class PG::BasicTypeMapBasedOnResult

database types.
This inserts a single row into copytable with type casts from ruby to
end
conn.put_copy_data [‘a’, 123, [5,4,3]]
conn.copy_data( “COPY copytable FROM STDIN”, row_encoder ) do |res|
row_encoder = PG::TextEncoder::CopyRow.new type_map: tm
tm = btm.build_column_map( res )
# Build a PG::TypeMapByColumn with encoders suitable for copytable.
btm = PG::BasicTypeMapBasedOnResult.new(conn)
# Build a type map for common ruby to database type encoders.
res = conn.exec( “SELECT * FROM copytable LIMIT 0” )
# Retrieve table OIDs per empty result set.
conn.exec( “CREATE TEMP TABLE copytable (t TEXT, i INT, ai INT[])” )
Example:
or #put_copy_data fields.
PG::TypeMapByColumn type map, which can subsequently be used to cast query bind parameters
PG::TypeMapByOid#build_column_map(result) can be used to generate a result independent
the type OID retrieved by a separate SQL query.
the given result OIDs, but encoders. So it can be used to type cast field values based on
This class works equal to PG::BasicTypeMapForResults, but does not define decoders for
PostgreSQL’s pg_type table in PG::BasicTypeMapBasedOnResult.new .
OIDs of supported type casts are not hard-coded in the sources, but are retrieved from the
to PostgreSQL.
Simple set of rules for type casting common PostgreSQL types from Ruby

def initialize(connection) 

def initialize(connection)
	@coder_maps = build_coder_maps(connection)
	# Populate TypeMapByOid hash with encoders
	@coder_maps.map{|f| f[:encoder].coders }.flatten.each do |coder|
		add_coder(coder)
	end
end