summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* [DSL] Suggest updating bundler on invalid optionsseg-unknown-keysSamuel Giddins2016-09-021-1/+2
|
* Auto merge of #4939 - JuanitoFatas:patch/mirror-equal, r=segiddinsHomu2016-09-021-0/+10
|\ | | | | | | | | | | Add a spec for Bundler::Settings::Mirror#== This PR adds a spec for `Mirror#==` method.
| * Add a spec for Bundler::Settings::Mirror#==JuanitoFatas2016-08-311-0/+10
| |
* | Auto merge of #4938 - JuanitoFatas:patch/settings-spec, r=segiddinsHomu2016-09-012-2/+13
|\ \ | | | | | | | | | | | | | | | | | | | | | Add some specs for Bundler::Settings#temporary This PR adds a spec and moved an alias in Settings class. followup of #4917.
| * | Add a spec for Bundler::Settings#temporaryJuanitoFatas2016-08-311-0/+12
| | |
| * | Move the alias to where it should beJuanitoFatas2016-08-311-2/+1
| |/
* | Auto merge of #4937 - bundler:default-travis-emails, r=segiddinsHomu2016-09-011-5/+0
|\ \ | |/ |/| | | | | | | Emails to the committer are better than emails to me :+1: None
| * back to default emailsdefault-travis-emailsAndre Arko2016-08-311-5/+0
|/
* Auto merge of #4917 - bundler:seg-settings-temporary, r=indirectHomu2016-08-302-6/+22
|\ | | | | | | [Settings] Allow temporarily overriding settings in-memory
| * [Settings] Allow temporarily overriding settings in-memoryseg-settings-temporarySamuel Giddins2016-08-262-6/+22
| |
* | Auto merge of #4921 - JuanitoFatas:doc/development-pry-byebug, r=indirectHomu2016-08-301-0/+2
|\ \ | | | | | | | | | | | | | | | | | | | | | Doc: Add a note on how to dive into the code with Pry Follow up of https://github.com/bundler/bundler/pull/4919#issuecomment-242706463. [Preview DEVELOPMENT.md](https://github.com/JuanitoFatas/bundler/blob/6667a9d9a1fd7f83b4d262ff00b933f8574a974f/DEVELOPMENT.md#development-setup)
| * | Add a note on how to dive into the code with Pry [ci skip]JuanitoFatas2016-08-261-0/+2
| | | | | | | | | | | | Follow up of https://github.com/bundler/bundler/pull/4919#issuecomment-242706463
* | | Auto merge of #4928 - bundler:seg-update-compact-index, r=segiddinsHomu2016-08-304-12/+17
|\ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | Update specs to use compact_index 0.11 Also removed the penalty for adding requirements to dependencies! \c @indirect
| * | | Speed up test gem installationseg-update-compact-indexSamuel Giddins2016-08-291-5/+9
| | | |
| * | | Update specs to use compact_index 0.11Samuel Giddins2016-08-294-7/+8
|/ / /
* | | Auto merge of #4923 - bundler:seg-apple-git-support, r=indirectHomu2016-08-282-0/+10
|\ \ \ | |/ / |/| | | | | [Env] Support running `bundle env` when not inside a bundle
| * | [Env] Support running `bundle env` when not inside a bundleseg-apple-git-supportSamuel Giddins2016-08-262-0/+10
|/ /
* | Auto merge of #4491 - bundler:seg-exec-file--zsh, r=indirectHomu2016-08-261-0/+36
|\ \ | |/ |/| | | | | | | [Exec] Add specs for $0 and __FILE__ when loaded Tests for #4488
| * [Exec] Add specs for $0 and __FILE__ when loadedseg-exec-file--zshSamuel Giddins2016-08-231-0/+36
| |
* | Auto merge of #4836 - bundler:seg-resolve-for-specific-platforms, r=indirectHomu2016-08-2627-60/+290
|\ \ | | | | | | | | | | | | | | | | | | | | | Resolve for specific platforms Closes #4295. This will require adding a bunch of tests, as well as figuring out how to put this new behavior behind a feature flag (thus fixing all of the existing tests).
| * | [Plugin] Don’t print GemfileErrors twiceseg-resolve-for-specific-platformsSamuel Giddins2016-08-251-1/+3
| | |
| * | [Definition] Validate platform in addition to rubySamuel Giddins2016-08-2511-10/+31
| | | | | | | | | | | | | | | # Conflicts: # lib/bundler/cli/doctor.rb
| * | [Definition] Only add current platform if not frozenSamuel Giddins2016-08-251-3/+7
| | |
| * | [LockfileParser] Ensure specs are consistently orderedSamuel Giddins2016-08-252-1/+15
| | |
| * | [SpecSet] Fix #for when installing with --deploymentSamuel Giddins2016-08-251-10/+13
| | |
| * | [Source] Always print the platform when installingSamuel Giddins2016-08-252-2/+2
| | |
| * | Add a spec for platform_specific resolution / runtimeSamuel Giddins2016-08-251-0/+95
| | |
| * | Automatically add platform to version in the spec’s buildersSamuel Giddins2016-08-251-1/+4
| | |
| * | Add a platform helper for x64_macSamuel Giddins2016-08-251-0/+4
| | |
| * | Add a helper to the_bundler to get a parsed lockfileSamuel Giddins2016-08-251-0/+5
| | |
| * | Refactor best-platform matchingSamuel Giddins2016-08-254-20/+53
| | |
| * | [Resolver] Remove required_by attribute from SpecGroupSamuel Giddins2016-08-251-2/+1
| | |
| * | [LazySpecification] Only materialize for the current platform whenSamuel Giddins2016-08-251-1/+2
| | | | | | | | | | | | :specific_platform is enabled
| * | Resolve for specific platformsSamuel Giddins2016-08-259-30/+76
|/ /
* | Auto merge of #4654 - bundler:aa-ruby-version-conflict-message, r=indirectHomu2016-08-258-40/+146
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add resolver-style error messages for required Ruby conflicts *Proposal:* When a gem in the Gemfile requires a Ruby version different than the Gemfile's `ruby`, print a conflict error message that looks more or less exactly like any other resolution conflict error. *Current behavior:* On `master` today, Bundler will ignore the gem's required_ruby_version until install-time, and raise an exception like "require_ruby-1.0 requires ruby version > 9000, which is incompatible with the current version`. After #4650 has been merged, Bundler will correctly reject gems that require a different version of Ruby, and instead print the error "Could not find `required_ruby` in any of the sources in your Gemfile". *Rationale for change:* I believe that error message is surprising and misleading. The source _does_ contain `required ruby`, and I think we'll get bug reports about how our error message is wrong.
| * | Update resolving specs for required ruby version refactoraa-ruby-version-conflict-messageSamuel Giddins2016-08-232-2/+5
| | |
| * | [Resolver] Remove ruby_version attributeSamuel Giddins2016-08-232-15/+18
| | |
| * | [Definition] Improve error message when locked ruby version fails to parseSamuel Giddins2016-08-231-2/+3
| | |
| * | Backport Requirement#exact? ffor RG < ~2.1Samuel Giddins2016-08-231-0/+5
| | |
| * | Update spec for consistency with how Gem.ruby_version is usedSamuel Giddins2016-08-231-1/+1
| | |
| * | [Definition] Handle non-exact ruby version requirementsSamuel Giddins2016-08-232-31/+61
| | |
| * | Fix resolving when the gemfile has a ruby patch requirementSamuel Giddins2016-08-233-5/+22
| | |
| * | [Definition] Only use RubyVersion objects to determine resolution behaviorSamuel Giddins2016-08-231-9/+17
| | |
| * | Use the resolver to resolve for ruby / rubygems versionsSamuel Giddins2016-08-233-19/+46
| | |
| * | test for a diagnostic error on ruby version failAndre Arko2016-08-231-2/+14
| |/
* | Auto merge of #4898 - JuanitoFatas:feature/generated-readme-github-url, ↵Homu2016-08-253-2/+32
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | r=segiddins Fill in git username to the generated gem's contributing section of README.md This Pull Request changes the GitHub URL in the Contributing section of README.md from generated gem (e.g. `bundle gem foo`). Why: I found myself change `[USERNAME]` to my git username so many times and I also help [some](https://github.com/suzuki86/rhymer/pull/1/files) [people](https://github.com/fastly/blockbuster/pull/12/files) [fix](https://github.com/abookyun/taiwan_validator/pull/2/files) [this](https://github.com/mkhairi/materialize-sass/pull/62/files). So I think this should be filled when new gem is generated. When git username not set, falls back to `[USERNAME]`. I don't really know how to make tests DRY, please advise, thank you.
| * | Fill in git username to the generated github urlJuanitoFatas2016-08-203-2/+32
| | | | | | | | | | | | | | | | | | Most of the developers should already properly configured "git config user.name". This save time and avoid generated README.md with wrong GitHub url.
* | | Auto merge of #4624 - chrismo:openssl_version_in_bundle_env, r=segiddinsHomu2016-08-252-0/+2
|\ \ \ | | | | | | | | | | | | Add OpenSSL::OPENSSL_VERSION to `bundle env`
| * | | Add OpenSSL::OPENSSL_VERSION to `bundle env`chrismo2016-08-092-0/+2
| | | |
* | | | Auto merge of #4878 - Zorbash:binstubs-standalone-flag, r=segiddinsHomu2016-08-241-1/+1
|\ \ \ \ | |_|_|/ |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Change standalone flag of binstubs cmd to boolean This flag was introduced in #4610, declared with the same code as the flag found in the `install` command. In the case of the `install` command, it makes sense to be an `Array` option, since one can specify groups. This doesn't hold true for the `binstubs` cmd where `--standalone` just stands to signify whether the generated stubs will require the Bundler runtime or not. In the current implementation even passing groups with the flag doesn't have any affect.