[GR-32618] Add truffle-regex-ignore-atomic-groups option.

PullRequest: truffleruby/3242

Author: jirkamarsik

spec/tags/truffle/options/truffle_regex_ignore_atomic_groups_tags.txt

@@ -0,0 +1 @@
+slow:The --truffle-regex-ignore-atomic-groups option works

spec/truffle/options/truffle_regex_ignore_atomic_groups_spec.rb

@@ -0,0 +1,21 @@
+# Copyright (c) 2022 Oracle and/or its affiliates. All rights reserved. This
+# code is released under a tri EPL/GPL/LGPL license. You can use it,
+# redistribute it and/or modify it under the terms of the:
+#
+# Eclipse Public License version 2.0, or
+# GNU General Public License version 2, or
+# GNU Lesser General Public License version 2.1.
+
+require_relative '../../ruby/spec_helper'
+
+describe "The --truffle-regex-ignore-atomic-groups option" do
+  it "works" do
+    out = ruby_exe('/foo(?>(?:A)+)Abar/ =~ "fooAAAAAbar"', options: "--experimental-options --truffle-regex-ignore-atomic-groups --compare-regex-engines", args: "2>&1")
+    out.should == <<~OUT
+      match_in_region(/foo(?>(?:A)+)Abar/, "fooAAAAAbar"@UTF-8, 0, 11, false, 0) enc=? gave
+          0 - 11
+      but we expected
+          NO MATCH
+    OUT
+  end
+end

src/main/java/org/truffleruby/core/regexp/TRegexCache.java

@@ -174,7 +174,12 @@ private static Object compileTRegex(RubyContext context, RubyRegexp regexp, bool
             return null;
         }
 
-        String regex = "Flavor=Ruby,Encoding=" + tRegexEncoding + "/" + processedRegexpSource + "/" + flags;
+        String ignoreAtomicGroups = context.getOptions().TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS
+                ? ",IgnoreAtomicGroups=true"
+                : "";
+
+        String regex = "Flavor=Ruby,Encoding=" + tRegexEncoding + ignoreAtomicGroups + "/" + processedRegexpSource +
+                "/" + flags;
         Source regexSource = Source
                 .newBuilder("regex", regex, "Regexp")
                 .mimeType("application/tregex")

src/main/java/org/truffleruby/options/Options.java

@@ -136,6 +136,8 @@ public class Options {
     public final boolean WARN_TRUFFLE_REGEX_COMPILE_FALLBACK;
     /** --warn-truffle-regex-match-fallback=false */
     public final boolean WARN_TRUFFLE_REGEX_MATCH_FALLBACK;
+    /** --truffle-regex-ignore-atomic-groups=false */
+    public final boolean TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS;
     /** --argv-globals=false */
     public final boolean ARGV_GLOBALS;
     /** --chomp-loop=false */
@@ -264,6 +266,7 @@ public Options(Env env, OptionValues options, LanguageOptions languageOptions) {
         USE_TRUFFLE_REGEX = options.get(OptionsCatalog.USE_TRUFFLE_REGEX_KEY);
         WARN_TRUFFLE_REGEX_COMPILE_FALLBACK = options.get(OptionsCatalog.WARN_TRUFFLE_REGEX_COMPILE_FALLBACK_KEY);
         WARN_TRUFFLE_REGEX_MATCH_FALLBACK = options.get(OptionsCatalog.WARN_TRUFFLE_REGEX_MATCH_FALLBACK_KEY);
+        TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS = options.get(OptionsCatalog.TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS_KEY);
         ARGV_GLOBALS = options.get(OptionsCatalog.ARGV_GLOBALS_KEY);
         CHOMP_LOOP = options.get(OptionsCatalog.CHOMP_LOOP_KEY);
         GETS_LOOP = options.get(OptionsCatalog.GETS_LOOP_KEY);
@@ -415,6 +418,8 @@ public Object fromDescriptor(OptionDescriptor descriptor) {
                 return WARN_TRUFFLE_REGEX_COMPILE_FALLBACK;
             case "ruby.warn-truffle-regex-match-fallback":
                 return WARN_TRUFFLE_REGEX_MATCH_FALLBACK;
+            case "ruby.truffle-regex-ignore-atomic-groups":
+                return TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS;
             case "ruby.argv-globals":
                 return ARGV_GLOBALS;
             case "ruby.chomp-loop":

src/options.yml

@@ -148,6 +148,7 @@ EXPERT:
     USE_TRUFFLE_REGEX: [use-truffle-regex, boolean, true, 'Use the Truffle regular expression engine when possible and fallback to Joni otherwise']
     WARN_TRUFFLE_REGEX_COMPILE_FALLBACK: [warn-truffle-regex-compile-fallback, boolean, false, 'Warn when a Ruby Regexp could not be compiled to a Truffle Regex and Joni is used instead']
     WARN_TRUFFLE_REGEX_MATCH_FALLBACK: [warn-truffle-regex-match-fallback, boolean, false, 'Warn every time Truffle Regex cannot be used for a Regexp match (and instead Joni is used)']
+    TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS: [truffle-regex-ignore-atomic-groups, boolean, false, 'Treat atomic groups (?>...) as ordinary groups (?:...) with Truffle Regex.']
 
 INTERNAL: # Options for debugging the TruffleRuby implementation
   EXPERIMENTAL:

src/shared/java/org/truffleruby/shared/options/OptionsCatalog.java

@@ -86,6 +86,7 @@ public class OptionsCatalog {
     public static final OptionKey<Boolean> USE_TRUFFLE_REGEX_KEY = new OptionKey<>(true);
     public static final OptionKey<Boolean> WARN_TRUFFLE_REGEX_COMPILE_FALLBACK_KEY = new OptionKey<>(false);
     public static final OptionKey<Boolean> WARN_TRUFFLE_REGEX_MATCH_FALLBACK_KEY = new OptionKey<>(false);
+    public static final OptionKey<Boolean> TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS_KEY = new OptionKey<>(false);
     public static final OptionKey<Boolean> ARGV_GLOBALS_KEY = new OptionKey<>(false);
     public static final OptionKey<Boolean> CHOMP_LOOP_KEY = new OptionKey<>(false);
     public static final OptionKey<Boolean> GETS_LOOP_KEY = new OptionKey<>(false);
@@ -691,6 +692,14 @@ public class OptionsCatalog {
             .usageSyntax("")
             .build();
 
+    public static final OptionDescriptor TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS = OptionDescriptor
+            .newBuilder(TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS_KEY, "ruby.truffle-regex-ignore-atomic-groups")
+            .help("Treat atomic groups (?>...) as ordinary groups (?:...) with Truffle Regex.")
+            .category(OptionCategory.EXPERT)
+            .stability(OptionStability.EXPERIMENTAL)
+            .usageSyntax("")
+            .build();
+
     public static final OptionDescriptor ARGV_GLOBALS = OptionDescriptor
             .newBuilder(ARGV_GLOBALS_KEY, "ruby.argv-globals")
             .help("Parse options in script argv into global variables (configured by the -s Ruby option)")
@@ -1433,6 +1442,8 @@ public static OptionDescriptor fromName(String name) {
                 return WARN_TRUFFLE_REGEX_COMPILE_FALLBACK;
             case "ruby.warn-truffle-regex-match-fallback":
                 return WARN_TRUFFLE_REGEX_MATCH_FALLBACK;
+            case "ruby.truffle-regex-ignore-atomic-groups":
+                return TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS;
             case "ruby.argv-globals":
                 return ARGV_GLOBALS;
             case "ruby.chomp-loop":
@@ -1658,6 +1669,7 @@ public static OptionDescriptor[] allDescriptors() {
             USE_TRUFFLE_REGEX,
             WARN_TRUFFLE_REGEX_COMPILE_FALLBACK,
             WARN_TRUFFLE_REGEX_MATCH_FALLBACK,
+            TRUFFLE_REGEX_IGNORE_ATOMIC_GROUPS,
             ARGV_GLOBALS,
             CHOMP_LOOP,
             GETS_LOOP,