Bumps [actions/cache](https://github.com/actions/cache) from 2.1.5 to 2.1.6.
- [Release notes](https://github.com/actions/cache/releases)
- [Commits](https://github.com/actions/cache/compare/v2.1.5...v2.1.6)
Signed-off-by: dependabot[bot] <support@github.com>

So that `mruby -e 'a'` does not print backtrace history, i.e.

```
% mruby -e 'a'
-e:1: undefined method 'a' (NoMethodError)
```

Instead of

```
% mruby -e 'a'
trace (most recent call last):
        [1] -e:1
-e:1: undefined method 'a' (NoMethodError)
```

Method names need not to be printed in backtrace history.

Use only `mrb_ary_entry` hereafter.

build(deps): bump github/super-linter from 3.17.1 to 4.0.2

Bumps [github/super-linter](https://github.com/github/super-linter) from 3.17.1 to 4.0.2.
- [Release notes](https://github.com/github/super-linter/releases)
- [Commits](https://github.com/github/super-linter/compare/v3.17.1...v4.0.2)
Signed-off-by: dependabot[bot] <support@github.com>

Instead of using copyrighted `strtod`, now use the public domain
implementation of `strtod` by Yasuhiro Matsumoto (@mattn). The function
has been renamed to `mrb_float_read()`; ref #5460 as well.

Fix typo in `src/fmt_fp.c` [ci skip]

Too many digits to fit in single precision float numbers, so that tests
fail when `MRB_USE_FLOAT32` defined.

The function was intended to be a utility function for `mruby-sprintf`.
The functionality was integrated into `sprintf.c`.

The original code can be found in `https://github.com/dhylands/format-float`.

Changes:
- support `double`
- support `#` (alt_form) modifier
- small refactoring

With major refactoring to prepare removing `snprintf(3) calls.

- `mrb_float_to_str()`
- `mrb_float_to_cstr()`

Both functions will be replaced to support new coming `format-float.c`.

`fmt_fp.c` does not support those specifiers. In addition, I believe no
one uses hexadecimal representation of float values.

Merge pull request #5456 from shuujii/support-load-64-bit-integer-from-mrb-format-with-MRB_32BIT-and-MRB_INT64

Support load 64-bit integer from mrb format with MRB_32BIT and MRB_INT64

This change allows `def hello = puts "Hello"` without parentheses.
This syntax has been introduced since Ruby3.1.

Consistent number conversion function names:
* `mrb_value` to immediate (C) value
  * `mrb_int()` -> `mrb_as_int()`
  * `mrb_to_flo()` -> `mrb_as_float()`
* `mrb_value` to `mrb_value` (converted)
  * `mrb_to_int()'
  * `mrb_Integer()` - removed
  * `mrb_Float()` -> `mrb_to_float`

Consistent function name (avoid `_flo` suffix):
* `mrb_div_flo()` -> `mrb_div_float`

To be consistent, functions with `rational_` prefix implements methods,
functions with `rat_` prefix are utility functions.

The Ruby version of `Array#rotate!` generated a rotated array and
replaced the receiver, but the C version rotates the receiver array
in-place. So the performance is improved a lot both in speed and memory
consumption. Look for the comments in `array.c` for the in-place rotating
algorithm, if you are interested.