Unquoted identifiers naming rule is incorrect

[jayzhan211]

The naming rule is in https://dev.mysql.com/doc/refman/8.0/en/identifiers.html

The rule for unquoted identifiers is incorrect.

• ASCII: [0-9,a-z,A-Z$_] (basic Latin letters, digits 0-9, dollar, underscore)
• Extended: U+0080 .. U+FFFF

In the following test, we can see that @ is unpermitted

sqlparser-rs/tests/sqlparser_mysql.rs

Line 1661 in 4472789

let sql = "SELECT 123col_$@123abc FROM \"table\"";

I think we need to fix the unquoted identifiers for MYSQL and HIVE

sqlparser-rs/src/tokenizer.rs

Lines 803 to 811 in 4472789

	if dialect_of!(self is MySqlDialect | HiveDialect) && exponent_part.is_empty() {
	let word =
	peeking_take_while(chars, |ch| self.dialect.is_identifier_part(ch));
	if !word.is_empty() {
	s += word.as_str();
	return Ok(Some(Token::make_word(s.as_str(), None)));
	}
	}

MYSQL

sqlparser-rs/src/dialect/mysql.rs

Lines 28 to 41 in 4472789

	fn is_identifier_start(&self, ch: char) -> bool {
	// See https://dev.mysql.com/doc/refman/8.0/en/identifiers.html.
	// Identifiers which begin with a digit are recognized while tokenizing numbers,
	// so they can be distinguished from exponent numeric literals.
	ch.is_alphabetic()
	|| ch == '_'
	|| ch == '$'
	|| ch == '@'
	|| ('\u{0080}'..='\u{ffff}').contains(&ch)
	}
	fn is_identifier_part(&self, ch: char) -> bool {
	self.is_identifier_start(ch) || ch.is_ascii_digit()
	}

HIVE

sqlparser-rs/src/dialect/hive.rs

Lines 28 to 36 in 4472789

	fn is_identifier_part(&self, ch: char) -> bool {
	ch.is_ascii_lowercase()
	|| ch.is_ascii_uppercase()
	|| ch.is_ascii_digit()
	|| ch == '_'
	|| ch == '$'
	|| ch == '{'
	|| ch == '}'
	}