[ruby/reline] Expand the scanned array to later case statement more

straightforward
(ruby/reline#526)

* Improve test coverage on Unicode.take_range

* Add test for Unicode.calculate_width

* Expand the scanned array to later case statement more straightforward

lib/reline/unicode.rb

@@ -40,11 +40,6 @@ class Reline::Unicode
   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
@@ -132,15 +127,14 @@ def self.calculate_width(str, allow_escape_code = false)
       width = 0
       rest = str.encode(Encoding::UTF_8)
       in_zero_width = false
-      rest.scan(WIDTH_SCANNER) do |gc|
+      rest.scan(WIDTH_SCANNER) do |non_printing_start, non_printing_end, csi, osc, gc|
         case
-        when gc[NON_PRINTING_START_INDEX]
+        when non_printing_start
           in_zero_width = true
-        when gc[NON_PRINTING_END_INDEX]
+        when non_printing_end
           in_zero_width = false
-        when gc[CSI_REGEXP_INDEX], gc[OSC_REGEXP_INDEX]
-        when gc[GRAPHEME_CLUSTER_INDEX]
-          gc = gc[GRAPHEME_CLUSTER_INDEX]
+        when csi, osc
+        when gc
           unless in_zero_width
             width += get_mbchar_width(gc)
           end
@@ -161,22 +155,21 @@ def self.split_by_width(str, max_width, encoding = str.encoding)
     rest = str.encode(Encoding::UTF_8)
     in_zero_width = false
     seq = String.new(encoding: encoding)
-    rest.scan(WIDTH_SCANNER) do |gc|
+    rest.scan(WIDTH_SCANNER) do |non_printing_start, non_printing_end, csi, osc, gc|
       case
-      when gc[NON_PRINTING_START_INDEX]
+      when non_printing_start
         in_zero_width = true
         lines.last << NON_PRINTING_START
-      when gc[NON_PRINTING_END_INDEX]
+      when non_printing_end
         in_zero_width = false
         lines.last << NON_PRINTING_END
-      when gc[CSI_REGEXP_INDEX]
-        lines.last << gc[CSI_REGEXP_INDEX]
-        seq << gc[CSI_REGEXP_INDEX]
-      when gc[OSC_REGEXP_INDEX]
-        lines.last << gc[OSC_REGEXP_INDEX]
-        seq << gc[OSC_REGEXP_INDEX]
-      when gc[GRAPHEME_CLUSTER_INDEX]
-        gc = gc[GRAPHEME_CLUSTER_INDEX]
+      when csi
+        lines.last << csi
+        seq << csi
+      when osc
+        lines.last << osc
+        seq << osc
+      when gc
         unless in_zero_width
           mbchar_width = get_mbchar_width(gc)
           if (width += mbchar_width) > max_width
@@ -204,18 +197,17 @@ def self.take_range(str, start_col, max_width, encoding = str.encoding)
     total_width = 0
     rest = str.encode(Encoding::UTF_8)
     in_zero_width = false
-    rest.scan(WIDTH_SCANNER) do |gc|
+    rest.scan(WIDTH_SCANNER) do |non_printing_start, non_printing_end, csi, osc, gc|
       case
-      when gc[NON_PRINTING_START_INDEX]
+      when non_printing_start
         in_zero_width = true
-      when gc[NON_PRINTING_END_INDEX]
+      when non_printing_end
         in_zero_width = false
-      when gc[CSI_REGEXP_INDEX]
-        chunk << gc[CSI_REGEXP_INDEX]
-      when gc[OSC_REGEXP_INDEX]
-        chunk << gc[OSC_REGEXP_INDEX]
-      when gc[GRAPHEME_CLUSTER_INDEX]
-        gc = gc[GRAPHEME_CLUSTER_INDEX]
+      when csi
+        chunk << csi
+      when osc
+        chunk << osc
+      when gc
         if in_zero_width
           chunk << gc
         else

test/reline/test_unicode.rb

@@ -29,6 +29,26 @@ def test_split_by_width
 
   def test_take_range
     assert_equal 'cdef', Reline::Unicode.take_range('abcdefghi', 2, 4)
+    assert_equal 'あde', Reline::Unicode.take_range('abあdef', 2, 4)
+    assert_equal 'zerocdef', Reline::Unicode.take_range("ab\1zero\2cdef", 2, 4)
+    assert_equal 'bzerocde', Reline::Unicode.take_range("ab\1zero\2cdef", 1, 4)
+    assert_equal "\e[31mcd\e[42mef", Reline::Unicode.take_range("\e[31mabcd\e[42mefg", 2, 4)
+    assert_equal "\e]0;1\acd", Reline::Unicode.take_range("ab\e]0;1\acd", 2, 3)
     assert_equal 'いう', Reline::Unicode.take_range('あいうえお', 2, 4)
   end
+
+  def test_calculate_width
+    assert_equal 9, Reline::Unicode.calculate_width('abcdefghi')
+    assert_equal 9, Reline::Unicode.calculate_width('abcdefghi', true)
+    assert_equal 7, Reline::Unicode.calculate_width('abあdef')
+    assert_equal 7, Reline::Unicode.calculate_width('abあdef', true)
+    assert_equal 14, Reline::Unicode.calculate_width("ab\1zero\2cdef")
+    assert_equal 6, Reline::Unicode.calculate_width("ab\1zero\2cdef", true)
+    assert_equal 19, Reline::Unicode.calculate_width("\e[31mabcd\e[42mefg")
+    assert_equal 7, Reline::Unicode.calculate_width("\e[31mabcd\e[42mefg", true)
+    assert_equal 12, Reline::Unicode.calculate_width("ab\e]0;1\acd")
+    assert_equal 4, Reline::Unicode.calculate_width("ab\e]0;1\acd", true)
+    assert_equal 10, Reline::Unicode.calculate_width('あいうえお')
+    assert_equal 10, Reline::Unicode.calculate_width('あいうえお', true)
+  end
 end