working subject parsing!
This commit is contained in:
parent
790aad1ff5
commit
a443a90738
1 changed files with 122 additions and 31 deletions
153
src/main.rs
153
src/main.rs
|
@ -1,3 +1,6 @@
|
||||||
|
use std::collections::HashMap;
|
||||||
|
use std::default::Default;
|
||||||
|
|
||||||
use nom::{
|
use nom::{
|
||||||
IResult,
|
IResult,
|
||||||
branch::alt,
|
branch::alt,
|
||||||
|
@ -5,9 +8,11 @@ use nom::{
|
||||||
character::complete::alphanumeric1,
|
character::complete::alphanumeric1,
|
||||||
character::complete::crlf,
|
character::complete::crlf,
|
||||||
character::complete::space0,
|
character::complete::space0,
|
||||||
|
character::complete::space1,
|
||||||
bytes::complete::tag,
|
bytes::complete::tag,
|
||||||
bytes::complete::take_while1,
|
bytes::complete::take_while1,
|
||||||
multi::many0,
|
multi::many0,
|
||||||
|
multi::fold_many0,
|
||||||
sequence::terminated,
|
sequence::terminated,
|
||||||
sequence::tuple,
|
sequence::tuple,
|
||||||
};
|
};
|
||||||
|
@ -39,7 +44,7 @@ fn perm_crlf(input: &str) -> IResult<&str, &str> {
|
||||||
fn fws(input: &str) -> IResult<&str, &str> {
|
fn fws(input: &str) -> IResult<&str, &str> {
|
||||||
let (input, _) = opt(terminated(space0, perm_crlf))(input)?;
|
let (input, _) = opt(terminated(space0, perm_crlf))(input)?;
|
||||||
// @FIXME: not implemented obs-FWS
|
// @FIXME: not implemented obs-FWS
|
||||||
space0(input)
|
space1(input)
|
||||||
}
|
}
|
||||||
|
|
||||||
/// Sequence of visible chars with the UTF-8 extension
|
/// Sequence of visible chars with the UTF-8 extension
|
||||||
|
@ -55,35 +60,107 @@ fn vchar_seq(input: &str) -> IResult<&str, &str> {
|
||||||
}
|
}
|
||||||
|
|
||||||
//--------------- HEADERS
|
//--------------- HEADERS
|
||||||
/// Optional Fields
|
/// Header section
|
||||||
///
|
///
|
||||||
/// Fields may appear in messages that are otherwise unspecified in this
|
/// See: https://www.rfc-editor.org/rfc/rfc5322.html#section-2.2
|
||||||
/// document. They MUST conform to the syntax of an optional-field.
|
#[derive(Debug, Default)]
|
||||||
/// This is a field name, made up of the printable US-ASCII characters
|
pub struct HeaderSection<'a> {
|
||||||
/// except SP and colon, followed by a colon, followed by any text that
|
pub subject: Option<String>,
|
||||||
/// conforms to the unstructured syntax.
|
pub from: Vec<String>,
|
||||||
///
|
pub optional: HashMap<&'a str, String>,
|
||||||
/// https://www.rfc-editor.org/rfc/rfc5322.html#section-3.6.8
|
|
||||||
///
|
|
||||||
/// ```abnf
|
|
||||||
/// optional-field = field-name ":" unstructured CRLF
|
|
||||||
/// field-name = 1*ftext
|
|
||||||
/// ftext = %d33-57 / ; Printable US-ASCII
|
|
||||||
/// %d59-126 ; characters not including
|
|
||||||
/// ; ":".
|
|
||||||
/// ```
|
|
||||||
#[derive(Debug, PartialEq)]
|
|
||||||
pub struct OptionalField<'a> {
|
|
||||||
pub name: &'a str,
|
|
||||||
pub body: String,
|
|
||||||
}
|
}
|
||||||
|
|
||||||
fn optional_field(input: &str) -> IResult<&str, OptionalField> {
|
fn header_section(input: &str) -> IResult<&str, HeaderSection> {
|
||||||
let (input, name) = take_while1(|c| c >= '\x21' && c <= '\x7E' && c != '\x3A')(input)?;
|
fold_many0(
|
||||||
let (input, _) = tag(":")(input)?;
|
header_field,
|
||||||
let (input, body) = unstructured(input)?;
|
HeaderSection::default,
|
||||||
|
|mut section, head| {
|
||||||
Ok((input, OptionalField { name, body }))
|
match head {
|
||||||
|
HeaderField::Subject(title) => {
|
||||||
|
section.subject = Some(title);
|
||||||
|
}
|
||||||
|
HeaderField::Optional(name, body) => {
|
||||||
|
section.optional.insert(name, body);
|
||||||
|
}
|
||||||
|
_ => unimplemented!(),
|
||||||
|
};
|
||||||
|
section
|
||||||
|
}
|
||||||
|
)(input)
|
||||||
|
}
|
||||||
|
|
||||||
|
enum HeaderField<'a> {
|
||||||
|
// 3.6.1. The Origination Date Field
|
||||||
|
Date,
|
||||||
|
|
||||||
|
// 3.6.2. Originator Fields
|
||||||
|
From,
|
||||||
|
Sender,
|
||||||
|
ReplyTo,
|
||||||
|
|
||||||
|
// 3.6.3. Destination Address Fields
|
||||||
|
To,
|
||||||
|
Cc,
|
||||||
|
Bcc,
|
||||||
|
|
||||||
|
// 3.6.4. Identification Fields
|
||||||
|
MessageID,
|
||||||
|
InReplyTo,
|
||||||
|
References,
|
||||||
|
|
||||||
|
// 3.6.5. Informational Fields
|
||||||
|
Subject(String),
|
||||||
|
Comments(String),
|
||||||
|
Keywords(Vec<&'a str>),
|
||||||
|
|
||||||
|
// 3.6.6. Resent Fields
|
||||||
|
ResentDate,
|
||||||
|
ResentFrom,
|
||||||
|
ResentSender,
|
||||||
|
ResentTo,
|
||||||
|
ResentCc,
|
||||||
|
ResentBcc,
|
||||||
|
ResentMessageID,
|
||||||
|
|
||||||
|
// 3.6.7. Trace Fields
|
||||||
|
Trace,
|
||||||
|
|
||||||
|
// 3.6.8. Optional Fields
|
||||||
|
Optional(&'a str, String)
|
||||||
|
}
|
||||||
|
|
||||||
|
/// Extract one header field
|
||||||
|
///
|
||||||
|
/// Derived grammar inspired by RFC5322 optional-field:
|
||||||
|
///
|
||||||
|
/// ```abnf
|
||||||
|
/// field = field-name ":" unstructured CRLF
|
||||||
|
/// field-name = 1*ftext
|
||||||
|
/// ftext = %d33-57 / ; Printable US-ASCII
|
||||||
|
/// %d59-126 ; characters not including
|
||||||
|
/// ; ":".
|
||||||
|
/// ```
|
||||||
|
fn header_field(input: &str) -> IResult<&str, HeaderField> {
|
||||||
|
// Extract field name
|
||||||
|
let (input, field_name) = take_while1(|c| c >= '\x21' && c <= '\x7E' && c != '\x3A')(input)?;
|
||||||
|
let (input, _) = tuple((tag(":"), space0))(input)?;
|
||||||
|
|
||||||
|
// Extract field body
|
||||||
|
match field_name {
|
||||||
|
"Date" => unimplemented!(),
|
||||||
|
//"From" => unimplemented!(),
|
||||||
|
"Sender" => unimplemented!(),
|
||||||
|
"Subject" => {
|
||||||
|
let (input, body) = unstructured(input)?;
|
||||||
|
let (input, _) = crlf(input)?;
|
||||||
|
Ok((input, HeaderField::Subject(body)))
|
||||||
|
},
|
||||||
|
_ => {
|
||||||
|
let (input, body) = unstructured(input)?;
|
||||||
|
let (input, _) = crlf(input)?;
|
||||||
|
Ok((input, HeaderField::Optional(field_name, body)))
|
||||||
|
}
|
||||||
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
/// Unstructured header field body
|
/// Unstructured header field body
|
||||||
|
@ -92,12 +169,26 @@ fn optional_field(input: &str) -> IResult<&str, OptionalField> {
|
||||||
/// unstructured = (*([FWS] VCHAR_SEQ) *WSP) / obs-unstruct
|
/// unstructured = (*([FWS] VCHAR_SEQ) *WSP) / obs-unstruct
|
||||||
/// ```
|
/// ```
|
||||||
fn unstructured(input: &str) -> IResult<&str, String> {
|
fn unstructured(input: &str) -> IResult<&str, String> {
|
||||||
let (input, _) = many0(tuple((opt(fws), vchar_seq)))(input)?;
|
let (input, r) = many0(tuple((opt(fws), vchar_seq)))(input)?;
|
||||||
let (input, _) = space0(input)?;
|
let (input, _) = space0(input)?;
|
||||||
Ok((input, "FIX ME".to_string()))
|
|
||||||
|
// Try to optimize for the most common cases
|
||||||
|
let body = match r.as_slice() {
|
||||||
|
[(None, content)] => content.to_string(),
|
||||||
|
[(Some(ws), content)] => ws.to_string() + content,
|
||||||
|
lines => lines.iter().fold(String::with_capacity(255), |mut acc, item| {
|
||||||
|
let (may_ws, content) = item;
|
||||||
|
match may_ws {
|
||||||
|
Some(ws) => acc + ws + content,
|
||||||
|
None => acc + content,
|
||||||
|
}
|
||||||
|
}),
|
||||||
|
};
|
||||||
|
|
||||||
|
Ok((input, body))
|
||||||
}
|
}
|
||||||
|
|
||||||
fn main() {
|
fn main() {
|
||||||
let header_fields = "Subject: Hello\r\n World\r\n";
|
let header = "Subject: Hello\r\n World\r\nFrom: <quentin@deuxfleurs.fr>\r\n\r\n";
|
||||||
println!("{:?}", optional_field(header_fields));
|
println!("{:?}", header_section(header));
|
||||||
}
|
}
|
||||||
|
|
Loading…
Reference in a new issue