I'm going to assert that this test should pass:
def test_jdbc_precision
Sample.delete_all()
assert_nil Sample.find_by_name('br1991a')
density = 5.8
Sample.new(:name=>'br1991a', :density=>density).save!
sample = Sample.find_by_name('br1991a')
assert_not_nil sample
assert_equal density, sample.density
end
Presently it doesn't:
1) Failure:
test_jdbc_precision(TestPrecision) [problem_rails.rb:25]:
<5.8> expected but was
<5.80000019073486>.
This is because the activerecord-jdbc adapter is casting the value from the database to a float (32 bit) and then to a RubyFloat which is backed by a java double type. To be consistent, activerecord-jdbc should just use the same representation that jruby uses internally (a double).
-Justin
On Wed, Jul 8, 2009 at 12:12 PM, Michael Campbell
<michael.campbell@...> wrote:
Uwe Kubosch wrote:
When using Float/Double, you should always be prepared for small
variations during handling.
If you want to store and retrieve a decimal number without it being
subject to rounding etc, you should use the DECIMAL column type. It
will be converted to a BigDecimal when loaded by ActiveRecord, and it
will not suffer any rounding/conversion changes.
For the OP, I think most unit test frameworks provide some sort of assertEquals(floatingpoint_value_1, floatingpoint_value_2, allowed_variance) type of call. If not, it's trivial to write.
