re2 1.17.0 → 1.17.1

package/README.md

@@ -343,6 +343,7 @@ console.log('re2_res : ' + re2_res); // prints: re2_res : abc,a,b,c
 
 ## Release history
 
+- 1.17.1 *Fix for `lastIndex` for U+10000 - U+10FFFF UTF characters. Thx, [omg](https://github.com/omg).*
 - 1.17.0 *Updated GYP, added support for Node 17, updated deps.*
 - 1.16.0 *Updated the compiler (thx, [Sergei Dyshel](https://github.com/sergei-dyshel)), updated GYP, removed support for Node 10, added support for Node 16, updated TS bindings (thx, [BannerBomb](https://github.com/BannerBomb)).*
 - 1.15.9 *Updated deps.*

package/lib/exec.cc

@@ -50,7 +50,10 @@ NAN_METHOD(WrappedRE2::Exec)
 			}
 			for (size_t n = re2->lastIndex; n; --n)
 			{
-				lastIndex += getUtf8CharSize(str.data[lastIndex]);
+				size_t s = getUtf8CharSize(str.data[lastIndex]);
+				lastIndex += s;
+				if (s == 4 && n >= 2) --n; // this utf8 character will take two utf16 characters
+				// the decrement above is protected to avoid an overflow of an unsigned integer
 			}
 		}
 	}

package/lib/match.cc

@@ -59,7 +59,10 @@ NAN_METHOD(WrappedRE2::Match)
 		{
 			for (size_t n = re2->lastIndex; n; --n)
 			{
-				lastIndex += getUtf8CharSize(a.data[lastIndex]);
+				size_t s = getUtf8CharSize(a.data[lastIndex]);
+				lastIndex += s;
+				if (s == 4 && n >= 2) --n; // this utf8 character will take two utf16 characters
+				// the decrement above is protected to avoid an overflow of an unsigned integer
 			}
 			anchor = RE2::ANCHOR_START;
 		}

package/lib/replace.cc

@@ -226,7 +226,10 @@ static Nan::Maybe<std::string> replace(WrappedRE2 *re2, const StrVal &replacee,
 			{
 				for (size_t n = re2->lastIndex; n; --n)
 				{
-					lastIndex += getUtf8CharSize(data[lastIndex]);
+					size_t s = getUtf8CharSize(data[lastIndex]);
+					lastIndex += s;
+					if (s == 4 && n >= 2) --n; // this utf8 character will take two utf16 characters
+					// the decrement above is protected to avoid an overflow of an unsigned integer
 				}
 			}
 		}
@@ -372,7 +375,10 @@ static Nan::Maybe<std::string> replace(WrappedRE2 *re2, const StrVal &replacee,
 			{
 				for (size_t n = re2->lastIndex; n; --n)
 				{
-					lastIndex += getUtf8CharSize(data[lastIndex]);
+					size_t s = getUtf8CharSize(data[lastIndex]);
+					lastIndex += s;
+					if (s == 4 && n >= 2) --n; // this utf8 character will take two utf16 characters
+					// the decrement above is protected to avoid an overflow of an unsigned integer
 				}
 			}
 		}

package/lib/test.cc

@@ -50,7 +50,10 @@ NAN_METHOD(WrappedRE2::Test)
 			}
 			for (size_t n = re2->lastIndex; n; --n)
 			{
-				lastIndex += getUtf8CharSize(str.data[lastIndex]);
+				size_t s = getUtf8CharSize(str.data[lastIndex]);
+				lastIndex += s;
+				if (s == 4 && n >= 2) --n; // this utf8 character will take two utf16 characters
+				// the decrement above is protected to avoid an overflow of an unsigned integer
 			}
 		}
 	}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "re2",
-  "version": "1.17.0",
+  "version": "1.17.1",
   "description": "Bindings for RE2: fast, safe alternative to backtracking regular expression engines.",
   "homepage": "https://github.com/uhop/node-re2",
   "bugs": "https://github.com/uhop/node-re2/issues",

package/tests/test_exec.js

@@ -284,6 +284,22 @@ unit.add(module, [
 		eval(t.TEST("re2.lastIndex === 6"));
   },
 
+	function test_execSupplemental(t) {
+		"use strict";
+
+		var re = new RE2("\\w+", "g");
+		var testString = "🤡🤡🤡 Hello clown world!";
+
+		var result = re.exec(testString);
+		eval(t.TEST("t.unify(result, ['Hello'])"));
+		
+		result = re.exec(testString);
+		eval(t.TEST("t.unify(result, ['clown'])"));
+
+		result = re.exec(testString);
+		eval(t.TEST("t.unify(result, ['world'])"));
+	},
+
   // Multiline test
 
   function test_execMultiline(t) {