- Previously I thought that comments were fine to discourage people from
wasting their time trying to bump things that used `undef` that Sorbet
didn't support. But RuboCop is better at this since it'll complain if
the comments are unnecessary.
- Suggested in https://github.com/Homebrew/brew/pull/18018#issuecomment-2283369501.
- I've gone for a mixture of `rubocop:disable` for the files that can't
be `typed: strict` (use of undef, required before everything else, etc)
and `rubocop:todo` for everything else that should be tried to make
strictly typed. There's no functional difference between the two as
`rubocop:todo` is `rubocop:disable` with a different name.
- And I entirely disabled the cop for the docs/ directory since
`typed: strict` isn't going to gain us anything for some Markdown
linting config files.
- This means that now it's easier to track what needs to be done rather
than relying on checklists of files in our big Sorbet issue:
```shell
$ git grep 'typed: true # rubocop:todo Sorbet/StrictSigil' | wc -l
268
```
- And this is confirmed working for new files:
```shell
$ git status
On branch use-rubocop-for-sorbet-strict-sigils
Untracked files:
(use "git add <file>..." to include in what will be committed)
Library/Homebrew/bad.rb
Library/Homebrew/good.rb
nothing added to commit but untracked files present (use "git add" to track)
$ brew style
Offenses:
bad.rb:1:1: C: Sorbet/StrictSigil: Sorbet sigil should be at least strict got true.
^^^^^^^^^^^^^
1340 files inspected, 1 offense detected
```
This is just a recipe for infinite loops. Plus, recursive references are
likely to be invalid, so we don't really gain much by attempting to
resolve them.[^1] (But we could if we made the logic here much more
complicated.)
Fixes a CI failure seen at Homebrew/homebrew-core#138323.
[^1]: See, for example, embree/embree#455.
We can add a small amount of logic to `#resolve_variable_name` that will
allow us to perform (limited) resolution of rpath references. This is
for informational purposes only: failing to resolve an `@rpath`
reference will not (and should not) result in `brew linkage` failures.
`dyld` will typically have more information than we do to resolve these
references, so not failing `brew linkage` when we fail to resolve an
`@rpath` reference is the right behaviour here.
As an example, before:
❯ brew linkage jpeg-turbo
System libraries:
/usr/lib/libSystem.B.dylib
@rpath-referenced libraries:
@rpath/libjpeg.8.dylib
@rpath/libturbojpeg.0.dylib
After:
❯ brew linkage jpeg-turbo
System libraries:
/usr/lib/libSystem.B.dylib
Homebrew libraries:
/usr/local/Cellar/jpeg-turbo/3.0.0/lib/libjpeg.8.dylib (jpeg-turbo)
/usr/local/Cellar/jpeg-turbo/3.0.0/lib/libturbojpeg.0.dylib (jpeg-turbo)
This implements the TODO that I left as a comment from #15722.
In order to simplify the handling of deleting an absolute rpath that has
a relative duplicate (or vice-versa, i.e., a relative rpath with an
absolute duplicate), we relocate all rpaths first in one pass and then
delete the duplicates in a separate pass.
We currently rely on some lower-level (but still public) methods from
ruby-macho since the standard `#delete_rpath` method changes the order
in which rpaths are resolved. We can switch back to using
`#delete_rpath` when Homebrew/ruby-macho#555 is merged and released.
We were rewriting dylib IDs and install names using `MachO::Tools`,
which doesn't update the state of the file in memory. This leads to
those changes being undone when we call `delete_rpath`.
We fix this by making sure the state of the file in memory always
matches the state of file on disk.
Closes#12832.
There have been a few instances I've noticed that we've been silently
installing binaries built for x86_64 on ARM. There's probably more that
I haven't found yet, so it seems useful to check this with an audit.
Rather than having to manually keep track of what version each thing in
here is and copy files around by hand on update let's use Bundler's
standalone mode and careful use of `.gitignore` to help us do it.
This means a `bundle update --standalone` will allow us to update all
gems in vendor.
We could consider vendoring other gems this way in future but I'd
suggest only doing this for gems with no dependencies or at least gems
with no native extensions. The only gem this applies to that we
currently use is `ruby-prof` and I'm not convinced it's widely used
enough to warrant vendoring for everyone. Perhaps that's another
criteria: it should be functionality that's used by non-developer
commands and/or normal Homebrew usage.
If people have `HOMEBREW_RUBY_PATH` set then things explode in a rather
confusing fashion. Instead, run `bundle` for them with the arguments
that they'd want.
Also, move `macho` requires into the module itself; it's a pain having
to do everything for Bundler before requiring `pathname` which is a core
Ruby class.
Homebrew's actually ended up using a fair few gems. While we want to
avoid Bundler at runtime (and this PR still does that, in fact uses
Bundler even less at runtime than it did before) writing our own version
to use at build-time seems redundant.
remove old mach.rb, replace with cctools_mach.rb and prune ELF case
move ArchitectureListExtension to separate file ELF support is
maintained in Linuxbrew
