CirrusSearch\Parser\QueryStringRegex\QueryStringRegexParser Class Reference

Full text query parser that uses regex to parse its token. More...

Inheritance diagram for CirrusSearch\Parser\QueryStringRegex\QueryStringRegexParser:

Collaboration diagram for CirrusSearch\Parser\QueryStringRegex\QueryStringRegexParser:

Public Member Functions
	__construct (\CirrusSearch\Parser\KeywordRegistry $keywordRegistry, Escaper $escaper, $qmarkStripLevel, ParsedQueryClassifiersRepository $classifierRepository, NamespacePrefixParser $namespacePrefixParser, ?int $maxQueryLen)
	parse (string $query)

Public Attributes
const	QUERY_LEN_HARD_LIMIT = 4096

Detailed Description

Full text query parser that uses regex to parse its token.

Far from being a state of the art parser it detects most of its tokens using regular expression. And make arbitrary decisions at tokenization.

The tokenizer will understand few token types:

• WHITESPACE: all unicode whitespace and control chars ([\pZ\pC]) the WHITESPACE token is ignored and never presented to the parser
• EOF: dummy type used to mark end of string
• BOOL_AND/BOOL_OR/BOOL_NOT: explicit boolean opeartors
• PARSED_NODE: complex type (usually part of the query)

PARSED_NODE is a type that groups:

• Keywords
• Phrase
• Words
• Wildcards/Prefix

Phrase does not have its own token " and is part the tokenization and is never exposed to the parser. Same for negation prefix (! and -), they are parsed at tokenization time.

NOTE that this parser is broken by design:

• no lexical context support, we first parse keywords
• no support for groupings (parenthesis)

Constructor & Destructor Documentation

__construct()

CirrusSearch\Parser\QueryStringRegex\QueryStringRegexParser::__construct	(	\CirrusSearch\Parser\KeywordRegistry	$keywordRegistry,
		Escaper	$escaper,
			$qmarkStripLevel,
		ParsedQueryClassifiersRepository	$classifierRepository,
		NamespacePrefixParser	$namespacePrefixParser,
		?int	$maxQueryLen )

Parameters

\CirrusSearch\Parser\KeywordRegistry	$keywordRegistry
Escaper	$escaper
string	$qmarkStripLevel	Level of question mark stripping to apply, either "all", "break", or "final"
ParsedQueryClassifiersRepository	$classifierRepository
NamespacePrefixParser	$namespacePrefixParser
int | null	$maxQueryLen	maximum length of the query in chars

See also

Util::stripQuestionMarks() for acceptable $qmarkStripLevel values

Member Function Documentation

parse()

CirrusSearch\Parser\QueryStringRegex\QueryStringRegexParser::parse

(

string

$query

)

Parameters

string

$query

Returns

\CirrusSearch\Parser\AST\ParsedQuery

Exceptions

SearchQueryParseException

Implements CirrusSearch\Parser\QueryParser.

---

The documentation for this class was generated from the following file:

• includes/Parser/QueryStringRegex/QueryStringRegexParser.php