e::equals_max_precision tests fail on i386

[jonassmedegaard]

With release v0.28.2, two tests fail on i386:

failures:

---- values::calculation::calc::constant::e::equals_max_precision stdout ----
thread 'values::calculation::calc::constant::e::equals_max_precision' panicked at 'assertion failed: `(left == right)`
  left: `"a {\n  b: false;\n}\n"`,
 right: `"a {\n  b: true;\n}\n"`', rsass/tests/spec/values/calculation/calc/constant.rs:32:9
stack backtrace:
   0: rust_begin_unwind
             at /usr/src/rustc-1.66.0/library/std/src/panicking.rs:575:5
   1: core::panicking::panic_fmt
             at /usr/src/rustc-1.66.0/library/core/src/panicking.rs:65:14
   2: core::panicking::assert_failed_inner
   3: core::panicking::assert_failed
             at /usr/src/rustc-1.66.0/library/core/src/panicking.rs:203:5
   4: spec::values::calculation::calc::constant::e::equals_max_precision
             at ./tests/spec/values/calculation/calc/constant.rs:32:9
   5: spec::values::calculation::calc::constant::e::equals_max_precision::{{closure}}
             at ./tests/spec/values/calculation/calc/constant.rs:31:5
   6: core::ops::function::FnOnce::call_once
             at /usr/src/rustc-1.66.0/library/core/src/ops/function.rs:251:5
   7: core::ops::function::FnOnce::call_once
             at /usr/src/rustc-1.66.0/library/core/src/ops/function.rs:251:5
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.

---- values::calculation::calc::constant::pi::equals_max_precision stdout ----
thread 'values::calculation::calc::constant::pi::equals_max_precision' panicked at 'assertion failed: `(left == right)`
  left: `"a {\n  b: false;\n}\n"`,
 right: `"a {\n  b: true;\n}\n"`', rsass/tests/spec/values/calculation/calc/constant.rs:333:9
stack backtrace:
   0: rust_begin_unwind
             at /usr/src/rustc-1.66.0/library/std/src/panicking.rs:575:5
   1: core::panicking::panic_fmt
             at /usr/src/rustc-1.66.0/library/core/src/panicking.rs:65:14
   2: core::panicking::assert_failed_inner
   3: core::panicking::assert_failed
             at /usr/src/rustc-1.66.0/library/core/src/panicking.rs:203:5
   4: spec::values::calculation::calc::constant::pi::equals_max_precision
             at ./tests/spec/values/calculation/calc/constant.rs:333:9
   5: spec::values::calculation::calc::constant::pi::equals_max_precision::{{closure}}
             at ./tests/spec/values/calculation/calc/constant.rs:332:5
   6: core::ops::function::FnOnce::call_once
             at /usr/src/rustc-1.66.0/library/core/src/ops/function.rs:251:5
   7: core::ops::function::FnOnce::call_once
             at /usr/src/rustc-1.66.0/library/core/src/ops/function.rs:251:5
note: Some details are omitted, run with `RUST_BACKTRACE=full` for a verbose backtrace.


failures:
    values::calculation::calc::constant::e::equals_max_precision
    values::calculation::calc::constant::pi::equals_max_precision

Full build log: https://buildd.debian.org/status/fetch.php?pkg=rsass&arch=i386&ver=0.28.2-1&stamp=1691158787&raw=0

[kaj]

This is similar to #168 . I'm not sure if it is a problem with 64-bit floats on a 32-bit architecture or if it may be a problem with my code for converting "text" floating point numbers to f64 floating point numbers. As I do that conversion through a num-rational value, that may also contribute.

Checking for exact equality in floating points numbers is often problematic. Maybe the tests (in sass-spec) should be modified to check that the value falls inside a suitably small range?