lql

README

Lily Query Language

A Solr like query language for the bluge index (beta)

Lily Query Language (LQL) for short is used to build a bluge search result. It's somewhat modeled after the Solr query language but is currently much more simple. In addition to building a query the language also lets you limit the result set.

Examples

# Find any match of string for a given asset
tag=U123

# Whitespace doesn't matter this is the same as above
tag  =   U123

# Additinally double quotes are used to escape invalid characters
"tag!=" = "U123"

# Not equal operator is also valid
tag!=U123

# Check greater than or less than
value>5

# use logical operators like and/or
value>5 and tag=U123 or value=10

# override operator precidence with brackets
value>5 and (tag=U123 or value=10)

# limit the field that is returned
tag=U123 returns tag

# have multiple returned fields
tag=U123 returns tag, value, test

# Find something between two values
value=[1.3,2.2]

# Use a regex query
tag=~U.*

# Invert your regex
tag!~U.*

Specification

Lexing

Everything in LQL is a string or an operator. Although under the hood the lexer will take a given string and convert it to a float64 for operators >, <, <=, >=, and =[,]. If it can not be parsed to a float64 an error will be returned.

strings - The tSTRING token defines a string in the LQL grammar. A string is defined as any valid character that is not a single character operator. Valid single character operators are =, !, (, ), ~, ,, [, ], \, and ". \ is reserved for escaping characters and a string that starts in an unescaped " must end in it. You only need to escape double quotes inside a quoted string.

operators - The following operators are defined in the LQL grammer. Multi-byte operators are and, or, returns, >=, <=, =~, and!~. Single-byte operators are =, !, (, ), ~, ,, [, ], \, and ".

Parser

The up to date grammer can be seen in parser.y.

Development

To rebuild parser.go you can run the following command

go run golang.org/x/tools/cmd/goyacc -l -o parser.go parser.y

Additionally if you want to add/delete/change something parser.y has been documented with lots of good external resouorces to read.

Before putting up a chanage against this make sure to run the following.

go generate ./... && go test -v ./...

Documentation

Overview

Code generated by goyacc -l -o parser.go parser.y. DO NOT EDIT.

Index

• func Parse(query []byte) (bluge.Query, []string, error)
• type ParseError
  • func (p *ParseError) Error() string
• type ParseNOP
  • func (p *ParseNOP) Error() string
• type UnknownParseError
  • func (u *UnknownParseError) Error() string

Functions

func Parse

func Parse(query []byte) (bluge.Query, []string, error)

Parse the passed in query

Types

type ParseError

type ParseError struct {
	// contains filtered or unexported fields
}

ParseError is for exposing known parse errors

func (*ParseError) Error

func (p *ParseError) Error() string

type ParseNOP

type ParseNOP struct{}

ParseNOP means you gave us nothing to do

func (*ParseNOP) Error

func (p *ParseNOP) Error() string

type UnknownParseError

type UnknownParseError struct{}

UnknownParseError if you get this something has gone very wrong

func (*UnknownParseError) Error

func (u *UnknownParseError) Error() string