09/05/2018

07:06 PM Ruby Feature #14473 (Closed): Add Range#subrange?

Applied in changeset trunk|r64640.
----------
range.c: Range#cover? accepts Range object. [Feature #14473]
* range.c (range_cover): add code for range argument.
If the argument is a Range, check it is or is not
covered b...

08/03/2018

03:04 AM Ruby Feature #14473: Add Range#subrange?

I am happy that match opinion with you.

I will commit based on your patch. It'll take a little while.

tarui (Masaya Tarui)

07/30/2018

07:31 AM Ruby Feature #14473: Add Range#subrange?

Thank you for your patch owst.
I reviewed it and feel the behavior below is strange.
~~~
$ ruby -e 'p (1..3.1).cover?(1...3)'
true
$ ruby -e 'p (1..3.1).cover?(1...4)'
Traceback (most recent call last):
1: from -e:1:i... tarui (Masaya Tarui)

07/18/2018

06:04 AM Ruby Feature #14473 (Assigned): Add Range#subrange?

tarui (Masaya Tarui)

07/12/2018

06:55 AM Ruby Feature #14473: Add Range#subrange?

As real-world use-case,
I want it(include,cover version) for guarding NArray's Index arguments.
NArray#[] can accept Integer and Range.
Currently, to limit the access range, We have to write complex code.
~~~ ruby
def check(idx,ra... tarui (Masaya Tarui)

06:40 AM Ruby Misc #14861: DevelopersMeeting20180718Japan

[Feature #14473] Add Range#subrange? (tarui)
tarui (Masaya Tarui)

05/11/2016

12:50 PM Ruby Bug #12274 (Closed): accessing to instance variable should be fast.

Applied in changeset r54976.
----------
* compile.c (iseq_compile_each): share InlineCache during same
instance variable accesses. Reducing memory consumption,
rasing cache hit rate and rasing branch prediction hit rate
are exp... tarui (Masaya Tarui)

04/13/2016

12:55 PM Ruby Bug #12274: accessing to instance variable should be fast.

> About the object check, is it not problematic to do ((struct RBasic*)obj)->klass if obj is a tagged integer (since klass is the second member, after flags)?
Thank you for pointing out.
I'll revive check.
tarui (Masaya Tarui)

01:32 AM Ruby Bug #12274: accessing to instance variable should be fast.

2016-04-13 5:41 GMT+09:00 <eregontp@gmail.com>:
> Issue #12274 has been updated by Benoit Daloze.
> ...
It is not for avoiding ID2SYM (In fact, it is calculated every time
:-), it is for sharing.
Please check the 0007 below
```
... tarui (Masaya Tarui)

01:13 AM Ruby Bug #12274: accessing to instance variable should be fast.

there are 2 parts of optimization.
* share inline cache between same symbol(at compile.c)
* inline fast pass only and cut useless check(RB_TYPE_P).(at insns.def)
We can skip st_lookup from the 2nd insns by sharing cache.
I... tarui (Masaya Tarui)