mail-chars 0.2.1

// ctext    = VCHAR - '(', ')',  '\\'
// dtext    = VCHAR - '[', '\\', ']'
// atext    = VCHAR - specials
// specials = '(' | ')' | '<' | '>' | '[' | ']' | ':' | ';' | '@' | '\\'| ',' | '.' | '"'

// ctl          = (!VCHAR) -  ' '
// ftext    = VCHAR - ':'
// WS       = ' ', '\t'
// Not Implemented as token_rfc2047 is too content dependent
//mod rfc2047 {
//    // token    = US-ASCII-CHAR - ' ', CTL, especial
//    // especial = '(' | ')' | '<' | '>' | '@' | ',' | ';' | ':' | '"' | '/'| '[' | ']' | '?' | '.' | '='
//    // encoded-text = VCHAR - '?', ' '
//    //                  in comment - '(', ')', '"'
//    //                  in phrase === ALPHA/DIGIT/ "!", "*", "+", "-", "/", "=", and "_"
//    //                  in ...
//
//}

/// `ctext` (rfc5322)
pub const CT: u8 = 0b00000001;
/// `dtext` (rfc5322)
pub const DT: u8 = 0b00000010;
/// `atext` (rfc5322)
pub const AT: u8 = 0b00000100;
/// `qtext` + `ws` (rfc5322) (or the content of quoted strings without
/// quoted-pairs)
pub const QC: u8 = 0b00001000;
/// Restricted-name-char (rfc6838)  (which is like a restricted token hence
/// RT).
pub const RT: u8 = 0b00010000;
/// `token` (rfc5322)
pub const TO: u8 = 0b00100000;
/// obs-NO-WS-CTL (rfc5322)
pub const NC: u8 = 0b01000000;
/// `token` (rfc7230) HTTP/1.1 media type token is only _mostly_ the same so it
/// needs it's own lookup. While it does not fit perfectly into the crate there
/// is one slot left (and while there are other things they tend to require
/// more than one slot)
pub const HT: u8 = 0b10000000;

pub static US_ASCII_LOOKUP: &[u8] = &[
    //0x00
    //0/8                    1/9                   2/A                   3/B                   4/C                   5/D                   6/E                      7/F
    0,                       NC,                   NC,                   NC,                   NC,                   NC,                   NC,                      NC,
    NC,                      CT|DT|QC,             0,                    NC,                   NC,                   0,                    NC,                      NC,
    //0x10
    NC,                      NC,                   NC,                   NC,                   NC,                   NC,                   NC,                      NC,
    NC,                      NC,                   NC,                   NC,                   NC,                   NC,                   NC,                      NC,
    //0x20
    CT|DT|QC,                CT|DT|AT|QC|RT|TO|HT, CT|DT,                CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|TO|HT,    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|TO|HT,
    DT|QC,                   DT|QC,                CT|DT|AT|QC|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|QC,             CT|DT|AT|QC|RT|TO|HT, CT|DT|QC|RT|TO|HT,       CT|DT|AT|QC,
    //0x30
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|QC,             CT|DT|QC,             CT|DT|QC,             CT|DT|AT|QC,          CT|DT|QC,                CT|DT|AT|QC,
    //0x40
    CT|DT|QC,                CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    //0x50
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|QC,                0,                    CT|QC,                CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    //0x60
    CT|DT|AT|QC|TO|HT,       CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    //0x70
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT,
    CT|DT|AT|QC|RT|TO|HT,    CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|RT|TO|HT, CT|DT|AT|QC|TO,       CT|DT|AT|QC|TO|HT,    CT|DT|AT|QC|TO,       CT|DT|AT|QC|TO|HT,       NC,
    //0x80
];

#[cfg(test)]
mod test {
    use super::*;

    fn is_qtext(ch: char) -> bool {
        match ch {
            //not ' ' [d:32]
            '!' |
            //not '"' [d:34]
            '#'...'[' |
            //not '\\' [d:92]
            ']'...'~' => true,
            _ => false
        }
    }

    fn is_tspecial(ch: char) -> bool {
        match ch {
            '(' | ')' | '<' | '>'  | '@' |
            ',' | ';' | ':' | '\\' | '"' |
            '/' | '[' | ']' | '?'  | '=' => true,
            _ => false
        }
    }

    fn is_restricted_char(ch: char) -> bool {
        match ch {
            '0'...'9' | 'A'...'Z' |
            'a'...'z' | '!' | '#' |
            '$' | '&' | '-' | '^' |
            '_' | '+' | '.'  => true,
            _ => false
        }
    }

    fn is_token(ch: char) -> bool {
        let bch = ch as u32;
        0x20 < bch && bch < 0x7f && !is_tspecial(ch)
    }

    fn is_qc(ch: char) -> bool {
        is_qtext(ch) || ch == ' ' || ch == '\t'
    }

    fn is_atext(ch: char ) -> bool {
        let bch = ch as u32;
        32 < bch && bch < 0x7f && match ch {
            '(' | ')' |
            '<' | '>' |
            '[' | ']' |
            ':' | ';' |
            '@' | '\\'|
            ',' | '.' |
            '"' => false,
            _ => true
        }
    }

    fn is_dtext_ws( ch: char ) -> bool {
        match ch as u32 {
            9 |
            32...90 |
            94...126 => true,
            _ => false
        }
    }

    fn is_ctext_ws(ch: char) -> bool {
        match ch {
            '\t' |
            ' '...'\'' |
            '*'...'[' |
            ']'...'~' => true,
            // obs-ctext
            _ => false
        }
    }

    fn is_no_ws_ctl(ch: char) -> bool {
        let bch = ch as u32;

        match bch {
            1...8 |
            11 | 12 |
            14...31 |
            127 => true,
            _ => false
        }
    }

    fn is_rfc7230_token(ch: char) -> bool {
        match ch {
            '!'  | '#' | '$' | '%' | '&' |
            '\'' | '*' | '+' | '-' | '.' |
            '^'  | '_' | '`' | '|' | '~' |
            'a'...'z' |
            'A'...'Z' |
            '0'...'9' => true,
            _ => false

        }

    }

    macro_rules! cmp {
        ($inp:expr, $res:expr, $mask:expr, $func:ident) => ({
            use std::str;
            assert!($inp < 0x80);
            let tmp = &[$inp as u8];
            let str = str::from_utf8(tmp).unwrap();
            let ch = str.chars().next().unwrap();
            let expected = $func(ch);
            assert!(($res & $mask != 0) == expected, "char 0x{:>02x} ({:?}) failed {} = {} [{:b}]",
                $inp, ch, stringify!($func), expected, $res);
        });
    }
    #[test]
    fn validate_lookup_table() {
        assert_eq!(US_ASCII_LOOKUP.len(), 128);
        for idx in 0..0x80 {
            let res = US_ASCII_LOOKUP[idx];

            cmp!(idx, res, QC, is_qc);
            cmp!(idx, res, RT, is_restricted_char);
            cmp!(idx, res, TO, is_token);
            cmp!(idx, res, AT, is_atext);
            cmp!(idx, res, DT, is_dtext_ws);
            cmp!(idx, res, CT, is_ctext_ws);
            cmp!(idx, res, HT, is_rfc7230_token);
            cmp!(idx, res, NC, is_no_ws_ctl);
        }
    }
}