Always issue deprecation warning when calling Regexp.new with 3rd positional argument

Previously, only certain values of the 3rd argument triggered a
deprecation warning.

First step for fix for bug #18797.  Support for the 3rd argument
will be removed after the release of Ruby 3.2.

Fix minor fallout discovered by the tests.

Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>

lib/racc/statetransitiontable.rb

@@ -216,7 +216,7 @@ module Racc
         end
         i = ii
       end
-      Regexp.compile(map, nil, 'n')
+      Regexp.compile(map, Regexp::NOENCODING)
     end
 
     def set_table(entries, dummy, tbl, chk, ptr)

re.c

@@ -3759,10 +3759,11 @@ struct reg_init_args {
 
 static VALUE reg_extract_args(int argc, VALUE *argv, struct reg_init_args *args);
 static VALUE reg_init_args(VALUE self, VALUE str, rb_encoding *enc, int flags);
+void rb_warn_deprecated_to_remove(const char *removal, const char *fmt, const char *suggest, ...);
 
 /*
  *  call-seq:
- *    Regexp.new(string, options = 0, n_flag = nil, timeout: nil) -> regexp
+ *    Regexp.new(string, options = 0, timeout: nil) -> regexp
  *    Regexp.new(regexp, timeout: nil) -> regexp
  *
  *  With argument +string+ given, returns a new regexp with the given string
@@ -3780,24 +3781,18 @@ static VALUE reg_init_args(VALUE self, VALUE str, rb_encoding *enc, int flags);
  *      Regexp.new('foo', 'im') # => /foo/im
  *
  *  - The logical OR of one or more of the constants
- *    Regexp::EXTENDED, Regexp::IGNORECASE, and Regexp::MULTILINE:
+ *    Regexp::EXTENDED, Regexp::IGNORECASE, Regexp::MULTILINE, and
+ *    Regexp::NOENCODING:
  *
  *      Regexp.new('foo', Regexp::IGNORECASE) # => /foo/i
  *      Regexp.new('foo', Regexp::EXTENDED)   # => /foo/x
  *      Regexp.new('foo', Regexp::MULTILINE)  # => /foo/m
+ *      Regexp.new('foo', Regexp::NOENCODING)  # => /foo/n
  *      flags = Regexp::IGNORECASE | Regexp::EXTENDED |  Regexp::MULTILINE
  *      Regexp.new('foo', flags)              # => /foo/mix
  *
  *  - +nil+ or +false+, which is ignored.
  *
- *  If optional argument +n_flag+ if it is a string starts with
- *  <code>'n'</code> or <code>'N'</code>, the encoding of +string+ is
- *  ignored and the new regexp encoding is fixed to +ASCII-8BIT+ or
- *  +US-ASCII+, by its content.
- *
- *      Regexp.new('foo', nil, 'n')     # => /foo/n
- *      Regexp.new("\u3042", nil, 'n')  # => /\xE3\x81\x82/n
- *
  *  If optional keyword argument +timeout+ is given,
  *  its float value overrides the timeout interval for the class,
  *  Regexp.timeout.
@@ -3841,7 +3836,7 @@ reg_extract_args(int argc, VALUE *argv, struct reg_init_args *args)
     VALUE str, src, opts = Qundef, n_flag = Qundef, kwargs;
     VALUE re = Qnil;
 
-    rb_scan_args(argc, argv, "12:", &src, &opts, &n_flag, &kwargs);
+    argc = rb_scan_args(argc, argv, "12:", &src, &opts, &n_flag, &kwargs);
 
     args->timeout = Qnil;
     if (!NIL_P(kwargs)) {
@@ -3852,6 +3847,10 @@ reg_extract_args(int argc, VALUE *argv, struct reg_init_args *args)
         rb_get_kwargs(kwargs, keywords, 0, 1, &args->timeout);
     }
 
+    if (argc == 3) {
+        rb_warn_deprecated_to_remove("3.3", "3rd argument to Regexp.new", "2nd argument");
+    }
+
     if (RB_TYPE_P(src, T_REGEXP)) {
         re = src;
 
@@ -3876,9 +3875,6 @@ reg_extract_args(int argc, VALUE *argv, struct reg_init_args *args)
                 enc = rb_ascii8bit_encoding();
                 flags |= ARG_ENCODING_NONE;
             }
-            else {
-                rb_category_warn(RB_WARN_CATEGORY_DEPRECATED, "encoding option is ignored - %s", kcode);
-            }
         }
         str = StringValue(src);
     }

spec/ruby/core/regexp/shared/new.rb

@@ -197,6 +197,7 @@ describe :regexp_new_string, shared: true do
     end
   end
 
+  ruby_version_is ""..."3.2" do
     it "ignores the third argument if it is 'e' or 'euc' (case-insensitive)" do
       -> {
         Regexp.send(@method, 'Hi', nil, 'e').encoding.should == Encoding::US_ASCII
@@ -240,6 +241,7 @@ describe :regexp_new_string, shared: true do
       Regexp.send(@method, str, nil, 'none').encoding.should == Encoding::BINARY
       Regexp.send(@method, str, nil, 'NONE').encoding.should == Encoding::BINARY
     end
+  end
 
   describe "with escaped characters" do
     it "raises a Regexp error if there is a trailing backslash" do
@@ -598,8 +600,10 @@ describe :regexp_new_regexp, shared: true do
       Regexp.send(@method, /Hi/n).encoding.should == Encoding::US_ASCII
     end
 
+    ruby_version_is ''...'3.2' do
       it "sets the encoding to source String's encoding if the Regexp literal has the 'n' option and the source String is not ASCII only" do
         Regexp.send(@method, Regexp.new("\\xff", nil, 'n')).encoding.should == Encoding::BINARY
       end
     end
+  end
 end

test/psych/test_yaml.rb

@@ -34,7 +34,7 @@ class Psych_Unit_Tests < Psych::TestCase
 
     # [ruby-core:34969]
     def test_regexp_with_n
-        assert_cycle(Regexp.new('',0,'n'))
+        assert_cycle(Regexp.new('',Regexp::NOENCODING))
     end
 	#
 	# Tests modified from 00basic.t in Psych.pm

test/ruby/test_regexp.rb

@@ -42,14 +42,14 @@ class TestRegexp < Test::Unit::TestCase
 
   def test_yoshidam_net_20041111_1
     s = "[\xC2\xA0-\xC3\xBE]"
-    r = assert_deprecated_warning(/ignored/) {Regexp.new(s, nil, "u")}
+    r = assert_deprecated_warning(/3\.3/) {Regexp.new(s, nil, "u")}
     assert_match(r, "\xC3\xBE")
   end
 
   def test_yoshidam_net_20041111_2
     assert_raise(RegexpError) do
       s = "[\xFF-\xFF]".force_encoding("utf-8")
-      assert_warning(/ignored/) {Regexp.new(s, nil, "u")}
+      assert_warning(/3\.3/) {Regexp.new(s, nil, "u")}
     end
   end
 
@@ -646,13 +646,37 @@ class TestRegexp < Test::Unit::TestCase
     assert_equal(/foo/, assert_no_warning(/ignored/) {Regexp.new(/foo/)})
     assert_equal(/foo/, assert_no_warning(/ignored/) {Regexp.new(/foo/, timeout: nil)})
 
-    assert_equal(Encoding.find("US-ASCII"), Regexp.new("b..", nil, "n").encoding)
-    assert_equal("bar", "foobarbaz"[Regexp.new("b..", nil, "n")])
-    assert_equal(//n, Regexp.new("", nil, "n"))
+    arg_encoding_none = //n.options # ARG_ENCODING_NONE is implementation defined value
 
-    arg_encoding_none = 32 # ARG_ENCODING_NONE is implementation defined value
+    assert_deprecated_warning('') do
+      assert_equal(Encoding.find("US-ASCII"), Regexp.new("b..", Regexp::NOENCODING).encoding)
+      assert_equal("bar", "foobarbaz"[Regexp.new("b..", Regexp::NOENCODING)])
+      assert_equal(//, Regexp.new(""))
+      assert_equal(//, Regexp.new("", timeout: 1))
+      assert_equal(//n, Regexp.new("", Regexp::NOENCODING))
+      assert_equal(//n, Regexp.new("", Regexp::NOENCODING, timeout: 1))
+
+      assert_equal(arg_encoding_none, Regexp.new("", Regexp::NOENCODING).options)
+    end
+
+    assert_deprecated_warning(/3\.3/) do
+      assert_equal(Encoding.find("US-ASCII"), Regexp.new("b..", nil, "n").encoding)
+    end
+    assert_deprecated_warning(/3\.3/) do
+      assert_equal(Encoding.find("US-ASCII"), Regexp.new("b..", nil, "n", timeout: 1).encoding)
+    end
+    assert_deprecated_warning(/3\.3/) do
+      assert_equal("bar", "foobarbaz"[Regexp.new("b..", nil, "n")])
+    end
+    assert_deprecated_warning(/3\.3/) do
+      assert_equal(//n, Regexp.new("", nil, "n"))
+    end
+    assert_deprecated_warning(/3\.3/) do
       assert_equal(arg_encoding_none, Regexp.new("", nil, "n").options)
+    end
+    assert_deprecated_warning(/3\.3/) do
       assert_equal(arg_encoding_none, Regexp.new("", nil, "N").options)
+    end
 
     assert_raise(RegexpError) { Regexp.new(")(") }
     assert_raise(RegexpError) { Regexp.new('[\\40000000000') }