1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158

use nom::IResult;
use nom::bytes::complete::{tag, take_while, take_while1};
use nom::sequence::tuple;
use nom::combinator::opt;
use nom::multi::many0;

fn is_ftext(character: u8) -> bool {
    character >= 33 && character <= 57 ||
    character >= 59 && character <= 126
}

fn is_text(character: u8) -> bool {
    character >= 1 && character <= 9 ||
    character >= 14 && character <= 127 ||
    character == 11 ||
    character == 12
}

fn unstructured_header_value(data: &[u8]) -> IResult<&[u8], &[u8]> {
    #[derive(PartialEq)]
    enum State {
        CarriageReturn,
        LineFeed,
        AnythingButWhitespace,
    }

    let mut state = State::CarriageReturn;
    for (mut idx, character) in data.iter().enumerate() {
        match state {
            State::CarriageReturn => if *character == b'\r' {
                state = State::LineFeed
            },
            State::LineFeed => if *character == b'\n' {
                state = State::AnythingButWhitespace
            } else {
                state = State::CarriageReturn
            },
            State::AnythingButWhitespace => if *character != b' ' && *character != b'\t' {
                idx -= 2;
                return Ok((&data[idx..], &data[..idx]))
            } else {
                state = State::CarriageReturn
            }
        }
    }

    if state == State::AnythingButWhitespace {
        Ok((&data[data.len() - 2..], &data[..data.len() - 2]))
    } else {
        Err(nom::Err::Incomplete(nom::Needed::Size(1)))
    }
}

fn field_name(data: &[u8]) -> IResult<&[u8], &[u8]> {
    take_while1(is_ftext)(data)
}

fn header_separator(data: &[u8]) -> IResult<&[u8], &[u8]> {
    let mut idx = 0;
    while idx < data.len() && (data[idx] == b' ' || data[idx] == b'\t') {
        idx += 1;
    }
    if idx < data.len() && data[idx] == b':' {
        idx += 1;
        Ok((&data[idx..], &data[..idx]))
    } else {
        Err(nom::Err::Failure((&data[idx..],nom::error::ErrorKind::CrLf)))
    }
}

fn optionnal_field(data: &[u8]) -> IResult<&[u8], (&[u8], &[u8])> {
    let result = tuple((field_name, header_separator, unstructured_header_value, tag("\r\n")))(data)?;
    Ok((result.0, ((result.1).0, (result.1).2)))
}
fn optionnal_field_with_separator(data: &[u8]) -> IResult<&[u8], (&[u8], &[u8], &[u8])> {
    let result = tuple((field_name, header_separator, unstructured_header_value, tag("\r\n")))(data)?;
    Ok((result.0, ((result.1).0, (result.1).1, (result.1).2)))
}

fn fields(data: &[u8]) -> IResult<&[u8], Vec<(&[u8], &[u8])>> {
    many0(optionnal_field)(data) // TODO obs-fields
}
fn fields_with_separators(data: &[u8]) -> IResult<&[u8], Vec<(&[u8], &[u8], &[u8])>> {
    many0(optionnal_field_with_separator)(data) // TODO obs-fields
}

fn body(data: &[u8]) -> IResult<&[u8], &[u8]> {
    let mut idx = 0;
    let mut line_lenght = 0;
    while idx < data.len() {
        if is_text(data[idx]) && line_lenght < 998 {
            idx += 1;
            line_lenght += 1
        } else if idx + 1 < data.len() && data[idx] == b'\r' && data[idx + 1] == b'\n' {
            idx += 2;
            line_lenght = 0;
        } else {
            return Err(nom::Err::Failure((&data[idx..],nom::error::ErrorKind::CrLf))); // todo
        }
    }
    Ok((&data[idx..], &data[..idx]))
}

/// Parse a raw email into a vec of headers and a body.
/// Keeps header values as they are in the mail (no unfolding).
pub fn parse_message(data: &[u8]) -> Result<(Vec<(&[u8], &[u8])>, Option<&[u8]>), ()> {
    if let Ok(result) = tuple((fields, opt(tuple((tag("\r\n"), body)))))(data) {
        Ok(((result.1).0, (result.1).1.map(|b| b.1)))
    } else {
        Err(())
    }
}

/// Like `parse_message()` but headers are a tuple of three values: (name, separator, value).  
///   
/// Example: `"HeaderName  : header value\r\n"` gives `("HeaderName", "  :", "header value")`
pub fn parse_message_with_separators(data: &[u8]) -> Result<(Vec<(&[u8], &[u8], &[u8])>, Option<&[u8]>), ()> {
    if let Ok(result) = tuple((fields_with_separators, opt(tuple((tag("\r\n"), body)))))(data) {
        Ok(((result.1).0, (result.1).1.map(|b| b.1)))
    } else {
        Err(())
    }
}

#[cfg(test)]
mod test {
    use super::{parse_message, header_separator};

    #[test]
    fn main_test() {
        use super::parse_message;
        let mail = include_bytes!("../mail.txt");
        
        println!("{:?}", parse_message(b"From: Once upon a time...\r\n There was a house\r\nTodo: But it has been destroyed\r\nEnd: That's the\r\n end.\r\n"));
        
        let (headers, body) = parse_message(mail).unwrap();
        for (name, value) in headers {
            println!("{}:{}", std::str::from_utf8(name).unwrap(), std::str::from_utf8(value).unwrap())
        }
        println!("{}", std::str::from_utf8(body.unwrap()).unwrap())
    }

    #[test]
    fn secondary_test() {
        let mail = b"A: X\r\nB : Y\t\r\n\tZ  \r\n\r\n C \r\nD \t E\r\n\r\n\r\n";
        
        let (headers, _body) = parse_message(mail).unwrap();
        
        assert_eq!(headers.len(), 2);
    }

    #[test]
    fn test_separator() {
        header_separator(b"   :  ").unwrap();
        header_separator(b"   :").unwrap();
        header_separator(b"   ").unwrap_err();
    }
}