[ruby/reline] Optimize regexp handling for width

The rendering time in IRB has been reduced as follows:

  start = Time.now

  [{"_id"=>"5f9072a4589a06d2d74b6028",
    "index"=>0,
    "guid"=>"6b3051e2-dbc7-4537-bdb9-6cd7bb5358a7",
    "isActive"=>true,
    "balance"=>"$1,442.84",
    "picture"=>"http://placehold.it/32x32",
    "age"=>34,
    "eyeColor"=>"blue",
    "name"=>{"first"=>"Ward", "last"=>"Levy"},
    "company"=>"HYPLEX",
    "email"=>"ward.levy@hyplex.us",
    "phone"=>"+1 (867) 568-3319",
    "address"=>"867 Cobek Court, Clara, Maryland, 3254",
    "about"=>
     "Exercitation eu ex aliqua sit. Pariatur aliquip incididunt sint id non consectetur ullamco Lorem ea mollit duis amet sint labore. Commodo laborum labore commodo officia in cillum adipisicing esse excepteur cupidatat adipisicing ut. Non esse incididunt voluptate aliquip cillum eu aute duis laboris sit et. Amet enim quis tempor occaecat excepteur exercitation excepteur deserunt amet cillum adipisicing.",
    "registered"=>"Monday, May 25, 2015 6:51 AM",
    "latitude"=>"16.001127",
    "longitude"=>"-72.377848",
    "tags"=>["dolore", "nostrud", "occaecat", "cillum", "nisi"],
    "range"=>[0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
    "friends"=>
     [{"id"=>0, "name"=>"Alison Bryant"},
      {"id"=>1, "name"=>"Ester Espinoza"},
      {"id"=>2, "name"=>"Sullivan Kane"}],
    "greeting"=>"Hello, Ward! You have 7 unread messages.",
    "favoriteFruit"=>"apple"}]

  puts "Duration: #{Time.now - start} seconds"

0.47sec -> 0.34sec

  start = Time.now

  "Exercitation eu ex aliqua sit. Pariatur aliquip incididunt sint id non consectetur ullamco Lorem ea mollit duis amet sint labore. Commodo laborum labore commodo officia in cillum adipisicing esse excepteur cupidatat adipisicing ut. Non esse incididunt voluptate aliquip cillum eu aute duis laboris sit et. Amet enim quis tempor occaecat excepteur exercitation excepteur deserunt amet cillum adipisicing."

  puts "Duration: #{Time.now - start} seconds"

0.11sec -> 0.08sec

  start = Time.now

  def each_top_level_statement
    initialize_input
    catch(:TERM_INPUT) do
      loop do
        begin
          prompt
          unless l = lex
            throw :TERM_INPUT if @line == ''
          else
            @line_no += l.count("\n")
            next if l == "\n"
            @line.concat l
            if @code_block_open or @ltype or @continue or @indent > 0
              next
            end
          end
          if @line != "\n"
            @line.force_encoding(@io.encoding)
            yield @line, @exp_line_no
          end
          break if @io.eof?
          @line = ''
          @exp_line_no = @line_no

          @indent = 0
        rescue TerminateLineInput
          initialize_input
          prompt
        end
      end
    end
  end

  puts "Duration: #{Time.now - start} seconds"

0.40sec -> 0.33sec

Co-authored-by: NARUSE, Yui <naruse@airemix.jp>

https://github.com/ruby/reline/commit/a9e39ddcc4

1 files changed, 23 insertions, 14 deletions

diff --git a/lib/reline/unicode.rb b/lib/reline/unicode.rb
index 5fcb8be7b3..09aff698b8 100644
--- a/lib/reline/unicode.rb
+++ b/lib/reline/unicode.rb
@@ -35,11 +35,16 @@ class Reline::Unicode
   }
   EscapedChars = EscapedPairs.keys.map(&:chr)
 
-  CSI_REGEXP = /\e\[[\d;]*[ABCDEFGHJKSTfminsuhl]/
-  OSC_REGEXP = /\e\]\d+(?:;[^;]+)*\a/
   NON_PRINTING_START = "\1"
   NON_PRINTING_END = "\2"
-  WIDTH_SCANNER = /\G(?:#{NON_PRINTING_START}|#{NON_PRINTING_END}|#{CSI_REGEXP}|#{OSC_REGEXP}|\X)/
+  CSI_REGEXP = /\e\[[\d;]*[ABCDEFGHJKSTfminsuhl]/
+  OSC_REGEXP = /\e\]\d+(?:;[^;]+)*\a/
+  WIDTH_SCANNER = /\G(?:(#{NON_PRINTING_START})|(#{NON_PRINTING_END})|(#{CSI_REGEXP})|(#{OSC_REGEXP})|(\X))/o
+  NON_PRINTING_START_INDEX = 0
+  NON_PRINTING_END_INDEX = 1
+  CSI_REGEXP_INDEX = 2
+  OSC_REGEXP_INDEX = 3
+  GRAPHEME_CLUSTER_INDEX = 4
 
   def self.get_mbchar_byte_size_by_first_char(c)
     # Checks UTF-8 character byte size
@@ -119,13 +124,14 @@ class Reline::Unicode
       rest = str.encode(Encoding::UTF_8)
       in_zero_width = false
       rest.scan(WIDTH_SCANNER) do |gc|
-        case gc
-        when NON_PRINTING_START
+        case
+        when gc[NON_PRINTING_START_INDEX]
           in_zero_width = true
-        when NON_PRINTING_END
+        when gc[NON_PRINTING_END_INDEX]
           in_zero_width = false
-        when CSI_REGEXP, OSC_REGEXP
-        else
+        when gc[CSI_REGEXP_INDEX], gc[OSC_REGEXP_INDEX]
+        when gc[GRAPHEME_CLUSTER_INDEX]
+          gc = gc[GRAPHEME_CLUSTER_INDEX]
           unless in_zero_width
             width += get_mbchar_width(gc)
           end
@@ -146,14 +152,17 @@ class Reline::Unicode
     rest = str.encode(Encoding::UTF_8)
     in_zero_width = false
     rest.scan(WIDTH_SCANNER) do |gc|
-      case gc
-      when NON_PRINTING_START
+      case
+      when gc[NON_PRINTING_START_INDEX]
         in_zero_width = true
-      when NON_PRINTING_END
+      when gc[NON_PRINTING_END_INDEX]
         in_zero_width = false
-      when CSI_REGEXP, OSC_REGEXP
-        lines.last << gc
-      else
+      when gc[CSI_REGEXP_INDEX]
+        lines.last << gc[CSI_REGEXP_INDEX]
+      when gc[OSC_REGEXP_INDEX]
+        lines.last << gc[OSC_REGEXP_INDEX]
+      when gc[GRAPHEME_CLUSTER_INDEX]
+        gc = gc[GRAPHEME_CLUSTER_INDEX]
         unless in_zero_width
           mbchar_width = get_mbchar_width(gc)
           if (width += mbchar_width) > max_width