[DOC] Adjust some new features wording/examples. #9183
Conversation
variable.c
Outdated
| * | ||
| * c.new # => #<MyClass(with description):0x0....> | ||
| * c.instance_method(:foo) # => #<UnboundMethod: MyClass(with description)#foo() ...> | ||
| * c.new.method(:foo) #<Method: MyClass(with description)#foo() ...> |
There was a problem hiding this comment.
Missing # => in this line
There was a problem hiding this comment.
Do we need an example regarding the method? It seems trivial. I don't think the code examples in the documentation should be more complicated than necessary.
There was a problem hiding this comment.
@mame My point was to demonstrate that all related objects are affected (as an emphasis on "why would this be useful").
Though we probably can just leave one of the two lines (instance_method vs method), did you mean that?
|
FYI: |
zverok
changed the title
zverok
force-pushed
the
adjust-new-features-docs
branch
from
3457741
to
a3473f4
Compare
|
@jeremyevans I tried to address all of your concerns. I also added 3 more small commits (listed under |
eval.c
Outdated
| * | ||
| * module M | ||
| * refine String do | ||
| * def indent(level) |
There was a problem hiding this comment.
Best remove the method definition here, it doesn't affect the result and is unrelated to the method being documented.
variable.c
Outdated
| @@ -189,15 +189,25 @@ is_constant_path(VALUE name) | |||
| * m.set_temporary_name(nil) # => #<Module:0x0000000102c68f38> | |||
| * m.name #=> nil | |||
| * | |||
| * n = Module.new | |||
| * n.set_temporary_name("fake_name") | |||
| * c = Class.new do | |||
There was a problem hiding this comment.
I'm with @mame here, the method already has significant examples, I don't think we need to add to them.
There was a problem hiding this comment.
OK. My point here was to demonstrate one of the cases where redefining Module#name and/or Module#inspect wouldn't fix the related object's #inspect to be meaningful (and adds to a justification "why somebody needs a new method to set a pretend-name"). But probably, it doesn't look important enough anyway.
zverok
force-pushed
the
adjust-new-features-docs
branch
from
a3473f4
to
3af3e18
Compare
|
@jeremyevans I think I fixed all the comments. So I rebased it from Thanks for the reviews! |
|
@jeremyevans Do you mean we need to have just one commit with all the changes (because I tried to structure it in a "commit-per-class" fashion)? |
|
Yes. |
(Part of working on this year's changelog.)
Some adjustments of the wording of the new features docs.
The changes are relatively small, so I packed them into one PR. If this is unacceptable, I can split in several (every change is its own commit), but I hope that it can be reviewed easily.
The changes in this PR:
:rescueevent.>>usage from class header example (as it is deprecated, we shoudn't encourage it with the initial example).Added Dec 12:
#refined_class#fchdirat all)