From 7415a8e8b261285dddc9e898ce341835bc718116 Mon Sep 17 00:00:00 2001
From: James Prior <jamesgr.prior@gmail.com>
Date: Thu, 9 May 2024 18:08:44 +0100
Subject: [PATCH 1/2] Map JS RegExp to I-Regexp

---
 src/path/functions/match.ts  | 53 ++++++++++++++++++++++++++++++++++--
 src/path/functions/search.ts | 40 ++++++++++++++++++++++++++-
 src/path/lex.ts              |  2 +-
 tests/path/cts               |  2 +-
 4 files changed, 91 insertions(+), 6 deletions(-)

diff --git a/src/path/functions/match.ts b/src/path/functions/match.ts
index 6a0c42d..ace9a99 100644
--- a/src/path/functions/match.ts
+++ b/src/path/functions/match.ts
@@ -58,9 +58,56 @@ export class Match implements FilterFunction {
 
   protected fullMatch(pattern: string): string {
     const parts: string[] = [];
-    if (!pattern.startsWith("^")) parts.push("^");
-    parts.push(pattern);
-    if (!pattern.endsWith("$")) parts.push("$");
+    let nonCaptureGroup = false;
+
+    if (!pattern.startsWith("^") && !pattern.startsWith("^(")) {
+      nonCaptureGroup = true;
+      parts.push("^(?:");
+    }
+    parts.push(this.mapRegexp(pattern));
+
+    if (nonCaptureGroup && !pattern.endsWith("$") && !pattern.endsWith(")$")) {
+      parts.push(")$");
+    }
+
+    return parts.join("");
+  }
+
+  // See https://datatracker.ietf.org/doc/html/rfc9485#name-ecmascript-regexps
+  protected mapRegexp(pattern: string): string {
+    let escaped = false;
+    let charClass = false;
+    const parts: string[] = [];
+    for (const ch of pattern) {
+      switch (ch) {
+        case ".":
+          if (!escaped && !charClass) {
+            parts.push("(?:(?![\r\n])\\P{Cs}|\\p{Cs}\\p{Cs})");
+          } else {
+            parts.push(ch);
+            escaped = false;
+          }
+          break;
+        case "\\":
+          escaped = true;
+          parts.push(ch);
+          break;
+        case "[":
+          charClass = true;
+          escaped = false;
+          parts.push(ch);
+          break;
+        case "]":
+          charClass = false;
+          escaped = false;
+          parts.push(ch);
+          break;
+        default:
+          escaped = false;
+          parts.push(ch);
+          break;
+      }
+    }
     return parts.join("");
   }
 }
diff --git a/src/path/functions/search.ts b/src/path/functions/search.ts
index 15c0b00..4ebca0d 100644
--- a/src/path/functions/search.ts
+++ b/src/path/functions/search.ts
@@ -48,7 +48,7 @@ export class Search implements FilterFunction {
     }
 
     try {
-      const re = new RegExp(pattern, "u");
+      const re = new RegExp(this.mapRegexp(pattern), "u");
       if (this.cacheSize > 0) this.#cache.set(pattern, re);
       return !!s.match(re);
     } catch (error) {
@@ -56,4 +56,42 @@ export class Search implements FilterFunction {
       return false;
     }
   }
+
+  // See https://datatracker.ietf.org/doc/html/rfc9485#name-ecmascript-regexps
+  protected mapRegexp(pattern: string): string {
+    let escaped = false;
+    let charClass = false;
+    const parts: string[] = [];
+    for (const ch of pattern) {
+      switch (ch) {
+        case ".":
+          if (!escaped && !charClass) {
+            parts.push("(?:(?![\r\n])\\P{Cs}|\\p{Cs}\\p{Cs})");
+          } else {
+            parts.push(ch);
+            escaped = false;
+          }
+          break;
+        case "\\":
+          escaped = true;
+          parts.push(ch);
+          break;
+        case "[":
+          charClass = true;
+          escaped = false;
+          parts.push(ch);
+          break;
+        case "]":
+          charClass = false;
+          escaped = false;
+          parts.push(ch);
+          break;
+        default:
+          escaped = false;
+          parts.push(ch);
+          break;
+      }
+    }
+    return parts.join("");
+  }
 }
diff --git a/src/path/lex.ts b/src/path/lex.ts
index ed7e68b..a76bd22 100644
--- a/src/path/lex.ts
+++ b/src/path/lex.ts
@@ -367,6 +367,7 @@ function lexInsideBracketedSelection(l: Lexer): StateFn | null {
     }
 
     if (!l.environment.strict && l.acceptMatchRun(l.environment.keysPattern)) {
+      // FIXME: fall back to legacy behavior if keysPattern is not the default
       switch (l.peek()) {
         case "'":
           l.ignore(); // ~
@@ -377,7 +378,6 @@ function lexInsideBracketedSelection(l: Lexer): StateFn | null {
           l.next();
           return lexDoubleQuoteKeyString(l);
         case "?":
-          l.ignore(); // ~
           l.next();
           l.emit(TokenKind.KEYS_FILTER);
           l.filterLevel += 1;
diff --git a/tests/path/cts b/tests/path/cts
index bb3ae8c..7c8e9bc 160000
--- a/tests/path/cts
+++ b/tests/path/cts
@@ -1 +1 @@
-Subproject commit bb3ae8c839a700702035038a299ceed97704fbe7
+Subproject commit 7c8e9bcd92f8ed8797331de02f488ebcb7856bec

From e32b90eb35fad79fa2f6a37499f9e4761af1df2c Mon Sep 17 00:00:00 2001
From: James Prior <jamesgr.prior@gmail.com>
Date: Thu, 9 May 2024 18:20:33 +0100
Subject: [PATCH 2/2] Update change log and bump version number

---
 CHANGELOG.md | 6 ++++++
 package.json | 2 +-
 2 files changed, 7 insertions(+), 1 deletion(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 127ca0d..be18177 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,5 +1,11 @@
 # JSON P3 Change Log
 
+## Version 1.3.1 (unreleased)
+
+**Fixes**
+
+- Fixed RegExp to I-Regex pattern mapping with the `match` and `search` filter functions. We now correctly match the special `.` character to everything other than `\r` and `\n`.
+
 ## Version 1.3.0
 
 **Fixes**
diff --git a/package.json b/package.json
index 689dfbe..0b187ee 100644
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "json-p3",
-  "version": "1.3.0",
+  "version": "1.3.1",
   "author": "James Prior",
   "license": "MIT",
   "description": "JSONPath, JSON Pointer and JSON Patch",