Switch parser over to prism

.github/workflows/main.yml

@@ -9,19 +9,20 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
+        os:
+        - ubuntu-latest
+        - macos-latest
+        - windows-latest
         ruby:
-        - '2.7.0'
-        - '3.0'
-        - '3.1'
         - '3.2'
         - '3.3'
         - '3.4'
         - truffleruby-head
+        exclude:
+        - os: windows-latest
+          ruby: truffleruby-head
     name: CI
-    runs-on: ubuntu-latest
-    env:
-      CI: true
-      # TESTOPTS: --verbose
+    runs-on: ${{ matrix.os }}
     steps:
     - uses: actions/checkout@master
     - uses: ruby/setup-ruby@v1
@@ -34,8 +35,6 @@ jobs:
   check:
     name: Check
     runs-on: ubuntu-latest
-    env:
-      CI: true
     steps:
     - uses: actions/checkout@master
     - uses: ruby/setup-ruby@v1

.rubocop.yml

@@ -5,7 +5,7 @@ AllCops:
   DisplayStyleGuide: true
   NewCops: enable
   SuggestExtensions: false
-  TargetRubyVersion: 2.7
+  TargetRubyVersion: 3.2
   Exclude:
   - '{.git,.github,.ruby-lsp,bin,coverage,doc,pkg,sorbet,spec,test/fixtures,vendor,tmp}/**/*'
   - test.rb
@@ -14,7 +14,7 @@ Gemspec/DevelopmentDependencies:
   Enabled: false
 
 Layout/LineLength:
-  Max: 80
+  Max: 100
 
 Lint/AmbiguousBlockAssociation:
   Enabled: false
@@ -70,6 +70,9 @@ Naming/MethodName:
 Naming/MethodParameterName:
   Enabled: false
 
+Naming/PredicateMethod:
+  Enabled: false
+
 Naming/RescuedExceptionsVariableName:
   PreferredName: error
 
@@ -79,12 +82,18 @@ Naming/VariableNumber:
 Security/Eval:
   Enabled: false
 
+Style/AccessModifierDeclarations:
+  Enabled: false
+
 Style/AccessorGrouping:
   Enabled: false
 
 Style/Alias:
   Enabled: false
 
+Style/BlockDelimiters:
+  Enabled: false
+
 Style/CaseEquality:
   Enabled: false
 
@@ -151,6 +160,9 @@ Style/Next:
 Style/NumericPredicate:
   Enabled: false
 
+Style/OptionalBooleanParameter:
+  Enabled: false
+
 Style/ParallelAssignment:
   Enabled: false
 

Gemfile

@@ -4,4 +4,8 @@ source "https://rubygems.org"
 
 gemspec
 
-gem "fiddle"
+gem "bundler"
+gem "minitest"
+gem "rake"
+gem "rubocop"
+gem "simplecov"

Gemfile.lock

@@ -2,23 +2,21 @@ PATH
   remote: .
   specs:
     syntax_tree (6.3.0)
-      prettier_print (>= 1.2.0)
+      prism
 
 GEM
   remote: https://rubygems.org/
   specs:
     ast (2.4.3)
     docile (1.4.1)
-    fiddle (1.1.8)
-    json (2.15.1)
+    json (2.15.2)
     language_server-protocol (3.17.0.5)
     lint_roller (1.1.0)
     minitest (5.26.0)
     parallel (1.27.0)
-    parser (3.3.9.0)
+    parser (3.3.10.0)
       ast (~> 2.4.1)
       racc
-    prettier_print (1.2.1)
     prism (1.6.0)
     racc (1.8.1)
     rainbow (3.1.1)
@@ -43,27 +41,24 @@ GEM
       docile (~> 1.1)
       simplecov-html (~> 0.11)
       simplecov_json_formatter (~> 0.1)
-    simplecov-html (0.13.1)
+    simplecov-html (0.13.2)
     simplecov_json_formatter (0.1.4)
     unicode-display_width (3.2.0)
       unicode-emoji (~> 4.1)
     unicode-emoji (4.1.0)
 
 PLATFORMS
-  arm64-darwin-21
-  ruby
-  x86_64-darwin-19
-  x86_64-darwin-21
+  arm64-darwin-23
+  x64-mingw-ucrt
   x86_64-linux
 
 DEPENDENCIES
   bundler
-  fiddle
   minitest
   rake
   rubocop
   simplecov
   syntax_tree!
 
 BUNDLED WITH
-   2.3.6
+   2.4.19

README.md

@@ -2,14 +2,12 @@
   <img alt="Syntax Tree" height="400px" src="./doc/logo.svg">
 </div>
 
-# SyntaxTree
+# Syntax Tree
 
 [![Build Status](https://github.com/ruby-syntax-tree/syntax_tree/actions/workflows/main.yml/badge.svg)](https://github.com/ruby-syntax-tree/syntax_tree/actions/workflows/main.yml)
 [![Gem Version](https://img.shields.io/gem/v/syntax_tree.svg)](https://rubygems.org/gems/syntax_tree)
 
-Syntax Tree is a suite of tools built on top of the internal CRuby parser. It provides the ability to generate a syntax tree from source, as well as the tools necessary to inspect and manipulate that syntax tree. It can be used to build formatters, linters, language servers, and more.
-
-It is built with only standard library dependencies. It additionally ships with a plugin system so that you can build your own syntax trees from other languages and incorporate these tools.
+Syntax Tree is fast Ruby parser built on top of the [prism](https://github.com/ruby/prism) Ruby parser. It is built with only standard library dependencies.
 
 - [Installation](#installation)
 - [CLI](#cli)
@@ -19,11 +17,9 @@ It is built with only standard library dependencies. It additionally ships with
   - [Configuration](#configuration)
   - [Globbing](#globbing)
 - [Language server](#language-server)
-  - [textDocument/formatting](#textdocumentformatting)
 - [Customization](#customization)
   - [Ignoring code](#ignoring-code)
   - [Plugins](#plugins)
-  - [Languages](#languages)
 - [Integration](#integration)
   - [Rake](#rake)
   - [RuboCop](#rubocop)
@@ -144,16 +140,28 @@ stree write --ignore-files='db/**/*.rb' '**/*.rb'
 
 ### Configuration
 
-Any of the above CLI commands can also read configuration options from a `.streerc` file in the directory where the commands are executed.
+All of the above commands accept additional configuration options. Those are:
+
+- `--print-width=?` - The print width is the suggested line length that should be used when formatting the source. Note that this is not a hard limit like a linter. Instead, it is used as a guideline for how long lines _should_ be. For example, if you have the following code:
+
+```ruby
+foo do
+  bar
+end
+```
+
+In this case, the formatter will see that the block fits into the print width and will rewrite it using the `{}` syntax. This will actually make the line longer than originally written. This is why it is helpful to think of it as a suggestion, rather than a limit.
+- `--preferred-quote=?` - The quote to use for string and character literals. This can be either `"` or `'`. It is "preferred" because in the case that the formatter encounters a string that contains interpolation or certain escape sequences, it will not attempt to change the quote style to avoid accidentally changing the semantic meaning of the code.
+- `--[no-]trailing-comma` - Whether or not to add trailing commas to multiline array literals, hash literals, and method calls that can support trailing commas.
 
-This should be a text file with each argument on a separate line.
+Any of the above CLI commands can also read configuration options from a `.streerc` file in the directory where the commands are executed. This should be a text file with each argument on a separate line.
 
 ```txt
 --print-width=100
---plugins=plugin/trailing_comma
+--trailing-comma
 ```
 
-If this file is present, it will _always_ be used for CLI commands. You can also pass options from the command line as in the examples above. The options in the `.streerc` file are passed to the CLI first, then the arguments from the command line. In the case of exclusive options (e.g. `--print-width`), this means that the command line options override what's in the config file. In the case of options that can take multiple inputs (e.g. `--plugins`), the effect is additive. That is, the plugins passed from the command line will be loaded _in addition to_ the plugins in the config file.
+If this file is present, it will _always_ be used for CLI commands. The options in the `.streerc` file are passed to the CLI first, then the arguments from the command line. In the case of exclusive options (e.g. `--print-width`), this means that the command line options override what's in the config file. In the case of options that can take multiple inputs (e.g. `--plugins`), the effect is additive. That is, the plugins passed from the command line will be loaded _in addition to_ the plugins in the config file.
 
 ### Globbing
 
@@ -220,32 +228,8 @@ To register plugins, define a file somewhere in your load path named `syntax_tre
 
 * `plugin/single_quotes` - This will change all of your string literals to use single quotes instead of the default double quotes.
 * `plugin/trailing_comma` - This will put trailing commas into multiline array literals, hash literals, and method calls that can support trailing commas.
-* `plugin/disable_auto_ternary` - This will prevent the automatic conversion of `if ... else` to ternary expressions.
-
-If you're using Syntax Tree as a library, you can require those files directly or manually pass those options to the formatter initializer through the `SyntaxTree::Formatter::Options` class.
-
-### Languages
-
-To register a new language, call:
-
-```ruby
-SyntaxTree.register_handler(".mylang", MyLanguage)
-```
-
-In this case, whenever the CLI encounters a filepath that ends with the given extension, it will invoke methods on `MyLanguage` instead of `SyntaxTree` itself. To make sure your object conforms to each of the necessary APIs, it should implement:
-
-* `MyLanguage.read(filepath)` - usually this is just an alias to `File.read(filepath)`, but if you need anything else that hook is here.
-* `MyLanguage.parse(source)` - this should return the syntax tree corresponding to the given source. Those objects should implement the `pretty_print` interface.
-* `MyLanguage.format(source)` - this should return the formatted version of the given source.
 
-Below are listed all of the "official" language plugins hosted under the same GitHub organization, which can be used as references for how to implement other plugins.
-
-* [bf](https://github.com/ruby-syntax-tree/syntax_tree-bf) for the [brainf*** language](https://esolangs.org/wiki/Brainfuck).
-* [css](https://github.com/ruby-syntax-tree/syntax_tree-css) for the [CSS stylesheet language](https://www.w3.org/Style/CSS/).
-* [haml](https://github.com/ruby-syntax-tree/syntax_tree-haml) for the [Haml template language](https://haml.info/).
-* [json](https://github.com/ruby-syntax-tree/syntax_tree-json) for the [JSON notation language](https://www.json.org/).
-* [rbs](https://github.com/ruby-syntax-tree/syntax_tree-rbs) for the [RBS type language](https://github.com/ruby/rbs).
-* [xml](https://github.com/ruby-syntax-tree/syntax_tree-xml) for the [XML markup language](https://www.w3.org/XML/).
+If you're using Syntax Tree as a library, you can require those files directly or manually pass those options to the formatter initializer through the `SyntaxTree::Options` class.
 
 ## Integration
 
@@ -256,12 +240,12 @@ Syntax Tree's goal is to seamlessly integrate into your workflow. To this end, i
 Syntax Tree ships with the ability to define [rake](https://github.com/ruby/rake) tasks that will trigger runs of the CLI. To define them in your application, add the following configuration to your `Rakefile`:
 
 ```ruby
-require "syntax_tree/rake_tasks"
+require "syntax_tree/rake"
 SyntaxTree::Rake::CheckTask.new
 SyntaxTree::Rake::WriteTask.new
 ```
 
-These calls will define `rake stree:check` and `rake stree:write` (equivalent to calling `stree check` and `stree write` with the CLI respectively). You can configure them by either passing arguments to the `new` method or by using a block.
+These calls will define `rake stree:check` and `rake stree:write` (equivalent to calling `stree check` and `stree write` with the CLI respectively). You can configure them by either passing arguments to the `new` method or by using a block. In addition to the regular configuration options used for the formatter, there are a few additional options specific to the rake tasks.
 
 #### `name`
 
@@ -292,26 +276,6 @@ SyntaxTree::Rake::WriteTask.new do |t|
 end
 ```
 
-#### `print_width`
-
-If you want to use a different print width from the default (80), you can pass that to the `print_width` field, as in:
-
-```ruby
-SyntaxTree::Rake::WriteTask.new do |t|
-  t.print_width = 100
-end
-```
-
-#### `plugins`
-
-If you're running Syntax Tree with plugins (either your own or the pre-built ones), you can pass that to the `plugins` field, as in:
-
-```ruby
-SyntaxTree::Rake::WriteTask.new do |t|
-  t.plugins = ["plugin/single_quotes"]
-end
-```
-
 ### RuboCop
 
 RuboCop and Syntax Tree serve different purposes, but there is overlap with some of RuboCop's functionality. Syntax Tree provides a RuboCop configuration file to disable rules that are redundant with Syntax Tree. To use this configuration file, add the following snippet to the top of your project's `.rubocop.yml`:

Rakefile

@@ -2,7 +2,7 @@
 
 require "bundler/gem_tasks"
 require "rake/testtask"
-require "syntax_tree/rake_tasks"
+require "syntax_tree/rake"
 
 Rake::TestTask.new(:test) do |t|
   t.libs << "test"
@@ -14,23 +14,7 @@ task default: :test
 
 configure = ->(task) do
   task.source_files =
-    FileList[
-      %w[
-        Gemfile
-        Rakefile
-        syntax_tree.gemspec
-        lib/**/*.rb
-        tasks/*.rake
-        test/*.rb
-      ]
-    ]
-
-  # Since Syntax Tree supports back to Ruby 2.7.0, we need to make sure that we
-  # format our code such that it's compatible with that version. This actually
-  # has very little effect on the output, the only change at the moment is that
-  # Ruby < 2.7.3 didn't allow a newline before the closing brace of a hash
-  # pattern.
-  task.target_ruby_version = Gem::Version.new("2.7.0")
+    FileList[%w[Gemfile Rakefile syntax_tree.gemspec lib/**/*.rb tasks/*.rake test/*.rb]]
 end
 
 SyntaxTree::Rake::CheckTask.new(&configure)

bin/profile

@@ -6,15 +6,15 @@ require "bundler/inline"
 gemfile do
   source "https://rubygems.org"
   gem "stackprof"
-  gem "prettier_print"
+  gem "prism"
 end
 
 $:.unshift(File.expand_path("../lib", __dir__))
 require "syntax_tree"
 
 StackProf.run(mode: :cpu, out: "tmp/profile.dump", raw: true) do
   Dir[File.join(RbConfig::CONFIG["libdir"], "**/*.rb")].each do |filepath|
-    SyntaxTree.format(SyntaxTree.read(filepath))
+    SyntaxTree.format_file(filepath)
   end
 end