Added support for double quotes for a term

README.md

@@ -9,13 +9,16 @@ This function converts a boolean query to a 2 dimensional array with all possibi
 Input                         | Output
 --------                      | ---------
 `a AND b`                     | `[[a, b]]`
+`a AND b`                     | `[[a, b]]`
 `a OR  b`                     | `[[a], [b]]`
 `a AND b AND c`               | `[[a, b, c]]`
 `a AND b OR  c`               | `[[a, b], [c]]`
 `a AND (b OR c)`              | `[[a, b], [a, c]]`
 `a AND (b OR c) AND (d OR e)` | `[[a, b, d], [a, b, e], [a, c, d], [a, c, e]]`
 
-Whereas `a`, `b` and `c` represent words, forming a complex query pattern.
+Whereas `a`, `b` and `c` represent terms, forming a complex query pattern.
+
+A term can be written as a single word, or a phrase can be represented by using double quotes to wrap multiple words (e.g. term `a` could be `"foo bar"`).
 
 This function works recursively trough all brackets and generates an array of all possible combinations
 of a matching query.

index.js

@@ -24,6 +24,79 @@ function _arraysAreEqual(arrA, arrB) {
   return true;
 }
 
+function parseBooleanQuery(searchPhrase) {
+
+	searchPhrase = escapeCharactersInQuotes(searchPhrase);
+
+	var permutations = _parseBooleanQuery(searchPhrase);
+
+	permutations = unescapeCharactersInQuotes(permutations);
+
+	return permutations;
+}
+
+//var defaultSplitTerm = 'AND';
+
+function injectOperatorBetweenTerms(searchPhrase) {
+  // Default to using AND
+	useAnd = (module.exports.defaultSplitTerm == 'AND');
+
+  // Remove leading and trailing whitespace
+	searchPhrase = searchPhrase.trim();
+
+	if(useAnd){
+		// replace all spaces with ' AND ', then remove any extra ANDs
+    searchPhrase = searchPhrase.replace(/ /gi, ' AND ');
+    searchPhrase = searchPhrase.replace(/ AND AND AND /gi, ' AND ');
+    searchPhrase = searchPhrase.replace(/ AND OR AND /gi, ' OR ');
+    searchPhrase = searchPhrase.replace(/\( AND /gi, '(');
+    searchPhrase = searchPhrase.replace(/ AND \)/gi, ')');
+	} else {
+		// replace all spaces with ' OR ', then remove any extra ORs
+    searchPhrase = searchPhrase.replace(/ /gi, ' OR ');
+    searchPhrase = searchPhrase.replace(/ OR AND OR /gi, ' AND ');
+    searchPhrase = searchPhrase.replace(/ OR OR OR /gi, ' OR ');
+    searchPhrase = searchPhrase.replace(/\( OR /gi, '(');
+    searchPhrase = searchPhrase.replace(/ OR \)/gi, ')');
+	}
+	return searchPhrase;
+}
+
+function escapeCharactersInQuotes(searchPhrase){
+  searchPhrase = searchPhrase.replace(/(".+?")/g, function(match, group1, offset, input_string) {
+     // remove spaces
+     var encoded = encodeURI(group1.trim());
+     // remove parenthesis
+     encoded = encoded.replace(/\(/g, '(');
+     encoded = encoded.replace(/\)/g, ')');
+     return encoded;
+  });
+	return searchPhrase;
+}
+
+function unescapeCharactersInQuotes(permutations){
+  var decodedPermutations = [];
+
+  permutations.forEach(function(element){
+      var decodedElement = [];
+      element.forEach(function(term){
+          // restore parenthesis that may have been encoded
+          var decoded = term.replace(/(/g, '(');
+          decoded = decoded.replace(/)/g, ')');
+          // restore spaces that may have been encoded
+          decoded = decodeURI(decoded);
+
+					// strip off quotes 
+  			  decoded = decoded.replace(/^"(.*)"$/, function(match, group1, offset, original){ 
+						return group1; 
+					});
+          decodedElement.push(decoded);
+      });
+      decodedPermutations.push(decodedElement);
+  });
+	return decodedPermutations;
+}
+
 // This function converts a boolean query to a 2 dimensional array.
 // a AND (b OR c)
 // Becomes:
@@ -34,14 +107,17 @@ function _arraysAreEqual(arrA, arrB) {
 // There are more efficient ways to match content to this query, though this is
 // the one that is most easy to maintain and limits risk of side-effects.
 // Especially when considering recursively nested queries.
-function parseBooleanQuery(searchPhrase) {
+function _parseBooleanQuery(searchPhrase) {
 
   // Remove outer brackets if they exist. EX: (a OR b) -> a OR b
   searchPhrase = removeOuterBrackets(searchPhrase);
 
   // remove double whitespaces
   searchPhrase = removeDoubleWhiteSpace(searchPhrase);
 
+	// Put ANDs inbetween all the terms that only have a space betwee them
+	searchPhrase = injectOperatorBetweenTerms(searchPhrase);
+
   // Split the phrase on the term 'OR', but don't do this on 'OR' that's in
   // between brackets. EX: a OR (b OR c) should not parse the `OR` in between b
   // and c.
@@ -68,7 +144,7 @@ function parseBooleanQuery(searchPhrase) {
       // If the string contains brackets, parse it recursively, and add it to
       // `nestedPaths`.
       if (containsBrackets(ands[i])) {
-        nestedPaths.push(parseBooleanQuery(ands[i]));
+        nestedPaths.push(_parseBooleanQuery(ands[i]));
       }
 
       // If it doesn't. Push the word to `andPath`.
@@ -302,5 +378,9 @@ module.exports = {
   removeDoubleWhiteSpace: removeDoubleWhiteSpace,
   removeOuterBrackets: removeOuterBrackets,
   parseBooleanQuery: parseBooleanQuery,
-  containsBrackets: containsBrackets
+  containsBrackets: containsBrackets,
+  escapeCharactersInQuotes: escapeCharactersInQuotes,
+  unescapeCharactersInQuotes: unescapeCharactersInQuotes,
+  injectOperatorBetweenTerms: injectOperatorBetweenTerms,
+	defaultSplitTerm: 'AND'
 };

test/test.js

@@ -89,7 +89,65 @@ describe('String functions', function() {
       assert.equal('a b c', bparser.removeDoubleWhiteSpace("a\nb\tc"));
     });
   });
+
+	describe('escapeCharactersInQuotes', function() {
+    it('Should modify inside quotes to have no spaces', function() {
+      assert.equal('a %22b%20c%22', bparser.escapeCharactersInQuotes('a "b c"'));
+    });
+
+    it('Should modify inside quotes to have no parenthesis', function() {
+      assert.equal('a %22(b-c)%22', bparser.escapeCharactersInQuotes('a "(b-c)"'));
+    });
+
+    it('Should modify multiple sets of quotes', function() {
+      assert.equal('a %22b%20c%22 d %22e%20f%22 g', bparser.escapeCharactersInQuotes('a "b c" d "e f" g'));
+    });
+
+		it('Should ignore dangling quotes', function() {
+      assert.equal('a "b c', bparser.escapeCharactersInQuotes('a "b c'));
+    });
+  });
+
+	describe('unescapeCharactersInQuotes', function() {
+    it('Should restore spaces in terms', function() {
+      assert.deepEqual([['a'],['b c']], bparser.unescapeCharactersInQuotes([['a'],['%22b%20c%22']]));
+    });
+
+    it('Should restore parenthesis in terms', function() {
+      assert.deepEqual([['a'],['(b-c)']], bparser.unescapeCharactersInQuotes([['a'],['%22(b-c)%22']]));
+    });
+
+    it('Should restore multiple sets of quotes', function() {
+      assert.deepEqual([['a'],['b c'],['d'],['e f'],['g']], bparser.unescapeCharactersInQuotes([['a'],['%22b%20c%22'],['d'],['%22e%20f%22'],['g']]));
+    });
+
+		it('Should ignore dangling quotes', function() {
+      assert.deepEqual([['a'],['"b'],['c']], bparser.unescapeCharactersInQuotes([['a'],['"b'],['c']]));
+    });
+  });
+
+  describe('injectOperatorBetweenTerms()', function() {
+    it('should add in additional ANDs to the searchPhrase by default', function() {
+      assert.equal('a AND b', bparser.injectOperatorBetweenTerms('a b'));
+      assert.equal('a AND b', bparser.injectOperatorBetweenTerms('a AND b'));
+      assert.equal('a OR b', bparser.injectOperatorBetweenTerms('a OR b'));
+      assert.equal('((a AND (b OR c)) AND (d AND e) AND (f OR g OR h)) OR i OR j', bparser.injectOperatorBetweenTerms(' ( ( a AND ( b OR c ) ) AND ( d AND e ) AND ( f OR g OR h ) ) OR i OR j '));
+      assert.equal('((a AND (b OR c)) AND (d AND e) AND (f OR g OR h)) OR i OR j', bparser.injectOperatorBetweenTerms('((a ( b OR c)) (d e) (f OR g OR h)) OR i OR j'));
+    });
 
+    it('should add in ORs to the searchPhrase, if specified', function() {
+			// Save off the old split term and override it to 'OR'
+			var oldSplitTerm = bparser.defaultSplitTerm;
+			bparser.defaultSplitTerm = 'OR';
+      assert.equal('a OR b', bparser.injectOperatorBetweenTerms('a b'));
+      assert.equal('a AND b', bparser.injectOperatorBetweenTerms('a AND b'));
+      assert.equal('a OR b', bparser.injectOperatorBetweenTerms('a OR b'));
+      assert.equal('((a AND (b OR c)) AND (d AND e) AND (f OR g OR h)) OR i OR j', bparser.injectOperatorBetweenTerms(' ( ( a AND ( b OR c ) ) AND ( d AND e ) AND ( f OR g OR h ) ) OR i OR j '));
+      assert.equal('((a AND (b OR c)) AND (d AND e) AND (f OR g OR h)) OR i OR j', bparser.injectOperatorBetweenTerms('((a AND (b c)) AND (d AND e) AND (f g h)) i j'));
+			// Restore the old split term
+			bparser.defaultSplitTerm = oldSplitTerm;
+    });
+  });
 });
 
 describe('query merging functions', function() {
@@ -177,6 +235,11 @@ describe('query merging functions', function() {
 });
 
 describe('parse function', function() {
+  it('Should parse a simple query without an operator', function() {
+    assert.deepEqual([['a', 'b']], bparser.parseBooleanQuery('a b'));
+    assert.deepEqual([['a', 'b','c']], bparser.parseBooleanQuery('a AND b c'));
+    assert.deepEqual([['a','b c']], bparser.parseBooleanQuery('a "b c"'));
+  });
   it('Should parse a simple query without any brackets', function() {
     assert.deepEqual([['a', 'b']], bparser.parseBooleanQuery('a AND b'));
     assert.deepEqual([['a'], ['b']], bparser.parseBooleanQuery('a OR b'));
@@ -189,6 +252,12 @@ describe('parse function', function() {
   it('Should parse a simple query a single depth of brackets', function() {
     assert.deepEqual([['a', 'c'], ['b', 'c']], bparser.parseBooleanQuery('(a OR b) AND c'));
   });
+  it('Should parse a simple query a query with quoted terms', function() {
+    assert.deepEqual([['a', 'c'], ['b', 'c']], bparser.parseBooleanQuery('("a" OR b) AND c'));
+  });
+  it('Should parse a more complex query a query with quoted terms', function() {
+    assert.deepEqual([['a b', 'e f'], ['c', 'e f']], bparser.parseBooleanQuery('("a b" OR c) AND "e f"'));
+  });
 
   // This resolves to issue #3 on github
   it('Should parse a query, where the final bracket is not related to the first bracket', function() {
@@ -220,5 +289,19 @@ describe('parse function', function() {
       recursiveSort(bparser.parseBooleanQuery(searchPhrase))
     );
   });
+  it('..long shot with quotes', function(){
+    var searchPhrase = '(("a " AND ("(b" OR "c)")) AND ("d AND" AND "e OR") AND ("(f)" OR g OR h)) OR i OR j';
+    assert.deepEqual(
+      recursiveSort(
+      [['a ','(b','d AND','e OR','(f)'],
+       ['a ','(b','d AND','e OR','g'],
+       ['a ','(b','d AND','e OR','h'],
+       ['a ','c)','d AND','e OR','(f)'],
+       ['a ','c)','d AND','e OR','g'],
+       ['a ','c)','d AND','e OR','h'],
+       ['i'],['j']]),
+      recursiveSort(bparser.parseBooleanQuery(searchPhrase))
+    );
+  });
 
 });