Support floats in spectest runner #445
Conversation
| @@ -9,7 +9,7 @@ add_test( | |||
| set_tests_properties( | |||
| fizzy/smoketests/spectests/default | |||
| PROPERTIES | |||
| PASS_REGULAR_EXPRESSION "PASSED 23, FAILED 0, SKIPPED 7" | |||
| PASS_REGULAR_EXPRESSION "PASSED 26, FAILED 1, SKIPPED 3" | |||
There was a problem hiding this comment.
1 failure is for float global, will be fixed in #446
test/spectests/spectests.cpp
Outdated
| template <> | ||
| fizzy::Value json_to_value<float>(const json& v) | ||
| { | ||
| return std::stoull(v.get<std::string>()); |
There was a problem hiding this comment.
Why using "to unsigned integer" for floating point values?
There was a problem hiding this comment.
wast2json puts them as 64-bit integers to json
| expected_failed: 9 | ||
| expected_skipped: 6336 | ||
| expected_passed: 5577 | ||
| expected_failed: 13 |
There was a problem hiding this comment.
4 new failures are all in globals.wast, will be fixed in #446
Codecov Report
@@ Coverage Diff @@
## master #445 +/- ##
==========================================
- Coverage 99.51% 99.51% -0.01%
==========================================
Files 52 52
Lines 14791 14785 -6
==========================================
- Hits 14719 14713 -6
Misses 72 72 |
| const auto arg_type = v.at("type").get<std::string>(); | ||
| if (arg_type != "i32" && arg_type != "i64" && arg_type != "f32" && arg_type != "f64") | ||
| { | ||
| skip("Unsupported value type '" + arg_type + "'."); |
There was a problem hiding this comment.
This is not covered by unit tests.
| } | ||
|
|
||
| // Values of all types are serialized to JSON as integers. | ||
| // Value type will handle correct conversions. |
There was a problem hiding this comment.
How? Why not keep the old json_to_value<int32_t>(arg.at("value")) way?
There was a problem hiding this comment.
It's just simpler without additional switch and template.
Presumably wast2json does the opposite conversion - typed value into unsigned integer. We could reinterpret_cast it back to correct type, but Value(correct_type_value) constructor does the same as Value(reinterpret_cast<unsigned long long>(correct_type_value)), which is the idea of Value type...
Depends on #424