feat($parse): Add the ability to define the identifier characters

lgalfaso · lgalfaso · commit 3e7fa19197c5 · 2016-03-26T20:41:01.000+01:00
Add the ability to define the identifier starts and identifier continue characters
diff --git a/src/ng/parse.js b/src/ng/parse.js
@@ -150,7 +150,7 @@ Lexer.prototype = {
         this.readString(ch);
       } else if (this.isNumber(ch) || ch === '.' && this.isNumber(this.peek())) {
         this.readNumber();
-      } else if (this.isIdent(ch)) {
+      } else if (this.isIdentifierStart(this.peekMultichar())) {
         this.readIdent();
       } else if (this.is(ch, '(){}[].,;:?')) {
         this.tokens.push({index: this.index, text: ch});
@@ -194,12 +194,49 @@ Lexer.prototype = {
             ch === '\n' || ch === '\v' || ch === '\u00A0');
   },
 
-  isIdent: function(ch) {
+  isIdentifierStart: function(ch) {
+    return this.options.isIdentifierStart ?
+        this.options.isIdentifierStart(ch, this.codePointAt(ch)) :
+        this.isValidIdentifierStart(ch);
+  },
+
+  isValidIdentifierStart: function(ch) {
     return ('a' <= ch && ch <= 'z' ||
             'A' <= ch && ch <= 'Z' ||
             '_' === ch || ch === '$');
   },
 
+  isIdentifierContinue: function(ch) {
+    return this.options.isIdentifierContinue ?
+        this.options.isIdentifierContinue(ch, this.codePointAt(ch)) :
+        this.isValidIdentifierContinue(ch);
+  },
+
+  isValidIdentifierContinue: function(ch, cp) {
+    return this.isValidIdentifierStart(ch, cp) || this.isNumber(ch);
+  },
+
+  codePointAt: function(ch) {
+    if (ch.length === 1) return ch.charCodeAt(0);
+    /*jshint bitwise: false*/
+    return (ch.charCodeAt(0) << 10) + ch.charCodeAt(1) - 0x35FDC00;
+    /*jshint bitwise: true*/
+  },
+
+  peekMultichar: function() {
+    var ch = this.text.charAt(this.index);
+    var peek = this.peek();
+    if (!peek) {
+      return ch;
+    }
+    var cp1 = ch.charCodeAt(0);
+    var cp2 = peek.charCodeAt(0);
+    if (cp1 >= 0xD800 && cp1 <= 0xDBFF && cp2 >= 0xDC00 && cp2 <= 0xDFFF) {
+      return ch + peek;
+    }
+    return ch;
+  },
+
   isExpOperator: function(ch) {
     return (ch === '-' || ch === '+' || this.isNumber(ch));
   },
@@ -248,12 +285,13 @@ Lexer.prototype = {
 
   readIdent: function() {
     var start = this.index;
+    this.index += this.peekMultichar().length;
     while (this.index < this.text.length) {
-      var ch = this.text.charAt(this.index);
-      if (!(this.isIdent(ch) || this.isNumber(ch))) {
+      var ch = this.peekMultichar();
+      if (!this.isIdentifierContinue(ch)) {
         break;
       }
-      this.index++;
+      this.index += ch.length;
     }
     this.tokens.push({
       index: start,
@@ -1183,7 +1221,13 @@ ASTCompiler.prototype = {
   },
 
   nonComputedMember: function(left, right) {
-    return left + '.' + right;
+    var SAFE_IDENTIFIER = /[$_a-zA-Z][$_a-zA-Z0-9]*/;
+    var UNSAFE_CHARACTERS = /[^$_a-zA-Z0-9]/g;
+    if (SAFE_IDENTIFIER.test(right)) {
+      return left + '.' + right;
+    } else {
+      return left  + '["' + right.replace(UNSAFE_CHARACTERS, this.stringEscapeFn) + '"]';
+    }
   },
 
   computedMember: function(left, right) {
@@ -1746,6 +1790,7 @@ function $ParseProvider() {
     'null': null,
     'undefined': undefined
   };
+  var identStart, identContinue;
 
   /**
    * @ngdoc method
@@ -1762,17 +1807,50 @@ function $ParseProvider() {
     literals[literalName] = literalValue;
   };
 
+ /**
+  * @ngdoc method
+  * @name $parseProvider#setIdentifierFns
+  * @description
+  *
+  * Allows defining the set of characters that are allowed in Angular expressions. The function
+  * `identifierStart` will get called to know if a given character is a valid character to be the
+  * first character for an identifier. The function `identifierContinue` will get called to know if
+  * a given character is a valid character to be a follow-up identifier character. The functions
+  * `identifierStart` and `identifierContinue` will receive as arguments the single character to be
+  * identifier and the character code point. These arguments will be `string` and `numeric`. Keep in
+  * mind that the `string` parameter can be two characters long depending on the character
+  * representation. It is expected for the function to return `true` or `false`, whether that
+  * character is allowed or not.
+  *
+  * Since this function will be called extensivelly, keep the implementation of these functions fast,
+  * as the performance of these functions have a direct impact on the expressions parsing speed.
+  *
+  * @param {function=} identifierStart The function that will decide whether the given character is
+  *   a valid identifier start character.
+  * @param {function=} identifierContinue The function that will decide whether the given character is
+  *   a valid identifier continue character.
+  */
+  this.setIdentifierFns = function(identifierStart, identifierContinue) {
+    identStart = identifierStart;
+    identContinue = identifierContinue;
+    return this;
+  };
+
   this.$get = ['$filter', function($filter) {
     var noUnsafeEval = csp().noUnsafeEval;
     var $parseOptions = {
           csp: noUnsafeEval,
           expensiveChecks: false,
-          literals: copy(literals)
+          literals: copy(literals),
+          isIdentifierStart: isFunction(identStart) && identStart,
+          isIdentifierContinue: isFunction(identContinue) && identContinue
         },
         $parseOptionsExpensive = {
           csp: noUnsafeEval,
           expensiveChecks: true,
-          literals: copy(literals)
+          literals: copy(literals),
+          isIdentifierStart: isFunction(identStart) && identStart,
+          isIdentifierContinue: isFunction(identContinue) && identContinue
         };
     var runningChecksEnabled = false;
 
diff --git a/test/ng/parseSpec.js b/test/ng/parseSpec.js
@@ -95,6 +95,37 @@ describe('parser', function() {
       expect(spaces).toEqual(noSpaces);
     });
 
+    it('should use callback functions to know when an identifier is valid', function() {
+      function getText(t) { return t.text; }
+      var isIdentifierStart = jasmine.createSpy('start');
+      var isIdentifierContinue = jasmine.createSpy('continue');
+      isIdentifierStart.and.returnValue(true);
+      var lex = new Lexer({csp: false, isIdentifierStart: isIdentifierStart, isIdentifierContinue: isIdentifierContinue});
+
+      isIdentifierContinue.and.returnValue(true);
+      var tokens = lex.lex('πΣε').map(getText);
+      expect(tokens).toEqual(['πΣε']);
+
+      isIdentifierContinue.and.returnValue(false);
+      tokens = lex.lex('πΣε').map(getText);
+      expect(tokens).toEqual(['π', 'Σ', 'ε']);
+    });
+
+    it('should send the unicode characters and code points', function() {
+      function getText(t) { return t.text; }
+      var isIdentifierStart = jasmine.createSpy('start');
+      var isIdentifierContinue = jasmine.createSpy('continue');
+      isIdentifierStart.and.returnValue(true);
+      isIdentifierContinue.and.returnValue(true);
+      var lex = new Lexer({csp: false, isIdentifierStart: isIdentifierStart, isIdentifierContinue: isIdentifierContinue});
+      var tokens = lex.lex('\uD801\uDC37\uD852\uDF62\uDBFF\uDFFF');
+      expect(isIdentifierStart).toHaveBeenCalledTimes(1);
+      expect(isIdentifierStart.calls.argsFor(0)).toEqual(['\uD801\uDC37', 0x10437]);
+      expect(isIdentifierContinue).toHaveBeenCalledTimes(2);
+      expect(isIdentifierContinue.calls.argsFor(0)).toEqual(['\uD852\uDF62', 0x24B62]);
+      expect(isIdentifierContinue.calls.argsFor(1)).toEqual(['\uDBFF\uDFFF', 0x10FFFF]);
+    });
+
     it('should tokenize undefined', function() {
       var tokens = lex("undefined");
       var i = 0;
