It's both unexpected and undesirable for `brew bundle (exec|env|sh)` to
filter the environment and makes these tools less useful.
Not filtering the environment, though, causes issues with the
`brew bundle sh` shell. Fix this up and, while we're here, also improve
the formatting for both `zsh` and `bash` (the default) to use nicer and
more consistent prompts and colours.
To simplify this, consolidate some logic in a new
`Utils::Shell.shell_with_prompt` method and add tests for it and a
similar notice for `brew bundle sh`.
Finally, avoid printing out the notice when `HOMEBREW_NO_ENV_HINTS` is
set.
`Livecheck::Strategy.page_headers` uses `Utils::Curl.curl_headers` but
the method only handles `HEAD` and `GET` requests. I recently added
`POST` support to livecheck but forgot to update `curl_headers` in the
process, so `livecheck` blocks using the `HeaderMatch` strategy along
with `post_form` or `post_json` will fail because curl doesn't allow
both `--head` and `--data`/`--json` arguments.
This addresses the issue by updating `curl_headers` to handle `POST`
requests and skip the `GET` retry logic.
This is particularly useful for third-party Python formulae that have a ton of resources, not all of which may adhere to homebrew/core's strict policies. See #19240 for context.
I've also added logic that ignores `--ignore-errors` on `homebrew/core`, although I personally think this new behavior is also useful for mainline formula creation.
Before: error out on a single non-conforming resource, zero resource blocks added to formula, scary stacktrace.
After: all conforming resources added, all non-conforming resources identified in comments, error message at end, `brew` exits non-zero without scary stacktrace:-
```
% brew update-python-resources --ignore-errors gromgit/test/auto-coder || echo OOPS
==> Retrieving PyPI dependencies for "auto-coder==0.1.243"...
==> Retrieving PyPI dependencies for excluded ""...
==> Getting PyPI info for "aiohappyeyeballs==2.4.4"
[200+ resource lines elided]
==> Getting PyPI info for "zhipuai==2.1.5.20250106"
==> Updating resource blocks
Error: Unable to resolve some dependencies. Please check /opt/homebrew/Library/Taps/gromgit/homebrew-test/Formula/auto-coder.rb for RESOURCE-ERROR comments.
OOPS
% brew cat gromgit/test/auto-coder | ggrep -C10 RESOURCE-ERROR
license "Apache-2.0"
depends_on "python@3.11"
# Additional dependency
# resource "" do
# url ""
# sha256 ""
# end
# RESOURCE-ERROR: Unable to resolve "azure-cognitiveservices-speech==1.42.0" (no suitable source distribution on PyPI)
# RESOURCE-ERROR: Unable to resolve "ray==2.42.0" (no suitable source distribution on PyPI)
resource "aiohappyeyeballs" do
url "e4373e888f/aiohappyeyeballs-2.4.4.tar.gz"
sha256 "5fdd7d87889c63183afc18ce9271f9b0a7d32c2303e394468dd45d514a757745"
end
resource "aiohttp" do
url "952d49c730/aiohttp-3.11.12.tar.gz"
sha256 "7603ca26d75b1b86160ce1bbe2787a0b706e592af5b2504e12caa88a217767b0"
end
```
Fixes:
Error: Parameter 'days': Expected type String, got type Integer with value 30
Caller: /opt/homebrew/Library/Homebrew/utils/analytics.rb:273
Definition: /opt/homebrew/Library/Homebrew/utils/analytics.rb:412 (Utils::Analytics.table_output)
This adds more tests to `curl_spec.rb` to increase test coverage.
This brings almost all of the methods that don't make network
requests up to 100% line and branch coverage (the exception being
some guards in `parse_curl_output` that shouldn't happen under
normal circumstances).
In the process of writing more tests for `parse_curl_response`, I
made some tweaks to remove checks for conditions that shouldn't ever
be true (e.g., `match["code"]` isn't optional, so it will be present
if `HTTP_STATUS_LINE_REGEX` matches) and to refactor some others. I
contributed this method a while back (9171eb2), so this is me coming
back to clarify some behavior.
This upgrades `utils/curl.rb` to `typed: strict`, which requires
a number of changes to pass `brew typecheck`. The most
straightforward are adding type signatures to methods, adding type
annotations (e.g., `T.let`) to variables that need them, and ensuring
that methods always use the expected return type.
I had to refactor areas where we call a `Utils::Curl` method and use
array destructuring on a `SystemCommand::Result` return value
(e.g., `output, errors, status = curl_output(...)`), as Sorbet
doesn't understand implicit array conversion. As suggested by Markus,
I've switched these areas to use `#stdout`, `#stderr`, and `#status`.
This requires the use of an intermediate variable (`result`) in some
cases but this was a fairly straightforward substitution.
I also had to refactor how `Cask::URL::BlockDSL::PageWithURL` works.
It currently uses `page.extend PageWithURL` to add a `url` attribute
but this reworks it to subclass `SimpleDelegator` and use an
`initialize` method instead. This achieves the same goal but in a way
that Sorbet can understand.
We need a way to escape systemd command lines properly as systemd treats
unrecognised escape sequences as separate literal characters. This helper
function does that.
Our autobump workflow sets the author and committer to the user who
triggered the workflow, defaulting to @BrewTestBot for scheduled runs.
This can be confusing for maintainers when GitHub shows up as
"Unverified" because the commit is signed with @BrewTestBot's key.[^1]
Let's fix that by configuring our autobump workflow to always commit as
@BrewTestBot, so that the committer matches the GPG signature. To do
that, we need to add support for setting `GIT_COMMITTER_NAME` and
`GIT_COMMITTER_EMAIL`.
[^1]: See, for example, Homebrew/homebrew-core#197234.
