From adf99f2fb1ffe1d89f936574549bb36a5437fee2 Mon Sep 17 00:00:00 2001
From: Anthony Fu <github@antfu.me>
Date: Thu, 12 Sep 2024 20:49:19 +0200
Subject: [PATCH] feat(engine-js): introduce `simulation` option

---
 packages/engine-javascript/src/index.ts       | 36 +++++++++++++------
 .../engine-javascript/test/verify.test.ts     |  2 +-
 2 files changed, 26 insertions(+), 12 deletions(-)

diff --git a/packages/engine-javascript/src/index.ts b/packages/engine-javascript/src/index.ts
index 77df016e3..9a639a74e 100644
--- a/packages/engine-javascript/src/index.ts
+++ b/packages/engine-javascript/src/index.ts
@@ -8,13 +8,22 @@ import { onigurumaToRegexp } from 'oniguruma-to-js'
 export interface JavaScriptRegexEngineOptions {
   /**
    * Whether to allow invalid regex patterns.
+   *
+   * @default false
    */
   forgiving?: boolean
 
+  /**
+   * Use JavaScript to simulate some unsupported regex features.
+   *
+   * @default true
+   */
+  simulation?: boolean
+
   /**
    * Cache for regex patterns.
    */
-  cache?: Map<string, RegExp | Error>
+  cache?: Map<string, RegExp | Error> | null
 
   /**
    * Custom pattern to RegExp constructor.
@@ -45,13 +54,18 @@ export class JavaScriptScanner implements PatternScanner {
 
   constructor(
     public patterns: string[],
-    public cache: Map<string, RegExp | Error>,
-    public forgiving: boolean,
-    public regexConstructor: (pattern: string) => RegExp = defaultJavaScriptRegexConstructor,
+    public options: JavaScriptRegexEngineOptions = {},
   ) {
+    const {
+      forgiving = false,
+      cache,
+      simulation = true,
+      regexConstructor = defaultJavaScriptRegexConstructor,
+    } = options
+
     this.contiguousAnchorSimulation = Array.from({ length: patterns.length }, () => false)
     this.regexps = patterns.map((p, idx) => {
-      if (p.startsWith('(^|\\G)') || p.startsWith('(\\G|^)'))
+      if (simulation && (p.startsWith('(^|\\G)') || p.startsWith('(\\G|^)')))
         this.contiguousAnchorSimulation[idx] = true
       const cached = cache?.get(p)
       if (cached) {
@@ -129,7 +143,7 @@ export class JavaScriptScanner implements PatternScanner {
         pending.push([i, match, offset])
       }
       catch (e) {
-        if (this.forgiving)
+        if (this.options.forgiving)
           continue
         throw e
       }
@@ -159,14 +173,14 @@ export class JavaScriptScanner implements PatternScanner {
  * @experimental
  */
 export function createJavaScriptRegexEngine(options: JavaScriptRegexEngineOptions = {}): RegexEngine {
-  const {
-    forgiving = false,
-    cache = new Map(),
-  } = options
+  const _options = {
+    cache: new Map(),
+    ...options,
+  }
 
   return {
     createScanner(patterns: string[]) {
-      return new JavaScriptScanner(patterns, cache, forgiving, options.regexConstructor)
+      return new JavaScriptScanner(patterns, _options)
     },
     createString(s: string) {
       return {
diff --git a/packages/engine-javascript/test/verify.test.ts b/packages/engine-javascript/test/verify.test.ts
index f04d9f22a..67e1b3099 100644
--- a/packages/engine-javascript/test/verify.test.ts
+++ b/packages/engine-javascript/test/verify.test.ts
@@ -28,7 +28,7 @@ for (const file of files) {
     for (const instance of instances) {
       i += 1
       describe(`instances ${i}`, () => {
-        const scanner = new JavaScriptScanner(instance.constractor[0], cache, false)
+        const scanner = new JavaScriptScanner(instance.constractor[0], { cache })
         let j = 0
         for (const execution of instance.executions) {
           j += 1