fix: minor bugs && example

lexer/lexer.go

@@ -1,7 +1,7 @@
 package lexer
 
 import (
-	//"fmt"
+	"fmt"
 	"regexp"
 )
 
@@ -23,7 +23,9 @@ func createLexer(source string) *lexer {
 		source:   source,
 		Tokens:   make([]Token, 0),
 		patterns: []regexPattern{
-			{regexp.MustCompile(`\s+`), defaultHandler(SPACE, " ")},
+			{regexp.MustCompile(`\s+`), skipHandler},
+			{regexp.MustCompile(`\+`), defaultHandler(PLUS, "+")},
+			{regexp.MustCompile(`\-`), defaultHandler(MINUS, "-")},
 			{regexp.MustCompile(`\:`), defaultHandler(COLON, ":")},
 			{regexp.MustCompile(`>=`), defaultHandler(MORE_EQUAL, ">=")},
 			{regexp.MustCompile(`>`), defaultHandler(MORE, ">")},
@@ -111,3 +113,27 @@ func commentHandler(lex *lexer, regex *regexp.Regexp) {
 		lex.incrementPosition(match[1])
 	}
 }
+
+func Tokenize(source string) []Token {
+	lex := createLexer(source)
+
+	for !lex.atEof() {
+		matched := false
+
+		for _, pattern := range lex.patterns {
+			location := pattern.regex.FindStringIndex(lex.currentString())
+			if location != nil && location[0] == 0 {
+				pattern.handler(lex, pattern.regex)
+				matched = true
+				break
+			}
+		}
+
+		if !matched {
+			panic(fmt.Sprintf("lexer error: unrecognized token near '%v'", lex.currentString()))
+		}
+	}
+
+	lex.push(Token{EOF, "EOF"})
+	return lex.Tokens
+}

lexer/tokens.go

@@ -1,5 +1,7 @@
 package lexer
 
+import "fmt"
+
 // что есть в запросе?
 // строковые литералы
 // двоеточия
@@ -17,7 +19,8 @@ package lexer
 type TokenType int
 
 const (
-	SYMBOL TokenType = iota
+	EOF TokenType = iota
+	SYMBOL
 	COLON
 	EXCLAMATION
 	EQUAL
@@ -31,6 +34,8 @@ const (
 	LESS_EQUAL
 	OPEN_BRACE
 	CLOSED_BRACE
+	MINUS
+	PLUS
 	COMMENT
 	PIPE
 	NUMBER
@@ -44,6 +49,73 @@ type Token struct {
 	value     string
 }
 
-func Parse(str string) {
-
+func TokenKindString(kind TokenType) string {
+	switch kind {
+	case EOF:
+		return "eof"
+	case SYMBOL:
+		return "symbol"
+	case NUMBER:
+		return "number"
+	case STRING_LITERAL:
+		return "string"
+	case FLOAT_NUMBER:
+		return "float"
+	case OPEN_BRACE:
+		return "open_paren"
+	case CLOSED_BRACE:
+		return "close_paren"
+	case EQUAL:
+		return "equals"
+	case NOT_EQUAL:
+		return "not_equals"
+	case NOT:
+		return "not"
+	case OR:
+		return "or"
+	case AND:
+		return "and"
+	case COLON:
+		return "colon"
+	case EXCLAMATION:
+		return "exclamation"
+	case MORE:
+		return "more"
+	case LESS:
+		return "less"
+	case MORE_EQUAL:
+		return "more_equal"
+	case LESS_EQUAL:
+		return "less_equal"
+	case COMMENT:
+		return "comment"
+	case PIPE:
+		return "pipe"
+	case MINUS:
+		return "minus"
+	case PLUS:
+		return "plus"
+	case SPACE:
+		return "space"
+	default:
+		return fmt.Sprintf("unknown(%d)", kind)
+	}
+}
+
+func (tk Token) IsOneOfMany(expectedTokens ...TokenType) bool {
+	for _, expected := range expectedTokens {
+		if expected == tk.tokenType {
+			return true
+		}
+	}
+
+	return false
+}
+
+func Debug(token Token) {
+	if token.tokenType == SYMBOL || token.tokenType == NUMBER || token.tokenType == STRING_LITERAL {
+		fmt.Printf("%s(%s)\n", TokenKindString(token.tokenType), token.value)
+	} else {
+		fmt.Printf("%s()\n", TokenKindString(token.tokenType))
+	}
 }

main.go

@@ -1,5 +1,10 @@
 package main
 
-func main() {
+import "github.com/e1lama/spl/lexer"
 
+func main() {
+	tokens := lexer.Tokenize(`index=security (action=failed OR action=denied or action=accepted) host!="localhost" | stats count by dest_ip | sort -count`)
+	for _, token := range tokens {
+		lexer.Debug(token)
+	}
 }