spellr 0.3.2 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e106c60133833af83e09be866a0ccdb5241aaa48363567146b01b908cb180a43
-  data.tar.gz: f2a08fee5d21140f628a8133b8bf6c7226e56b6452911a7846fee34ac22a9968
+  metadata.gz: 3725eb0cbde87e7f9ba2e6c8cfbd3c64143e3518e9d835ae284cd44214063f61
+  data.tar.gz: c7949e1f234014929bc8d57455127ca0c00e907a152cd346778621991c15f390
 SHA512:
-  metadata.gz: 0d562f849cf66c85c6591fff5830856104bd5f3bc5176e0f247aa1f3f441e6c6b93afcbf9b6ac7fe982717cf9824d4df445ff318626024d78a1c2d81bbb76fa4
-  data.tar.gz: 61abf55e8e008ae65386c763c039e55df0f634db23402741c15c92bc9154603be2464f793da88fedae9e56fe0d5419eadca28d47e64cf82a7cd809521ac49c98
+  metadata.gz: b076fef5196b5d997ef375de33f575aa83a351d6529755fa26c4de9aeab6b43162ca7b76668ec171405b57e7dec733228d10aaa860f1ebc975aa84e859ace642
+  data.tar.gz: 248adcf239ec2cc0c8d271d45f295a1aeeb2661ae3ba1c5852e671d98726f776ba1691bda4dec39647f9b03596c8282bf0cfac0a9ec50e6428e2f2de24837778

data/CHANGELOG.md

@@ -1,3 +1,14 @@
+# v0.4.0 (unreleased)
+- LOTS of performance improvements. it's about 4 times faster
+- significantly better key heuristic matching, with configurable weight (`key_heuristic_weight`).
+- Update FastIgnore dependency.
+- Change the yml format slightly. `ignore` is now `excludes`. `only` is now `includes`
+  I feel like this makes more sense for the way the config is merged. and the right time to do it is when you'll probably have to tweak it anyway because:
+- the `only`/`includes` items are now parsed using FastIgnore's gitignore inspired allow list format
+  (see https://github.com/robotdana/fast_ignore#using-an-includes-list)
+  Mostly it's the same just more flexible, though there may need to be some small adjustments.
+- the cli arguments are now also managed using FastIgnore's rules, fixing issues with absolute paths and paths beginning with `./` also it's a LOT faster when just checking a single file, basically instant. so that's nice.
+
 # v0.3.2
 - add automatic rubygems and dockerhub deploy
 

data/Gemfile.lock

@@ -1,8 +1,8 @@
 PATH
   remote: .
   specs:
-    spellr (0.3.2)
-      fast_ignore
+    spellr (0.4.0)
+      fast_ignore (~> 0.4.0)
 
 GEM
   remote: https://rubygems.org/
@@ -10,11 +10,11 @@ GEM
     ast (2.4.0)
     coderay (1.1.2)
     diff-lcs (1.3)
-    fast_ignore (0.3.2)
+    fast_ignore (0.4.0)
     jaro_winkler (1.5.3)
     method_source (0.9.2)
     parallel (1.17.0)
-    parser (2.6.4.1)
+    parser (2.6.5.0)
       ast (~> 2.4.0)
     pry (0.12.2)
       coderay (~> 1.1.0)
@@ -28,23 +28,25 @@ GEM
     rspec-core (3.8.2)
       rspec-support (~> 3.8.0)
     rspec-eventually (0.2.2)
-    rspec-expectations (3.8.4)
+    rspec-expectations (3.8.5)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.8.0)
-    rspec-mocks (3.8.1)
+    rspec-mocks (3.8.2)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.8.0)
-    rspec-support (3.8.2)
-    rubocop (0.74.0)
+    rspec-support (3.8.3)
+    rubocop (0.75.0)
       jaro_winkler (~> 1.5.1)
       parallel (~> 1.10)
       parser (>= 2.6)
       rainbow (>= 2.2.2, < 4.0)
       ruby-progressbar (~> 1.7)
       unicode-display_width (>= 1.4.0, < 1.7)
-    rubocop-rspec (1.35.0)
-      rubocop (>= 0.60.0)
+    rubocop-rspec (1.36.0)
+      rubocop (>= 0.68.1)
     ruby-progressbar (1.10.1)
+    terminal-table (1.8.0)
+      unicode-display_width (~> 1.1, >= 1.1.1)
     tty_string (0.1.0)
     unicode-display_width (1.6.0)
 
@@ -60,6 +62,7 @@ DEPENDENCIES
   rubocop
   rubocop-rspec
   spellr!
+  terminal-table
   tty_string
 
 BUNDLED WITH

data/bin/fetch_wordlist/english

@@ -40,7 +40,9 @@ class SCOWLDownloader
 
         @options[:spelling] = a
       end
-      opts.on('-p', '--[no-]programming', 'include common programming terms like grep (default true)') { |h| @options[:special] = :hacker if h }
+      opts.on('-p', '--[no-]programming', 'include common programming terms like grep (default true)') do |h|
+        @options[:special] = :hacker if h
+      end
       opts.on_tail('-h', '--help') do
         warn opts.to_s
         exit 1

data/lib/.spellr.yml

@@ -1,14 +1,17 @@
 ---
 word_minimum_length: 3
+key_heuristic_weight: 5
+key_minimum_length: 6
 
-ignore: # this list is parsed with the .gitignore format
-  - /.git/
+excludes: # this list is parsed with the .gitignore format
+  - .git/
   - .spellr_wordlists/
   - .DS_Store
   - Gemfile.lock
   - .rspec_status
   - '*.png'
   - '*.jpg'
+  - '*.jpeg'
   - '*.gif'
   - '*.ico'
   - .gitkeep
@@ -28,7 +31,7 @@ languages:
     generate: "fetch english"
     # TODO: don't generate the ruby file until you actually need one
   ruby:
-    only: # Filtered using File.fn_match
+    includes: # Filtered using gitignore format
       - '*.rb'
       - '*.rake'
       - '*.gemspec'
@@ -42,7 +45,7 @@ languages:
     hashbangs:
       - ruby
   html:
-    only:
+    includes:
       - '*.html'
       - '*.hml'
       - '*.jsx'
@@ -59,7 +62,7 @@ languages:
       - '*.sass'
       - '*.less'
   js:
-    only:
+    includes:
       - '*.html'
       - '*.hml'
       - '*.jsx'
@@ -71,22 +74,22 @@ languages:
       - '*.erb'
       - '*.json'
   shell:
-    only:
+    includes:
       - '*.sh'
       - Dockerfile
     hashbangs:
       - bash
       - sh
   dockerfile:
-    only:
+    includes:
       - Dockerfile
   css:
-    only:
+    includes:
       - '*.css'
       - '*.sass'
       - '*.scss'
   xml:
-    only:
+    includes:
       - '*.xml'
       - '*.html'
       - '*.haml'

data/lib/spellr/backports.rb

@@ -0,0 +1,19 @@
+# frozen_string_literal: true
+
+class Array
+  unless RUBY_VERSION >= '2.4'
+    def sum
+      reduce(0) do |total, value|
+        total + if block_given?
+          yield value
+        else
+          value
+        end
+      end
+    end
+  end
+end
+
+class String
+  alias_method :match?, :match unless RUBY_VERSION >= '2.4'
+end

data/lib/spellr/check.rb

@@ -36,21 +36,24 @@ module Spellr
 
     private
 
-    def check_file(file, start_at: nil, wordlists: Spellr.config.wordlists_for(file)) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength, Metrics/LineLength
-      Spellr::Tokenizer.new(file, start_at: start_at).each_token do |token|
-        next if wordlists.any? { |d| d.include?(token) }
-
-        start_at = token.location
-        reporter.call(token)
-        @exit_code = 1
+    def check_file(file, start_at: nil, wordlists: Spellr.config.wordlists_for(file)) # rubocop:disable Metrics/AbcSize, Metrics/MethodLength
+      restart_token = catch(:check_file_from) do
+        Spellr::Tokenizer.new(file, start_at: start_at).each_token do |token|
+          next if wordlists.any? { |d| d.include?(token) }
+
+          start_at = token.location
+          reporter.call(token)
+          @exit_code = 1
+        end
+        nil
+      end
+      if restart_token
+        wordlist_arg = restart_token.replacement ? { wordlists: wordlists } : {} # new wordlist cache when adding a word
+        check_file(file, start_at: restart_token.location, **wordlist_arg)
       end
-    rescue Spellr::DidReplacement => e # Yeah this is exceptions for control flow, but it makes sense to me
-      check_file(file, start_at: e.token.location, wordlists: wordlists)
-    rescue Spellr::DidAdd => e
-      check_file(file, start_at: e.token.location) # don't cache the wordlists
     rescue InvalidByteSequence
       # sometimes files are binary
-      puts "Skipped unreadable file: #{file}" unless Spellr.config.quiet?
+      warn "Skipped unreadable file: #{file}" unless Spellr.config.quiet?
     end
   end
 end

data/lib/spellr/config.rb

@@ -2,9 +2,12 @@
 
 require_relative '../spellr'
 require_relative 'config_loader'
+require_relative 'language'
+require_relative 'reporter'
+require 'pathname'
 
 module Spellr
-  class Config
+  class Config # rubocop:disable Metrics/ClassLength
     attr_writer :reporter
     attr_reader :config_file
     attr_accessor :quiet
@@ -33,35 +36,69 @@ module Spellr
     end
 
     def word_minimum_length
-      @config[:word_minimum_length]
+      @word_minimum_length ||= @config[:word_minimum_length]
     end
 
-    def only
-      @config[:only] || []
+    def key_heuristic_weight
+      @key_heuristic_weight ||= @config[:key_heuristic_weight]
     end
 
-    def ignored
-      @config[:ignore]
+    def key_minimum_length
+      @key_minimum_length ||= @config[:key_minimum_length]
+    end
+
+    def includes
+      return @includes if defined?(@includes)
+
+      if @config[:only]
+        warn <<~WARNING
+          \e[33mSpellr: `only:` yaml key with a list of fnmatch rules is deprecated.
+          Please use `includes:` instead, which uses gitignore-inspired rules.
+          see github.com/robotdana/fast_ignore#using-an-includes-list for details\e[0m
+        WARNING
+      end
+
+      @includes = (@config[:includes] || []) + (@config[:only] || [])
+    end
+
+    def excludes
+      return @excludes if defined?(@excludes)
+
+      if @config[:ignore]
+        warn <<~WARNING
+          \e[33mSpellr: `ignore:` yaml key is deprecated.
+          Please use `excludes:` instead.\e[0m
+        WARNING
+      end
+
+      @excludes = (@config[:excludes] || []) + (@config[:ignore] || [])
     end
 
     def color
       @config[:color]
     end
 
-    def clear_cache
+    def clear_cache # rubocop:disable Metrics/CyclomaticComplexity, Metrics/PerceivedComplexity
       remove_instance_variable(:@wordlists) if defined?(@wordlists)
       remove_instance_variable(:@languages) if defined?(@languages)
       remove_instance_variable(:@errors) if defined?(@errors)
+      remove_instance_variable(:@word_minimum_length) if defined?(@word_minimum_length)
+      remove_instance_variable(:@key_heuristic_weight) if defined?(@key_heuristic_weight)
+      remove_instance_variable(:@key_minimum_length) if defined?(@key_minimum_length)
+      remove_instance_variable(:@excludes) if defined?(@excludes)
+      remove_instance_variable(:@includes) if defined?(@includes)
     end
 
     def languages
-      require_relative 'language'
-
       @languages ||= @config[:languages].map do |key, args|
         Spellr::Language.new(key, args)
       end
     end
 
+    def pwd
+      @pwd ||= Pathname.pwd
+    end
+
     def languages_for(file)
       languages.select { |l| l.matches?(file) }
     end
@@ -109,8 +146,6 @@ module Spellr
     end
 
     def default_reporter
-      require_relative 'reporter'
-
       Spellr::Reporter.new
     end
   end

data/lib/spellr/file.rb

@@ -2,6 +2,8 @@
 
 require 'pathname'
 
+# TODO: maybe just extend pathname if you have to
+
 module Spellr
   class File < Pathname
     def self.wrap(file)
@@ -9,19 +11,16 @@ module Spellr
     end
 
     def hashbang
-      return if extname != ''
-      return unless first_line&.start_with?('#!')
+      @hashbang ||= begin
+        return if extname != ''
+        return unless first_line&.start_with?('#!')
 
-      first_line
+        first_line
+      end
     end
 
     def first_line
       @first_line ||= each_line.first
     end
-
-    def fnmatch?(pattern)
-      relative_path_from(Pathname.pwd).fnmatch?(pattern, ::File::FNM_DOTMATCH) ||
-        Pathname.new(basename).fnmatch?(pattern, ::File::FNM_DOTMATCH)
-    end
   end
 end

data/lib/spellr/file_list.rb

@@ -12,22 +12,27 @@ module Spellr
       @patterns = patterns
     end
 
-    def config_only?(file)
-      Spellr.config.only.empty? || Spellr.config.only.any? { |o| file.fnmatch?(o) }
-    end
-
-    def cli_only?(file)
-      @patterns.empty? || @patterns.any? { |p| file.fnmatch?(p) }
+    # anchored patterns are significantly faster on large codebases
+    def cli_patterns
+      @patterns.map do |pattern|
+        if pattern.match?(%r{^([/~*]|\.{1,2}/)})
+          pattern
+        else
+          "/#{pattern}"
+        end
+      end
     end
 
     def each
-      # TODO: handle no gitignore
       gitignore = ::File.join(Dir.pwd, '.gitignore')
       gitignore = nil unless ::File.exist?(gitignore)
-      FastIgnore.new(rules: Spellr.config.ignored, gitignore: gitignore).each do |file|
+
+      FastIgnore.new(
+        ignore_rules: Spellr.config.excludes,
+        include_rules: Spellr.config.includes + cli_patterns,
+        gitignore: gitignore
+      ).each do |file|
         file = Spellr::File.new(file)
-        next unless cli_only?(file)
-        next unless config_only?(file)
 
         yield(file)
       end

data/lib/spellr/interactive.rb

@@ -61,6 +61,7 @@ module Spellr
       return unless global_skips.include?(token.to_s) ||
         global_insensitive_skips.include?(token.normalize)
 
+      puts "Automatically skipped #{red(token)}"
       self.total_skipped += 1
     end
 
@@ -71,7 +72,8 @@ module Spellr
 
       token.replace(global_replacement)
       self.total_fixed += 1
-      raise Spellr::DidReplacement, token
+      puts "Automatically replaced #{red(token)} with #{green(global_replacement)}"
+      throw :check_file_from, token
     end
 
     def clear_current_line
@@ -112,6 +114,7 @@ module Spellr
     def handle_skip(token)
       self.total_skipped += 1
       yield token if block_given?
+      puts "Skipped #{red(token)}"
     end
 
     # TODO: handle more than 16 options
@@ -133,7 +136,8 @@ module Spellr
 
         wl.add(token)
         self.total_added += 1
-        raise Spellr::DidAdd, token
+        puts "Added #{red(token)} to #{wl.name} wordlist"
+        throw :check_file_from, token
       else
         handle_add(token)
       end
@@ -152,7 +156,8 @@ module Spellr
         token.replace(full_replacement)
         yield replacement if block_given?
         self.total_fixed += 1
-        raise Spellr::DidReplacement, token
+        puts "Replaced #{red(token.chomp)} with #{green(replacement.chomp)}"
+        throw :check_file_from, token
       end
     rescue Interrupt
       puts '^C again to exit'