After some discussion, we decided to change the DSL to get rid of
the `plist_name` and `service_name` methods which aren't meaningful
for most users.
The new DSL looks like this:
```rb
service do
name macos: "name", linux: "name"
end
```
I also updated some specs here to reflect these changes.
There was some talk about maybe deprecating `plist_name` and `service_name`
but I think that's outside of the scope of this PR so I'm leaving
them as is for now. One benefit of this is that everything here is backwards
compatible.
It now checks for two new edge cases.
1. Are the required methods defined in the block?
2. Is the `run` command defined if we're using methods other
than `service_name` or `plist_name`?
It has also been updated to only get the list of method call nodes
once instead of doing it repeatedly for each cellar path check.
- Now that we detect correct stanza _grouping_ within `on_*` blocks in
Casks (PR 15211), correct stanza _ordering_ in `on_*` blocks was the
next logical step. For example, `url` has to come after `version` and
`sha256` in an `on_macos` or `on_intel` block for consistency with the
top-level stanza order we enforce elsewhere.
- Still not doing the nested `on_os` inside `on_arch`, that felt
excessive for an edge case that isn't present in any actual real
Casks we have. I removed the test with that specific TODO.
- This test tests nothing. And the TODO comment is wrong.
- We _could_ fix it, but it's a very edgy edge case which pertains to
`livecheck` blocks which currently don't have stanza grouping or
ordering cop support. If we decide in the future to add these, we can
add this back too (provided I remember).
- Also I think I may have got confused with the stanza grouping vs.
stanza ordering cops when writing this, rendering this test more
useless.
- Since moving `comments_hash` to `Stanza`, we've been using the wrong
kind of "comments": the comments for the _stanza_, not the comments
for the entire Cask.
- Add a test to ensure this actually works. There was previously an
infinite loop here due to the bad `comments`, visible in a `StanzaOrder`
cop test, which I speculatively added a failing test for. Turns out
that supporting nested stanza _ordering_ (vs. just grouping) is a
whole separate piece of work (there are multiple TODOs there already),
so I've backed that out and will do that separately.
- A variant of this was an ancient TODO from 2018 (with `if/else` blocks).
- Now in 2023 we have `on_*` blocks within Casks that are very common.
- The most common stanzas present inside `on_*` blocks are `version`,
`sha256` and `url`. So I feel like it's worth keeping a consistent
style for these inside and outside `on_*` blocks.
- Fixing the test expected output was unbelievably tedious.
- There's been debate about this setting being `false` but in
https://github.com/Homebrew/brew/pull/15136#issuecomment-1500063225
we decided that it was worth using the default since RuboCop behaviour changed
so we'd have had to do some horrible things to keep it as `false` -
https://github.com/Homebrew/brew/pull/15136#issuecomment-1500037278 -
and multiple maintainers specify the `--display-cop-names` option to
`brew style` themselves since it's clearer what's gone wrong.
The preferred method is opt_bin because that works
with the API and is more portable (works between versions).
Also removed the last example from the docs of `bin/"name"`
from the service block section
Handle good things like:
```ruby
url "https://example.org/download",
verified: "example.org/download" # This is fine.
```
And bad things like:
```ruby
url "https://example.org/",
verified: "example.org" # This should end with a slash.
```
- Previously this components order cop only checked for correct stanza
order inside `on_*` blocks. This commit extends this cop to also check
for correct stanza order inside `head` and `resource` blocks. This is
a positive change since it standardizes the order of stanzas in all of
the places, making formulae more readable.
- Fixes issue 14017.
- This, ie Mojave first, is more common in real Casks than the
alternative of newest to oldest ie Ventura first.
- Doing it this way reduces the number of offenses from ~500 to ~200.
- Complaining about only `on_arm` and `on_intel` was too restrictive
since casks can have many `on_system` blocks (`on_#{arch}` and
`on_#{os}`).
- We're a bit of the way there, anyway. Still doesn't support stanza
ordering within blocks, but that's for another time (there's a
separate issue that's been open for a while - 14017).
- These were previously being manually fixed which is time maintainers
could have spent fixing more important problems.
- I don't work with Casks much at all, so I was unsure as to what the
existing "arch" and "on_arch_conditional" parts were, if they're
deprecated or if things were eventually going to migrate to
`on_#{arch}` blocks?
- This skips over stanza names that are not overrideable in `on_*`
blocks, with the positive side effect that `on_*` blocks themselves
aren't in the list so we can get rid of another conditional.
- Stanzas overrideable in blocks are defined in `Cask::DSL` by each of
the methods calling `set_unique_stanza`.
- This came up in Cask `simply-fortran`:
```
Scanning /opt/homebrew/Library/Taps/homebrew/homebrew-cask/Casks/simply-fortran.rb
send_node: s(:send, nil, :arch), send_node.parent: s(:begin,
s(:send, nil, :arch)), send_node.parent.parent: (regexp
(str "href=.*?simplyfortran[._-]v?(\\d+(?:\\.\\d+)+)")
(begin
(send nil :arch))
(str "\\.dmg")
(regopt :i))
Casks/simply-fortran.rb:2:3: C: Cask/NoOverrides: Do not use a top-level arch stanza as the default. Add it to an on_{system} block instead.
Use :or_older or :or_newer to specify a range of macOS versions.
arch arm: "-arm64", intel: "-x86_64"
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1 file inspected, 1 offense detected
```
- This passes the previously failing test for `on_*` blocks with
`livecheck` blocks with multiple stanzas inside them (eg `url` and
`strategy`) that weren't being correctly skipped because we weren't
detecting high enough up the ancestry.
- The Cask `little-snitch4` in `Homebrew/homebrew-cask-versions` was
failing and it took me a while to figure out _how_. Add a test for
easier further debugging (and to prevent breakage once the bug is
fixed).
```
❯ brew tests --only=rubocops/cask/no_overrides
Randomized with seed 29917
1 process for 1 spec, ~ 1 spec per process
F
Failures:
1) RuboCop::Cop::Cask::NoOverrides when there are livecheck blocks within `on_*` blocks, ignore their contents does not report any offenses
Failure/Error: DEFAULT_FAILURE_NOTIFIER = lambda { |failure, _opts| raise failure }
expected `[#<RuboCop::Cop::Offense:0x000000012de636c8 @severity=#<RuboCop::Cop::Severity:0x000000012de636a0 @name=:convention>, @location=#<Parser::Source::Range (string) 244...273>, @message="Do not use a top-level `url` stanza as the default. Add it to an `on_{system}` block instead.\nUse `:or_older` or `:or_newer` to specify a range of macOS versions.\n", @cop_name="Cask/NoOverrides", @status=:unsupported, @corrector=nil>].empty?` to be truthy, got false
Shared Example Group: "does not report any offenses" called from ./test/rubocops/cask/no_overrides_spec.rb:77
# ./test/rubocops/cask/shared_examples/cask_cop.rb:24:in `expect_no_offenses'
# ./test/rubocops/cask/shared_examples/cask_cop.rb:7:in `block (2 levels) in <module:CaskCop>'
Took 2 seconds
Tests Failed
```
- The Cask `sip`, to give a random example, was failing this RuboCop
because it has a `livecheck` block within an `on_*` block and the
livecheck block and the top-level Cask both have `url` stanzas. This
is a legitimate use of `livecheck` blocks because the cask software
download URL and the livecheck version check URL are not the same
thing, so let's skip over `livecheck` blocks and their contents.
- I wrote this because I was concerned that I'd done something wrong
since `brew style --only=Cask/NoOverrides` in
`$(brew --repo homebrew/cask)` didn't report any offenses. Turns out
I'd just missed the `.` off the end of the command to target the
current directory! But, the cost of writing the test is sunk now,
and more tests can't hurt?
- In the event that there's only one common stanza within the `on_*`
blocks (eg, `url`) with a generic `version` that doesn't change per-OS,
let's not force adding `version` to each `on_*` block as well.
- As discussed in
https://github.com/Homebrew/brew/pull/14976#issuecomment-1474544569
and further comments, this is needed because in order to enforce the
order of `on_{arch,system}` blocks we need to have everything
consistently within one of those blocks.
- We previously allowed overrides where the top-level `version` stanza
would be the default, unless on an OS that had an `on_system` block
with a `version` specified. But this breaks down when we try to order
the `on_system` blocks because if a `url` at the top-level has a
`version` interpolated in it, then the `version` stanza needs to be
above the `url` stanza. But it could be that `version` is OS-specific.
- Let's stop allowing overrides and require that everything be in an
`on_system` block. This will make it easier to enforce the order of
`on_system` blocks in a future PR (14976).
- Apparently the "verified" parameter in the URL (present when a Cask's
download URL is not the same as its homepage) shouldn't have the
protocol (`https`, `http`) at the front.
- Removing this has happened manually in the past, so here's an
autocorrecting RuboCop for it.
