Raise a RegexpError for empty or too large \u{} escapes

natalie-lang · Sep 20, 2024 · 8a7666c · 8a7666c
1 parent bde3657
commit 8a7666c
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 6 deletions.
diff --git a/spec/core/regexp/shared/new.rb b/spec/core/regexp/shared/new.rb
@@ -459,15 +459,11 @@ def obj.to_int() ScratchPad.record(:called) end
     end
 
     it "raises a RegexpError if the \\u{} escape is empty" do
-      NATFIXME "raises a RegexpError if the \\u{} escape is empty", exception: SpecFailedException do
-        -> { Regexp.send(@method, "\\" + "u{}") }.should raise_error(RegexpError, Regexp.new(Regexp.escape("invalid Unicode list: /\\u{}/")))
-      end
+      -> { Regexp.send(@method, "\\" + "u{}") }.should raise_error(RegexpError, Regexp.new(Regexp.escape("invalid Unicode list: /\\u{}/")))
     end
 
     it "raises a RegexpError if more than six hexadecimal digits are given" do
-      NATFIXME "raises a RegexpError if more than six hexadecimal digits are given", exception: SpecFailedException do
-        -> { Regexp.send(@method, "\\" + "u{0ffffff}") }.should raise_error(RegexpError, Regexp.new(Regexp.escape("invalid Unicode range: /\\u{0ffffff}/")))
-      end
+      -> { Regexp.send(@method, "\\" + "u{0ffffff}") }.should raise_error(RegexpError, Regexp.new(Regexp.escape("invalid Unicode range: /\\u{0ffffff}/")))
     end
 
     it "returns a Regexp with US-ASCII encoding if only 7-bit ASCII characters are present regardless of the input String's encoding" do

diff --git a/src/regexp_object.cpp b/src/regexp_object.cpp
@@ -302,9 +302,11 @@ static String prepare_pattern_for_onigmo(Env *env, const StringObject *pattern,
                 // Convert \u{dd} to \udddd
                 case '{': {
                     c = next_char();
+                    size_t num_digits = 0;
                     do {
                         long codepoint = 0;
                         while (isxdigit(c)) {
+                            num_digits++;
                             codepoint *= 16;
                             if (isdigit(c))
                                 codepoint += c - '0';
@@ -323,6 +325,11 @@ static String prepare_pattern_for_onigmo(Env *env, const StringObject *pattern,
                         while (c == ' ')
                             c = next_char();
                     } while (c != '}');
+                    if (num_digits == 0) {
+                        env->raise("RegexpError", "invalid Unicode list: /{}/", pattern->string());
+                    } else if (num_digits > 6) {
+                        env->raise("RegexpError", "invalid Unicode range: /{}/", pattern->string());
+                    }
                     break;
                 }