paquito 0.7.0 → 0.8.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ae7b47980ce61a6b1c3d650217d29a5c196df0c69470a36b74f643d0ded5334c
-  data.tar.gz: 06a85f164bcf5a0e472b4ebf41ce1a9680f71557663ed3176cb7e0df368e460b
+  metadata.gz: 2cfefa4171fc3e9f968368e99c134e864d493c6651ed21aff0e1bc299a95fb7f
+  data.tar.gz: e92cacde6d2aa0867b740f99a001d3ff109f4c88eddb7d89594305a579122af3
 SHA512:
-  metadata.gz: 7fcc3454d84b6cfac7ffe5f3b33780fc23aa609cad375db0e35b43a28582fc3b5dbb6913fc1b3e991b1f034418abe9dbd8c0d673f28e591fc8dee001eefd7c1c
-  data.tar.gz: 2baa0d2518993e2e140f5e989ede70a0d9aa19edb2361414bb6375a6161db0424fcbf7ab03f8f8e531ff0cef9160ea2fa0708afb345b6d3f2d2b3dcdeec332fa
+  metadata.gz: 3e032eb7438d07371862c7f9015b8c0a2f72dc8a93004387f1b151f75d06614dba89bc97da21d3eb83399f1c86909f1dc0bbd5702ec8cc4fef1f5001f512a3da
+  data.tar.gz: e1bd239ec600c28dc5e322b8a8a95daeb539aaada4c9167cf046f5ee5716eb3a6231487b663b534d0db8cb6e1ce2ce53d1dd1b6706184b57f36890edbcbfa663

data/.github/workflows/ci.yml

@@ -23,7 +23,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        ruby: [ ruby-head, '3.0', '2.7', '2.6' ]
+        ruby: [ ruby-head, '3.1', '3.0', '2.7', '2.6' ]
     steps:
       - name: Checkout
         uses: actions/checkout@v3

data/.rubocop.yml

@@ -3,3 +3,6 @@ inherit_gem:
 
 AllCops:
   TargetRubyVersion: 2.6
+
+Style/ClassMethodsDefinitions:
+  Enabled: false

data/CHANGELOG.md

@@ -1,5 +1,7 @@
 # Unreleased
 
+* Introduce `SingleBytePrefixVersionWithStringBypass` (#18, #20).
+
 # 0.7.0
 
 * Make ActiveRecordCoder now encode wether records were persisted or not (#12).

data/Gemfile.lock

@@ -1,18 +1,18 @@
 PATH
   remote: .
   specs:
-    paquito (0.7.0)
+    paquito (0.8.0)
       msgpack (>= 1.5.2)
 
 GEM
   remote: https://rubygems.org/
   specs:
-    activemodel (6.1.5)
-      activesupport (= 6.1.5)
-    activerecord (6.1.5)
-      activemodel (= 6.1.5)
-      activesupport (= 6.1.5)
-    activesupport (6.1.5)
+    activemodel (6.1.6.1)
+      activesupport (= 6.1.6.1)
+    activerecord (6.1.6.1)
+      activemodel (= 6.1.6.1)
+      activesupport (= 6.1.6.1)
+    activesupport (6.1.6.1)
       concurrent-ruby (~> 1.0, >= 1.0.2)
       i18n (>= 1.6, < 2)
       minitest (>= 5.1)
@@ -21,38 +21,38 @@ GEM
     ast (2.4.2)
     benchmark-ips (2.10.0)
     byebug (11.1.3)
-    concurrent-ruby (1.1.9)
-    i18n (1.10.0)
+    concurrent-ruby (1.1.10)
+    i18n (1.12.0)
       concurrent-ruby (~> 1.0)
-    minitest (5.15.0)
-    msgpack (1.5.4)
-    parallel (1.21.0)
-    parser (3.1.1.0)
+    minitest (5.16.3)
+    msgpack (1.6.0)
+    parallel (1.22.1)
+    parser (3.1.2.1)
       ast (~> 2.4.1)
     rainbow (3.1.1)
     rake (13.0.6)
-    regexp_parser (2.2.1)
+    regexp_parser (2.5.0)
     rexml (3.2.5)
-    rubocop (1.25.1)
+    rubocop (1.30.1)
       parallel (~> 1.10)
       parser (>= 3.1.0.0)
       rainbow (>= 2.2.2, < 4.0)
       regexp_parser (>= 1.8, < 3.0)
-      rexml
-      rubocop-ast (>= 1.15.1, < 2.0)
+      rexml (>= 3.2.5, < 4.0)
+      rubocop-ast (>= 1.18.0, < 2.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 1.4.0, < 3.0)
-    rubocop-ast (1.16.0)
+    rubocop-ast (1.21.0)
       parser (>= 3.1.1.0)
     rubocop-shopify (2.5.0)
       rubocop (~> 1.25)
     ruby-progressbar (1.11.0)
-    sorbet-runtime (0.5.9742)
-    sqlite3 (1.4.2)
-    tzinfo (2.0.4)
+    sorbet-runtime (0.5.10365)
+    sqlite3 (1.4.4)
+    tzinfo (2.0.5)
       concurrent-ruby (~> 1.0)
-    unicode-display_width (2.1.0)
-    zeitwerk (2.5.4)
+    unicode-display_width (2.2.0)
+    zeitwerk (2.6.0)
 
 PLATFORMS
   ruby

data/README.md

@@ -1,6 +1,6 @@
 # Paquito
 
-`Paquito` provies utility classes to define optimized and evolutive serializers.
+`Paquito` provides utility classes to define optimized and evolutive serializers.
 
 ## Installation
 
@@ -62,6 +62,16 @@ coder.dump([1]) # => "\x01[1]"
 coder.load("\x00---\n:foo: 42") # => { foo: 42 }
 ```
 
+### `SingleBytePrefixVersionWithStringBypass`
+
+Works like `Paquito::SingleBytePrefixVersion` except that versions `253`, `254` and `255` are reserved for serializing strings
+in an optimized way.
+
+When the object to serialize is an `UTF-8`, `ASCII` or `BINARY` string, rather than invoking the underlying serializer, it simply
+prepends a single byte to the string which indicates the encoding.
+
+The larger the string the larger the speed gain is, e.g. for a 1MB string, it's over 500x faster than going through `MessagePack` or `Marshal`.
+
 ### `CommentPrefixVersion`
 
 Similar to the single byte prefix, but meant to be human readable and to allow for migrating unversioned payloads.

data/benchmark/string-bypass.rb

@@ -0,0 +1,35 @@
+#!/usr/bin/env ruby
+# frozen_string_literal: true
+
+require "bundler/setup"
+require "paquito"
+require "benchmark/ips"
+
+CODEC = Paquito::CodecFactory.build([])
+VERSIONED = Paquito::SingleBytePrefixVersion.new(0, 0 => CODEC)
+BYPASS = Paquito::SingleBytePrefixVersionWithStringBypass.new(0, 0 => CODEC)
+
+[100, 10_000, 1_000_000].each do |size|
+  string = Random.bytes(size).freeze
+  msgpack_payload = VERSIONED.dump(string).freeze
+  bypass_payload = BYPASS.dump(string).freeze
+  marshal_payload = Marshal.dump(string).freeze
+
+  puts " === Read #{size}B ==="
+  Benchmark.ips do |x|
+    x.report("marshal") { Marshal.load(marshal_payload) }
+    x.report("msgpack") { VERSIONED.load(msgpack_payload) }
+    x.report("bypass") { BYPASS.load(bypass_payload) }
+    x.compare!(order: :baseline)
+  end
+
+  puts " === Write #{size}B ==="
+  Benchmark.ips do |x|
+    x.report("marshal") { Marshal.dump(string) }
+    x.report("msgpack") { VERSIONED.dump(string) }
+    x.report("bypass") { BYPASS.dump(string) }
+    x.compare!(order: :baseline)
+  end
+
+  puts
+end

data/lib/paquito/single_byte_prefix_version.rb

@@ -3,8 +3,8 @@
 module Paquito
   class SingleBytePrefixVersion
     def initialize(current_version, coders)
-      @current_version = current_version
-      @coders = coders.transform_values { |c| Paquito.cast(c) }
+      @current_version = validate_version(current_version)
+      @coders = coders.transform_keys { |v| validate_version(v) }.transform_values { |c| Paquito.cast(c) }
       @current_coder = coders.fetch(current_version)
     end
 
@@ -23,5 +23,15 @@ module Paquito
       end
       coder.load(payload.byteslice(1..-1))
     end
+
+    private
+
+    def validate_version(version)
+      unless (0..255).cover?(version)
+        raise ArgumentError, "Invalid version #{version.inspect}, versions must be an integer between 0 and 255"
+      end
+
+      version
+    end
   end
 end

data/lib/paquito/single_byte_prefix_version_with_string_bypass.rb

@@ -0,0 +1,57 @@
+# frozen_string_literal: true
+
+module Paquito
+  class SingleBytePrefixVersionWithStringBypass < SingleBytePrefixVersion
+    UTF8_VERSION = 255
+    BINARY_VERSION = 254
+    ASCII_VERSION = 253
+
+    def dump(object)
+      if object.class == String # We don't want to match subclasses
+        case object.encoding
+        when Encoding::UTF_8
+          UTF8_VERSION.chr(Encoding::BINARY) << object
+        when Encoding::BINARY
+          BINARY_VERSION.chr(Encoding::BINARY) << object
+        when Encoding::US_ASCII
+          ASCII_VERSION.chr(Encoding::BINARY) << object
+        else
+          super
+        end
+      else
+        super
+      end
+    end
+
+    def load(payload)
+      payload_version = payload.getbyte(0)
+      unless payload_version
+        raise UnsupportedCodec, "Missing version byte."
+      end
+
+      case payload_version
+      when UTF8_VERSION
+        payload.byteslice(1..-1).force_encoding(Encoding::UTF_8)
+      when BINARY_VERSION
+        payload.byteslice(1..-1).force_encoding(Encoding::BINARY)
+      when ASCII_VERSION
+        payload.byteslice(1..-1).force_encoding(Encoding::US_ASCII)
+      else
+        coder = @coders.fetch(payload_version) do
+          raise UnsupportedCodec, "Unsupported packer version #{payload_version}"
+        end
+        coder.load(payload.byteslice(1..-1))
+      end
+    end
+
+    private
+
+    def validate_version(version)
+      unless (0..252).cover?(version)
+        raise ArgumentError, "Invalid version #{version.inspect}, versions must be an integer between 0 and 252"
+      end
+
+      version
+    end
+  end
+end

data/lib/paquito/types.rb

@@ -160,7 +160,7 @@ module Paquito
 
           packer.write(value.to_h)
         end,
-        unpacker: ->(unpacker) { HashWithIndifferentAccess.new(unpacker.read) },
+        unpacker: ->(unpacker) { ActiveSupport::HashWithIndifferentAccess.new(unpacker.read) },
         recursive: true,
       },
       "ActiveSupport::TimeWithZone" => {
@@ -206,11 +206,11 @@ module Paquito
     class << self
       def register(factory, types)
         types.each do |type|
-          name = type.name
-
           # Up to Rails 7 ActiveSupport::TimeWithZone#name returns "Time"
-          if name == "Time" && defined?(ActiveSupport::TimeWithZone)
-            name = "ActiveSupport::TimeWithZone" if type == ActiveSupport::TimeWithZone
+          name = if defined?(ActiveSupport::TimeWithZone) && type == ActiveSupport::TimeWithZone
+            "ActiveSupport::TimeWithZone"
+          else
+            type.name
           end
 
           type_attributes = TYPES.fetch(name)

data/lib/paquito/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Paquito
-  VERSION = "0.7.0"
+  VERSION = "0.8.0"
 end

data/lib/paquito.rb

@@ -16,6 +16,7 @@ require "paquito/safe_yaml"
 require "paquito/conditional_compressor"
 require "paquito/cache_entry_coder"
 require "paquito/single_byte_prefix_version"
+require "paquito/single_byte_prefix_version_with_string_bypass"
 require "paquito/comment_prefix_version"
 require "paquito/types"
 require "paquito/codec_factory"

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: paquito
 version: !ruby/object:Gem::Version
-  version: 0.7.0
+  version: 0.8.0
 platform: ruby
 authors:
 - Jean Boussier
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2022-08-08 00:00:00.000000000 Z
+date: 2022-11-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: msgpack
@@ -41,6 +41,7 @@ files:
 - README.md
 - Rakefile
 - benchmark/msgpack-pooling.rb
+- benchmark/string-bypass.rb
 - bin/console
 - bin/setup
 - dev.yml
@@ -57,6 +58,7 @@ files:
 - lib/paquito/safe_yaml.rb
 - lib/paquito/serialized_column.rb
 - lib/paquito/single_byte_prefix_version.rb
+- lib/paquito/single_byte_prefix_version_with_string_bypass.rb
 - lib/paquito/struct.rb
 - lib/paquito/translate_errors.rb
 - lib/paquito/typed_struct.rb