The addresses for packed inline symbols reference `mrb->symbuf` that
could be overridden by the later call of `mrb_sym2name_len`. Since
file names in call stack information are kept as symbols, keeping the
address in the C structures could cause problems like #4342.

This changes small incompatible changes in function prototypes:
* `mrb_parser_get_filename`: return value changed to `mrb_sym`.
* `mrb_debug_get_filename`: add `mrb_state*` as a first argument.
* `mrb_debug_get_line`: ditto.

I believe above functions are almost internal, and no third-party
mrbgem use them.

Fix warning: '*' interpreted as argument prefix

Add `pass` and `flunk` to `test/assert.rb`

Fix missing assertions in `mruby-math` test

fixes build on OpenBSD.

fixes build on OpenBSD.

Use `mrb_sym2str` in implementation of `Symbol#to_s`

* rename `sym2name` to `sym2name_len`.
* `MRB_API` -> `static`

Remove unused variable in `each_backtrace()`

For short symbols with alpha numeric characters, `mrb_sym2name_len()`
returns the same buffer `mrb->symbuf`. Some occasion, we forget the fact
that the second call could overwrite the result of first call of the
function.

We have prepared the static function `sym2name()` which specifies the
buffer region for inline packed symbols and use the function in
`mrb_sym_to_s`.

Fix dealing with infinity and NaN in `test/assert.rb:assert_float`

`assert_float` is always passed when expected value and/or actual value are
infinity or NaN. This behavior seems unintentional.

Before this patch:

    assert_float(Float::INFINITY, 1.0)  #=> pass
    assert_float(-Float::INFINITY, 1)   #=> pass
    assert_float(1, 1/0)                #=> pass
    assert_float(1, -1/0)               #=> pass
    assert_float(1.0, Float::NAN)       #=> pass
    assert_float(Float::NAN, 1)         #=> pass

After this patch:

    assert_float(Float::INFINITY, 1.0)  #=> fail: Expected 1.0 to be Infinity.
    assert_float(-Float::INFINITY, 1)   #=> fail: Expected 1 to be -Infinity.
    assert_float(1, 1/0)                #=> fail: Expected Infinity to be 1.
    assert_float(1, -1/0)               #=> fail: Expected -Infinity to be 1.
    assert_float(1.0, Float::NAN)       #=> fail: Expected NaN to be 1.0.
    assert_float(Float::NAN, 1)         #=> fail: Expected 1 to be NaN.

Functions to add prototypes to headers:
* mrb_ary_splice()
* mrb_notimplement()
* mrb_vformat()
* mrb_cstr_to_dbl()
* mrb_cstr_to_inum()

Functions to be made `static` (`MRB_API` was not needed):
* mrb_mod_module_function()
* mrb_obj_hash()
* mrb_str_len_to_inum()

Functions to remove `MRB_API` from definitions (referenced from within `libmruby`):
* mrb_mod_cv_defined()
* mrb_mod_cv_get()
* mrb_f_send()

Remove implementation of `Symbol#===`

For reducing program size.

Fix arguments spec in `src/proc.c`

Refactor `t_print` for test

Simplify `assert_step` in `test/t/numeric.rb`

Merge pull request #4334 from shuujii/remove-redundant-content-in-assertion-failure-message-and-diff

Remove redundant content in assertion failure message and diff

Fix `Float#eql?`

Based on minitest RubyGem.

Example of before this patch:

    - Assertion[1] Failed: Expected 1 to be 2
       Expected: 2
         Actual: 1
    - Assertion[2] Failed: Expected [1, 3] to include 2
       Collection: [1, 3]
           Object: 2

Example of after this patch:

    - Assertion[1]
       Expected: 2
         Actual: 1
    - Assertion[2]
       Expected [1, 3] to include 2.

Merge pull request #4333 from shuujii/use-FrozenError-instead-of-RuntimeError-in-frozen-obj-mod-test

Use `FrozenError` instead of `RuntimeError` in frozen object modification test

Use Rake instead of MiniRake on AppVeyor