Class Sequel::Postgres::PGRange::Parser
In: lib/sequel/extensions/pg_range.rb
Parent: Object

Creates callable objects that convert strings into PGRange instances.

Methods

call   new  

Attributes

converter  [R]  A callable object to convert the beginning and ending of the range into the appropriate ruby type.
db_type  [R]  The database range type for this parser (e.g. ‘int4range’), automatically setting the db_type for the returned PGRange instances.

Public Class methods

Set the db_type and converter on initialization.

[Source]

    # File lib/sequel/extensions/pg_range.rb, line 89
89:         def initialize(db_type, converter=nil)
90:           @db_type = db_type.to_s.dup.freeze if db_type
91:           @converter = converter
92:         end

Public Instance methods

Parse the range type input string into a PGRange value.

[Source]

     # File lib/sequel/extensions/pg_range.rb, line 95
 95:         def call(string)
 96:           if string == 'empty'
 97:             return PGRange.empty(db_type)
 98:           end
 99: 
100:           raise(InvalidValue, "invalid or unhandled range format: #{string.inspect}") unless matches = /\A(\[|\()("((?:\\"|[^"])*)"|[^"]*),("((?:\\"|[^"])*)"|[^"]*)(\]|\))\z/.match(string)
101: 
102:           exclude_begin = matches[1] == '('
103:           exclude_end = matches[6] == ')'
104: 
105:           # If the input is quoted, it needs to be unescaped.  Also, quoted input isn't
106:           # checked for emptiness, since the empty quoted string is considered an 
107:           # element that happens to be the empty string, while an unquoted empty string
108:           # is considered unbounded.
109:           #
110:           # While PostgreSQL allows pure escaping for input (without quoting), it appears
111:           # to always use the quoted output form when characters need to be escaped, so
112:           # there isn't a need to unescape unquoted output.
113:           if beg = matches[3]
114:             beg.gsub!(/\\(.)/, '\1')
115:           else
116:             beg = matches[2] unless matches[2].empty?
117:           end
118:           if en = matches[5]
119:             en.gsub!(/\\(.)/, '\1')
120:           else
121:             en = matches[4] unless matches[4].empty?
122:           end
123: 
124:           if c = converter
125:             beg = c.call(beg) if beg
126:             en = c.call(en) if en
127:           end
128: 
129:           PGRange.new(beg, en, :exclude_begin=>exclude_begin, :exclude_end=>exclude_end, :db_type=>db_type)
130:         end

[Validate]