character_set 1.5.0 → 1.6.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 9622bc20bbdb48f8deff84dbed9e800e6bc500a6a08a27e7b3aea2ea651cd278
-  data.tar.gz: 5853e8d5be7e9a1963419aa4f9fbc631148fe5bef45aa185b9117d32b44aa959
+  metadata.gz: e216e6c199ac9443cda9180a9e35d5ed92b50b45c12e7f64f45d74ecd2cf08d6
+  data.tar.gz: 5f3634d426dc33875d6c197ce75466544d97808b1e8b1858ac56d93422b226e8
 SHA512:
-  metadata.gz: 2cc2a60b9388a2e3beef66da20aa8205cc501980a7dc66f2716c66f7e999a083927b27a761e6b932b6d5c16b8e5968f8e04370ecf3c999326f378f60bfa3cedc
-  data.tar.gz: a2a8d1f9ac6cdf6302af98662fc3efda4b8c6fe003c7cdc853a61a64f9c7a596b1bbd7a79dca19081b8ce2576f9c3d848869141b164c145e22befaaffec8b265
+  metadata.gz: d24cfaa40b6e4e472e1f76cc8b6f7f3f1282e6830c0cbf76c4810c0f6f365c7419a19816d0b741cee99eb428dae03fc1d60eecab7d1ba6d210015f0cf2d5ee14
+  data.tar.gz: 2bd7ea63b286e106358293b1428a687374d0cd2cdc985b2da5b5cf1f45c6c541cb0ddde5b06477243cf4011065cfac7fa6bb8a521fb144a750c90039d268f03b

data/.gitattributes

@@ -1,3 +1,3 @@
 *.cps linguist-detectable=false
 benchmarks/* linguist-detectable=false
-spec/ruby-spec/* linguist-vendored
+spec/* linguist-detectable=false

data/.github/workflows/tests.yml

@@ -1,6 +1,10 @@
 name: tests
 
-on: [push, pull_request]
+on:
+  push:
+  pull_request:
+  schedule:
+    - cron: '11 11 14 * *' # at 11:11 am on the 14th of every month
 
 jobs:
   build:
@@ -8,7 +12,7 @@ jobs:
 
     strategy:
       matrix:
-        ruby: [ '2.2', '2.7', '3.0', 'ruby-head', 'jruby-head' ]
+        ruby: [ '2.2', '2.7', '3.0', '3.1', 'ruby-head', 'jruby-head' ]
 
     steps:
       - uses: actions/checkout@v2

data/BENCHMARK.md

@@ -1,86 +1,90 @@
-Results of `rake:benchmark` on ruby 3.0.0p0 (2020-12-25 revision 95aff21468) [x86_64-darwin19]
+Results of `rake:benchmark` on ruby 3.2.0dev (2022-02-14T14:35:54Z master 26187a8520) [arm64-darwin21]
 
 ```
 Counting non-letters
 
-CharacterSet#count_in:  9472902.2 i/s
-        String#count:  2221799.9 i/s - 4.26x slower
+CharacterSet#count_in: 14794607.9 i/s
+        String#count:  3875939.3 i/s - 3.82x slower
 ```
 ```
 Detecting non-whitespace
 
- CharacterSet#cover?: 12388427.2 i/s
-       Regexp#match?:  7901676.8 i/s - 1.57x slower
+ CharacterSet#cover?: 17448329.0 i/s
+       Regexp#match?: 13089358.1 i/s - 1.33x slower
 ```
 ```
 Detecting non-letters
 
- CharacterSet#cover?: 12263689.1 i/s
-       Regexp#match?:  4940889.9 i/s - 2.48x slower
+ CharacterSet#cover?: 17565596.9 i/s
+       Regexp#match?:  7951108.0 i/s - 2.21x slower
 ```
 ```
-Removing whitespace
+Removing ASCII whitespace
 
-CharacterSet#delete_in:  2406722.6 i/s
-         String#gsub:   235760.3 i/s - 10.21x slower
+CharacterSet#delete_in:  6306078.2 i/s
+           String#tr:  4734401.0 i/s - 1.33x slower
+         String#gsub:   211631.8 i/s - 29.80x slower
 ```
 ```
 Removing whitespace, emoji and umlauts
 
-CharacterSet#delete_in:  1653607.6 i/s
-         String#gsub:   272782.9 i/s - 6.06x slower
+CharacterSet#delete_in:  5984149.6 i/s
+           String#tr:   363643.1 i/s - 16.46x slower
+         String#gsub:   317201.7 i/s - 18.87x slower
 ```
 ```
 Removing non-whitespace
 
-CharacterSet#keep_in:  2671038.2 i/s
-         String#gsub:   242551.0 i/s - 11.01x slower
+CharacterSet#keep_in:  7650925.6 i/s
+         String#gsub:   207374.6 i/s - 36.89x slower
+           String#tr:       12.3 i/s - 619745.60x slower
 ```
 ```
-Extracting emoji
+Keeping only emoji
 
-CharacterSet#keep_in:  1726496.5 i/s
-         String#gsub:   215609.2 i/s - 8.01x slower
+CharacterSet#keep_in:  7272940.1 i/s
+         String#gsub:   177993.8 i/s - 40.86x slower
+           String#tr:       12.3 i/s - 590222.71x slower
 ```
 ```
 Extracting emoji to an Array
 
-   CharacterSet#scan:  2373856.1 i/s
-         String#scan:   480000.5 i/s - 4.95x slower
+   CharacterSet#scan:  2978285.0 i/s
+         String#scan:   865793.8 i/s - 3.44x slower
 ```
 ```
 Detecting whitespace
 
-CharacterSet#used_by?: 11988328.7 i/s
-       Regexp#match?:  6758146.8 i/s - 1.77x slower
+CharacterSet#used_by?: 17292338.4 i/s
+       Regexp#match?: 11705563.9 i/s - 1.48x slower
 ```
 ```
 Detecting emoji in a large string
 
-CharacterSet#used_by?:   288223.3 i/s
-       Regexp#match?:   102384.2 i/s - 2.82x slower
+CharacterSet#used_by?:   340444.1 i/s
+       Regexp#match?:   180549.8 i/s - 1.89x slower
 ```
 ```
 Adding entries
 
-    CharacterSet#add:  2538251.2 i/s
-       SortedSet#add:   443925.9 i/s - 5.72x slower
+    CharacterSet#add:  4951781.4 i/s
+       SortedSet#add:  1019637.9 i/s - 4.86x slower
 ```
 ```
 Removing entries
 
- CharacterSet#delete:  2487620.8 i/s
-    SortedSet#delete:   628816.1 i/s - 3.96x slower
+ CharacterSet#delete:  5006337.6 i/s
+    SortedSet#delete:  3922752.2 i/s - same-ish
 ```
 ```
 Merging entries
 
-  CharacterSet#merge:      551.6 i/s
-     SortedSet#merge:        1.4 i/s - 393.59x slower
+  CharacterSet#merge:      661.8 i/s
+     SortedSet#merge:        3.9 i/s - 167.82x slower
 ```
 ```
 Getting the min and max
 
- CharacterSet#minmax:   636890.7 i/s
-    SortedSet#minmax:      254.1 i/s - 2506.20x slower
+ CharacterSet#minmax:  1212462.2 i/s
+    SortedSet#minmax:      844.4 i/s - 1435.93x slower
 ```

data/CHANGELOG.md

@@ -4,6 +4,21 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/)
 and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.html).
 
+## [1.6.0] - 2022-02-16
+
+### Added
+
+- `::of` now supports both `String` and `Regexp` arguments
+
+### Fixed
+
+- fixed segfault during `String` manipulation on Ruby 3.2.0-dev
+- improved performance for `String` manipulation
+- allow usage in Ractors
+  - predefined sets must be pre-initialized for this, though
+  - e.g. `CharacterSet.ascii`, `keep_character_set(:ascii)` etc.
+  - call them once in the main Ractor to trigger initialization
+
 ## [1.5.0] - 2021-12-05
 
 ### Added

data/Gemfile

@@ -4,3 +4,17 @@ git_source(:github) {|repo_name| "https://github.com/#{repo_name}" }
 
 # Specify your gem's dependencies in character_set.gemspec
 gemspec
+
+gem 'benchmark-ips', '~> 2.7'
+gem 'get_process_mem', '~> 0.2.3'
+gem 'rake', '~> 13.0'
+gem 'rake-compiler', '~> 1.1'
+gem 'range_compressor', '~> 1.0'
+gem 'regexp_parser', '~> 2.1'
+gem 'regexp_property_values', '~> 1.0'
+gem 'rspec', '~> 3.8'
+if RUBY_VERSION.to_f >= 2.7
+  gem 'codecov', '~> 0.2.12'
+  gem 'gouteur', '~> 1.0.0'
+  gem 'rubocop', '~> 1.8'
+end

data/README.md

@@ -5,16 +5,17 @@
 [![Build Status](https://github.com/jaynetics/character_set/workflows/gouteur/badge.svg)](https://github.com/jaynetics/character_set/actions)
 [![codecov](https://codecov.io/gh/jaynetics/character_set/branch/master/graph/badge.svg)](https://codecov.io/gh/jaynetics/character_set)
 
-This is a C-extended Ruby gem to work with sets of Unicode codepoints. It can read and write these sets in various formats and implements the stdlib `Set` interface for them.
+This is a C-extended Ruby gem to work with sets of Unicode codepoints.
 
-It also offers an alternate paradigm of `String` processing which grants much better performance than `Regexp` and `String` methods from the stdlib where applicable (see [benchmarks](./BENCHMARK.md)).
+It can [read](#parseinitialize) and [write](#write) sets of codepoints in various formats and it implements the stdlib `Set` interface for them.
+
+It also offers a [way of scrubbing and scanning characters in Strings](#interact-with-strings) that is more semantic and consistently offers better performance than `Regexp` and `String` methods from the stdlib for this (see [benchmarks](./BENCHMARK.md)).
 
 Many parts can be used independently, e.g.:
 - `CharacterSet::Character`
 - `CharacterSet::ExpressionConverter`
 - `CharacterSet::Parser`
 - `CharacterSet::Writer`
-- [`RangeCompressor`](https://github.com/jaynetics/range_compressor)
 
 ## Usage
 
@@ -42,9 +43,10 @@ CharacterSet.parse('[a-c]')
 CharacterSet.parse('\U00000061-\U00000063')
 ```
 
-If the gems [`regexp_parser`](https://github.com/ammar/regexp_parser) and [`regexp_property_values`](https://github.com/jaynetics/regexp_property_values) are installed, `::of_regexp` and `::of_property` can also be used. `::of_regexp` can handle intersections, negations, and set nesting. Regexp's `i`-flag is ignored; call `#case_insensitive` on the result if needed.
+If the gems [`regexp_parser`](https://github.com/ammar/regexp_parser) and [`regexp_property_values`](https://github.com/jaynetics/regexp_property_values) are installed, `Regexp` and unicode property names can also be read. Regexp intersections, negations, and set nesting are covered, but the `i`-flag is ignored; call `#case_insensitive` on the result if needed.
 
 ```ruby
+CharacterSet.of(/./) # => #<CharacterSet (size: 1112064)>
 CharacterSet.of_property('Thai') # => #<CharacterSet (size: 86)>
 
 require 'character_set/core_ext/regexp_ext'
@@ -145,6 +147,7 @@ CharacterSet['1', 'A'].case_insensitive # => CharacterSet['1', 'A', 'a']
 ```
 
 ### Write
+
 ```ruby
 set = CharacterSet['a', 'b', 'c', 'j', '-']
 
@@ -211,6 +214,6 @@ CharacterSet['a', 'ü', '🤩'].member_in_plane?(7) # => false
 CharacterSet::Character.new('a').plane # => 0
 ```
 
-### Contributions
+## Contributions
 
 Feel free to send suggestions, point out issues, or submit pull requests.

data/Rakefile

@@ -147,8 +147,11 @@ namespace :benchmark do
       f.puts "Results of `rake:benchmark` on #{RUBY_DESCRIPTION}", ''
 
       $store_comparison_results.each do |caption, result|
-        f.puts '```', caption, '',
-               result.strip.gsub(/(same-ish).*$/, '\1').lines[1..-1], '```'
+        f.puts '```',
+               caption,
+               '',
+               result.strip.gsub(/ \(±[^)]+\) /, '').gsub(/(same-ish).*$/, '\1').lines[1..-1],
+               '```'
       end
     end
   end

data/benchmarks/delete_in.rb

@@ -2,24 +2,28 @@ require_relative './shared'
 
 str = 'Lorem     ipsum       et      dolorem'
 rx = /\s/
+trt = "\t\n\v\f\r\s"
 cs = CharacterSet.whitespace
 
 benchmark(
-  caption: 'Removing whitespace',
+  caption: 'Removing ASCII whitespace',
   cases: {
     'String#gsub'            => -> { str.gsub(rx, '') },
+    'String#tr'              => -> { str.tr(trt, '') },
     'CharacterSet#delete_in' => -> { cs.delete_in(str) },
   }
 )
 
 str = 'Lörem ipsüm ⛷ et dölörem'
 rx = /[\s\p{emoji}äüö]/
+trt = "\t\n\v\f\r\s😀-🙏äüö"
 cs = CharacterSet.whitespace + CharacterSet.emoji + CharacterSet['ä', 'ö', 'ü']
 
 benchmark(
   caption: 'Removing whitespace, emoji and umlauts',
   cases: {
     'String#gsub'            => -> { str.gsub(rx, '') },
+    'String#tr'              => -> { str.tr(trt, '') },
     'CharacterSet#delete_in' => -> { cs.delete_in(str) },
   }
 )

data/benchmarks/keep_in.rb

@@ -2,24 +2,28 @@ require_relative './shared'
 
 str = 'Lorem ipsum et dolorem'
 rx = /\S/
+trt = "\u{0080}-\u{10FFFF}" # approximation
 cs = CharacterSet.whitespace
 
 benchmark(
   caption: 'Removing non-whitespace',
   cases: {
     'String#gsub'          => -> { str.gsub(rx, '') },
+    'String#tr'            => -> { str.tr(trt, '') },
     'CharacterSet#keep_in' => -> { cs.keep_in(str) },
   }
 )
 
 str = 'Lorem ipsum ⛷ et dolorem'
 rx = /\p{^emoji}/
+trt = "\u0000-\u{1F599}\u{1F650}-\u{10FFFF}"
 cs = CharacterSet.emoji
 
 benchmark(
-  caption: 'Extracting emoji',
+  caption: 'Keeping only emoji',
   cases: {
     'String#gsub'          => -> { str.gsub(rx, '') },
+    'String#tr'            => -> { str.tr(trt, '') },
     'CharacterSet#keep_in' => -> { cs.keep_in(str) },
   }
 )

data/character_set.gemspec

@@ -28,18 +28,4 @@ Gem::Specification.new do |s|
   if RUBY_VERSION.to_f >= 3.0 && !RUBY_PLATFORM[/java/i]
     s.add_dependency 'sorted_set', '~> 1.0'
   end
-
-  s.add_development_dependency 'benchmark-ips', '~> 2.7'
-  s.add_development_dependency 'get_process_mem', '~> 0.2.3'
-  s.add_development_dependency 'rake', '~> 13.0'
-  s.add_development_dependency 'rake-compiler', '~> 1.1'
-  s.add_development_dependency 'range_compressor', '~> 1.0'
-  s.add_development_dependency 'regexp_parser', '~> 2.1'
-  s.add_development_dependency 'regexp_property_values', '~> 1.0'
-  s.add_development_dependency 'rspec', '~> 3.8'
-  if RUBY_VERSION.to_f >= 2.7
-    s.add_development_dependency 'codecov', '~> 0.2.12'
-    s.add_development_dependency 'gouteur', '~> 1.0.0'
-    s.add_development_dependency 'rubocop', '~> 1.8'
-  end
 end

data/ext/character_set/character_set.c

@@ -82,7 +82,11 @@ static const rb_data_type_t cs_type = {
         .dsize = cs_memsize,
     },
     .data = NULL,
+#ifdef RUBY_TYPED_FROZEN_SHAREABLE
+    .flags = RUBY_TYPED_FREE_IMMEDIATELY | RUBY_TYPED_FROZEN_SHAREABLE,
+#else
     .flags = RUBY_TYPED_FREE_IMMEDIATELY,
+#endif
 };
 
 static inline VALUE
@@ -315,9 +319,9 @@ cs_method_minmax(VALUE self)
     cs_cp cp, alen, blen;                                        \
     cs_ar *acps, *bcps;                                          \
     struct cs_data *new_data;                                    \
-    new_cs = cs_alloc(RBASIC(self)->klass, &new_data);           \
     acps = cs_fetch_cps(cs_a, &alen);                            \
     bcps = cs_fetch_cps(cs_b, &blen);                            \
+    new_cs = cs_alloc(RBASIC(self)->klass, &new_data);           \
     for (cp = 0; cp < UNICODE_CP_COUNT; cp++)                    \
     {                                                            \
       if (tst_cp(acps, alen, cp) comp_op tst_cp(bcps, blen, cp)) \
@@ -705,8 +709,7 @@ cs_method_ranges(VALUE self)
 
       if (!previous_cp_num) {
         current_start = cp_num;
-      } else if (previous_cp_num + 2 != cp_num)
-      {
+      } else if (previous_cp_num + 2 != cp_num) {
         // gap found, finalize previous range
         rb_ary_push(ranges, rb_range_new(current_start, current_end, 0));
         current_start = cp_num;
@@ -1047,17 +1050,14 @@ raise_arg_err_unless_string(VALUE val)
 }
 
 static VALUE
-cs_class_method_of(int argc, VALUE *argv, VALUE self)
+cs_class_method_of_string(VALUE self, VALUE string)
 {
   VALUE new_cs;
   struct cs_data *new_data;
-  int i;
+
+  raise_arg_err_unless_string(string);
   new_cs = cs_alloc(self, &new_data);
-  for (i = 0; i < argc; i++)
-  {
-    raise_arg_err_unless_string(argv[i]);
-    each_cp(argv[i], add_str_cp_to_arr, 0, 0, new_data, 0);
-  }
+  each_cp(string, add_str_cp_to_arr, 0, 0, new_data, 0);
   return new_cs;
 }
 
@@ -1138,116 +1138,76 @@ cs_method_used_by_p(VALUE self, VALUE str)
   return only_uses_other_cps == Qfalse ? Qtrue : Qfalse;
 }
 
-static void
-cs_str_buf_cat(VALUE str, const char *ptr, long len)
-{
-  long total, olen;
-  char *sptr;
-
-  RSTRING_GETMEM(str, sptr, olen);
-  sptr = RSTRING(str)->as.heap.ptr;
-  olen = RSTRING(str)->as.heap.len;
-  total = olen + len;
-  memcpy(sptr + olen, ptr, len);
-  RSTRING(str)->as.heap.len = total;
-}
-
-#ifndef TERM_FILL
-#define TERM_FILL(ptr, termlen)                     \
-  do                                                \
-  {                                                 \
-    char *const term_fill_ptr = (ptr);              \
-    const int term_fill_len = (termlen);            \
-    *term_fill_ptr = '\0';                          \
-    if (__builtin_expect(!!(term_fill_len > 1), 0)) \
-      memset(term_fill_ptr, 0, term_fill_len);      \
-  } while (0)
-#endif
-
-static void
-cs_str_buf_terminate(VALUE str, rb_encoding *enc)
-{
-  char *ptr;
-  long len;
-
-  ptr = RSTRING(str)->as.heap.ptr;
-  len = RSTRING(str)->as.heap.len;
-  TERM_FILL(ptr + len, rb_enc_mbminlen(enc));
-}
-
+// partially based on rb_str_delete_bang
 static inline VALUE
 cs_apply_to_str(VALUE set, VALUE str, int delete, int bang)
 {
   cs_ar *cps;
-  cs_cp len;
-  rb_encoding *str_enc;
-  VALUE orig_len, new_str_buf;
-  int cp_len;
-  unsigned int str_cp;
-  const char *ptr, *end;
+  cs_cp cs_len;
+  VALUE orig_str_len;
+
+  rb_encoding *enc;
+  char *s, *send, *t;
+  int ascompat, cr;
 
   raise_arg_err_unless_string(str);
 
-  cps = cs_fetch_cps(set, &len);
+  orig_str_len = RSTRING_LEN(str);
 
-  orig_len = RSTRING_LEN(str);
-  if (orig_len < 1) // empty string, will never change
+  if (orig_str_len == 0)
   {
-    if (bang)
-    {
-      return Qnil;
-    }
-    return rb_str_dup(str);
+    return bang ? Qnil : str;
   }
 
-  new_str_buf = rb_str_buf_new(orig_len + 30); // len + margin
-  str_enc = rb_enc_get(str);
-  rb_enc_associate(new_str_buf, str_enc);
-  rb_str_modify(new_str_buf);
-  ENC_CODERANGE_SET(new_str_buf, rb_enc_asciicompat(str_enc) ? ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID);
-
-  ptr = RSTRING_PTR(str);
-  end = RSTRING_END(str);
+  if (!bang)
+  {
+    str = rb_str_dup(str);
+  }
 
-  if (single_byte_optimizable(str))
+  cps = cs_fetch_cps(set, &cs_len);
+  rb_str_modify(str);
+  enc = rb_enc_get(str);
+  ascompat = rb_enc_asciicompat(enc);
+  s = t = RSTRING_PTR(str);
+  send = RSTRING_END(str);
+  cr = ascompat ? ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID;
+  while (s < send)
   {
-    while (ptr < end)
+    unsigned int c;
+    int clen;
+
+    if (ascompat && (c = *(unsigned char *)s) < 0x80)
     {
-      str_cp = *ptr & 0xff;
-      if ((!tst_cp(cps, len, str_cp)) == delete)
+      if (tst_cp(cps, cs_len, c) != delete)
       {
-        cs_str_buf_cat(new_str_buf, ptr, 1);
+        if (t != s)
+          *t = c;
+        t++;
       }
-      ptr++;
+      s++;
     }
-  }
-  else // likely to be multibyte string
-  {
-    while (ptr < end)
+    else
     {
-      str_cp = rb_enc_codepoint_len(ptr, end, &cp_len, str_enc);
-      if ((!tst_cp(cps, len, str_cp)) == delete)
+      c = rb_enc_codepoint_len(s, send, &clen, enc);
+
+      if (tst_cp(cps, cs_len, c) != delete)
       {
-        cs_str_buf_cat(new_str_buf, ptr, cp_len);
+        if (t != s)
+          rb_enc_mbcput(c, t, enc);
+        t += clen;
+        if (cr == ENC_CODERANGE_7BIT)
+          cr = ENC_CODERANGE_VALID;
       }
-      ptr += cp_len;
+      s += clen;
     }
   }
 
-  cs_str_buf_terminate(new_str_buf, str_enc);
+  rb_str_set_len(str, t - RSTRING_PTR(str));
+  ENC_CODERANGE_SET(str, cr);
 
-  if (bang)
+  if (bang && (RSTRING_LEN(str) == (long)orig_str_len)) // string unchanged
   {
-    if (RSTRING_LEN(new_str_buf) == (long)orig_len) // string unchanged
-    {
-      return Qnil;
-    }
-    rb_str_shared_replace(str, new_str_buf);
-  }
-  else
-  {
-    RB_OBJ_WRITE(new_str_buf, &(RBASIC(new_str_buf))->klass, rb_obj_class(str));
-    str = new_str_buf;
+    return Qnil;
   }
 
   return str;
@@ -1289,6 +1249,10 @@ cs_method_allocated_length(VALUE self)
 
 void Init_character_set()
 {
+#ifdef HAVE_RB_EXT_RACTOR_SAFE
+  rb_ext_ractor_safe(true);
+#endif
+
   VALUE cs = rb_define_class("CharacterSet", rb_cObject);
 
   rb_define_alloc_func(cs, cs_method_allocate);
@@ -1343,7 +1307,7 @@ void Init_character_set()
   // `CharacterSet`-specific methods
 
   rb_define_singleton_method(cs, "from_ranges", cs_class_method_from_ranges, -2);
-  rb_define_singleton_method(cs, "of", cs_class_method_of, -1);
+  rb_define_singleton_method(cs, "of_string", cs_class_method_of_string, 1);
 
   rb_define_method(cs, "ranges", cs_method_ranges, 0);
   rb_define_method(cs, "sample", cs_method_sample, -1);

data/lib/character_set/core_ext/string_ext.rb

@@ -2,7 +2,7 @@ class CharacterSet
   module CoreExt
     module StringExt
       def character_set
-        CharacterSet.of(self)
+        CharacterSet.of_string(self)
       end
 
       {

data/lib/character_set/predefined_sets.rb

@@ -22,6 +22,17 @@ class CharacterSet
     alias valid                    unicode
 
     def build_from_cps_file(path)
+      if defined?(Ractor) && Ractor.current != Ractor.main
+        raise <<-EOS.gsub(/^ */, '')
+          CharacterSet's predefined sets are lazy-loaded.
+          Pre-load them to use them in Ractors. E.g.:
+
+          CharacterSet.ascii # pre-load
+          Ractor.new { CharacterSet.ascii.size }.take # => 128
+          Ractor.new { 'abc'.keep_character_set(:ascii) }.take # => 'abc'
+        EOS
+      end
+
       File.readlines(path).inject(new) do |set, line|
         range_start, range_end = line.split(',')
         set.merge((range_start.to_i(16))..(range_end.to_i(16)))

data/lib/character_set/ruby_fallback/character_set_methods.rb

@@ -6,13 +6,9 @@ class CharacterSet
           new(Array(ranges).flat_map(&:to_a))
         end
 
-        def of(*strings)
-          new_set = new
-          strings.each do |str|
-            raise ArgumentError, 'pass a String' unless str.respond_to?(:codepoints)
-            str.codepoints.each { |cp| new_set << cp }
-          end
-          new_set
+        def of_string(str)
+          raise ArgumentError, 'pass a String' unless str.respond_to?(:codepoints)
+          str.codepoints.each_with_object(new) { |cp, set| set << cp }
         end
       end
 

data/lib/character_set/shared_methods.rb

@@ -15,6 +15,12 @@ class CharacterSet
             new(Array(args))
           end
 
+          def of(*args)
+            args.map do |arg|
+              arg.is_a?(Regexp) ? of_regexp(arg) : of_string(arg)
+            end.reduce(:merge) || new
+          end
+
           def parse(string)
             codepoints = Parser.codepoints_from_bracket_expression(string)
             result = new(codepoints)

data/lib/character_set/version.rb

@@ -1,3 +1,3 @@
 class CharacterSet
-  VERSION = '1.5.0'
+  VERSION = '1.6.0'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: character_set
 version: !ruby/object:Gem::Version
-  version: 1.5.0
+  version: 1.6.0
 platform: ruby
 authors:
 - Janosch Müller
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-12-05 00:00:00.000000000 Z
+date: 2022-02-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: sorted_set
@@ -24,160 +24,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.0'
-- !ruby/object:Gem::Dependency
-  name: benchmark-ips
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '2.7'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '2.7'
-- !ruby/object:Gem::Dependency
-  name: get_process_mem
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.2.3
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.2.3
-- !ruby/object:Gem::Dependency
-  name: rake
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '13.0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '13.0'
-- !ruby/object:Gem::Dependency
-  name: rake-compiler
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.1'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.1'
-- !ruby/object:Gem::Dependency
-  name: range_compressor
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.0'
-- !ruby/object:Gem::Dependency
-  name: regexp_parser
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '2.1'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '2.1'
-- !ruby/object:Gem::Dependency
-  name: regexp_property_values
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.0'
-- !ruby/object:Gem::Dependency
-  name: rspec
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '3.8'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '3.8'
-- !ruby/object:Gem::Dependency
-  name: codecov
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.2.12
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.2.12
-- !ruby/object:Gem::Dependency
-  name: gouteur
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 1.0.0
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 1.0.0
-- !ruby/object:Gem::Dependency
-  name: rubocop
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.8'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '1.8'
 description:
 email:
 - janosch84@gmail.com
@@ -269,7 +115,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.3.0.dev
+rubygems_version: 3.4.0.dev
 signing_key:
 specification_version: 4
 summary: Build, read, write and compare sets of Unicode codepoints.