sorbet-eraser 0.4.0 → 0.5.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 3859ee512cb01ed5b84df15af3dedaa629c30d122e80a7436821841f87184562
-  data.tar.gz: 8392610b8fb03ea78d2572b684c2ccf9b4806f082043e3b2614550a370eeea13
+  metadata.gz: 420576226d037471f75759599ea53397cd251dd546431b24e9d3cf556285902a
+  data.tar.gz: 7c6d09a382019245254f55d0a9eff1b4cd2430d066c364efde59567ada2f56c0
 SHA512:
-  metadata.gz: 02ee32fc8bcf3d1951ddd9577a4f46c181a01f196eb6c2af34bb8bb0b68a0ebc3fb28515d0ca1e391d2d1ee305d3012aafd943918df643dd2a3a44c5272a38b2
-  data.tar.gz: fa7b5e80e3fda49d50df015532110f6b4050450c5348f4d9f0ac6f40c8e2b411882a5348b38e305419b74e0bf6e833b887997bb6757e440fb3acc552e286d7e4
+  metadata.gz: 5478636318a06fe00d279813bd724659d6653efef7364ea583e11566e455cf9528e8c14338dfdab11298e329225ca9ef2c0847df4125b0edea09f4ca5a953452
+  data.tar.gz: 26d79c3b183ef5f7a629205b94e61a09862db83ca50884d68e1e7d22954b1543355787e7fd798223ff8b483bdbdcbb5c14a49959d72f19a12bfcedd13f8f51e8

data/.gitignore

@@ -6,3 +6,5 @@
 /pkg/
 /spec/reports/
 /tmp/
+
+test.rb

data/CHANGELOG.md

@@ -6,6 +6,15 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/) a
 
 ## [Unreleased]
 
+## [0.5.0] - 2023-07-13
+
+### Added
+
+- Replace `typed: strict` comments with empty comments.
+- Replace all `typed:` sigil comments with empty comments instead of `typed: ignore`. Do this because `typed: ignore` longer than other options, which can cause issues with byte ranges, and violates an assumption by this gem that it is only erasing, not adding content.
+- Add a `--verify` option to the CLI to ensure output is valid Ruby.
+- Enhance `sorbet/eraser/autoload` to hook into `load_iseq` even if bootsnap is not present.
+
 ## [0.4.0] - 2023-07-03
 
 ### Added
@@ -41,7 +50,8 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/) a
 
 - Require MFA for releasing.
 
-[unreleased]: https://github.com/kddnewton/sorbet-eraser/compare/v0.4.0...HEAD
+[unreleased]: https://github.com/kddnewton/sorbet-eraser/compare/v0.4.1...HEAD
+[0.4.1]: https://github.com/kddnewton/sorbet-eraser/compare/v0.4.0...v0.4.1
 [0.4.0]: https://github.com/kddnewton/sorbet-eraser/compare/v0.3.1...v0.4.0
 [0.3.1]: https://github.com/kddnewton/sorbet-eraser/compare/v0.3.0...v0.3.1
 [0.3.0]: https://github.com/kddnewton/sorbet-eraser/compare/v0.2.0...v0.3.0

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    sorbet-eraser (0.4.0)
+    sorbet-eraser (0.5.0)
 
 GEM
   remote: https://rubygems.org/

data/README.md

@@ -27,7 +27,7 @@ end
 will be transformed into
 
 ```ruby
-             
+#
 
 class HelloWorld
                
@@ -95,30 +95,30 @@ If you used any runtime structures like `T::Struct` or `T::Enum` you'll need a r
 
 Below is a table of the status of each `sorbet-runtime` construct and its current support status.
 
-| Construct                                           | Status | Replacement       |
-| --------------------------------------------------- | ------ | ----------------- |
-| `# typed: foo`                                      | ✅     | `# typed: ignore` |
-| `extend T::*`                                       | ✅     | Shimmed           |
-| `abstract!`, `final!`, `interface!`, `sealed!`      | ✅     | Shimmed           |
-| `mixes_in_class_methods(*)`, `requires_ancestor(*)` | ✅     | Shimmed           |
-| `type_member(*)`, `type_template(*)`                | ✅     | Shimmed           |
-| `class Foo < T::Enum`                               | ✅     | Shimmed           |
-| `class Foo < T::InexactStruct`                      | 🛠     | Shimmed           |
-| `class Foo < T::Struct`                             | 🛠     | Shimmed           |
-| `class Foo < T::ImmutableStruct`                    | 🛠     | Shimmed           |
-| `include T::Props`                                  | 🛠     | Shimmed           |
-| `include T::Props::Serializable`                    | 🛠     | Shimmed           |
-| `include T::Props::Constructor`                     | 🛠     | Shimmed           |
-| `sig`                                               | ✅     | Removed           |
-| `T.absurd(foo)`                                     | ✅     | Shimmed           |
-| `T.assert_type!(foo, bar)`                          | ✅     | `foo`             |
-| `T.bind(self, foo)`                                 | ✅     | `self`            |
-| `T.cast(foo, bar)`                                  | ✅     | `foo`             |
-| `T.let(foo, bar)`                                   | ✅     | `foo`             |
-| `T.must(foo)`                                       | ✅     | `foo`             |
-| `T.reveal_type(foo)`                                | ✅     | `foo`             |
-| `T.type_alias { foo }`                              | ✅     | Shimmed           |
-| `T.unsafe(foo)`                                     | ✅     | `foo`             |
+| Construct                                           | Status | Replacement |
+| --------------------------------------------------- | ------ | ----------- |
+| `# typed: foo`                                      | ✅      | `#        ` |
+| `extend T::*`                                       | ✅      | Shimmed     |
+| `abstract!`, `final!`, `interface!`, `sealed!`      | ✅      | Shimmed     |
+| `mixes_in_class_methods(*)`, `requires_ancestor(*)` | ✅      | Shimmed     |
+| `type_member(*)`, `type_template(*)`                | ✅      | Shimmed     |
+| `class Foo < T::Enum`                               | ✅      | Shimmed     |
+| `class Foo < T::InexactStruct`                      | 🛠      | Shimmed     |
+| `class Foo < T::Struct`                             | 🛠      | Shimmed     |
+| `class Foo < T::ImmutableStruct`                    | 🛠      | Shimmed     |
+| `include T::Props`                                  | 🛠      | Shimmed     |
+| `include T::Props::Serializable`                    | 🛠      | Shimmed     |
+| `include T::Props::Constructor`                     | 🛠      | Shimmed     |
+| `sig`                                               | ✅      | Removed     |
+| `T.absurd(foo)`                                     | ✅      | Shimmed     |
+| `T.assert_type!(foo, bar)`                          | ✅      | `foo`       |
+| `T.bind(self, foo)`                                 | ✅      | `self`      |
+| `T.cast(foo, bar)`                                  | ✅      | `foo`       |
+| `T.let(foo, bar)`                                   | ✅      | `foo`       |
+| `T.must(foo)`                                       | ✅      | `foo`       |
+| `T.reveal_type(foo)`                                | ✅      | `foo`       |
+| `T.type_alias { foo }`                              | ✅      | Shimmed     |
+| `T.unsafe(foo)`                                     | ✅      | `foo`       |
 
 In the above table, for `Status`:
 

data/lib/sorbet/eraser/autoload.rb

@@ -2,21 +2,22 @@
 
 require "sorbet/eraser"
 
-# Hook into bootsnap so that before the source is compiled through
-# RubyVM::InstructionSequence it gets erased through the eraser.
-if RubyVM::InstructionSequence.method_defined?(:load_iseq)
-  load_iseq, = RubyVM::InstructionSequence.method(:load_iseq).source_location
-
-  if load_iseq.include?("/bootsnap/")
-    module Sorbet::Eraser::Patch
-      def input_to_storage(contents, filepath)
-        erased = Sorbet::Eraser.erase(contents)
-        RubyVM::InstructionSequence.compile(erased, filepath, filepath).to_binary
-      rescue SyntaxError
-        raise ::Bootsnap::CompileCache::Uncompilable, "syntax error"
-      end
+if RubyVM::InstructionSequence.method_defined?(:load_iseq) &&
+   RubyVM::InstructionSequence.method(:load_iseq).source_location[0].include?("/bootsnap/")
+  # If the load_iseq method is defined by bootsnap, then we need to override it.
+  module Sorbet::Eraser::Patch
+    def input_to_storage(contents, filepath)
+      super(Sorbet::Eraser.erase(contents), filepath)
     end
+  end
 
-    Bootsnap::CompileCache::ISeq.singleton_class.prepend(Sorbet::Eraser::Patch)
+  Bootsnap::CompileCache::ISeq.singleton_class.prepend(Sorbet::Eraser::Patch)
+else
+  # Otherwise if the method isn't defined by bootsnap, then we'll define it
+  # ourselves.
+  def (RubyVM::InstructionSequence).load_iseq(filepath)
+    contents = File.read(filepath)
+    erased = Sorbet::Eraser.erase(contents)
+    RubyVM::InstructionSequence.compile(erased, filepath, filepath)
   end
 end

data/lib/sorbet/eraser/cli.rb

@@ -7,9 +7,10 @@ module Sorbet
     class CLI
       POOL_SIZE = 4
 
-      attr_reader :filepaths
+      attr_reader :verify, :filepaths
 
-      def initialize(filepaths)
+      def initialize(verify, filepaths)
+        @verify = verify
         @filepaths = filepaths
       end
 
@@ -28,20 +29,36 @@ module Sorbet
                 break if filepath == :eoq
                 process(filepath)
               end
-            end
+            end.tap { |thread| thread.abort_on_exception = true }
           end
 
         workers.each(&:join)
       end
 
       def self.start(argv)
-        new(argv.flat_map { |pattern| Dir.glob(pattern) }).start
+        verify = false
+
+        if argv.first == "--verify"
+          verify = true
+          argv.shift
+        end
+
+        filepaths = []
+        argv.each { |pattern| filepaths.concat(Dir.glob(pattern)) }
+
+        new(verify, filepaths).start
       end
 
       private
 
       def process(filepath)
-        File.write(filepath, Eraser.erase(File.read(filepath)))
+        contents = Eraser.erase(File.read(filepath))
+
+        if verify && Ripper.sexp_raw(contents).nil?
+          warn("Internal error while parsing #{filepath}")
+        else
+          File.write(filepath, contents)
+        end
       rescue Parser::ParsingError => error
         warn("Could not parse #{filepath}: #{error}")
       rescue => error

data/lib/sorbet/eraser/parser.rb

@@ -136,18 +136,18 @@ module Sorbet
 
       # Better location information for aref.
       def on_aref(recv, arg)
-        rend = arg.range.end + source[arg.range.end..].index("]") + 1
+        rend = arg.range.end + source.byteslice(arg.range.end..).index("]") + 1
         Node.new(:aref, [recv, arg], recv.range.begin...rend)
       end
 
       # Better location information for arg_paren.
       def on_arg_paren(arg)
         if arg
-          rbegin = source[..arg.range.begin].rindex("(")
-          rend = arg.range.end + source[arg.range.end..].index(")") + 1
+          rbegin = source.byteslice(..arg.range.begin).rindex("(")
+          rend = arg.range.end + source.byteslice(arg.range.end..).index(")") + 1
           Node.new(:arg_paren, [arg], rbegin...rend)
         else
-          segment = source[..loc]
+          segment = source.byteslice(..loc)
           Node.new(:arg_paren, [arg], segment.rindex("(")...(segment.rindex(")") + 1))
         end
       end
@@ -159,11 +159,11 @@ module Sorbet
       def on_array(arg)
         case arg&.event
         when nil
-          segment = source[..loc]
+          segment = source.byteslice(..loc)
           Node.new(:array, [arg], segment.rindex("[")...(segment.rindex("]") + 1))
         when :qsymbols, :qwords, :symbols, :words
-          rbegin = source[...arg.range.begin].rindex(LISTS.fetch(arg.event))
-          rend = source[arg.range.end..].index(TERMINATORS.fetch(source[rbegin + 2]) { source[rbegin + 2] }) + arg.range.end + 1
+          rbegin = source.byteslice(...arg.range.begin).rindex(LISTS.fetch(arg.event))
+          rend = source.byteslice(arg.range.end..).index(TERMINATORS.fetch(source.byteslice(rbegin + 2)) { source.byteslice(rbegin + 2) }) + arg.range.end + 1
           Node.new(:array, [arg], rbegin...rend)
         else
           Node.new(:array, [arg], arg.range)
@@ -173,14 +173,14 @@ module Sorbet
       # Better location information for brace_block.
       def on_brace_block(params, body)
         if params || body.range
-          rbegin = source[...(params || body).range.begin].rindex("{")
+          rbegin = source.byteslice(...(params || body).range.begin).rindex("{")
 
           rend = body.range&.end || params.range.end
-          rend = rend + source[rend..].index("}") + 1
+          rend = rend + source.byteslice(rend..).index("}") + 1
 
           Node.new(:brace_block, [params, body], rbegin...rend)
         else
-          segment = source[..loc]
+          segment = source.byteslice(..loc)
           Node.new(:brace_block, [params, body], segment.rindex("{")...(segment.rindex("}") + 1))
         end
       end
@@ -188,14 +188,14 @@ module Sorbet
       # Better location information for do_block.
       def on_do_block(params, body)
         if params || body.range
-          rbegin = source[...(params || body).range.begin].rindex("do")
+          rbegin = source.byteslice(...(params || body).range.begin).rindex("do")
 
           rend = body.range&.end || params.range.end
-          rend = rend + source[rend..].index("end") + 3
+          rend = rend + source.byteslice(rend..).index("end") + 3
 
           Node.new(:do_block, [params, body], rbegin...rend)
         else
-          segment = source[..loc]
+          segment = source.byteslice(..loc)
           Node.new(:do_block, [params, body], segment.rindex("do")...(segment.rindex("end") + 3))
         end
       end
@@ -205,7 +205,7 @@ module Sorbet
         if arg
           Node.new(:hash, [arg], arg.range)
         else
-          segment = source[..loc]
+          segment = source.byteslice(..loc)
           Node.new(:hash, [arg], segment.rindex("{")...(segment.rindex("}") + 1))
         end
       end
@@ -239,7 +239,7 @@ module Sorbet
         if heredoc = heredocs.find { |(_, _, end_arg)| end_arg }
           Node.new(:string_literal, [arg], heredocs.delete(heredoc)[0])
         else
-          Node.new(:string_literal, [arg], arg.range)
+          Node.new(:string_literal, [arg], (arg.range.begin - 1)...(arg.range.end + 1))
         end
       end
 

data/lib/sorbet/eraser/patterns.rb

@@ -13,14 +13,11 @@ module Sorbet
         end
 
         def erase(source)
-          original = source[range]
-          replaced = replace(original)
+          encoding = source.encoding
+          source.force_encoding(Encoding::ASCII_8BIT)
 
-          # puts "Replacing #{original} (len=#{original.length}) " \
-          #      "with #{replaced} (len=#{replaced.length})"
-
-          source[range] = replaced
-          source
+          source[range] = replace(source[range])
+          source.force_encoding(encoding)
         end
 
         def blank(segment)
@@ -57,11 +54,11 @@ module Sorbet
           # We can't really rely on regex here because commas have semantic
           # meaning and you might have some in the value of the first argument.
           comma = metadata.fetch(:comma)
-          pre, post = 0..comma, (comma + 1)..-1
+          pre, post = 0...comma, comma..-1
 
           replacement[pre] =
-            replacement[pre].gsub(/(T\s*\.(?:assert_type!|bind|cast|let)\(\s*)(.+)(\s*,)(.*)/m) do
-              "#{blank($1)}#{$2}#{blank($3)}#{$4}"
+            replacement[pre].gsub(/(T\s*\.(?:assert_type!|bind|cast|let)\(\s*)(.+)/m) do
+              "#{blank($1)}#{$2}"
             end
 
           replacement[post] = blank(replacement[post])
@@ -100,24 +97,26 @@ module Sorbet
         end
       end
 
-      # typed: ignore
-      # typed: false
-      # typed: true
-      # typed: strong
+      # typed: ignore => #
+      # typed: false => #
+      # typed: true => #
+      # typed: strict => #
+      # typed: strong => #
       class TypedCommentPattern < Pattern
         def replace(segment)
-          segment.gsub(/(\A#\s*typed:\s*)(?:ignore|false|true|strong)(\s*)\z/) do
-            "#{$1}ignore#{$2}"
+          segment.gsub(/\A#(\s*typed:\s*(?:ignore|false|true|strict|strong)(\s*))\z/) do
+            "##{blank($1)}"
           end
         end
       end
 
       def on_comment(comment)
         super.tap do |node|
-          if lineno == 1 && comment.match?(/\A#\s*typed:\s*(?:ignore|false|true|strong)\s*\z/)
+          if lineno == 1 && comment.match?(/\A#\s*typed:\s*(?:ignore|false|true|strict|strong)\s*\z/)
             # typed: ignore
             # typed: false
             # typed: true
+            # typed: strict
             # typed: strong
             patterns << TypedCommentPattern.new(node.range)
           end

data/lib/sorbet/eraser/version.rb

@@ -2,6 +2,6 @@
 
 module Sorbet
   module Eraser
-    VERSION = "0.4.0"
+    VERSION = "0.5.0"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: sorbet-eraser
 version: !ruby/object:Gem::Version
-  version: 0.4.0
+  version: 0.5.0
 platform: ruby
 authors:
 - Kevin Newton
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2023-07-03 00:00:00.000000000 Z
+date: 2023-07-13 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -89,7 +89,7 @@ licenses:
 - MIT
 metadata:
   bug_tracker_uri: https://github.com/kddnewton/sorbet-eraser/issues
-  changelog_uri: https://github.com/kddnewton/sorbet-eraser/blob/v0.4.0/CHANGELOG.md
+  changelog_uri: https://github.com/kddnewton/sorbet-eraser/blob/v0.5.0/CHANGELOG.md
   source_code_uri: https://github.com/kddnewton/sorbet-eraser
   rubygems_mfa_required: 'true'
 post_install_message: