and defaults to `Digest::MD5`.

Replaced calls to `::Digest::MD5.hexdigest` with calls to `ActiveSupport::Digest.hexdigest`.

This basically reverts c4d1a4ef

This reverts commit 3420a145, reversing
changes made to afb66a5a.

Stop complaining about not being able to digest dynamic template partial names and simplify error logging to a single line when not

The current code base is not uniform. After some discussion,
we have chosen to go with double quotes by default.

- `partial` option is not used anymore, this was removed in https://github.com/rails/rails/pull/23724

Fix digesting templates with identical logical names when requesting a format other than the first default

Leftover from an earlier commit, chose to fix because I was
just down in this file.

Per request digest caches were added before we had hooks into
different units of work that Rails performs.

As such the most reliable way to prevent stale cache digests
was with a middleware. The middleware prevented staleness in
Action Controller requests.

However, the executor is superior because it should also prevent
staleness when running just Active Job jobs or broadcasting
through Action Cable's server.

we know that all child templates will be considered as "partials", so
the only one that needs detection by name is the root node

Only the root node has injected dependencies, so we don't need to care
about them at lower levels.  This change pushes the injected
dependencies up to where the user passed them in.

this is for backwards compatibility.  We should remove this after
5-0-stable is created because digest calculation should be amortized.
Caching digests of subtrees will speed up digests for other trees that
share the same children.  However, this will also increase memory, and
after the app warms up, then those subtrees will never be reused which
means that memory is wasted.

This moves digest calculation cache on to the details key object.
Before, the digest cache was a class level ivar, and one of the keys was
the hash value of the details key object:

  https://github.com/rails/rails/blob/13c4cc3b5aea02716b7459c0da641438077f5236/actionview/lib/action_view/digestor.rb#L28

An object's hash value is not unique, so it's possible for this cache
key to produce colliding keys with no resolution.  This commit move
cache on to the details key object itself, so we know that the digests
are always unique per details key object.

this commit removes unused code and changes the monitor to a mutex.
Since the digest doesn't recurse on itself anymore, we can just use a
mutex

also remove the EMPTY node since we won't need it

also add an EMPTY sentinel node

this lets us leverage Ruby's kwarg handling (exceptions for missing
params, etc) ASAP which allows us to skip active support method calls
and make sure the exception stack is closer to where the user called the
methods.

we can use kwargs in this case to avoid values_at and except calls on
the options hash

Wasn't removed in 57ac7774.

then we can stop scattering nil checks (e.g. `try`) through the class.

The thread_safe gem is being deprecated and all its code has been merged
into the concurrent-ruby gem. The new class, Concurrent::Map, is exactly
the same as its predecessor except for fixes to two bugs discovered
during the merge.