use code unit without u flag

acornjs · Feb 14, 2018 · 7a008b1 · 7a008b1
1 parent 68fd78e
commit 7a008b1
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 6 deletions.
diff --git a/src/regexp.js b/src/regexp.js
@@ -96,15 +96,16 @@ export class RegExpValidationState {
     this.parser.raise(this.start, `Invalid regular expression: /${this.source}/: ${message}`)
   }
 
-  // Node.js 0.12/0.10 don't support String.prototype.codePointAt().
-  codePointAt(i) {
+  // If u flag is given, this returns the code point at the index (it combines a surrogate pair).
+  // Otherwise, this returns the code unit of the index (can be a part of a surrogate pair).
+  at(i) {
     const s = this.source
     const l = s.length
     if (i >= l) {
       return -1
     }
     const c = s.charCodeAt(i)
-    if (c <= 0xD7FF || c >= 0xE000 || i + 1 >= l) {
+    if (!this.switchU || c <= 0xD7FF || c >= 0xE000 || i + 1 >= l) {
       return c
     }
     return (c << 10) + s.charCodeAt(i + 1) - 0x35FDC00
@@ -117,18 +118,18 @@ export class RegExpValidationState {
       return l
     }
     const c = s.charCodeAt(i)
-    if (c <= 0xD7FF || c >= 0xE000 || i + 1 >= l) {
+    if (!this.switchU || c <= 0xD7FF || c >= 0xE000 || i + 1 >= l) {
       return i + 1
     }
     return i + 2
   }
 
   current() {
-    return this.codePointAt(this.pos)
+    return this.at(this.pos)
   }
 
   lookahead() {
-    return this.codePointAt(this.nextIndex(this.pos))
+    return this.at(this.nextIndex(this.pos))
   }
 
   advance() {

diff --git a/test/tests-regexp.js b/test/tests-regexp.js
@@ -943,6 +943,11 @@ testFail("/[🌷-🌸]/", "Invalid regular expression: /[🌷-🌸]/: Range out
 testFail("/[🌷-🌸]/", "Invalid regular expression: /[🌷-🌸]/: Range out of order in character class (1:1)", { ecmaVersion: 2015 })
 testFail("/[🌷-🌸]/u", "Invalid regular expression flag (1:1)", { ecmaVersion: 5 })
 test("/[🌷-🌸]/u", {}, { ecmaVersion: 2015 })
+testFail("/[\\u0000-🌸-\\u0000]/", "Invalid regular expression: /[\\u0000-🌸-\\u0000]/: Range out of order in character class (1:1)", { ecmaVersion: 2015 })
+testFail("/[\\u0000-\\ud83c\\udf38-\\u0000]/", "Invalid regular expression: /[\\u0000-\\ud83c\\udf38-\\u0000]/: Range out of order in character class (1:1)", { ecmaVersion: 2015 })
+test("/[\\u0000-🌸-\\u0000]/u", {}, { ecmaVersion: 2015 })
+test("/[\\u0000-\\u{1f338}-\\u0000]/u", {}, { ecmaVersion: 2015 })
+test("/[\\u0000-\\ud83c\\udf38-\\u0000]/u", {}, { ecmaVersion: 2015 })
 testFail("/[🌸-🌷]/", "Invalid regular expression: /[🌸-🌷]/: Range out of order in character class (1:1)", { ecmaVersion: 5 })
 testFail("/[🌸-🌷]/", "Invalid regular expression: /[🌸-🌷]/: Range out of order in character class (1:1)", { ecmaVersion: 2015 })
 testFail("/[🌸-🌷]/u", "Invalid regular expression flag (1:1)", { ecmaVersion: 5 })