CREATE ANALYZER
Define a new fulltext analyzer.
Table of Contents
Synopsis
CREATE ANALYZER analyzer_name EXTENDS parent_analyzer_name
WITH ( override_parameter [= value] [, ... ] )
CREATE ANALYZER analyzer_name (
[ TOKENIZER
{
tokenizer_name
| custom_name WITH ( type = tokenizer_name, tokenizer_parameter [= value] [, ... ] )
}
]
[ TOKEN_FILTERS (
{
token_filter_name
| custom_name WITH ( type = token_filter_name, token_filter_parameter [= value] [, ... ] )
}
[, ... ]
)
]
[ CHAR_FILTERS (
{
char_filter_name
| custom_name WITH ( type = char_filter_name, char_filter_parameter [= value] [, ... ] )
}
[, ... ]
)
]
)
Description
CREATE ANALYZER speifies a whole analyzer chain for use in fulltext searches. It is possible to extend an existing analyzer or define a new analyzer chain from scratch. For examples and detailed explanation see Create a Custom Analyzer.
Parameters
analyzer_name: | The globally unique name of the analzyer being created. |
---|---|
parent_analyzer_name: | The name of the analyzer to inherit defaults from. |
override_parameter: | The name of a parameter of the parent analyzer which should be assigned a new value to. |
tokenizer_name: | The name of a builtin tokenizer to be used. |
tokenizer_parameter: | A name of a parameter for a given tokenizer. |
token_filter_name: | The name of a builtin token filter to be used. |
token_filter_parameter: | A name of a parameter for a given token filter. |
char_filter_name: | The name of a builtin char filter to be used. |
char_filter_parameter: | A name of a parameter for a given char filter. |
custom_name: | A custom unqiue name needed when defining custom tokenizers/token_filter/char_filter. |