ruby-core@ruby-lang.org archive (unofficial mirror)
 help / color / mirror / Atom feed
From: "AMomchilov (Alexander Momchilov) via ruby-core" <ruby-core@ml.ruby-lang.org>
To: ruby-core@ml.ruby-lang.org
Cc: "AMomchilov (Alexander Momchilov)" <noreply@ruby-lang.org>
Subject: [ruby-core:117832] [Ruby master Feature#20300] Hash: set value and get pre-existing value in one call
Date: Sat, 11 May 2024 16:21:55 +0000 (UTC)	[thread overview]
Message-ID: <redmine.journal-108239.20240511162154.51722@ruby-lang.org> (raw)
In-Reply-To: redmine.issue-20300.20240226024229.51722@ruby-lang.org

Issue #20300 has been updated by AMomchilov (Alexander Momchilov).


mame (Yusuke Endoh) wrote in #note-24:

> It was originally intended as a method to improve the efficiency of `Set#add?`, but the use case was shifted to a method for thread safety. This history makes the use case less persuasive.

The immediate need and use case is for that performance improvement to Set.

The discussion of thread safety is just an incidental benefit of this approach, as compared to the count-based alternative that was proposed.

> If it is just an internal method for efficiency, a long and verbose name may be preferred.

I would be okay with this, if there’s no appetite for a public, short-named API for this. Perhaps some underscored name that communicates that this is “library internal” to the stdlib.

> You may want to explain the concrete example of the use case of thread safety for Hash value exchange, this proposed API is really sufficient for that example use case,

I don’t think I can make a better case than the concrete implementation I’ve already proposed, and its immediate application to the Set performance problem I’ve identified and fixed, using this API.

If others want to chime in with the use cases in they have in mind, I’d love to hear those.

> and what API and name are given to similar feature in other languages.

Let’s first decide if we want this API, in *any* form. Once we agree on that, I’d be happy to dive into details like naming, and comparisons to other languages.

----------------------------------------
Feature #20300: Hash: set value and get pre-existing value in one call
https://bugs.ruby-lang.org/issues/20300#change-108239

* Author: AMomchilov (Alexander Momchilov)
* Status: Open
----------------------------------------
When using a Hash, sometimes you want to set a new value, **and** see what was already there. Today, you **have** to do this in two steps:

```ruby
h = { k: "old value" }

# 1. Do a look-up for `:k`.
old_value = h[:k]
# 2. Do another look-up for `:k`, even though we just did that!
h[:k] = "new value"

use(old_value)
```

This requires two separate `Hash` look-ups for `:k`. This is fine for symbols, but is expensive if computing `#hash` or `#eql?` is expensive for the key. It's impossible to work around this today from pure Ruby code.

One example use case is `Set#add?`. See https://bugs.ruby-lang.org/issues/20301 for more details.

I propose adding `Hash#exchange_value`, which has semantics are similar to this Ruby snippet:

```ruby
class Hash
  # Exact method name TBD.
  def exchange_value(key, new_value)
    old_value = self[key]
    self[key] = new_value
    old_value
  end
end
```

... except it'll be implemented in C, with modifications to `tbl_update` that achieves this with a hash-lookup. 

I'm opening to alternative name suggestions. @nobu came up with `exchange_value`, which I think is great.

Here's a PR with a PoC implementation: https://github.com/ruby/ruby/pull/10092

```ruby
h = { k: "old value" }

# Does only a single hash look-up
old_value = h.exchange_value(:k, "new value")

use(old_value)
```



-- 
https://bugs.ruby-lang.org/
 ______________________________________________
 ruby-core mailing list -- ruby-core@ml.ruby-lang.org
 To unsubscribe send an email to ruby-core-leave@ml.ruby-lang.org
 ruby-core info -- https://ml.ruby-lang.org/mailman3/postorius/lists/ruby-core.ml.ruby-lang.org/

      parent reply	other threads:[~2024-05-11 16:22 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-26  2:42 [ruby-core:116940] [Ruby master Feature#20300] Hash: set value and get pre-existing value in one call AMomchilov (Alexander Momchilov) via ruby-core
2024-02-26  4:18 ` [ruby-core:116943] " nobu (Nobuyoshi Nakada) via ruby-core
2024-02-26 10:16 ` [ruby-core:116947] " Eregon (Benoit Daloze) via ruby-core
2024-02-26 10:38 ` [ruby-core:116948] " rubyFeedback (robert heiler) via ruby-core
2024-02-26 16:21 ` [ruby-core:116958] " MaxLap (Maxime Lapointe) via ruby-core
2024-02-27 16:42 ` [ruby-core:116973] " matheusrich (Matheus Richard) via ruby-core
2024-02-27 16:46 ` [ruby-core:116975] " AMomchilov (Alexander Momchilov) via ruby-core
2024-02-27 17:34 ` [ruby-core:116978] " matheusrich (Matheus Richard) via ruby-core
2024-02-27 21:30 ` [ruby-core:116981] " AMomchilov (Alexander Momchilov) via ruby-core
2024-02-27 21:39 ` [ruby-core:116982] " matheusrich (Matheus Richard) via ruby-core
2024-03-14  5:22 ` [ruby-core:117136] " nobu (Nobuyoshi Nakada) via ruby-core
2024-03-14 10:12 ` [ruby-core:117156] " mame (Yusuke Endoh) via ruby-core
2024-03-14 15:05 ` [ruby-core:117178] " Eregon (Benoit Daloze) via ruby-core
2024-03-14 15:06 ` [ruby-core:117179] " Eregon (Benoit Daloze) via ruby-core
2024-03-15  1:49 ` [ruby-core:117193] " AMomchilov (Alexander Momchilov) via ruby-core
2024-03-15  2:46 ` [ruby-core:117194] " shyouhei (Shyouhei Urabe) via ruby-core
2024-04-17  4:30 ` [ruby-core:117545] " nobu (Nobuyoshi Nakada) via ruby-core
2024-04-17 10:57 ` [ruby-core:117560] " mame (Yusuke Endoh) via ruby-core
2024-05-11 16:21 ` AMomchilov (Alexander Momchilov) via ruby-core [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-list from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.ruby-lang.org/en/community/mailing-lists/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=redmine.journal-108239.20240511162154.51722@ruby-lang.org \
    --to=ruby-core@ruby-lang.org \
    --cc=noreply@ruby-lang.org \
    --cc=ruby-core@ml.ruby-lang.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).