Inverse bracket expression is broken

[nihei9]

I made the following file:

{
	"name": "test",
	"entries": [
		{
			"kind": "not_paren",
			"pattern": "[^()]+"
		}
	]
}

Then I got the following result:

$ echo -n '()' | maleeni lex test.json
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"(","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_paren","row":0,"col":1,"lexeme":")","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"","eof":true,"invalid":false}

Even though the inverse expression contains ), maleeni v0.6.0 recognizes ) as a valid token. The expectation is that ) will be a part of an invalid token.

[nihei9]

Possibly this issue occurs when characters to be excluded are adjacent to each other.

Example of characters to be adjacent to each other

Lexical specification:

{
	"name": "test",
	"entries": [
		{
			"kind": "not_digit",
			"pattern": "[^0123456789]+"
		}
	]
}

Result:

$ echo -n '0123456789' | maleeni lex test.json
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"0","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_digit","row":0,"col":1,"lexeme":"1","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":2,"lexeme":"2","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_digit","row":0,"col":3,"lexeme":"3","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":4,"lexeme":"4","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_digit","row":0,"col":5,"lexeme":"5","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":6,"lexeme":"6","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_digit","row":0,"col":7,"lexeme":"7","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":8,"lexeme":"8","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_digit","row":0,"col":9,"lexeme":"9","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"","eof":true,"invalid":false}

Example of characters not to be adjacent to each other

Lexical specification:

{
	"name": "test",
	"entries": [
		{
			"kind": "not_odd",
			"pattern": "[^13579]+"
		}
	]
}

Result:

$ echo -n '13579' | maleeni lex test.json
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"13579","eof":false,"invalid":true}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"","eof":true,"invalid":false}

$ echo -n '02468' | maleeni lex test.json
{"mode_id":1,"mode_name":"default","kind_id":1,"mode_kind_id":1,"kind_name":"not_odd","row":0,"col":0,"lexeme":"02468","eof":false,"invalid":false}
{"mode_id":1,"mode_name":"default","kind_id":0,"mode_kind_id":0,"kind_name":"","row":0,"col":0,"lexeme":"","eof":true,"invalid":false}