feat: add & and | logical operators

__tests__/tokenizer_test.ts

@@ -19,87 +19,130 @@ describe('Tokenizer', () => {
   })
 
   describe('Groups', () => {
-    it('Should tokenize single-word group', () => {
+    test('Should tokenize single-word group', () => {
       const reader = new StringReader('(word)')
       const tokenizer = new Tokenizer(reader)
+      const tokens = tokenizer.read()
 
-      let token = tokenizer.consume()
-      expect(token.token).toBe(Token.group)
-      expect(token.value).toBe('(')
+      expect(tokens[0].token).toBe(Token.group)
+      expect(tokens[0].value).toBe('(')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.word)
-      expect(token.value).toBe('word')
+      expect(tokens[1].token).toBe(Token.word)
+      expect(tokens[1].value).toBe('word')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.group)
-      expect(token.value).toBe(')')
+      expect(tokens[2].token).toBe(Token.group)
+      expect(tokens[2].value).toBe(')')
     })
 
-    it('Should tokenize logical operator OR group', () => {
-      const reader = new StringReader('(word OR word)')
-      const tokenizer = new Tokenizer(reader)
+    describe('logical operator OR group', () => {
+      test('should parse OR separator', () => {
+        const reader = new StringReader('(word OR word)')
+        const tokenizer = new Tokenizer(reader)
+        const tokens = tokenizer.read()
 
-      let token = tokenizer.consume()
-      expect(token.token).toBe(Token.group)
-      expect(token.value).toBe('(')
+        expect(tokens[0].token).toBe(Token.group)
+        expect(tokens[0].value).toBe('(')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.word)
-      expect(token.value).toBe('word')
+        expect(tokens[1].token).toBe(Token.word)
+        expect(tokens[1].value).toBe('word')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.whitespace)
-      expect(token.value).toBe(' ')
+        expect(tokens[2].token).toBe(Token.whitespace)
+        expect(tokens[2].value).toBe(' ')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.operator)
-      expect(token.value).toBe('or')
+        expect(tokens[3].token).toBe(Token.operator)
+        expect(tokens[3].value).toBe('or')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.whitespace)
-      expect(token.value).toBe(' ')
+        expect(tokens[4].token).toBe(Token.whitespace)
+        expect(tokens[4].value).toBe(' ')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.word)
-      expect(token.value).toBe('word')
+        expect(tokens[5].token).toBe(Token.word)
+        expect(tokens[5].value).toBe('word')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.group)
-      expect(token.value).toBe(')')
+        expect(tokens[6].token).toBe(Token.group)
+        expect(tokens[6].value).toBe(')')
+      })
+
+      test('should parse | separator', () => {
+        const reader = new StringReader('(word | word)')
+        const tokenizer = new Tokenizer(reader)
+        const tokens = tokenizer.read()
+
+        expect(tokens[0].token).toBe(Token.group)
+        expect(tokens[0].value).toBe('(')
+
+        expect(tokens[1].token).toBe(Token.word)
+        expect(tokens[1].value).toBe('word')
+
+        expect(tokens[2].token).toBe(Token.whitespace)
+        expect(tokens[2].value).toBe(' ')
+
+        expect(tokens[3].token).toBe(Token.operator)
+        expect(tokens[3].value).toBe('|')
+
+        expect(tokens[4].token).toBe(Token.whitespace)
+        expect(tokens[4].value).toBe(' ')
+
+        expect(tokens[5].token).toBe(Token.word)
+        expect(tokens[5].value).toBe('word')
+
+        expect(tokens[6].token).toBe(Token.group)
+        expect(tokens[6].value).toBe(')')
+      })
     })
 
-    it('Should tokenize logical operator AND group', () => {
-      const reader = new StringReader('(word AND word)')
-      const tokenizer = new Tokenizer(reader)
+    describe('logical operator AND group', () => {
+      test('should parse AND separator', () => {
+        const reader = new StringReader('(word AND word)')
+        const tokenizer = new Tokenizer(reader)
+        const tokens = tokenizer.read()
 
-      let token = tokenizer.consume()
-      expect(token.token).toBe(Token.group)
-      expect(token.value).toBe('(')
+        expect(tokens[0].token).toBe(Token.group)
+        expect(tokens[0].value).toBe('(')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.word)
-      expect(token.value).toBe('word')
+        expect(tokens[1].token).toBe(Token.word)
+        expect(tokens[1].value).toBe('word')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.whitespace)
-      expect(token.value).toBe(' ')
+        expect(tokens[2].token).toBe(Token.whitespace)
+        expect(tokens[2].value).toBe(' ')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.operator)
-      expect(token.value).toBe('and')
+        expect(tokens[3].token).toBe(Token.operator)
+        expect(tokens[3].value).toBe('and')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.whitespace)
-      expect(token.value).toBe(' ')
+        expect(tokens[4].token).toBe(Token.whitespace)
+        expect(tokens[4].value).toBe(' ')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.word)
-      expect(token.value).toBe('word')
+        expect(tokens[5].token).toBe(Token.word)
+        expect(tokens[5].value).toBe('word')
 
-      token = tokenizer.consume()
-      expect(token.token).toBe(Token.group)
-      expect(token.value).toBe(')')
+        expect(tokens[6].token).toBe(Token.group)
+        expect(tokens[6].value).toBe(')')
+      })
+      test('should parse & separator', () => {
+        const reader = new StringReader('(word & word)')
+        const tokenizer = new Tokenizer(reader)
+        const tokens = tokenizer.read()
+
+        expect(tokens[0].token).toBe(Token.group)
+        expect(tokens[0].value).toBe('(')
+
+        expect(tokens[1].token).toBe(Token.word)
+        expect(tokens[1].value).toBe('word')
+
+        expect(tokens[2].token).toBe(Token.whitespace)
+        expect(tokens[2].value).toBe(' ')
+
+        expect(tokens[3].token).toBe(Token.operator)
+        expect(tokens[3].value).toBe('&')
+
+        expect(tokens[4].token).toBe(Token.whitespace)
+        expect(tokens[4].value).toBe(' ')
+
+        expect(tokens[5].token).toBe(Token.word)
+        expect(tokens[5].value).toBe('word')
+
+        expect(tokens[6].token).toBe(Token.group)
+        expect(tokens[6].value).toBe(')')
+      })
     })
   })
 })

src/tokenizer.ts

@@ -56,19 +56,23 @@ export class Tokenizer implements InputReader<TokenValue> {
         }
 
         if (this.isAlphanumeric(nextChar.charCodeAt(0))) {
-          if (nextChar.toUpperCase() === 'O' && this.reader.peek(1) === 'R') {
+          if (this.confirmExactWord('OR')) {
             return this.consumeOr()
           }
-          if (
-            nextChar.toUpperCase() === 'A' &&
-            this.reader.peek(1) === 'N' &&
-            this.reader.peek(2) === 'D'
-          ) {
+          if (this.confirmExactWord('AND')) {
             return this.consumeAnd()
           }
           return this.consumeWord()
         }
 
+        if (nextChar === '|') {
+          return this.consumeOr()
+        }
+
+        if (nextChar === '&') {
+          return this.consumeAnd()
+        }
+
         if (nextChar === '(' || nextChar === ')') {
           // this.state = TokenizerState.inGroup
           return this.consumeGroup()
@@ -84,24 +88,62 @@ export class Tokenizer implements InputReader<TokenValue> {
         throw new Error('bad state')
     }
   }
+
   consumeAnd(): TokenValue {
-    this.reader.consume()
-    this.reader.consume()
-    this.reader.consume()
+    let value = ''
+    if (this.confirmExactWord('AND')) {
+      this.consumeExactWord('AND')
+      value = 'and'
+    } else if (this.confirmExactWord('&')) {
+      this.consumeExactWord('&')
+      value = '&'
+    }
     return {
-      value: 'and',
+      value,
       token: Token.operator,
     }
   }
+
   consumeOr(): TokenValue {
-    this.reader.consume()
-    this.reader.consume()
+    let value = ''
+    if (this.confirmExactWord('OR')) {
+      this.consumeExactWord('OR')
+      value = 'or'
+    } else if (this.confirmExactWord('|')) {
+      this.consumeExactWord('|')
+      value = '|'
+    }
     return {
-      value: 'or',
+      value,
       token: Token.operator,
     }
   }
 
+  confirmExactWord(word: string) {
+    let nextChar = this.reader.peek()
+    for (let i = 0; i < word.length; i++) {
+      if (nextChar !== word[i]) {
+        return false
+      }
+      nextChar = this.reader.peek(i + 1)
+    }
+    return true
+  }
+
+  consumeExactWord(word: string) {
+    if (this.confirmExactWord(word)) {
+      this.consumeReader(word.length)
+    } else {
+      throw new Error("Can't find exact word: " + word)
+    }
+  }
+
+  consumeReader(times = 1) {
+    for (let i = 0; i < times; i++) {
+      this.reader.consume()
+    }
+  }
+
   private consumeGroup(): TokenValue {
     return {
       value: this.reader.consume(),
@@ -166,10 +208,6 @@ export class Tokenizer implements InputReader<TokenValue> {
     return tokens
   }
 
-  // public read() {
-  //   throw new Error('Method not implemented.')
-  // }
-
   private isWhitespace(nextChar: string) {
     return ' \t\n\r'.includes(nextChar)
   }