re2 1.16.0 → 1.17.3

package/.github/actions/{linux-alpine-node-15 → linux-alpine-node-17}/Dockerfile

@@ -1,4 +1,4 @@
-FROM node:15-alpine
+FROM node:17-alpine
 
 RUN apk add --no-cache python3 make gcc g++
 

package/.github/actions/linux-alpine-node-17/action.yml

@@ -0,0 +1,7 @@
+name: 'Create a binary artifact for Node 17 on Alpine Linux'
+description: 'Create a binary artifact for Node 17 on Alpine Linux using musl'
+runs:
+  using: 'docker'
+  image: 'Dockerfile'
+  args:
+    - ${{inputs.node-version}}

package/.github/dependabot.yml

@@ -5,7 +5,7 @@
 
 version: 2
 updates:
-  - package-ecosystem: "nvm" # See documentation for possible values
+  - package-ecosystem: "npm" # See documentation for possible values
     directory: "/" # Location of package manifests
     schedule:
       interval: "weekly"

package/.github/workflows/build.yml

@@ -31,7 +31,7 @@ jobs:
     strategy:
       matrix:
         os: [windows-latest, macOS-latest]
-        node-version: [12, 14, 15, 16]
+        node-version: [12, 14, 16, 17]
 
     steps:
     - uses: actions/checkout@v2
@@ -73,7 +73,7 @@ jobs:
 
     strategy:
       matrix:
-        node-version: [12, 14, 15, 16]
+        node-version: [12, 14, 16, 17]
 
     steps:
     - uses: actions/checkout@v2
@@ -152,8 +152,8 @@ jobs:
       env:
         GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
 
-  build-linux-alpine-node-15:
-    name: Node.js 15 on Alpine Linux
+  build-linux-alpine-node-16:
+    name: Node.js 16 on Alpine Linux
     needs: create-release
     runs-on: ubuntu-latest
     continue-on-error: true
@@ -175,12 +175,12 @@ jobs:
           Linux-Alpine-node-
           Linux-Alpine-
     - name: Install, test, and create artifact
-      uses: ./.github/actions/linux-alpine-node-15/
+      uses: ./.github/actions/linux-alpine-node-16/
       env:
         GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}
 
-  build-linux-alpine-node-16:
-    name: Node.js 16 on Alpine Linux
+  build-linux-alpine-node-17:
+    name: Node.js 17 on Alpine Linux
     needs: create-release
     runs-on: ubuntu-latest
     continue-on-error: true
@@ -202,6 +202,6 @@ jobs:
           Linux-Alpine-node-
           Linux-Alpine-
     - name: Install, test, and create artifact
-      uses: ./.github/actions/linux-alpine-node-16/
+      uses: ./.github/actions/linux-alpine-node-17/
       env:
         GITHUB_TOKEN: ${{secrets.GITHUB_TOKEN}}

package/.github/workflows/tests.yml

@@ -14,7 +14,7 @@ jobs:
     strategy:
       matrix:
         os: [ubuntu-latest, windows-latest, macOS-latest]
-        node-version: [12, 14, 15, 16]
+        node-version: [12, 14, 16, 17]
 
     steps:
     - uses: actions/checkout@v2

package/README.md

@@ -13,6 +13,10 @@ at his [Implementing Regular Expressions](http://swtch.com/~rsc/regexp/) page.
 (see [Syntax](https://github.com/google/re2/wiki/Syntax)),
 but it lacks two features: backreferences and lookahead assertions. See below for more details.
 
+`RE2` always works in the [Unicode mode](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/RegExp/unicode), which means that all matches that use character codes are interpret as Unicode code points, not as binary values of UTF-16.
+See `RE2.unicodeWarningLevel` below for more details.
+
+
 `RE2` object emulates standard `RegExp` making it a practical drop-in replacement in most cases.
 `RE2` is extended to provide `String`-based regular expression methods as well. To help to convert
 `RegExp` objects to `RE2` its constructor can take `RegExp` directly honoring all properties.
@@ -343,6 +347,10 @@ console.log('re2_res : ' + re2_res); // prints: re2_res : abc,a,b,c
 
 ## Release history
 
+- 1.17.3 *Fixed bug with zero-length replacements.*
+- 1.17.2 *Added support for the enhanced local mirroring by updating [install-artifact-from-github](https://github.com/uhop/install-artifact-from-github).*
+- 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.*
 - 1.15.8 *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 +226,12 @@ 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
 				}
 			}
 		}
@@ -242,28 +247,30 @@ static Nan::Maybe<std::string> replace(WrappedRE2 *re2, const StrVal &replacee,
 	while (lastIndex <= size && re2->regexp.Match(str, lastIndex, size, anchor, &groups[0], groups.size()))
 	{
 		noMatch = false;
+		auto offset = match.data() - data;
 		if (!re2->global && re2->sticky)
 		{
-			re2->lastIndex += replacee.isBuffer ? match.data() - data + match.size() - lastIndex : getUtf16Length(data + lastIndex, match.data() + match.size());
+			re2->lastIndex += replacee.isBuffer ? offset + match.size() - lastIndex : getUtf16Length(data + lastIndex, match.data() + match.size());
+		}
+		if (match.data() == data || offset > static_cast<long>(lastIndex))
+		{
+			result += std::string(data + lastIndex, offset - lastIndex);
 		}
+		result += replace(replacer, replacer_size, groups, str, namedGroups);
 		if (match.size())
 		{
-			if (match.data() == data || match.data() - data > static_cast<long>(lastIndex))
-			{
-				result += std::string(data + lastIndex, match.data() - data - lastIndex);
-			}
-			result += replace(replacer, replacer_size, groups, str, namedGroups);
-			lastIndex = match.data() - data + match.size();
+			lastIndex = offset + match.size();
+		}
+		else if (offset < size)
+		{
+			auto sym_size = getUtf8CharSize(data[offset]);
+			result.append(data + offset, sym_size);
+			lastIndex = offset + sym_size;
 		}
 		else
 		{
-			result += replace(replacer, replacer_size, groups, str, namedGroups);
-			size_t sym_size = getUtf8CharSize(data[lastIndex]);
-			if (lastIndex < size)
-			{
-				result.append(data + lastIndex, sym_size);
-			}
-			lastIndex += sym_size;
+			lastIndex = size;
+			break;
 		}
 		if (!re2->global)
 		{
@@ -292,7 +299,7 @@ static Nan::Maybe<std::string> replace(WrappedRE2 *re2, const StrVal &replacee,
 
 inline Nan::Maybe<std::string> replace(const Nan::Callback *replacer, const std::vector<re2::StringPiece> &groups, const re2::StringPiece &str, const v8::Local<v8::Value> &input, bool useBuffers, const std::map<std::string, int> &namedGroups)
 {
-	std::vector<v8::Local<v8::Value>> argv;
+	std::vector<v8::Local<v8::Value> > argv;
 
 	auto context = Nan::GetCurrentContext();
 
@@ -372,7 +379,12 @@ 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
 				}
 			}
 		}
@@ -390,38 +402,35 @@ static Nan::Maybe<std::string> replace(WrappedRE2 *re2, const StrVal &replacee,
 	while (lastIndex <= size && re2->regexp.Match(str, lastIndex, size, anchor, &groups[0], groups.size()))
 	{
 		noMatch = false;
+		auto offset = match.data() - data;
 		if (!re2->global && re2->sticky)
 		{
-			re2->lastIndex += replacee.isBuffer ? match.data() - data + match.size() - lastIndex : getUtf16Length(data + lastIndex, match.data() + match.size());
+			re2->lastIndex += replacee.isBuffer ? offset + match.size() - lastIndex : getUtf16Length(data + lastIndex, match.data() + match.size());
+		}
+		if (match.data() == data || offset > static_cast<long>(lastIndex))
+		{
+			result += std::string(data + lastIndex, offset - lastIndex);
 		}
+		const auto part = replace(replacer, groups, str, input, useBuffers, namedGroups);
+		if (part.IsNothing())
+		{
+			return part;
+		}
+		result += part.FromJust();
 		if (match.size())
 		{
-			if (match.data() == data || match.data() - data > static_cast<long>(lastIndex))
-			{
-				result += std::string(data + lastIndex, match.data() - data - lastIndex);
-			}
-			const auto part = replace(replacer, groups, str, input, useBuffers, namedGroups);
-			if (part.IsNothing())
-			{
-				return part;
-			}
-			result += part.FromJust();
-			lastIndex = match.data() - data + match.size();
+			lastIndex = offset + match.size();
+		}
+		else if (offset < size)
+		{
+			auto sym_size = getUtf8CharSize(data[offset]);
+			result.append(data + offset, sym_size);
+			lastIndex = offset + sym_size;
 		}
 		else
 		{
-			const auto part = replace(replacer, groups, str, input, useBuffers, namedGroups);
-			if (part.IsNothing())
-			{
-				return part;
-			}
-			result += part.FromJust();
-			size_t sym_size = getUtf8CharSize(data[lastIndex]);
-			if (lastIndex < size)
-			{
-				result.append(data + lastIndex, sym_size);
-			}
-			lastIndex += sym_size;
+			lastIndex = size;
+			break;
 		}
 		if (!re2->global)
 		{

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": "1.16.0",
+  "version": "1.17.3",
   "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",
@@ -10,9 +10,9 @@
     "test": "tests"
   },
   "dependencies": {
-    "install-artifact-from-github": "^1.2.0",
-    "nan": "^2.14.2",
-    "node-gyp": "^8.0.0"
+    "install-artifact-from-github": "^1.3.0",
+    "nan": "^2.15.0",
+    "node-gyp": "^8.4.1"
   },
   "devDependencies": {
     "heya-unit": "^0.3.0"
@@ -20,7 +20,7 @@
   "scripts": {
     "test": "node tests/tests.js",
     "save-to-github": "save-to-github-cache --artifact build/Release/re2.node",
-    "install": "install-from-cache --artifact build/Release/re2.node --host-var RE2_DOWNLOAD_MIRROR || npm run rebuild",
+    "install": "install-from-cache --artifact build/Release/re2.node --host-var RE2_DOWNLOAD_MIRROR --skip-path-var RE2_DOWNLOAD_SKIP_PATH --skip-ver-var RE2_DOWNLOAD_SKIP_VER || npm run rebuild",
     "verify-build": "node scripts/verify-build.js",
     "rebuild": "node-gyp rebuild"
   },
@@ -35,6 +35,6 @@
     "text processing",
     "PCRE alternative"
   ],
-  "author": "Eugene Lazutkin <eugene.lazutkin@gmail.com> (http://lazutkin.com/)",
+  "author": "Eugene Lazutkin <eugene.lazutkin@gmail.com> (https://lazutkin.com/)",
   "license": "BSD-3-Clause"
 }

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) {

package/tests/test_general.js

@@ -209,6 +209,12 @@ unit.add(module, [
 		eval(t.TEST("s3.length === 1"));
 		eval(t.TEST("RE2.getUtf8Length(s3) === 3"));
 
+		var s4 = "🤡";
+
+		eval(t.TEST("s4.length === 2"));
+		eval(t.TEST("RE2.getUtf8Length(s4) === 4"));
+		eval(t.TEST("RE2.getUtf16Length(Buffer.from(s4, 'utf8')) === s4.length"));
+
 		var b3 = new Buffer([0xF0]);
 
 		eval(t.TEST("b3.length === 1"));