] = "0000\0";
/* Group Number Lookup Table */
static char soundTable[26] =
{0, /* A */
'1', /* B */
'2', /* C */
'3', /* D */
0, /* E */
'1', /* F */
'2', /* G */
0, /* H */
0, /* I */
'2', /* J */
'2', /* K */
'4', /* L */
'5', /* M */
'5', /* N */
0, /* O */
'1', /* P */
'2', /* Q */
'6', /* R */
'2', /* S */
'3', /* T */
0, /* U */
'1', /* V */
0, /* W */
'2', /* X */
0, /* Y */
'2'}; /* Z */
/* Make sure the word is not too large from the start. */
if ( strlen( inword ) >= MAXWORDLEN )
{
fw->error = STEM_WORD_TOO_BIG;
return fw;
}
/* make working copy */
strcpy( word, inword );
#ifdef _DEBUG
/* Debug to console */
printf("# %15s: %s ", "soundex.c", word);
#endif
/* Make sure it actually starts with a letter */
if(!isalpha((int)((unsigned char)word[0])))
{
fw->error = STEM_NOT_ALPHA;
return fw;
}
#ifdef _DEBUG
/* Debug to console */
printf("isalpha, ");
#endif
/* Get string length and make sure its at least 3 characters */
if((n = (int)strlen(word)) < 3)
{
fw->error = STEM_TOO_SMALL;
return fw;
}
#ifdef _DEBUG
/* Debug to console */
printf("=>3, ");
#endif
/* If looks like a 4 digit soundex code we don't want to touch it. */
/* Humm. Just because it looks like a duck, doesn't mean it is one
* The source is suppose to not be soundex, so this doesn't make a lot of sense. - moseely */
#ifdef skip_section
if((n = (int)strlen(word)) == 4){
if( isdigit( (int)(unsigned char)word[1] )
&& isdigit( (int)(unsigned char)word[2] )
&& isdigit( (int)(unsigned char)word[3] ) )
return STEM_OK; /* Hum, probably not right */
}
#endif
/* Convert chars to lower case and strip non-letter chars */
j = 0;
for (i = 0; i < n; i++) {
u = tolower((unsigned char)word[i]);
if ((u > 96) && (u < 123)) {
word[j] = u;
j++;
}
}
/* terminate string */
word[j] = 0;
/* String length again */
n = strlen(word);
soundCode[0] = word[0];
/* remember first char */
l = soundTable[((word[0]) - 97)];
j = 1;
/* build soundex string */
for (i = 1; i < n && j < 4; i++) {
u = soundTable[((word[i]) - 97)];
if (u != l) {
if (u != 0) {
soundCode[(int) j++] = u;
}
l = u;
}
}
fw->free_strings = 1; /* flag that we are creating a string */
fw->string_list[0] = estrdup( soundCode );
return fw;
}
