-
Notifications
You must be signed in to change notification settings - Fork 113
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
No DidYouMean constants in Travis CI #117
Comments
gem 'did_you_mean'
Update 2: Further investigation revealed that the first update may not work in some cases. Please see the comment below for more details. |
@yuki24, thank you for looking into my gem's travis log. I tried your suggestion, but now error happens before running the test.
In ruby 2.6, the error was:
So, I tried adding So, I have two confusions now:
|
@styd I'll look into your repository since it's open and I should be able to look into it. |
@yuki24, thank you for taking the trouble to help me. |
We've been having the same issue since we upgraded to Ruby 2.6 (and thus rubygems 3.0). Basically if the version defined in your The reason you'd want to have it locally installed (e.g. when using bundler's
You can reproduce this issue with the $ docker run -it --rm ruby:2.6.2-slim bash
$ gem list did_you_mean | grep did_you_mean
did_you_mean (1.3.0)
$ printf "source 'https://rubygems.org'\ngem 'did_you_mean', '1.3.0'" > Gemfile
$ bundle install --path bundle
Fetching gem metadata from https://rubygems.org/...
Resolving dependencies...
Using bundler 1.17.2
Using did_you_mean 1.3.0
Bundle complete! 1 Gemfile dependency, 2 gems now installed.
Bundled gems are installed into `./bundle` As you can see However if you try this with $ docker run -it --rm ruby:2.6.2-slim bash
$ gem list minitest | grep minitest
did_you_mean (5.11.3)
$ printf "source 'https://rubygems.org'\ngem 'minitest', '5.11.3'" > Gemfile
$ bundle install --path bundle
Fetching gem metadata from https://rubygems.org/.............
Resolving dependencies...
Using bundler 1.17.2
Fetching minitest 5.11.3
Installing minitest 5.11.3
Bundle complete! 1 Gemfile dependency, 2 gems now installed.
Bundled gems are installed into `./bundle` If you use bundler 2.0 it also does not install $ docker run -it --rm ruby:2.6.2-slim bash
$ gem install bundler -v 2.0.1
Fetching bundler-2.0.1.gem
Successfully installed bundler-2.0.1
1 gem installed
$ gem list did_you_mean | grep did_you_mean
did_you_mean (1.3.0)
$ printf "source 'https://rubygems.org'\ngem 'did_you_mean', '1.3.0'" > Gemfile
$ bundle install --path bundle
Fetching gem metadata from https://rubygems.org/...
Resolving dependencies...
Using bundler 2.0.1
Using did_you_mean 1.3.0
Bundle complete! 1 Gemfile dependency, 2 gems now installed.
Bundled gems are installed into `./bundle` However, downgrading $ docker run -it --rm ruby:2.6.2-slim bash
$ gem update --system 2.7.9
Updating rubygems-update
Fetching rubygems-update-2.7.9.gem
Successfully installed rubygems-update-2.7.9
Installing RubyGems 2.7.9
Bundler 1.16.6 installed
RubyGems 2.7.9 installed
Regenerating binstubs
------------------------------------------------------------------------------
RubyGems installed the following executables:
/usr/local/bin/gem
/usr/local/bin/bundle
RubyGems system software updated
$ gem list did_you_mean | grep did_you_mean
did_you_mean (1.3.0)
$ printf "source 'https://rubygems.org'\ngem 'did_you_mean', '1.3.0'" > Gemfile
$ bundle install --path bundle
Fetching gem metadata from https://rubygems.org/...
Resolving dependencies...
Using bundler 1.16.6
Fetching did_you_mean 1.3.0
Installing did_you_mean 1.3.0
Bundle complete! 1 Gemfile dependency, 2 gems now installed.
Bundled gems are installed into `/bundle` We have been using a downgraded $ docker run -it --rm ruby:2.6.2-slim bash
$ rm /usr/local/lib/ruby/gems/2.6.0/specifications/did_you_mean-1.3.0.gemspec
$ gem list did_you_mean | grep did_you_mean # returns nothing as it's no longer found
$ printf "source 'https://rubygems.org'\ngem 'did_you_mean', '1.3.0'" > Gemfile
$ bundle install --path bundle
Fetching gem metadata from https://rubygems.org/...
Resolving dependencies...
Using bundler 1.17.2
Fetching did_you_mean 1.3.0
Installing did_you_mean 1.3.0
Bundle complete! 1 Gemfile dependency, 2 gems now installed.
Bundled gems are installed into `./bundle` To me this looks like an issue with rubygems 3.0. But the weird thing is that it only seems to affect I have yet to try to track down what exactly causes the difference (either in |
@jeroenj Thanks for reporting. Something is definitely off here and I'll look into it. |
Okay, I thought there were two distinct issues here but turns out it's actually just one issue that awfully affects Bundler's
This commit in Rubygems rubygems/rubygems@42bba27#diff-05477d38d75c98664b0e7dec6eb73b45R823 seems to have introduced this behavior by accident, but we can't just revert it otherwise it may break other things. As @jeroenj pointed out, Rubygems 2.7.9 does not have this issue becaue the commit mentioned above is not part of the version 2.7.9. The other workaround I came up with is to specify # In Gemfile
gem 'did_you_mean' RUBYOPT='--disable-did_you_mean' bundle install --path bundle I think we should promote DYM up to a default gem to avoid confusions like this. It's very odd that it gets loaded with no clear notice despite the fact that it can be missing at any time. |
What an investigation. 👏 Thanks for looking into this this quickly. I’d indeed think promoting |
Sad to report back here, but bundler 2.0.2 broke the I was able to track it down to this change: https://github.com/bundler/bundler/pull/7067/files#diff-d08082204d5579c541f2c1206ffeb652R4 If I understand this (and the original issue (#117 (comment))) correctly this is caused by (the vendored version of) thor now requiring It seems that thor (master) no longer requires did you mean: rails/thor@8c57f1d. So a future version of bundler (with an updated version of thor) might no longer have that issue. So if I understand it all correctly none of this would be an issue of |
Deploying to production failed during `bundle install` because of an incompatibility between did_you_mean-1.3.0 + ruby-2.6 + rubygems-3.0 + bundler-2.0. ref: `https://github.com/yuki24/did_you_mean/issues/117#issuecomment-482565387`
Agreed having |
I'm also +1 to promote did_you_mean to default gems. |
Me too, for what's it's worth. |
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
I opened a PR for it, though I'm not sure if I did it correctly. Feel free to leave comments over there ^^^^ |
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
At the moment, there are some problems with regard to bundler + did_you_mean because of did_you_mean being a bundled gem. Since the vendored version of thor inside bundler and ruby itself explicitly requires did_you_mean, it can become difficult to load it when using Bundler.setup. See this issue: ruby/did_you_mean#117 (comment) for more details.
closing in favor of ruby/ruby#2689. |
…ecification.stubs_for` The rationale is that: * The change has caused realworld issues. See for example ruby/did_you_mean#117 and specifically [this comment](ruby/did_you_mean#117 (comment)) for a great explanation of the issue it caused for `did_you_mean`. * The change also causes problems for our development workflows. For example, because of it, our `bundler` specs cannot currently be run with `bin/rake` and we have to use `bin/rspec` or `bin/parallel_spec` directly. The explanation for this is: - Our specs install test dependencies to `tmp` before running specs. - `rake` is one of these test dependencies. - Before installing each test dependency, we check whether it has matching installed specs: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/bundler/spec/support/rubygems_ext.rb#L109-L114. - Normally, if `rake` has not yet been installed to `tmp`, this check fails and `rake` is installed, but since the loaded specs are now added to `Gem::Specification.stubs` and `rake`'s specification _is_ loaded because we're running through `bin/rake`, the check incorrectly assumes that `rake` is already installed to `tmp` and skips installation. - At a later point the specs check whether `rake` is actually installed and fail if it's not: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/bundler/spec/support/builders.rb#L372-L383 Essentially, both of the issues are the same. If at runtime we change the location of gems, we'll _want_ to not consider loaded specifications when dealing with the new gem location, because the loaded specifications have not been loaded from there. Loaded specifications is something different from installed stub specifications and those should not be mixed. The PR still seemed to have fixed an issue, so I did my archaeology job and investigated the original issue to double check if reverting is ok. The logs for the original error can be found here: https://ci.appveyor.com/project/rubygems/rubygems/build/1172/job/ogubyucpljcv22ux. So I installed ruby 2.4.4, checked out the commit reference before the offending PR, and the exact error reproduced. 🎉 ``` $ rake test /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:231:in `search_for': Unable to resolve dependency: user requested 'bundler (= 1.16.2)' (Gem::UnsatisfiableDependencyError) from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:283:in `block in sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `each' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `sort_by' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `with_index' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:52:in `block in sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:69:in `with_no_such_dependency_error_handling' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:51:in `sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:165:in `initial_state' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:106:in `start_resolution' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:64:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolver.rb:42:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:188:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/request_set.rb:396:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/request_set.rb:408:in `resolve_current' from /home/deivid/Code/rubygems/lib/rubygems.rb:243:in `finish_resolve' from /home/deivid/Code/rubygems/lib/rubygems/rdoc.rb:13:in `<top (required)>' from /home/deivid/Code/rubygems/lib/rubygems/core_ext/kernel_require.rb:54:in `require' from /home/deivid/Code/rubygems/lib/rubygems/core_ext/kernel_require.rb:54:in `require' from /home/deivid/Code/rubygems/lib/rubygems/test_case.rb:1563:in `<top (required)>' from /home/deivid/Code/rubygems/test/rubygems/test_bundled_ca.rb:2:in `require' from /home/deivid/Code/rubygems/test/rubygems/test_bundled_ca.rb:2:in `<top (required)>' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:15:in `require' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:15:in `block in <main>' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:4:in `select' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:4:in `<main>' rake aborted! Command failed with status (1) Tasks: TOP => test ``` Now the explanation of the error: * Rubygems base `TestCase` class requires `bundler` because some tests use `bundler`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/test_case.rb#L26 * That `require` (our custom rubygems require) would activate the default bundler spec (1.16.1 for ruby 2.4.4) but then overwrite it with a 1.16.2 version (the locally provided bundler those days) due to [this old hack](https://github.com/rubygems/bundler/blob/9f7bf0ac3ab8d995e3a274cec3c292a5203f4534/lib/bundler/version.rb#L7-L23). * Rubygems base `TestCase` class requires `rubygems/rdoc`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/test_case.rb#L1536 * And that file ends up calling `Gem.finish_resolve`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/rdoc.rb#L13 * `Gem.finish_resolve` adds the currently loaded specs to the resolution: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems.rb#L235 * That means it would try to resolve bundler 1.16.2, but no specification for that version was installed since the default was 1.16.1. That explains why upgrading to rubygems 2.7.7 fixed the issue, since it provided bundler 1.16.2 by default so there was not bundler version discrepancy. After understanding the error, I conclude that: * Only this part of the original patch was actually needed to resolve the error, not any of the changes in `Gem::Specification.stubs` and `Gem::Specification.stubs_for`: ```diff diff --git a/lib/rubygems/test_case.rb b/lib/rubygems/test_case.rb index f1cd3d274c..92c848e870 100644 --- a/lib/rubygems/test_case.rb +++ b/lib/rubygems/test_case.rb @@ -13,6 +13,15 @@ else require 'rubygems' end +# If bundler gemspec exists, add to stubs +bundler_gemspec = File.expand_path("../../../bundler/bundler.gemspec", __FILE__) +if File.exist?(bundler_gemspec) + Gem::Specification.dirs.unshift File.dirname(bundler_gemspec) + Gem::Specification.class_variable_set :@@stubs, nil + Gem::Specification.stubs + Gem::Specification.dirs.shift +end + begin gem 'minitest' rescue Gem::LoadError ``` So, I propose to revert adding loaded specification to `Gem::Specification.stubs` and `Gem::Specification.stubs_for` because I think it's safe, it fixes the issues caused by their addition, and it simplifies `Gem::Specification` code, which is already complicated enough.
…ecification.stubs_for` The rationale is that: * The change has caused realworld issues. See for example ruby/did_you_mean#117 and specifically [this comment](ruby/did_you_mean#117 (comment)) for a great explanation of the issue it caused for `did_you_mean`. * The change also causes problems for our development workflows. For example, because of it, our `bundler` specs cannot currently be run with `bin/rake` and we have to use `bin/rspec` or `bin/parallel_spec` directly. The explanation for this is: - Our specs install test dependencies to `tmp` before running specs. - `rake` is one of these test dependencies. - Before installing each test dependency, we check whether it has matching installed specs: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/bundler/spec/support/rubygems_ext.rb#L109-L114. - Normally, if `rake` has not yet been installed to `tmp`, this check fails and `rake` is installed, but since the loaded specs are now added to `Gem::Specification.stubs` and `rake`'s specification _is_ loaded because we're running through `bin/rake`, the check incorrectly assumes that `rake` is already installed to `tmp` and skips installation. - At a later point the specs check whether `rake` is actually installed and fail if it's not: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/bundler/spec/support/builders.rb#L372-L383 Essentially, both of the issues are the same. If at runtime we change the location of gems, we'll _want_ to not consider loaded specifications when dealing with the new gem location, because the loaded specifications have not been loaded from there. Loaded specifications is something different from installed stub specifications and those should not be mixed. The PR still seemed to have fixed an issue, so I did my archaeology job and investigated the original issue to double check if reverting is ok. The logs for the original error can be found here: https://ci.appveyor.com/project/rubygems/rubygems/build/1172/job/ogubyucpljcv22ux. So I installed ruby 2.4.4, checked out the commit reference before the offending PR, and the exact error reproduced. 🎉 ``` $ rake test /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:231:in `search_for': Unable to resolve dependency: user requested 'bundler (= 1.16.2)' (Gem::UnsatisfiableDependencyError) from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:283:in `block in sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `each' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `sort_by' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `with_index' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:52:in `block in sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:69:in `with_no_such_dependency_error_handling' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:51:in `sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:165:in `initial_state' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:106:in `start_resolution' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:64:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolver.rb:42:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:188:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/request_set.rb:396:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/request_set.rb:408:in `resolve_current' from /home/deivid/Code/rubygems/lib/rubygems.rb:243:in `finish_resolve' from /home/deivid/Code/rubygems/lib/rubygems/rdoc.rb:13:in `<top (required)>' from /home/deivid/Code/rubygems/lib/rubygems/core_ext/kernel_require.rb:54:in `require' from /home/deivid/Code/rubygems/lib/rubygems/core_ext/kernel_require.rb:54:in `require' from /home/deivid/Code/rubygems/lib/rubygems/test_case.rb:1563:in `<top (required)>' from /home/deivid/Code/rubygems/test/rubygems/test_bundled_ca.rb:2:in `require' from /home/deivid/Code/rubygems/test/rubygems/test_bundled_ca.rb:2:in `<top (required)>' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:15:in `require' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:15:in `block in <main>' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:4:in `select' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:4:in `<main>' rake aborted! Command failed with status (1) Tasks: TOP => test ``` Now the explanation of the error: * Rubygems base `TestCase` class requires `bundler` because some tests use `bundler`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/test_case.rb#L26 * That `require` (our custom rubygems require) would activate the default bundler spec (1.16.1 for ruby 2.4.4) but then overwrite it with a 1.16.2 version (the locally provided bundler those days) due to [this old hack](https://github.com/rubygems/bundler/blob/9f7bf0ac3ab8d995e3a274cec3c292a5203f4534/lib/bundler/version.rb#L7-L23). * Rubygems base `TestCase` class requires `rubygems/rdoc`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/test_case.rb#L1536 * And that file ends up calling `Gem.finish_resolve`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/rdoc.rb#L13 * `Gem.finish_resolve` adds the currently loaded specs to the resolution: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems.rb#L235 * That means it would try to resolve bundler 1.16.2, but no specification for that version was installed since the default was 1.16.1. That explains why upgrading to rubygems 2.7.7 fixed the issue, since it provided bundler 1.16.2 by default so there was not bundler version discrepancy. After understanding the error, I conclude that: * Only this part of the original patch was actually needed to resolve the error, not any of the changes in `Gem::Specification.stubs` and `Gem::Specification.stubs_for`: ```diff diff --git a/lib/rubygems/test_case.rb b/lib/rubygems/test_case.rb index f1cd3d274c..92c848e870 100644 --- a/lib/rubygems/test_case.rb +++ b/lib/rubygems/test_case.rb @@ -13,6 +13,15 @@ else require 'rubygems' end +# If bundler gemspec exists, add to stubs +bundler_gemspec = File.expand_path("../../../bundler/bundler.gemspec", __FILE__) +if File.exist?(bundler_gemspec) + Gem::Specification.dirs.unshift File.dirname(bundler_gemspec) + Gem::Specification.class_variable_set :@@stubs, nil + Gem::Specification.stubs + Gem::Specification.dirs.shift +end + begin gem 'minitest' rescue Gem::LoadError ``` So, I propose to revert adding loaded specification to `Gem::Specification.stubs` and `Gem::Specification.stubs_for` because I think it's safe, it fixes the issues caused by their addition, and it simplifies `Gem::Specification` code, which is already complicated enough.
….stubs` and `Gem::Specification.stubs_for` The rationale is that: * The change has caused realworld issues. See for example ruby/did_you_mean#117 and specifically [this comment](ruby/did_you_mean#117 (comment)) for a great explanation of the issue it caused for `did_you_mean`. * The change also causes problems for our development workflows. For example, because of it, our `bundler` specs cannot currently be run with `bin/rake` and we have to use `bin/rspec` or `bin/parallel_spec` directly. The explanation for this is: - Our specs install test dependencies to `tmp` before running specs. - `rake` is one of these test dependencies. - Before installing each test dependency, we check whether it has matching installed specs: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/bundler/spec/support/rubygems_ext.rb#L109-L114. - Normally, if `rake` has not yet been installed to `tmp`, this check fails and `rake` is installed, but since the loaded specs are now added to `Gem::Specification.stubs` and `rake`'s specification _is_ loaded because we're running through `bin/rake`, the check incorrectly assumes that `rake` is already installed to `tmp` and skips installation. - At a later point the specs check whether `rake` is actually installed and fail if it's not: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/bundler/spec/support/builders.rb#L372-L383 Essentially, both of the issues are the same. If at runtime we change the location of gems, we'll _want_ to not consider loaded specifications when dealing with the new gem location, because the loaded specifications have not been loaded from there. Loaded specifications is something different from installed stub specifications and those should not be mixed. The PR still seemed to have fixed an issue, so I did my archaeology job and investigated the original issue to double check if reverting is ok. The logs for the original error can be found here: https://ci.appveyor.com/project/rubygems/rubygems/build/1172/job/ogubyucpljcv22ux. So I installed ruby 2.4.4, checked out the commit reference before the offending PR, and the exact error reproduced. 🎉 ``` $ rake test /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:231:in `search_for': Unable to resolve dependency: user requested 'bundler (= 1.16.2)' (Gem::UnsatisfiableDependencyError) from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:283:in `block in sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `each' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `sort_by' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `with_index' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:277:in `sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:52:in `block in sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:69:in `with_no_such_dependency_error_handling' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/delegates/specification_provider.rb:51:in `sort_dependencies' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:165:in `initial_state' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:106:in `start_resolution' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolution.rb:64:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/resolver/molinillo/lib/molinillo/resolver.rb:42:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/resolver.rb:188:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/request_set.rb:396:in `resolve' from /home/deivid/Code/rubygems/lib/rubygems/request_set.rb:408:in `resolve_current' from /home/deivid/Code/rubygems/lib/rubygems.rb:243:in `finish_resolve' from /home/deivid/Code/rubygems/lib/rubygems/rdoc.rb:13:in `<top (required)>' from /home/deivid/Code/rubygems/lib/rubygems/core_ext/kernel_require.rb:54:in `require' from /home/deivid/Code/rubygems/lib/rubygems/core_ext/kernel_require.rb:54:in `require' from /home/deivid/Code/rubygems/lib/rubygems/test_case.rb:1563:in `<top (required)>' from /home/deivid/Code/rubygems/test/rubygems/test_bundled_ca.rb:2:in `require' from /home/deivid/Code/rubygems/test/rubygems/test_bundled_ca.rb:2:in `<top (required)>' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:15:in `require' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:15:in `block in <main>' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:4:in `select' from /home/deivid/.rbenv/versions/2.4.4/lib/ruby/gems/2.4.0/gems/rake-12.0.0/lib/rake/rake_test_loader.rb:4:in `<main>' rake aborted! Command failed with status (1) Tasks: TOP => test ``` Now the explanation of the error: * Rubygems base `TestCase` class requires `bundler` because some tests use `bundler`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/test_case.rb#L26 * That `require` (our custom rubygems require) would activate the default bundler spec (1.16.1 for ruby 2.4.4) but then overwrite it with a 1.16.2 version (the locally provided bundler those days) due to [this old hack](https://github.com/rubygems/bundler/blob/9f7bf0ac3ab8d995e3a274cec3c292a5203f4534/lib/bundler/version.rb#L7-L23). * Rubygems base `TestCase` class requires `rubygems/rdoc`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/test_case.rb#L1536 * And that file ends up calling `Gem.finish_resolve`: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems/rdoc.rb#L13 * `Gem.finish_resolve` adds the currently loaded specs to the resolution: https://github.com/rubygems/rubygems/blob/2bbcdcde08b90d4ef03da8fb1f7a8a3313e13bb8/lib/rubygems.rb#L235 * That means it would try to resolve bundler 1.16.2, but no specification for that version was installed since the default was 1.16.1. That explains why upgrading to rubygems 2.7.7 fixed the issue, since it provided bundler 1.16.2 by default so there was not bundler version discrepancy. After understanding the error, I conclude that: * Only this part of the original patch was actually needed to resolve the error, not any of the changes in `Gem::Specification.stubs` and `Gem::Specification.stubs_for`: ```diff diff --git a/lib/rubygems/test_case.rb b/lib/rubygems/test_case.rb index f1cd3d274c..92c848e870 100644 --- a/lib/rubygems/test_case.rb +++ b/lib/rubygems/test_case.rb @@ -13,6 +13,15 @@ else require 'rubygems' end +# If bundler gemspec exists, add to stubs +bundler_gemspec = File.expand_path("../../../bundler/bundler.gemspec", __FILE__) +if File.exist?(bundler_gemspec) + Gem::Specification.dirs.unshift File.dirname(bundler_gemspec) + Gem::Specification.class_variable_set :@@stubs, nil + Gem::Specification.stubs + Gem::Specification.dirs.shift +end + begin gem 'minitest' rescue Gem::LoadError ``` So, I propose to revert adding loaded specification to `Gem::Specification.stubs` and `Gem::Specification.stubs_for` because I think it's safe, it fixes the issues caused by their addition, and it simplifies `Gem::Specification` code, which is already complicated enough. rubygems/rubygems@5269cd617c
Hi @yuki24!
I was trying to use DYM for my gem's KeyError. So, I added this code:
It's working in my local computer, but failed in Travis CI, saying
uninitialized constant DidYouMean::SPELL_CHECKERS
.So I added this code to print DidYouMean methods and constants inside the module:
To my surprise, it was printing these lines in Travis CI:
This only happens in CI. It's fine in my local computer.
Do you know what causes this?
The text was updated successfully, but these errors were encountered: