1
/* Copyright (C) 2005 MySQL AB
3
This program is free software; you can redistribute it and/or modify
4
it under the terms of the GNU General Public License as published by
5
the Free Software Foundation; version 2 of the License.
7
This program is distributed in the hope that it will be useful,
8
but WITHOUT ANY WARRANTY; without even the implied warranty of
9
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
10
GNU General Public License for more details.
12
You should have received a copy of the GNU General Public License
13
along with this program; if not, write to the Free Software
14
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
16
#ifndef _my_plugin_ftparser_h
17
#define _my_plugin_ftparser_h
19
/*************************************************************************
20
API for Full-text parser plugin. (MYSQL_FTPARSER_PLUGIN)
23
#define MYSQL_FTPARSER_INTERFACE_VERSION 0x0100
25
/* Parsing modes. Set in MYSQL_FTPARSER_PARAM::mode */
26
enum enum_ftparser_mode
29
Fast and simple mode. This mode is used for indexing, and natural
32
The parser is expected to return only those words that go into the
33
index. Stopwords or too short/long words should not be returned. The
34
'boolean_info' argument of mysql_add_word() does not have to be set.
36
MYSQL_FTPARSER_SIMPLE_MODE= 0,
39
Parse with stopwords mode. This mode is used in boolean searches for
42
The parser is not allowed to ignore words in this mode. Every word
43
should be returned, including stopwords and words that are too short
44
or long. The 'boolean_info' argument of mysql_add_word() does not
47
MYSQL_FTPARSER_WITH_STOPWORDS= 1,
50
Parse in boolean mode. This mode is used to parse a boolean query string.
52
The parser should provide a valid MYSQL_FTPARSER_BOOLEAN_INFO
53
structure in the 'boolean_info' argument to mysql_add_word().
54
Usually that means that the parser should recognize boolean operators
55
in the parsing stream and set appropriate fields in
56
MYSQL_FTPARSER_BOOLEAN_INFO structure accordingly. As for
57
MYSQL_FTPARSER_WITH_STOPWORDS mode, no word should be ignored.
58
Instead, use FT_TOKEN_STOPWORD for the token type of such a word.
60
MYSQL_FTPARSER_FULL_BOOLEAN_INFO= 2
64
Token types for boolean mode searching (used for the type member of
65
MYSQL_FTPARSER_BOOLEAN_INFO struct)
67
FT_TOKEN_EOF: End of data.
68
FT_TOKEN_WORD: Regular word.
69
FT_TOKEN_LEFT_PAREN: Left parenthesis (start of group/sub-expression).
70
FT_TOKEN_RIGHT_PAREN: Right parenthesis (end of group/sub-expression).
71
FT_TOKEN_STOPWORD: Stopword.
74
enum enum_ft_token_type
78
FT_TOKEN_LEFT_PAREN= 2,
79
FT_TOKEN_RIGHT_PAREN= 3,
84
This structure is used in boolean search mode only. It conveys
85
boolean-mode metadata to the MySQL search engine for every word in
86
the search query. A valid instance of this structure must be filled
87
in by the plugin parser and passed as an argument in the call to
88
mysql_add_word (the callback function in the MYSQL_FTPARSER_PARAM
89
structure) when a query is parsed in boolean mode.
91
type: The token type. Should be one of the enum_ft_token_type values.
93
yesno: Whether the word must be present for a match to occur:
95
<0 Must not be present
96
0 Neither; the word is optional but its presence increases the relevance
97
With the default settings of the ft_boolean_syntax system variable,
98
>0 corresponds to the '+' operator, <0 corrresponds to the '-' operator,
99
and 0 means neither operator was used.
101
weight_adjust: A weighting factor that determines how much a match
102
for the word counts. Positive values increase, negative - decrease the
103
relative word's importance in the query.
105
wasign: The sign of the word's weight in the query. If it's non-negative
106
the match for the word will increase document relevance, if it's
107
negative - decrease (the word becomes a "noise word", the less of it the
110
trunc: Corresponds to the '*' operator in the default setting of the
111
ft_boolean_syntax system variable.
114
typedef struct st_mysql_ftparser_boolean_info
116
enum enum_ft_token_type type;
121
/* These are parser state and must be removed. */
124
} MYSQL_FTPARSER_BOOLEAN_INFO;
127
The following flag means that buffer with a string (document, word)
128
may be overwritten by the caller before the end of the parsing (that is
129
before st_mysql_ftparser::deinit() call). If one needs the string
130
to survive between two successive calls of the parsing function, she
131
needs to save a copy of it. The flag may be set by MySQL before calling
132
st_mysql_ftparser::parse(), or it may be set by a plugin before calling
133
st_mysql_ftparser_param::mysql_parse() or
134
st_mysql_ftparser_param::mysql_add_word().
136
#define MYSQL_FTFLAGS_NEED_COPY 1
139
An argument of the full-text parser plugin. This structure is
140
filled in by MySQL server and passed to the parsing function of the
141
plugin as an in/out parameter.
143
mysql_parse: A pointer to the built-in parser implementation of the
144
server. It's set by the server and can be used by the parser plugin
145
to invoke the MySQL default parser. If plugin's role is to extract
146
textual data from .doc, .pdf or .xml content, it might extract
147
plaintext from the content, and then pass the text to the default
148
MySQL parser to be parsed.
150
mysql_add_word: A server callback to add a new word. When parsing
151
a document, the server sets this to point at a function that adds
152
the word to MySQL full-text index. When parsing a search query,
153
this function will add the new word to the list of words to search
154
for. The boolean_info argument can be NULL for all cases except
155
when mode is MYSQL_FTPARSER_FULL_BOOLEAN_INFO.
157
ftparser_state: A generic pointer. The plugin can set it to point
158
to information to be used internally for its own purposes.
160
mysql_ftparam: This is set by the server. It is used by MySQL functions
161
called via mysql_parse() and mysql_add_word() callback. The plugin
162
should not modify it.
164
cs: Information about the character set of the document or query string.
166
doc: A pointer to the document or query string to be parsed.
168
length: Length of the document or query string, in bytes.
170
flags: See MYSQL_FTFLAGS_* constants above.
172
mode: The parsing mode. With boolean operators, with stopwords, or
173
nothing. See enum_ftparser_mode above.
176
typedef struct st_mysql_ftparser_param
178
int (*mysql_parse)(struct st_mysql_ftparser_param *,
179
char *doc, int doc_len);
180
int (*mysql_add_word)(struct st_mysql_ftparser_param *,
181
char *word, int word_len,
182
MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info);
183
void *ftparser_state;
185
struct charset_info_st *cs;
189
enum enum_ftparser_mode mode;
190
} MYSQL_FTPARSER_PARAM;
193
Full-text parser descriptor.
195
interface_version is, e.g., MYSQL_FTPARSER_INTERFACE_VERSION.
196
The parsing, initialization, and deinitialization functions are
197
invoked per SQL statement for which the parser is used.
200
struct st_mysql_ftparser
202
int interface_version;
203
int (*parse)(MYSQL_FTPARSER_PARAM *param);
204
int (*init)(MYSQL_FTPARSER_PARAM *param);
205
int (*deinit)(MYSQL_FTPARSER_PARAM *param);
added
removed
include/mysql/plugin_ftparser.h
