diff options
| author | Thomas Van Lenten <thomasvl@google.com> | 2017-06-16 12:30:08 -0400 |
|---|---|---|
| committer | Thomas Van Lenten <thomasvl@google.com> | 2017-06-16 15:24:33 -0400 |
| commit | dd19b876d4c9a604946f6c4e39cc4eac5f12cfb9 (patch) | |
| tree | dabceb4c021d2916092a080c02f880facd222e94 /conformance | |
| parent | 91bf623aa145cd1576451ed3764a64a089384ac4 (diff) | |
| download | protobuf-dd19b876d4c9a604946f6c4e39cc4eac5f12cfb9.tar.gz protobuf-dd19b876d4c9a604946f6c4e39cc4eac5f12cfb9.tar.bz2 protobuf-dd19b876d4c9a604946f6c4e39cc4eac5f12cfb9.zip | |
Raise the number of digits used for floats.
About 1.5% of all IEEE754 single-precision numbers require nine
decimal digits to represent accurately.
Diffstat (limited to 'conformance')
| -rw-r--r-- | conformance/conformance_test.cc | 1 | ||||
| -rw-r--r-- | conformance/failure_list_ruby.txt | 1 |
2 files changed, 2 insertions, 0 deletions
diff --git a/conformance/conformance_test.cc b/conformance/conformance_test.cc index 0dd7787c..a899435d 100644 --- a/conformance/conformance_test.cc +++ b/conformance/conformance_test.cc @@ -759,6 +759,7 @@ bool ConformanceTestSuite::RunSuite(ConformanceTestRunner* runner, }); TestValidDataForType(FieldDescriptor::TYPE_FLOAT, { {flt(0.1), "0.1"}, + {flt(1.00000075e-36), "1.00000075e-36"}, {flt(3.402823e+38), "3.402823e+38"}, // 3.40282347e+38 {flt(1.17549435e-38f), "1.17549435e-38"} }); diff --git a/conformance/failure_list_ruby.txt b/conformance/failure_list_ruby.txt index 1de6c439..4bac533e 100644 --- a/conformance/failure_list_ruby.txt +++ b/conformance/failure_list_ruby.txt @@ -199,5 +199,6 @@ Required.ProtobufInput.DoubleFieldNormalizeQuietNan.JsonOutput Required.ProtobufInput.DoubleFieldNormalizeSignalingNan.JsonOutput Required.ProtobufInput.FloatFieldNormalizeQuietNan.JsonOutput Required.ProtobufInput.FloatFieldNormalizeSignalingNan.JsonOutput +Required.ProtobufInput.ValidDataRepeated.FLOAT.JsonOutput Required.TimestampProtoInputTooLarge.JsonOutput Required.TimestampProtoInputTooSmall.JsonOutput |