Some definitions for full-text indices. More...

#include <math.h>
#include <mysql/plugin.h>
#include <cstdint>
#include "my_tree.h"
#include "mysql/strings/m_ctype.h"
#include "storage/myisam/fulltext.h"
#include "storage/myisam/queues.h"

Classes
struct	FT_WORD
	FTS query token. More...

struct	_st_ft_seg_iterator

Macros
#define	FT_MAX_WORD_LEN_FOR_SORT 31

#define	FTPARSER_MEMROOT_ALLOC_SIZE 65536

#define	LWS_FOR_QUERY LWS_TF

#define	LWS_IN_USE LWS_LOG

#define	PRENORM_IN_USE PRENORM_AVG

#define	NORM_IN_USE NORM_PIVOT

#define	GWS_IN_USE GWS_PROB

#define	LWS_TF (count)

#define	LWS_BINARY (count > 0)

#define	LWS_SQUARE (count * count)

#define	LWS_LOG (count ? (log((double)count) + 1) : 0)

#define	PRENORM_NONE (p->weight)

#define	PRENORM_MAX (p->weight / docstat.max)

#define	PRENORM_AUG (0.4 + 0.6 * p->weight / docstat.max)

#define	PRENORM_AVG (p->weight / docstat.sum * docstat.uniq)

#define	PRENORM_AVGLOG ((1 + log(p->weight)) / (1 + log(docstat.sum / docstat.uniq)))

#define	NORM_NONE (1)

#define	NORM_SUM (docstat.nsum)

#define	NORM_COS (sqrt(docstat.nsum2))

#define	PIVOT_VAL (0.0115)

#define	NORM_PIVOT (1 + PIVOT_VAL * docstat.uniq)

#define	GWS_NORM (1 / sqrt(sum2))

#define	GWS_GFIDF (sum / doc_cnt)

#define	GWS_IDF log(aio->info->state->records / doc_cnt)

#define	GWS_IDF1 log((double)aio->info->state->records / doc_cnt)

#define	GWS_PROB

#define	GWS_FREQ (1.0 / doc_cnt)

#define	GWS_SQUARED pow(log((double)aio->info->state->records / doc_cnt), 2)

#define	GWS_CUBIC pow(log((double)aio->info->state->records / doc_cnt), 3)

#define	GWS_ENTROPY (1 - (suml / sum - log(sum)) / log(aio->info->state->records))

#define	FTB_YES (ft_boolean_syntax[0])

#define	FTB_EGAL (ft_boolean_syntax[1])

#define	FTB_NO (ft_boolean_syntax[2])

#define	FTB_INC (ft_boolean_syntax[3])

#define	FTB_DEC (ft_boolean_syntax[4])

#define	FTB_LBR (ft_boolean_syntax[5])

#define	FTB_RBR (ft_boolean_syntax[6])

#define	FTB_NEG (ft_boolean_syntax[7])

#define	FTB_TRUNC (ft_boolean_syntax[8])

#define	FTB_LQUOT (ft_boolean_syntax[10])

#define	FTB_RQUOT (ft_boolean_syntax[11])

Typedefs
typedef struct _st_ft_seg_iterator	FT_SEG_ITERATOR

Functions
bool	true_word_char (int c, uint8_t ch)

int	is_stopword (char *word, uint len)

uint	_ft_make_key (MI_INFO , uint, uchar , FT_WORD *, my_off_t)

uchar	ft_get_word (const CHARSET_INFO , uchar , uchar , FT_WORD , MYSQL_FTPARSER_BOOLEAN_INFO )

uchar	ft_simple_get_word (const CHARSET_INFO , uchar , const uchar , FT_WORD *, bool)

void	_mi_ft_segiterator_init (MI_INFO , uint, const uchar , FT_SEG_ITERATOR *)

void	_mi_ft_segiterator_dummy_init (const uchar , uint, FT_SEG_ITERATOR )

uint	_mi_ft_segiterator (FT_SEG_ITERATOR *)

void	ft_parse_init (TREE , const CHARSET_INFO )

int	ft_parse (TREE , uchar , int, struct st_mysql_ftparser parser, MYSQL_FTPARSER_PARAM , MEM_ROOT *)

FT_WORD *	ft_linearize (TREE , MEM_ROOT )

FT_WORD *	_mi_ft_parserecord (MI_INFO , uint, const uchar , MEM_ROOT *)

uint	_mi_ft_parse (TREE , MI_INFO , uint, const uchar , MYSQL_FTPARSER_PARAM , MEM_ROOT *)

FT_INFO *	ft_init_nlq_search (MI_INFO , uint, uchar , uint, uint, uchar *)

FT_INFO *	ft_init_boolean_search (MI_INFO , uint, uchar , uint, const CHARSET_INFO *)

int	ft_nlq_read_next (FT_INFO , char )

float	ft_nlq_find_relevance (FT_INFO , uchar , uint)

void	ft_nlq_close_search (FT_INFO *)

float	ft_nlq_get_relevance (FT_INFO *)

my_off_t	ft_nlq_get_docid (FT_INFO *)

void	ft_nlq_reinit_search (FT_INFO *)

int	ft_boolean_read_next (FT_INFO , char )

float	ft_boolean_find_relevance (FT_INFO , uchar , uint)

void	ft_boolean_close_search (FT_INFO *)

float	ft_boolean_get_relevance (FT_INFO *)

my_off_t	ft_boolean_get_docid (FT_INFO *)

void	ft_boolean_reinit_search (FT_INFO *)

MYSQL_FTPARSER_PARAM *	ftparser_alloc_param (MI_INFO *info)

MYSQL_FTPARSER_PARAM *	ftparser_call_initializer (MI_INFO *info, uint keynr, uint paramnr)

void	ftparser_call_deinitializer (MI_INFO *info)

Variables
const struct _ft_vft	_ft_vft_nlq

const struct _ft_vft	_ft_vft_boolean

Detailed Description

Some definitions for full-text indices.

Macro Definition Documentation

◆ FT_MAX_WORD_LEN_FOR_SORT

#define FT_MAX_WORD_LEN_FOR_SORT 31

◆ FTB_DEC

#define FTB_DEC (ft_boolean_syntax[4])

◆ FTB_EGAL

#define FTB_EGAL (ft_boolean_syntax[1])

◆ FTB_INC

#define FTB_INC (ft_boolean_syntax[3])

◆ FTB_LBR

#define FTB_LBR (ft_boolean_syntax[5])

◆ FTB_LQUOT

#define FTB_LQUOT (ft_boolean_syntax[10])

◆ FTB_NEG

#define FTB_NEG (ft_boolean_syntax[7])

◆ FTB_NO

#define FTB_NO (ft_boolean_syntax[2])

◆ FTB_RBR

#define FTB_RBR (ft_boolean_syntax[6])

◆ FTB_RQUOT

#define FTB_RQUOT (ft_boolean_syntax[11])

◆ FTB_TRUNC

#define FTB_TRUNC (ft_boolean_syntax[8])

◆ FTB_YES

#define FTB_YES (ft_boolean_syntax[0])

◆ FTPARSER_MEMROOT_ALLOC_SIZE

#define FTPARSER_MEMROOT_ALLOC_SIZE 65536

◆ GWS_CUBIC

#define GWS_CUBIC pow(log((double)aio->info->state->records / doc_cnt), 3)

◆ GWS_ENTROPY

#define GWS_ENTROPY (1 - (suml / sum - log(sum)) / log(aio->info->state->records))

◆ GWS_FREQ

#define GWS_FREQ (1.0 / doc_cnt)

◆ GWS_GFIDF

#define GWS_GFIDF (sum / doc_cnt)

◆ GWS_IDF

#define GWS_IDF log(aio->info->state->records / doc_cnt)

◆ GWS_IDF1

#define GWS_IDF1 log((double)aio->info->state->records / doc_cnt)

◆ GWS_IN_USE

#define GWS_IN_USE GWS_PROB

◆ GWS_NORM

#define GWS_NORM (1 / sqrt(sum2))

◆ GWS_PROB

#define GWS_PROB

Value:

  ((aio->info->state->records > doc_cnt)                                \
       ? log(((double)(aio->info->state->records - doc_cnt)) / doc_cnt) \
       : 0)

◆ GWS_SQUARED

#define GWS_SQUARED pow(log((double)aio->info->state->records / doc_cnt), 2)

◆ LWS_BINARY

#define LWS_BINARY (count > 0)

◆ LWS_FOR_QUERY

#define LWS_FOR_QUERY LWS_TF

◆ LWS_IN_USE

#define LWS_IN_USE LWS_LOG

◆ LWS_LOG

#define LWS_LOG (count ? (log((double)count) + 1) : 0)

◆ LWS_SQUARE

#define LWS_SQUARE (count * count)

◆ LWS_TF

#define LWS_TF (count)

◆ NORM_COS

#define NORM_COS (sqrt(docstat.nsum2))

◆ NORM_IN_USE

#define NORM_IN_USE NORM_PIVOT

◆ NORM_NONE

#define NORM_NONE (1)

◆ NORM_PIVOT

#define NORM_PIVOT (1 + PIVOT_VAL * docstat.uniq)

◆ NORM_SUM

#define NORM_SUM (docstat.nsum)

◆ PIVOT_VAL

#define PIVOT_VAL (0.0115)

◆ PRENORM_AUG

#define PRENORM_AUG (0.4 + 0.6 * p->weight / docstat.max)

◆ PRENORM_AVG

#define PRENORM_AVG (p->weight / docstat.sum * docstat.uniq)

◆ PRENORM_AVGLOG

#define PRENORM_AVGLOG ((1 + log(p->weight)) / (1 + log(docstat.sum / docstat.uniq)))

◆ PRENORM_IN_USE

#define PRENORM_IN_USE PRENORM_AVG

◆ PRENORM_MAX

#define PRENORM_MAX (p->weight / docstat.max)

◆ PRENORM_NONE

#define PRENORM_NONE (p->weight)

Typedef Documentation

◆ FT_SEG_ITERATOR

typedef struct _st_ft_seg_iterator FT_SEG_ITERATOR

Function Documentation

◆ _ft_make_key()

uint _ft_make_key	(	MI_INFO *	info,
		uint	keynr,
		uchar *	keybuf,
		FT_WORD *	wptr,
		my_off_t	filepos
	)

◆ _mi_ft_parse()

uint _mi_ft_parse	(	TREE *	parsed,
		MI_INFO *	info,
		uint	keynr,
		const uchar *	record,
		MYSQL_FTPARSER_PARAM *	param,
		MEM_ROOT *	mem_root
	)

◆ _mi_ft_parserecord()

FT_WORD * _mi_ft_parserecord	(	MI_INFO *	info,
		uint	keynr,
		const uchar *	record,
		MEM_ROOT *	mem_root
	)

◆ _mi_ft_segiterator()

uint _mi_ft_segiterator ( FT_SEG_ITERATOR * ftsi )

◆ _mi_ft_segiterator_dummy_init()

void _mi_ft_segiterator_dummy_init	(	const uchar *	record,
		uint	len,
		FT_SEG_ITERATOR *	ftsi
	)

◆ _mi_ft_segiterator_init()

void _mi_ft_segiterator_init	(	MI_INFO *	info,
		uint	keynr,
		const uchar *	record,
		FT_SEG_ITERATOR *	ftsi
	)

◆ ft_boolean_close_search()

void ft_boolean_close_search ( FT_INFO * ftb_base )

◆ ft_boolean_find_relevance()

float ft_boolean_find_relevance	(	FT_INFO *	ftb_base,
		uchar *	record,
		uint	length
	)

◆ ft_boolean_get_docid()

my_off_t ft_boolean_get_docid ( FT_INFO * )

◆ ft_boolean_get_relevance()

float ft_boolean_get_relevance ( FT_INFO * ftb_base )

◆ ft_boolean_read_next()

int ft_boolean_read_next	(	FT_INFO *	ftb_base,
		char *	record
	)

◆ ft_boolean_reinit_search()

void ft_boolean_reinit_search ( FT_INFO * ftb )

◆ ft_get_word()

uchar ft_get_word	(	const CHARSET_INFO *	cs,
		uchar **	start,
		uchar *	end,
		FT_WORD *	word,
		MYSQL_FTPARSER_BOOLEAN_INFO *	param
	)

◆ ft_init_boolean_search()

FT_INFO * ft_init_boolean_search	(	MI_INFO *	info,
		uint	keynr,
		uchar *	query,
		uint	query_len,
		const CHARSET_INFO *	cs
	)

◆ ft_init_nlq_search()

FT_INFO * ft_init_nlq_search	(	MI_INFO *	info,
		uint	keynr,
		uchar *	query,
		uint	query_len,
		uint	flags,
		uchar *	record
	)

◆ ft_linearize()

FT_WORD * ft_linearize	(	TREE *	wtree,
		MEM_ROOT *	mem_root
	)

◆ ft_nlq_close_search()

void ft_nlq_close_search ( FT_INFO * handler )

◆ ft_nlq_find_relevance()

float ft_nlq_find_relevance	(	FT_INFO *	handler_base,
		uchar *	record,
		uint	length
	)

◆ ft_nlq_get_docid()

my_off_t ft_nlq_get_docid ( FT_INFO * )

◆ ft_nlq_get_relevance()

float ft_nlq_get_relevance ( FT_INFO * handler_base )

◆ ft_nlq_read_next()

int ft_nlq_read_next	(	FT_INFO *	handler_base,
		char *	record
	)

◆ ft_nlq_reinit_search()

void ft_nlq_reinit_search ( FT_INFO * handler_base )

◆ ft_parse()

int ft_parse	(	TREE *	wtree,
		uchar *	doc,
		int	doclen,
		struct st_mysql_ftparser *	parser,
		MYSQL_FTPARSER_PARAM *	param,
		MEM_ROOT *	mem_root
	)

◆ ft_parse_init()

void ft_parse_init	(	TREE *	wtree,
		const CHARSET_INFO *	cs
	)

◆ ft_simple_get_word()

uchar ft_simple_get_word	(	const CHARSET_INFO *	cs,
		uchar **	start,
		const uchar *	end,
		FT_WORD *	word,
		bool	skip_stopwords
	)

◆ ftparser_alloc_param()

MYSQL_FTPARSER_PARAM * ftparser_alloc_param ( MI_INFO * info )

◆ ftparser_call_deinitializer()

void ftparser_call_deinitializer ( MI_INFO * info )

◆ ftparser_call_initializer()

MYSQL_FTPARSER_PARAM * ftparser_call_initializer	(	MI_INFO *	info,
		uint	keynr,
		uint	paramnr
	)

◆ is_stopword()

int is_stopword	(	char *	word,
		uint	len
	)

◆ true_word_char()

bool true_word_char	(	int	c,
		uint8_t	ch
	)

inline

Variable Documentation

◆ _ft_vft_boolean

const struct _ft_vft _ft_vft_boolean

extern

◆ _ft_vft_nlq

const struct _ft_vft _ft_vft_nlq