Class: Sequel::Postgres::PGRow::Splitter

Inherits:
StringScanner
  • Object
show all
Defined in:
lib/sequel/extensions/pg_row.rb

Overview

This parser-like class splits the PostgreSQL row-valued/composite type output string format into an array of strings. Note this class makes no attempt to handle all input formats that PostgreSQL will accept, it only handles the output format that PostgreSQL uses.

Constant Summary collapse

OPEN_PAREN = 
/\(/.freeze
CLOSE_PAREN = 
/\)/.freeze
UNQUOTED_RE = 
/[^,)]*/.freeze
SEP_RE = 
/[,)]/.freeze
QUOTE_RE = 
/"/.freeze
QUOTE_SEP_RE = 
/"[,)]/.freeze
QUOTED_RE = 
/(\\.|""|[^"])*/.freeze
REPLACE_RE = 
/\\(.)|"(")/.freeze
REPLACE_WITH = 
'\1\2'.freeze

Instance Method Summary collapse

Instance Method Details

#parse ⇒ Object

Split the stored string into an array of strings, handling the different types of quoting.



235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
 
# File 'lib/sequel/extensions/pg_row.rb', line 235

def parse
  return @result if @result
  values = []
  skip(OPEN_PAREN)
  if skip(CLOSE_PAREN)
    values << nil
  else
    until eos?
      if skip(QUOTE_RE)
        values << scan(QUOTED_RE).gsub(REPLACE_RE, REPLACE_WITH)
        skip(QUOTE_SEP_RE)
      else
        v = scan(UNQUOTED_RE)
        values << (v unless v.empty?)
        skip(SEP_RE)
      end
    end
  end
  values
end