My experience recently playing around with our locking behaviour is
that, while mostly seamless and not seen by users, it's leaks
implementation details a bit too heavily.
As a result, the following improvements are in this commit:
- Ensure that, whenever possible, we tell the user the actual command
that is holding a given lock instead of the lock name (an internal
implementation detail)
- Make the locking error output a little more consistent and user
friendly
- Add a `DownloadLock` class to simplify locking downloads
- Add a `HOMEBREW_LOCK_CONTEXT` variable to allow adding additional
context for logging error messages
- Lock paths and leave deciding how this translates to lock names up
to the locking code itself
- Lock the Cellar/Caskroom paths explicitly rather than implicitly
Co-authored-by: Carlo Cabrera <30379873+carlocab@users.noreply.github.com>
This provides a decent speedup:
```
$ hyperfine 'git checkout master; brew help' 'git checkout help_bash; brew help'
Benchmark 1: git checkout master; brew help
Time (mean ± σ): 506.4 ms ± 50.9 ms [User: 223.7 ms, System: 99.9 ms]
Range (min … max): 454.6 ms … 634.1 ms 10 runs
Benchmark 2: git checkout help_bash; brew help
Time (mean ± σ): 109.5 ms ± 57.1 ms [User: 1
```
and compares favourably to `pip3 help`:
```
$ hyperfine 'brew help' 'pip3 help'
Benchmark 1: brew help
Time (mean ± σ): 72.9 ms ± 15.9 ms [User: 4.9 ms, System: 6.3 ms]
Range (min … max): 53.6 ms … 126.6 ms 31 runs
Benchmark 2: pip3 help
Time (mean ± σ): 171.5 ms ± 6.1 ms [User: 131.6 ms, System: 24.7 ms]
Range (min … max): 164.2 ms … 189.3 ms 15 runs
Summary
brew help ran
2.35 ± 0.52 times faster than pip3 help
```
This improves the load time of most brew commands. For an example of
one of the simplest commands this speeds up:
Without Bootsnap:
```
$ hyperfine 'git checkout master; brew help' 'git checkout optimise_requires; brew help'
Benchmark 1: git checkout master; brew help
Time (mean ± σ): 525.0 ms ± 35.8 ms [User: 229.9 ms, System: 113.1 ms]
Range (min … max): 465.3 ms … 576.6 ms 10 runs
Benchmark 2: git checkout optimise_requires; brew help
Time (mean ± σ): 383.3 ms ± 25.1 ms [User: 133.0 ms, System: 72.1 ms]
Range (min … max): 353.0 ms … 443.6 ms 10 runs
Summary
git checkout optimise_requires; brew help ran
1.37 ± 0.13 times faster than git checkout master; brew help
```
With Bootsnap:
```
$ hyperfine 'git checkout master; brew help' 'git checkout optimise_requires; brew help'
Benchmark 1: git checkout master; brew help
Time (mean ± σ): 386.0 ms ± 30.9 ms [User: 130.2 ms, System: 93.8 ms]
Range (min … max): 359.5 ms … 469.3 ms 10 runs
Benchmark 2: git checkout optimise_requires; brew help
Time (mean ± σ): 330.2 ms ± 32.4 ms [User: 93.4 ms, System: 73.0 ms]
Range (min … max): 302.9 ms … 413.9 ms 10 runs
Summary
git checkout optimise_requires; brew help ran
1.17 ± 0.15 times faster than git checkout master; brew help
```
This turns the ability to replace common paths with placeholders
into a mixin that can be used with both Casks and Formulae.
The idea here is to make formula hash generation more consistent.
Instead of doing so literally whenever we query for a formula, Instead
do so only when we're in an auto-updateable command.
This better fits the existing behaviour while still updating when it's
most important to do so.
