Fixes #1848 by sharing recursive dependency resolution between
dependencies and requirements.

Coincidentally, this also fixes the errors introduced by #1784 that
necessitated it being reverted in #1797.

also allow higher stable versions with the same version prefix

When the first parameter to inreplace was an array, and the replacement
failed, InreplaceError would end up crashing with an undefined method
exception because the order of operations resulted in super not being
passed the value of the entire inject block.

Xcode can be installed anywhere but for most people it's in
`/Applications/Xcode.app` so just look there if `xcode-select` isn't
helpful before looking at Spotlight which can return weird results on
e.g. backup disks.

Fixes #1587.

Xcode can be installed anywhere but for most people it's in
`/Applications/Xcode.app` so just look there by default before looking
at Spotlight which can return weird results on e.g. backup disks.

We want to prefer the newer/versioned formulae when possible but this
is preferable to a random cached formula or, worse, no formula.

This means `brew info foo` will never complain that `foo` is a missing
formula.

Which, in turn, provides it for `brew info --json=v1` so other tools
such as e.g. `brew bundle` can make use of this information.

Unfortunately `brew linkapps` cannot behave nicely with e.g. Spotlight
using either aliases or symlinks and Homebrew formulae do not build
"proper" `.app` bundles that can be relocated. Instead, please consider
using `brew cask` and migrate formulae using `.app`s to casks.

`virtualenv_install_with_resources` will automatically define and
install a `setuptools` resource so this is unnecessary.

References https://github.com/Homebrew/homebrew-core/pull/8570

If it's not there, produce a nicer error.

Fixes #1805.

Since /tmp (the default HOMEBREW_TEMP) is a symlink to /private/tmp,
some build systems (like Parrot's) will attempt to use the realpath
instead of the literal /tmp we supply it with. This breaks the relocation
code, which only tested the literal HOMEBREW_TEMP and not its realpath.

This allows us to detect if homepages such as e.g. `aiccu` which
blocks `curl` are up or not.

This guards against this being a non-symlink in which case it will fail.

For example:
This is perl, v5.10.1 (*) built for x86_64-linux-thread-multi

Fixes #1776.

If any known formula had a dependency on an untapped tap,
Formula#recursive_dependencies would throw an exception, which would be
caught by the outer exception handler, causing the rest of the
dependencies for that formula to be skipped and incomplete output to be
generated.

To fix this, I added a check to avoid analysing the dependencies of
formulae from uninstalled taps.

Additionally, I removed the aforementioned outer exception handler added
in 5fdb89ae, because the only other
place that should be capable of throwing such an exception is the
statement that was surrounded by another wider exception handler in
Homebrew/legacy-homebrew#40682.

Previously, this returned a String, but a Tap instance seems much more
sensible.

I couldn't find anywhere this method was actually used, so the change
shouldn't break anything.

otherwise whenever prefix is unversioned, prefix_Linked will be false

versioned prefix is a directory, but unversioned is a symlink, so
realpath it before computing abv

This had to be added in #1750 to work around special-casing for tabs
generated with Homebrew versions < 1.1.6. Now that 1.1.6 is the current
version, we can remove this hack.

This should apply only for new formulae but we should start gradually
phasing it out for older ones too.

See https://github.com/Homebrew/brew/pull/1771#issuecomment-270179479.

Don't use Homebrew/versions in a test and remove them both from the
OFFICIAL_TAPS list (i.e. `brew search`).

Imports from homebrew/versions are migrated from that tap and then
renamed immediately when they hit homebrew/core. This did not trigger
our previous rename detection so address these to improve the output and
handle migration correctly.

When reproducing issues with software that hasn’t been bottled yet on
your version of macOS it can sometimes be helpful to use `or_later`
bottle functionality i.e. just use the bottle for the latest version of
macOS available. This maps well to the existing `--force-bottle`
argument so it will now act as if the latest bottle has a `or_later`
ending.