| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
| |
|
|\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
r=colby-swandale
check permissions in doctor command
Thanks so much for the contribution!
To make reviewing this PR a bit easier, please fill out answers to the following questions.
### What was the end-user problem that led to this PR?
The problem was...
#5786
> We should have a check in bundle doctor for the file/folder permissions in the Bundler home directory.
>We should print a warning if there are any files/folders that is owned by another user but is readable/writable but prints an error when a file cannot be read or written to.
### What is your fix for the problem, implemented in this PR?
Created private method ```check_home_permissions``` that will print a warning if there are any files/folders that are owned by another user but are readable/writable, and print an error when the
bundler home dir contains a file cannot be read or written to
### Why did you choose this fix out of the possible options?
I chose this fix because it's what was requested in the open issue.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
r=colby-swandale
Added `--gemfile` option to `bundle update`
Thanks so much for the contribution!
To make reviewing this PR a bit easier, please fill out answers to the following questions.
### What was the end-user problem that led to this PR?
The problem was that `BUNDLE_GEMFILE` is not respected when `.bundle/config` specifies an alternate Gemfile. However, my specific issue is that `bundle install --gemfile Gemfile2` is an option, but `bundle update` won't let me specify an alternate Gemfile.
### What was your diagnosis of the problem?
My diagnosis was that Bundler copies `BUNDLE_GEMFILE` environment variable to `BUNDLE_ORIG_GEMFILE` and always uses `.bundle/config`. Simplest solution was to add parity to `bundle update` rather than diagnose why environment variables don't override `.bundle/config` settings across the board.
### What is your fix for the problem, implemented in this PR?
My fix is to add the `--gemfile` option to `bundle update` for parity with `bundle install`.
### Why did you choose this fix out of the possible options?
I chose this fix because this allows installing and updating alternative Gemfiles without untangling environment variables. It's the most direct use case that I'm having this issue with.
Ideally, the environment variables specified for a command should be respected and override settings in `.bundle/config` (https://github.com/bundler/bundler/issues/6270).
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
[Init] Check if the current directory is writeable
### What was the end-user problem that led to this PR?
The problem was that when running `bundle init` inside a directory which is not writable by the current user (e.g. `/` as demonstrated in #6219) Bundler prints out an `EACCES` error with a huge backtrace. In the mentioned PR @segiddins suggested to print out a better error message. This PR addresses that.
### What was your diagnosis of the problem?
See [this comment on said PR](https://github.com/bundler/bundler/issues/6219#issuecomment-359002919).
### What is your fix for the problem, implemented in this PR?
My fix is simple: adding a check whether the current directory is writeable before trying to create `gems.rb`/`Gemfile`. If that's not the case, print out an error and exit.
### Why did you choose this fix out of the possible options?
I chose this fix because... it was really simple to implement.
|
| | | | |
|
| | | | |
|
| | | | |
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Fix failing spec in the bundle exec system gem spec
### What was your diagnosis of the problem?
There is an issue where a spec was failing because it was activating the version of Bundler inside of RubyGems and not the development version that we're working on.
This would cause the Lockfile to raise an error because we were generating a Bundler 2 lockfile in our tests but Bundler 1 was being activated making it complain.
### What is your fix for the problem, implemented in this PR?
Add the Bundler development dir in the Ruby include argument.
### Why did you choose this fix out of the possible options?
This error is raising in CI and is preventing PRs from being merged.
See https://travis-ci.org/bundler/bundler/jobs/333434747
|
|/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
There was an issue where a spec was failing because it was activating
the version of bundler inside of RubyGems and not the development
version that we're working on.
This would cause the Lockfile to raise an error because we were generating
a Bundler 2 lockfile but Bundler 1 was being activated.
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Update contributing documentation to use new tags
Github has defined a couple of tags to use for classifying issues that are suitable for new contributors to fix/contribute towards. We have started using these tags in Github so we should start linking to them in our documentation as well.
|
|/ / / /
| | | |
| | | | |
Github has defined a couple of tags to use for classifying issues that are suitable for new contributors to fix/contribute towards. We have started using these tags in Github so we should start linking to them in our documentation as well.
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Add base error class to new gems.
Closes #6260.
Room for discussion:
- Which error class to use (`StandardError` makes sense to me)
- What formatting to use (the three-lines-with-comment seemed nicest to me)
- Whether or not using the flag to provide a different error base class is useful, and if it should validate the user's choice or not (I threw it in because it seemed harmless; is it? a boolean flag would work fine too)
---
### What was the end-user problem that led to this PR?
Libraries don't always follow best practice from discussion in linked issue.
### What was your diagnosis of the problem?
Bundler could encourage best practice by adding it to the gem scaffold.
### What is your fix for the problem, implemented in this PR?
I added a base error class to the templates, and provided a flag to change/disable this behaviour.
### Why did you choose this fix out of the possible options?
Like any best-practice-by-default, this could ruin someones workflow/go against someone's preferences so I made it as configurable as possible.
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | | |
Closes #6260.
|
|\ \ \ \ \
| |/ / / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
add the new team members of the bundler core team to the author list in the Bundler gemspec
Just adding myself and Hiroshi-san to the list of authors in the Bundler gemspec
\cc @indirect
|
| | | | |
| | | | |
| | | | |
| | | | | |
Bundler gemspec
|
|\ \ \ \ \
| |_|/ / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
handle gzip corruption errors in the compact index client
### What was the end-user problem that led to this PR?
Bundler will raise an exception and crash if the Gzip that the Compact Client Index downloaded was corrupt.
### What was your diagnosis of the problem?
See #6261
### What is your fix for the problem, implemented in this PR?
Handle the exception and allow Bundler to continue without crashing
|
| |/ / / |
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
[FriendlyErrors] Fallback to the original error if the friendly message raises
### What was the end-user problem that led to this PR?
This would have made diagnosing https://github.com/bundler/bundler/issues/6220 much easier
### What was your diagnosis of the problem?
My diagnosis was we do a lot in our error handler, so it needs a fallback
### What is your fix for the problem, implemented in this PR?
My fix falls back to just re-raising the original error
|
| |/ / /
| | | |
| | | |
| | | | |
This would have made diagnosing https://github.com/bundler/bundler/issues/6220 much easier
|
|\ \ \ \
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
[man] fix typo on bundle-binstubs
Typo found while i was packaging 1.16.1 for Debian
|
| | | | | |
|
|\ \ \ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Rename Bundler.frozen? to avoid Object method conflict
### What was the end-user problem that led to this PR?
The problem was `Bundler` overrode the `Object#frozen?` method to mean something else, breaking anything that checks that method to determine if an object is frozen.
Closes #6252.
### What was your diagnosis of the problem?
My diagnosis was the method name had to be changed.
|
| | | | | | |
|
|\ \ \ \ \ \
| |/ / / / /
|/| | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
[Source::RubyGems] Ensure the bindir exists before installing
### What was the end-user problem that led to this PR?
The problem was occasional failures on CI w/ ruby 2.5
### What was your diagnosis of the problem?
My diagnosis was that there was a race condition in the rubygems installer, see https://github.com/rubygems/rubygems/pull/2148 for the fix there.
### What is your fix for the problem, implemented in this PR?
My fix will ensure that directory is created before we get into that rubygems code path
|
|/ / / / /
| | | | |
| | | | |
| | | | | |
This will help avoid a potential race condition in RubyGems
|
|\ \ \ \ \
| |_|/ / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
[CLI] Skip the outdated bundler check when MD5 is not available
### What was the end-user problem that led to this PR?
The problem was Bundler would try to do the outdated version check on FIPS systems, leading to an exception.
Closes #6032.
### What was your diagnosis of the problem?
My diagnosis was we needed to skip the check when MD5 is unavailable.
### Why did you choose this fix out of the possible options?
I chose this fix because using `#available?` would make network requests, which is not acceptable.
|
| | | | | |
|
|\ \ \ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Make `bundle clean` clean extension directories
### What was the end-user problem that led to this PR?
The problem was that `bundle clean` command doesn't remove gem extensions (#5596)
### What was your diagnosis of the problem?
I've looked into `Bundler::Runtime#clean` and realized that extension dirs are not removed
### What is your fix for the problem, implemented in this PR?
My fix is to tweak `Bundler::Runtime#clean` to remove extensions dirs
as well.
### Why did you choose this fix out of the possible options?
I chose this fix because I didn't see any other option.
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
`bundle clean` command doesn't remove gem extensions, because
extensions reside out of gem directories.
This change tweaks `Bundler::Runtime#clean` to remove extensions dirs
as well.
|
|\ \ \ \ \ \
| |/ / / / /
|/| | | | |
| | | | | |
| | | | | |
| | | | | | |
Fixed typos
Found by misspell.
|
|/ / / / / |
|
|\ \ \ \ \
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Update unencrypted http links to https
To avoid human-in-the-middle attacks
|
| | |/ / /
| |/| | |
| | | | | |
To avoid human-in-the-middle attacks
|