add support for credentials usage in combination with other mirroring
features. previously `HOMEBREW_DOCKER_REGISTRY_TOKEN` and
`HOMEBREW_DOCKER_REGISTRY_BASIC_AUTH_TOKEN` where only used in combination
with `HOMEBREW_ARTIFACT_DOMAIN` which affects *all* curl download strategies.
`HOMEBREW_BOTTLE_DOMAIN` is only used for bottle artifacts, whose mirror might
also require credentials.
this change removes the requirement of using `HOMEBREW_ARTIFACT_DOMAIN`.
related to #13226
- Rename use of "preinstall" to "auto update". The original "preinstall"
naming came from the fact that we used to only auto-update before
`brew install` but now that it's many commands: this is more confusing
than useful.
- Add `HOMEBREW_NO_UPDATE_REPORT_ONLY_INSTALLED` and remove
`HOMEBREW_UPDATE_REPORT_ONLY_INSTALLED`; the latter is now the default
and the prior provides an opt-out for better output, performance and
avoiding reading potentially untrusted formulae.
- Add `HOMEBREW_UPDATE_FORMULA_VERSION_CHECKS` and don't check formula
versions by default for better performance by default.
Co-authored-by: Eric Knibbe <3324775+EricFromCanada@users.noreply.github.com>
Co-authored-by: Sam Ford <1584702+samford@users.noreply.github.com>
At the moment, `#use_homebrew_curl?` can only be true for a
`homepage` or `stable`/cask `url` with `using: :homebrew_curl`. If
the checked URL differs from these URLs, livecheck won't use brewed
curl. This limitation prevents livecheck from using brewed curl for a
`livecheck` block URL that's a string literal (not a symbol for a
`#checkable_url` like `:stable`, `:head`, `:url`). `libzip` was the
original formula referenced in the related brew issue and it meets
this criterion, so it doesn't appear to be handled by the existing
`#use_homebrew_curl?` implementation.
Additionally, the existing behavior can cause livecheck to
unnecessarily use brewed curl for a completely different website
(e.g., `cubelib`, `otf2`). For example, if the `stable` URL has
`using: :homebrew_curl` and the `livecheck` block has `url
:homepage`, livecheck will use brewed curl when checking the
`homepage`. If these are completely different domains/servers, it's
unlikely that we would need to use brewed curl when checking the
`homepage`, so this particular behavior may not be beneficial.
This commit reimplements `use_homebrew_curl?` to apply brewed curl
when the checked URL's root domain is the same as the root domain of
an aforementioned formula/cask URL with `using: :homebrew_curl`. For
example, this looser approach would allow a `livecheck` block
checking `https://www.example.com/downloads/` to use brewed curl if
the `stable` URL was `https://downloads.example.com/example.zip` with
`using: :homebrew_curl`. These could be different servers but, based
on related formulae, this looseness is necessary for the moment.
This approach aims to resolve both issues, allowing brewed curl to be
applied to a slightly broader range of URLs (i.e., not limited to
just the `#checkable_urls`) while also helping to avoid unnecessarily
applying brewed curl when it's less likely to be useful (completely
different domains). Neither approach is perfect but this one may be
more useful in the interim time.
Depending on how this looser approach works in practice, we may want
to consider returning to a stricter approach once we have something
like `using: :homebrew_curl` in `livecheck` blocks (this is
forthcoming). Being explicit in a `livecheck` block is the most
reliable approach (i.e., only use brewed curl when needed), so we
could favor that and pare down the automated approach to only what's
needed to support implicit checks (i.e., with no `livecheck` block).
Of course, it's also possible to drop the automated approach entirely
and simply require a `livecheck` block in this scenario but we can
decide on how to handle this when the time comes.
A sha256 hash of the previous download is stored and compared with
new downloads before updating :latest casks. This prevents unnecessary
reinstalls when the cask hasn't been updated.
Move download path to cask from installer to prevent unnecessary
redownloads of casks.
At the moment, `#use_homebrew_curl?` can only be true for a
`homepage` or `stable`/cask `url` with `using: :homebrew_curl`. If
the checked URL differs from these URLs, livecheck won't use brewed
curl. This limitation prevents livecheck from using brewed curl for a
`livecheck` block URL that's a string literal (not a symbol for a
`#checkable_url` like `:stable`, `:head`, `:url`). `libzip` was the
original formula referenced in the related brew issue and it meets
this criterion, so it doesn't appear to be handled by the existing
`#use_homebrew_curl?` implementation.
Additionally, the existing behavior can cause livecheck to
unnecessarily use brewed curl for a completely different website
(e.g., `cubelib`, `otf2`). For example, if the `stable` URL has
`using: :homebrew_curl` and the `livecheck` block has `url
:homepage`, livecheck will use brewed curl when checking the
`homepage`. If these are completely different domains/servers, it's
unlikely that we would need to use brewed curl when checking the
`homepage`, so this particular behavior may not be beneficial.
This commit reimplements `use_homebrew_curl?` to apply brewed curl
when the checked URL's root domain is the same as the root domain of
an aforementioned formula/cask URL with `using: :homebrew_curl`. For
example, this looser approach would allow a `livecheck` block
checking `https://www.example.com/downloads/` to use brewed curl if
the `stable` URL was `https://downloads.example.com/example.zip` with
`using: :homebrew_curl`. These could be different servers but, based
on related formulae, this looseness is necessary for the moment.
This approach aims to resolve both issues, allowing brewed curl to be
applied to a slightly broader range of URLs (i.e., not limited to
just the `#checkable_urls`) while also helping to avoid unnecessarily
applying brewed curl when it's less likely to be useful (completely
different domains). Neither approach is perfect but this one may be
more useful in the interim time.
Depending on how this looser approach works in practice, we may want
to consider returning to a stricter approach once we have something
like `using: :homebrew_curl` in `livecheck` blocks (this is
forthcoming). Being explicit in a `livecheck` block is the most
reliable approach (i.e., only use brewed curl when needed), so we
could favor that and pare down the automated approach to only what's
needed to support implicit checks (i.e., with no `livecheck` block).
Of course, it's also possible to drop the automated approach entirely
and simply require a `livecheck` block in this scenario but we can
decide on how to handle this when the time comes.
Before `#parse_curl_output` was introduced and related methods were
updated to use it, `#url_protected_by_cloudflare?` and
`#url_protected_by_incapsula?` were checking a string of all the
headers from a response and using a regex to check related header
values.
However, when `#curl_http_content_headers_and_checksum` was updated
to use `#parse_curl_output` internally, the `:headers` value became
a hash generated by `#parse_curl_response`. The `#url_protected_by_*`
methods were updated to work with the hash value but this wasn't able
to fully replicate the previous behavior because
`#parse_curl_response` was only keeping the last instance of a given
header (maintaining pre-existing behavior). This is an issue for
these methods because they check `Set-Cookie` headers and there can
be multiple instances of this header in a response.
This commit updates these methods to handle an array of strings in
addition to the existing string support. This change ensures that
these methods properly check all `Set-Cookie` headers, effectively
reinstating the previous behavior.
Past that, this updates one of the early return values in
`#url_protected_by_cloudflare?` to be `false` instead of an implicit
`nil`. After adding a type signature to this method, it became clear
that it wasn't always returning a boolean value and this fixes it.
`Curl#parse_curl_response` only includes the last instance of a given
header in its `:headers` hash (replicating pre-existing behavior).
This is a problem for headers like `Set-Cookie`, which can appear more
than once in a response.
This commit addresses the issue by collecting duplicate headers into
an array instead. Headers that only appear once in the response will
still have a string value but headers that appear more than once will
be an array of strings. Whenever headers from `#parse_curl_response`
are used (directly or indirectly), it's important to conditionally
handle the expected types.
The problem occurred when calling `brew reinstall` on a cask
with an out of date caskfile.
To solve the problem Cask::Installer#uninstall_existing_cask has been changed to
catch a possible CaskInvalidError when trying to load outdated
caskfiles using Cask::CaskLoader#load.
Follow up adter #13124
I made the choice to convert the pr variable to an integer
at the very end and adjust the tests.
It would be maybe more consistent to work with an integer
everywhere, but this needs a more careful analysis and we
are in a hurry to fix the homberew-core upload CI
Fixes:
2022-04-11T20:19:34.1395885Z [31mError:[0m : Variable $pr of type Int! was provided invalid value
2022-04-11T20:19:34.1398279Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/utils/github/api.rb:261:in `open_graphql'
2022-04-11T20:19:34.1399774Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/utils/github.rb:310:in `get_workflow_run'
2022-04-11T20:19:34.1403699Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:418:in `block (4 levels) in pr_pull'
2022-04-11T20:19:34.1405233Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:417:in `each'
2022-04-11T20:19:34.1406723Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:417:in `block (3 levels) in pr_pull'
2022-04-11T20:19:34.1408112Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/vendor/portable-ruby/2.6.8/lib/ruby/2.6.0/fileutils.rb:128:in `chdir'
2022-04-11T20:19:34.1408986Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/vendor/portable-ruby/2.6.8/lib/ruby/2.6.0/fileutils.rb:128:in `cd'
2022-04-11T20:19:34.1409813Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:400:in `block (2 levels) in pr_pull'
2022-04-11T20:19:34.1410671Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/vendor/portable-ruby/2.6.8/lib/ruby/2.6.0/tmpdir.rb:93:in `mktmpdir'
2022-04-11T20:19:34.1411495Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:399:in `block in pr_pull'
2022-04-11T20:19:34.1412250Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:388:in `each'
2022-04-11T20:19:34.1413056Z /home/linuxbrew/.linuxbrew/Homebrew/Library/Homebrew/dev-cmd/pr-pull.rb:388:in `pr_pull'
- The `brew uninstall` command has `--zap`, so let's make `brew
reinstall` have parity here for a better user experience. (Requested
in issue 12983.)
- It feels weird that to get my new reinstall test to pass I had to add
`--zap` to `cask/cmd/install.rb`, not `cask/cmd/reinstall.rb` to get
the tests to pass. But the `brew reinstall --cask caffeine --zap`
command worked fine all the time. The CLI argument parser from the
test run was complaining about not knowing what `zap` was. As a
result, `--zap` now shows up as a switch in `brew install --help`
which I'm not 100% convinced is the desired UX. But I've edited the
description accordingly to specify that it will only work on
`reinstall` operations (and `--zap` on `install` is a no-op).
```
issyl0 at pictor in /opt/homebrew on reinstall-cask-zap
❯ brew reinstall --cask caffeine --zap
==> Downloading https://github.com/IntelliScape/caffeine/releases/download/1.1.3/Caffeine.dmg
Already downloaded: /Users/issyl0/Library/Caches/Homebrew/downloads/3d6ccfdd3b8d0ab37d1c2468d6e69078c2d31d3b12bf51947c4db21e5f376af2--Caffeine.dmg
==> Implied `brew uninstall --cask caffeine`
==> Backing App 'Caffeine.app' up to '/opt/homebrew/Caskroom/caffeine/1.1.3/Caffeine.app'
==> Removing App '/Applications/Caffeine.app'
==> Dispatching zap stanza
==> Trashing files:
~/Library/Application Support/com.intelliscapesolutions.caffeine
~/Library/Preferences/com.intelliscapesolutions.caffeine.plist
~/Library/Caches/com.intelliscapesolutions.caffeine
~/Library/HTTPStoages/com.intelliscapesolutions.caffeine.binarycookies
==> Removing all staged versions of Cask 'caffeine'
==> Installing Cask caffeine
==> Moving App 'Caffeine.app' to '/Applications/Caffeine.app'
🍺 caffeine was successfully installed!
```
