class Sequel::MySQL::Dataset

Dataset class for MySQL datasets accessed via the native driver.

Public Instance Methods

fetch_rows(sql) { |s| ... } click to toggle source

Yield all rows matching this dataset. If the dataset is set to split multiple statements, yield arrays of hashes one per statement instead of yielding results for all statements as hashes.

# File lib/sequel/adapters/mysql.rb, line 295
def fetch_rows(sql)
  execute(sql) do |r|
    i = -1
    cps = db.conversion_procs
    cols = r.fetch_fields.map do |f| 
      # Pretend tinyint is another integer type if its length is not 1, to
      # avoid casting to boolean if Sequel::MySQL.convert_tinyint_to_bool
      # is set.
      type_proc = f.type == 1 && cast_tinyint_integer?(f) ? cps[2] : cps[f.type]
      [output_identifier(f.name), type_proc, i+=1]
    end
    @columns = cols.map{|c| c.first}
    if opts[:split_multiple_result_sets]
      s = []
      yield_rows(r, cols){|h| s << h}
      yield s
    else
      yield_rows(r, cols){|h| yield h}
    end
  end
  self
end
graph(*) click to toggle source

Don't allow graphing a dataset that splits multiple statements

Calls superclass method Sequel::Dataset#graph
# File lib/sequel/adapters/mysql.rb, line 319
def graph(*)
  raise(Error, "Can't graph a dataset that splits multiple result sets") if opts[:split_multiple_result_sets]
  super
end
split_multiple_result_sets() click to toggle source

Makes each yield arrays of rows, with each array containing the rows for a given result set. Does not work with graphing. So you can submit SQL with multiple statements and easily determine which statement returned which results.

Modifies the row_proc of the returned dataset so that it still works as expected (running on the hashes instead of on the arrays of hashes). If you modify the row_proc afterward, note that it will receive an array of hashes instead of a hash.

# File lib/sequel/adapters/mysql.rb, line 333
def split_multiple_result_sets
  raise(Error, "Can't split multiple statements on a graphed dataset") if opts[:graph]
  ds = clone(:split_multiple_result_sets=>true)
  ds.row_proc = proc{|x| x.map{|h| row_proc.call(h)}} if row_proc
  ds
end

Private Instance Methods

cast_tinyint_integer?(field) click to toggle source

Whether a tinyint field should be casted as an integer. By default, casts to integer if the field length is not 1. Can be overwritten to make tinyint casting dataset dependent.

# File lib/sequel/adapters/mysql.rb, line 345
def cast_tinyint_integer?(field)
  field.length != 1
end
execute(sql, opts=OPTS, &block) click to toggle source

Set the :type option to :select if it hasn't been set.

Calls superclass method Sequel::Dataset#execute
# File lib/sequel/adapters/mysql.rb, line 350
def execute(sql, opts=OPTS, &block)
  super(sql, {:type=>:select}.merge(opts), &block)
end
literal_string_append(sql, v) click to toggle source

Handle correct quoting of strings using ::MySQL.quote.

# File lib/sequel/adapters/mysql.rb, line 355
def literal_string_append(sql, v)
  sql << "'"
  sql << ::Mysql.quote(v)
  sql << "'"
end
yield_rows(r, cols) { |h| ... } click to toggle source

Yield each row of the given result set r with columns cols as a hash with symbol keys

# File lib/sequel/adapters/mysql.rb, line 363
def yield_rows(r, cols)
  while row = r.fetch_row
    h = {}
    cols.each{|n, p, i| v = row[i]; h[n] = (v && p) ? p.call(v) : v}
    yield h
  end
end