[#92891] Question: ruby 2.7.0-preview1 also upgrades bundler to 2.1.0.pre.1? — Al Snow <jasnow@...>
Tried the new 2.7.0-preview1 upgrade to Ruby and see that bundler is also u=
5 messages
2019/05/30
[#92892] Re: Question: ruby 2.7.0-preview1 also upgrades bundler to 2.1.0.pre.1?
— SHIBATA Hiroshi <hsbt@...>
2019/05/30
Bundler 2.1.0.pree.1 is the expected version.
[#92893] Re: Question: ruby 2.7.0-preview1 also upgrades bundler to 2.1.0.pre.1?
— Al Snow <jasnow@...>
2019/05/30
[ruby-core:92798] [Ruby trunk Feature#15868] Implement `File.absolute_path?`
From:
shevegen@...
Date:
2019-05-23 13:32:19 UTC
List:
ruby-core #92798
Issue #15868 has been updated by shevegen (Robert A. Heiler). I have no particular pro/con opinion on the suggestion itself, but I should= state that since Pathname was mentioned - I myself use File.* related methods exclusively. I used to use pathname in th= e past but I sort of gave up on it eventually, primarily because File.* seems to work just fine (and FileUtils); partially= also because of having to do require 'pathname' and my general impression that pathname is more clumsy to work with than Fi= le (but this is a subjective opinion). So this comment from me here is mostly about pathname to File.* related methods - t= o me, in the current use cases, I could happily use File.* but it would be unlikely for me to go back to when I used to use= pathname. So when File.* and Pathname is compared, I would like to point out that the= way how ruby users use either of them may be different rather than equal/synonymous. (I am not sure if this comment i= s very useful but I wanted to point this out at the least once.) Having said that, I personally have not had a need for File.absolute_path?(= ) yet. I use File.absolute_path() a lot, though, often because I have to deal/handle symlinks through ruby, includin= g re-symlinking, removing old symlinks etc.... I can say that this may be useful, but personally I honestly have not yet h= ad a need for it. As for the documentation - I guess it could easily be made more accurate or= slightly lengthier in this case; I assume that this may depend a lot on the operating system, since windows is mentio= ned here. Perhaps the documentation can add a sentence about present windows support - I have worked only very litt= le with windows, but the ruby code I write tends to work very well on windows out of the box, without even me trying t= o do much at all (even ruby-gtk stuff works on windows, which is pretty cool). ---------------------------------------- Feature #15868: Implement `File.absolute_path?` https://bugs.ruby-lang.org/issues/15868#change-78172 * Author: deivid (David Rodr=EDguez) * Status: Open * Priority: Normal * Assignee: = * Target version: = ---------------------------------------- Currently there's no way to check whether a path is absolute or not in a wa= y that works accross OSs. The pathname library has the #absolute? method, b= ut that only checks whether the path starts with a slash, which is not appr= opriate for Windows. I thought of reimplementing it as something like File.absolute_path(self) = =3D=3D self, but that would mean accessing the filesystem, which I don't th= ink we want here. I also thought of implementing the "windows letter checks" in the pathname'= s library, but then I saw that those are already implemented in file.c, so = I thought it would be a good idea to expose those. So I propose to add File= .absolute_path? for this. If this is accepted, I can do a follow-up PR to change Pathname#absolute? t= o delegate to File.absolute_path?. What do you think? I attach a patch to add `File.absolute_path?` here (I also opened a PR on G= ithub: https://github.com/ruby/ruby/pull/2198). = ---Files-------------------------------- 0001-Add-File.absolute_path.patch (3.24 KB) -- = https://bugs.ruby-lang.org/ Unsubscribe: <mailto:ruby-core-request@ruby-lang.org?subject=3Dunsubscribe> <http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>