Deuxfleurs/eml-codec
15
1
Fork 0
Code Issues 10 Pull requests Projects Releases 2 Packages Wiki Activity

Compare commits

base: Deuxfleurs:main
Branches Tags
Deuxfleurs:main
Deuxfleurs:lifetime
Deuxfleurs:wip-cow
Deuxfleurs:wip-datetime
Deuxfleurs:wip
Deuxfleurs:0.1.1
Deuxfleurs:0.1.0
..
compare: Deuxfleurs:0.1.0
Branches Tags
Deuxfleurs:lifetime
Deuxfleurs:main
Deuxfleurs:wip-cow
Deuxfleurs:wip-datetime
Deuxfleurs:wip
Deuxfleurs:0.1.1
Deuxfleurs:0.1.0

No commits in common. "main" and "0.1.0" have entirely different histories.
main ... 0.1.0

21 changed files with 441 additions and 849 deletions
Show stats Expand all files Collapse all files

2
Cargo.lock generated
View file

@ -70,7 +70,7 @@ checksum = "e496a50fda8aacccc86d7529e2c1e0892dbd0f898a6b5645b5561b89c3210efa"
[[package]]
name = "eml-codec"
version = "0.1.2"
version = "0.1.0"
dependencies = [
"base64",
"chrono",

12
Cargo.toml
View file

@ -1,16 +1,9 @@
[package]
name = "eml-codec"
version = "0.1.2"
version = "0.1.0"
edition = "2021"
license = "GPL-3.0-or-later"
repository = "https://git.deuxfleurs.fr/Deuxfleurs/eml-codec"
description = "Email enCOder DECoder in Rust. Support Internet Message Format and MIME (RFC 822, 5322, 2045, 2046, 2047, 2048, 2049)."
documentation = "https://docs.rs/eml-codec"
readme = "README.md"
exclude = [
"doc/",
"resources/",
]
# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
[lib]
@ -30,4 +23,3 @@ encoding_rs = "0.8"
[dev-dependencies]
walkdir = "2"

2
README.md
View file

@ -18,7 +18,7 @@ Content-Type: text/plain; charset=us-ascii
This is the plain text body of the message. Note the blank line
between the header information and the body of the message."#;
let (_, email) = eml_codec::parse_message(input).unwrap();
let (_, email) = eml_codec::email(input).unwrap();
println!(
"{} just sent you an email with subject \"{}\"",
email.imf.from[0].to_string(),

4
examples/simple.rs
View file

@ -10,7 +10,7 @@ This is the plain text body of the message. Note the blank line
between the header information and the body of the message."#;
// if you are only interested in email metadata/headers
let (_, imf) = eml_codec::parse_imf(input).unwrap();
let (_, imf) = eml_codec::imf(input).unwrap();
println!(
"{} just sent you an email with subject \"{}\"",
imf.from[0].to_string(),
@ -18,7 +18,7 @@ between the header information and the body of the message."#;
);
// if you like to also parse the body/content
let (_, email) = eml_codec::parse_message(input).unwrap();
let (_, email) = eml_codec::email(input).unwrap();
println!(
"{} raw message is:\n{}",
email.imf.from[0].to_string(),

98
src/header.rs
View file

@ -1,62 +1,55 @@
use crate::text::misc_token::{unstructured, Unstructured};
use crate::text::whitespace::{foldable_line, obs_crlf};
use nom::{
branch::alt,
bytes::complete::{tag, take_while1},
bytes::complete::{tag, tag_no_case, take_while1},
character::complete::space0,
combinator::{into, recognize},
multi::many0,
combinator::map,
multi::{fold_many0},
sequence::{pair, terminated, tuple},
IResult,
};
use std::fmt;
use crate::text::misc_token::unstructured;
use crate::text::whitespace::{foldable_line, obs_crlf};
#[derive(PartialEq, Clone)]
pub struct Kv2<'a>(pub &'a [u8], pub &'a [u8]);
impl<'a> From<(&'a [u8], &'a [u8])> for Kv2<'a> {
fn from(pair: (&'a [u8], &'a [u8])) -> Self {
Self(pair.0, pair.1)
}
}
impl<'a> fmt::Debug for Kv2<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_tuple("header::Kv2")
.field(&String::from_utf8_lossy(self.0))
.field(&String::from_utf8_lossy(self.1))
.finish()
}
#[derive(Debug, PartialEq)]
pub enum CompField<'a, T> {
Known(T),
Unknown(Kv<'a>),
Bad(&'a [u8]),
}
#[derive(Debug, PartialEq, Clone)]
pub enum Field<'a> {
Good(Kv2<'a>),
Bad(&'a [u8]),
}
impl<'a> From<Kv2<'a>> for Field<'a> {
fn from(kv: Kv2<'a>) -> Self {
Self::Good(kv)
}
}
impl<'a> From<&'a [u8]> for Field<'a> {
fn from(bad: &'a [u8]) -> Self {
Self::Bad(bad)
pub struct Kv<'a>(pub &'a [u8], pub Unstructured<'a>);
pub fn header<'a, T>(
fx: impl Fn(&'a [u8]) -> IResult<&'a [u8], T> + Copy,
) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], (Vec::<T>, Vec::<Kv>, Vec<&'a [u8]>)> {
move |input| {
terminated(
fold_many0(
alt((
map(fx, CompField::Known),
map(opt_field, |(k, v)| CompField::Unknown(Kv(k, v))),
map(foldable_line, CompField::Bad),
)),
|| (Vec::<T>::new(), Vec::<Kv>::new(), Vec::<&'a [u8]>::new()),
|(mut known, mut unknown, mut bad), item| {
match item {
CompField::Known(v) => known.push(v),
CompField::Unknown(v) => unknown.push(v),
CompField::Bad(v) => bad.push(v),
};
(known, unknown, bad)
}
),
obs_crlf,
)(input)
}
}
/// Parse headers as key/values
pub fn header_kv(input: &[u8]) -> IResult<&[u8], Vec<Field>> {
terminated(
many0(alt((into(correct_field), into(foldable_line)))),
obs_crlf,
)(input)
}
pub fn field_any(input: &[u8]) -> IResult<&[u8], &[u8]> {
terminated(
take_while1(|c| (0x21..=0x7E).contains(&c) && c != 0x3A),
tuple((space0, tag(b":"), space0)),
)(input)
pub fn field_name<'a>(name: &'static [u8]) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], &'a [u8]> {
move |input| terminated(tag_no_case(name), tuple((space0, tag(b":"), space0)))(input)
}
/// Optional field
@ -68,6 +61,15 @@ pub fn field_any(input: &[u8]) -> IResult<&[u8], &[u8]> {
/// %d59-126 ; characters not including
/// ; ":".
/// ```
pub fn correct_field(input: &[u8]) -> IResult<&[u8], Kv2> {
terminated(into(pair(field_any, recognize(unstructured))), obs_crlf)(input)
pub fn opt_field(input: &[u8]) -> IResult<&[u8], (&[u8], Unstructured)> {
terminated(
pair(
terminated(
take_while1(|c| (0x21..=0x7E).contains(&c) && c != 0x3A),
tuple((space0, tag(b":"), space0)),
),
unstructured,
),
obs_crlf,
)(input)
}

127
src/imf/field.rs
View file

@ -1,14 +1,21 @@
use chrono::{DateTime, FixedOffset};
use nom::combinator::map;
use nom::{
branch::alt,
combinator::map,
sequence::{preceded, terminated},
IResult,
};
use crate::header;
use crate::header::{field_name, header};
use crate::imf::address::{address_list, mailbox_list, nullable_address_list, AddressList};
use crate::imf::datetime::section as date;
use crate::imf::identification::{msg_id, msg_list, MessageID, MessageIDList};
use crate::imf::mailbox::{mailbox, AddrSpec, MailboxList, MailboxRef};
use crate::imf::mime::{version, Version};
use crate::imf::trace::{received_log, return_path, ReceivedLog};
use crate::imf::Imf;
use crate::text::misc_token::{phrase_list, unstructured, PhraseList, Unstructured};
use crate::text::whitespace::obs_crlf;
#[derive(Debug, PartialEq)]
pub enum Field<'a> {
@ -42,34 +49,94 @@ pub enum Field<'a> {
MIMEVersion(Version),
}
impl<'a> TryFrom<&header::Field<'a>> for Field<'a> {
type Error = ();
fn try_from(f: &header::Field<'a>) -> Result<Self, Self::Error> {
let content = match f {
header::Field::Good(header::Kv2(key, value)) => {
match key.to_ascii_lowercase().as_slice() {
b"date" => map(date, Field::Date)(value),
b"from" => map(mailbox_list, Field::From)(value),
b"sender" => map(mailbox, Field::Sender)(value),
b"reply-to" => map(address_list, Field::ReplyTo)(value),
b"to" => map(address_list, Field::To)(value),
b"cc" => map(address_list, Field::Cc)(value),
b"bcc" => map(nullable_address_list, Field::Bcc)(value),
b"message-id" => map(msg_id, Field::MessageID)(value),
b"in-reply-to" => map(msg_list, Field::InReplyTo)(value),
b"references" => map(msg_list, Field::References)(value),
b"subject" => map(unstructured, Field::Subject)(value),
b"comments" => map(unstructured, Field::Comments)(value),
b"keywords" => map(phrase_list, Field::Keywords)(value),
b"return-path" => map(return_path, Field::ReturnPath)(value),
b"received" => map(received_log, Field::Received)(value),
b"mime-version" => map(version, Field::MIMEVersion)(value),
_ => return Err(()),
}
}
_ => return Err(()),
};
content.map(|(_, content)| content).or(Err(()))
pub fn field(input: &[u8]) -> IResult<&[u8], Field> {
terminated(
alt((
preceded(field_name(b"date"), map(date, Field::Date)),
preceded(field_name(b"from"), map(mailbox_list, Field::From)),
preceded(field_name(b"sender"), map(mailbox, Field::Sender)),
preceded(field_name(b"reply-to"), map(address_list, Field::ReplyTo)),
preceded(field_name(b"to"), map(address_list, Field::To)),
preceded(field_name(b"cc"), map(address_list, Field::Cc)),
preceded(field_name(b"bcc"), map(nullable_address_list, Field::Bcc)),
preceded(field_name(b"message-id"), map(msg_id, Field::MessageID)),
preceded(field_name(b"in-reply-to"), map(msg_list, Field::InReplyTo)),
preceded(field_name(b"references"), map(msg_list, Field::References)),
preceded(field_name(b"subject"), map(unstructured, Field::Subject)),
preceded(field_name(b"comments"), map(unstructured, Field::Comments)),
preceded(field_name(b"keywords"), map(phrase_list, Field::Keywords)),
preceded(
field_name(b"return-path"),
map(return_path, Field::ReturnPath),
),
preceded(field_name(b"received"), map(received_log, Field::Received)),
preceded(
field_name(b"mime-version"),
map(version, Field::MIMEVersion),
),
)),
obs_crlf,
)(input)
}
pub fn imf(input: &[u8]) -> IResult<&[u8], Imf> {
map(header(field), |(known, unknown, bad)| {
let mut imf = Imf::from_iter(known);
imf.header_ext = unknown;
imf.header_bad = bad;
imf
})(input)
}
#[cfg(test)]
mod tests {
use super::*;
use crate::imf::address::*;
use crate::imf::mailbox::*;
use crate::text::misc_token::*;
use chrono::{FixedOffset, TimeZone};
#[test]
fn test_header() {
let fullmail = b"Date: 7 Mar 2023 08:00:00 +0200
From: someone@example.com
To: someone_else@example.com
Subject: An RFC 822 formatted message
This is the plain text body of the message. Note the blank line
between the header information and the body of the message.";
assert_eq!(
imf(fullmail),
Ok((
&b"This is the plain text body of the message. Note the blank line\nbetween the header information and the body of the message."[..],
Imf {
date: Some(FixedOffset::east_opt(2 * 3600).unwrap().with_ymd_and_hms(2023, 3, 7, 8, 0, 0).unwrap()),
from: vec![MailboxRef {
name: None,
addrspec: AddrSpec {
local_part: LocalPart(vec![LocalPartToken::Word(Word::Atom(&b"someone"[..]))]),
domain: Domain::Atoms(vec![&b"example"[..], &b"com"[..]]),
}
}],
to: vec![AddressRef::Single(MailboxRef {
name: None,
addrspec: AddrSpec {
local_part: LocalPart(vec![LocalPartToken::Word(Word::Atom(&b"someone_else"[..]))]),
domain: Domain::Atoms(vec![&b"example"[..], &b"com"[..]]),
}
})],
subject: Some(Unstructured(vec![
UnstrToken::Plain(&b"An"[..]),
UnstrToken::Plain(&b"RFC"[..]),
UnstrToken::Plain(&b"822"[..]),
UnstrToken::Plain(&b"formatted"[..]),
UnstrToken::Plain(&b"message"[..]),
])),
..Imf::default()
}
)),
)
}
}

80
src/imf/mod.rs
View file

@ -1,4 +1,5 @@
/// Parse and represent IMF (Internet Message Format) headers (RFC822, RFC5322)
pub mod address;
pub mod datetime;
pub mod field;
@ -7,15 +8,13 @@ pub mod mailbox;
pub mod mime;
pub mod trace;
use nom::{combinator::map, IResult};
use crate::header;
use crate::imf::address::AddressRef;
use crate::imf::field::Field;
use crate::imf::identification::MessageID;
use crate::imf::mailbox::{AddrSpec, MailboxRef};
use crate::imf::mime::Version;
use crate::imf::trace::ReceivedLog;
use crate::header;
use crate::text::misc_token::{PhraseList, Unstructured};
use chrono::{DateTime, FixedOffset};
@ -51,6 +50,19 @@ pub struct Imf<'a> {
// MIME
pub mime_version: Option<Version>,
// Junk
pub header_ext: Vec<header::Kv<'a>>,
pub header_bad: Vec<&'a [u8]>,
}
impl<'a> Imf<'a> {
pub fn with_opt(mut self, opt: Vec<header::Kv<'a>>) -> Self {
self.header_ext = opt; self
}
pub fn with_bad(mut self, bad: Vec<&'a [u8]>) -> Self {
self.header_bad = bad; self
}
}
//@FIXME min and max limits are not enforced,
@ -80,65 +92,3 @@ impl<'a> FromIterator<Field<'a>> for Imf<'a> {
})
}
}
pub fn imf(input: &[u8]) -> IResult<&[u8], Imf> {
map(header::header_kv, |fields| {
fields
.iter()
.flat_map(Field::try_from)
.into_iter()
.collect::<Imf>()
})(input)
}
#[cfg(test)]
mod tests {
use super::*;
use crate::imf::address::*;
use crate::imf::mailbox::*;
use crate::text::misc_token::*;
use chrono::{FixedOffset, TimeZone};
#[test]
fn test_header() {
let fullmail = b"Date: 7 Mar 2023 08:00:00 +0200
From: someone@example.com
To: someone_else@example.com
Subject: An RFC 822 formatted message
This is the plain text body of the message. Note the blank line
between the header information and the body of the message.";
assert_eq!(
imf(fullmail),
Ok((
&b"This is the plain text body of the message. Note the blank line\nbetween the header information and the body of the message."[..],
Imf {
date: Some(FixedOffset::east_opt(2 * 3600).unwrap().with_ymd_and_hms(2023, 3, 7, 8, 0, 0).unwrap()),
from: vec![MailboxRef {
name: None,
addrspec: AddrSpec {
local_part: LocalPart(vec![LocalPartToken::Word(Word::Atom(&b"someone"[..]))]),
domain: Domain::Atoms(vec![&b"example"[..], &b"com"[..]]),
}
}],
to: vec![AddressRef::Single(MailboxRef {
name: None,
addrspec: AddrSpec {
local_part: LocalPart(vec![LocalPartToken::Word(Word::Atom(&b"someone_else"[..]))]),
domain: Domain::Atoms(vec![&b"example"[..], &b"com"[..]]),
}
})],
subject: Some(Unstructured(vec![
UnstrToken::Plain(&b"An"[..]),
UnstrToken::Plain(&b"RFC"[..]),
UnstrToken::Plain(&b"822"[..]),
UnstrToken::Plain(&b"formatted"[..]),
UnstrToken::Plain(&b"message"[..]),
])),
..Imf::default()
}
)),
)
}
}

19
src/lib.rs
View file

@ -15,10 +15,7 @@ pub mod header;
/// Low-level email-specific text-based representation for data
pub mod text;
/// Manipulate buffer of bytes
mod pointers;
use nom::{combinator::into, IResult};
use nom::IResult;
/// Parse a whole email including its (MIME) body
///
@ -49,17 +46,15 @@ use nom::{combinator::into, IResult};
/// This is the plain text body of the message. Note the blank line
/// between the header information and the body of the message."#;
///
/// let (_, email) = eml_codec::parse_message(input).unwrap();
/// let (_, email) = eml_codec::email(input).unwrap();
/// println!(
/// "{} raw message is:\n{}",
/// email.imf.from[0].to_string(),
/// String::from_utf8_lossy(email.child.as_text().unwrap().body),
/// );
/// ```
pub fn parse_message(input: &[u8]) -> IResult<&[u8], part::composite::Message> {
into(part::composite::message(mime::MIME::<
mime::r#type::DeductibleMessage,
>::default()))(input)
pub fn email(input: &[u8]) -> IResult<&[u8], part::composite::Message> {
part::composite::message(mime::MIME::<mime::r#type::Message>::default())(input)
}
/// Only extract the headers of the email that are part of the Internet Message Format spec
@ -92,13 +87,13 @@ pub fn parse_message(input: &[u8]) -> IResult<&[u8], part::composite::Message> {
/// This is the plain text body of the message. Note the blank line
/// between the header information and the body of the message."#;
///
/// let (_, imf) = eml_codec::parse_imf(input).unwrap();
/// let (_, imf) = eml_codec::imf(input).unwrap();
/// println!(
/// "{} just sent you an email with subject \"{}\"",
/// imf.from[0].to_string(),
/// imf.subject.unwrap().to_string(),
/// );
/// ```
pub fn parse_imf(input: &[u8]) -> IResult<&[u8], imf::Imf> {
imf::imf(input)
pub fn imf(input: &[u8]) -> IResult<&[u8], imf::Imf> {
imf::field::imf(input)
}

6
src/mime/charset.rs
View file

@ -77,12 +77,6 @@ impl<'a> From<&'a [u8]> for EmailCharset {
}
}
impl ToString for EmailCharset {
fn to_string(&self) -> String {
self.as_str().into()
}
}
impl EmailCharset {
pub fn as_str(&self) -> &'static str {
use EmailCharset::*;

67
src/mime/field.rs
View file

@ -1,10 +1,16 @@
use nom::combinator::map;
use nom::{
branch::alt,
combinator::map,
sequence::{preceded, terminated},
IResult,
};
use crate::header;
use crate::header::{field_name};
use crate::imf::identification::{msg_id, MessageID};
use crate::mime::mechanism::{mechanism, Mechanism};
use crate::mime::r#type::{naive_type, NaiveType};
use crate::text::misc_token::{unstructured, Unstructured};
use crate::text::whitespace::obs_crlf;
#[derive(Debug, PartialEq)]
pub enum Content<'a> {
@ -41,38 +47,38 @@ impl<'a> Content<'a> {
}
}
impl<'a> TryFrom<&header::Field<'a>> for Content<'a> {
type Error = ();
fn try_from(f: &header::Field<'a>) -> Result<Self, Self::Error> {
let content = match f {
header::Field::Good(header::Kv2(key, value)) => match key
.to_ascii_lowercase()
.as_slice()
{
b"content-type" => map(naive_type, Content::Type)(value),
b"content-transfer-encoding" => map(mechanism, Content::TransferEncoding)(value),
b"content-id" => map(msg_id, Content::ID)(value),
b"content-description" => map(unstructured, Content::Description)(value),
_ => return Err(()),
},
_ => return Err(()),
};
/*
pub fn to_mime<'a, T: WithDefaultType>(list: Vec<Content<'a>>) -> AnyMIMEWithDefault<'a, T> {
list.into_iter().collect::<AnyMIMEWithDefault<T>>()
}*/
//@TODO check that the full value is parsed, otherwise maybe log an error ?!
content.map(|(_, content)| content).or(Err(()))
}
pub fn content(input: &[u8]) -> IResult<&[u8], Content> {
terminated(
alt((
preceded(field_name(b"content-type"), map(naive_type, Content::Type)),
preceded(
field_name(b"content-transfer-encoding"),
map(mechanism, Content::TransferEncoding),
),
preceded(field_name(b"content-id"), map(msg_id, Content::ID)),
preceded(
field_name(b"content-description"),
map(unstructured, Content::Description),
),
)),
obs_crlf,
)(input)
}
#[cfg(test)]
mod tests {
use super::*;
use crate::header;
//use crate::mime::charset::EmailCharset;
use crate::header::{header};
use crate::mime::charset::EmailCharset;
use crate::mime::r#type::*;
use crate::text::misc_token::MIMEWord;
use crate::text::quoted::QuotedString;
/*
#[test]
fn test_content_type() {
let (rest, content) =
@ -82,15 +88,15 @@ mod tests {
if let Content::Type(nt) = content {
assert_eq!(
nt.to_type(),
AnyType::Text(Deductible::Explicit(Text {
charset: Deductible::Explicit(EmailCharset::UTF_8),
AnyType::Text(Text {
charset: EmailCharset::UTF_8,
subtype: TextSubtype::Plain,
})),
}),
);
} else {
panic!("Expected Content::Type, got {:?}", content);
}
}*/
}
#[test]
fn test_header() {
@ -110,10 +116,7 @@ This is a multipart message.
.as_bytes();
assert_eq!(
map(header::header_kv, |k| k
.iter()
.flat_map(Content::try_from)
.collect())(fullmail),
map(header(content), |(k, _, _)| k)(fullmail),
Ok((
&b"This is a multipart message.\n\n"[..],
vec![

14
src/mime/mechanism.rs
View file

@ -18,20 +18,6 @@ pub enum Mechanism<'a> {
Base64,
Other(&'a [u8]),
}
impl<'a> ToString for Mechanism<'a> {
fn to_string(&self) -> String {
use Mechanism::*;
let buf: &[u8] = match self {
_7Bit => b"7bit",
_8Bit => b"8bit",
Binary => b"binary",
QuotedPrintable => b"quoted-printable",
Base64 => b"base64",
Other(x) => x,
};
String::from_utf8_lossy(buf).to_string()
}
}
pub fn mechanism(input: &[u8]) -> IResult<&[u8], Mechanism> {
use Mechanism::*;

82
src/mime/mod.rs
View file

@ -10,34 +10,33 @@ pub mod mechanism;
/// Content-Type representation
pub mod r#type;
use std::fmt;
use std::marker::PhantomData;
use crate::header;
use crate::imf::identification::MessageID;
use crate::mime::field::Content;
use crate::mime::mechanism::Mechanism;
use crate::mime::r#type::{AnyType, NaiveType};
use crate::header;
use crate::text::misc_token::Unstructured; //Multipart, Message, Text, Binary};
#[derive(Debug, PartialEq, Clone)]
pub struct MIME<'a, T> {
pub interpreted_type: T,
pub fields: NaiveMIME<'a>,
pub interpreted: T,
pub parsed: NaiveMIME<'a>
}
impl<'a> Default for MIME<'a, r#type::DeductibleText> {
impl<'a> Default for MIME<'a, r#type::Text> {
fn default() -> Self {
Self {
interpreted_type: r#type::DeductibleText::default(),
fields: NaiveMIME::default(),
interpreted: r#type::Text::default(),
parsed: NaiveMIME::default(),
}
}
}
impl<'a> Default for MIME<'a, r#type::DeductibleMessage> {
impl<'a> Default for MIME<'a, r#type::Message> {
fn default() -> Self {
Self {
interpreted_type: r#type::DeductibleMessage::default(),
fields: NaiveMIME::default(),
interpreted: r#type::Message::default(),
parsed: NaiveMIME::default(),
}
}
}
@ -45,20 +44,10 @@ impl<'a> Default for MIME<'a, r#type::DeductibleMessage> {
#[derive(Debug, PartialEq, Clone)]
pub enum AnyMIME<'a> {
Mult(MIME<'a, r#type::Multipart>),
Msg(MIME<'a, r#type::DeductibleMessage>),
Txt(MIME<'a, r#type::DeductibleText>),
Msg(MIME<'a, r#type::Message>),
Txt(MIME<'a, r#type::Text>),
Bin(MIME<'a, r#type::Binary>),
}
impl<'a> AnyMIME<'a> {
pub fn fields(&self) -> &NaiveMIME<'a> {
match self {
Self::Mult(v) => &v.fields,
Self::Msg(v) => &v.fields,
Self::Txt(v) => &v.fields,
Self::Bin(v) => &v.fields,
}
}
}
impl<'a, T: WithDefaultType> From<AnyMIMEWithDefault<'a, T>> for AnyMIME<'a> {
fn from(a: AnyMIMEWithDefault<'a, T>) -> Self {
@ -66,32 +55,21 @@ impl<'a, T: WithDefaultType> From<AnyMIMEWithDefault<'a, T>> for AnyMIME<'a> {
}
}
#[derive(PartialEq, Default, Clone)]
#[derive(Debug, PartialEq, Default, Clone)]
pub struct NaiveMIME<'a> {
pub ctype: Option<NaiveType<'a>>,
pub transfer_encoding: Mechanism<'a>,
pub id: Option<MessageID<'a>>,
pub description: Option<Unstructured<'a>>,
pub kv: Vec<header::Field<'a>>,
pub raw: &'a [u8],
}
impl<'a> fmt::Debug for NaiveMIME<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("NaiveMime")
.field("ctype", &self.ctype)
.field("transfer_encoding", &self.transfer_encoding)
.field("id", &self.id)
.field("description", &self.description)
.field("kv", &self.kv)
.field("raw", &String::from_utf8_lossy(self.raw))
.finish()
}
pub header_ext: Vec<header::Kv<'a>>,
pub header_bad: Vec<&'a [u8]>,
}
impl<'a> FromIterator<Content<'a>> for NaiveMIME<'a> {
fn from_iter<I: IntoIterator<Item = Content<'a>>>(it: I) -> Self {
it.into_iter()
.fold(NaiveMIME::default(), |mut section, field| {
it.into_iter().fold(
NaiveMIME::default(),
|mut section, field| {
match field {
Content::Type(v) => section.ctype = Some(v),
Content::TransferEncoding(v) => section.transfer_encoding = v,
@ -99,29 +77,25 @@ impl<'a> FromIterator<Content<'a>> for NaiveMIME<'a> {
Content::Description(v) => section.description = Some(v),
};
section
})
},
)
}
}
impl<'a> NaiveMIME<'a> {
pub fn with_kv(mut self, fields: Vec<header::Field<'a>>) -> Self {
self.kv = fields;
self
pub fn with_opt(mut self, opt: Vec<header::Kv<'a>>) -> Self {
self.header_ext = opt; self
}
pub fn with_raw(mut self, raw: &'a [u8]) -> Self {
self.raw = raw;
self
pub fn with_bad(mut self, bad: Vec<&'a [u8]>) -> Self {
self.header_bad = bad; self
}
pub fn to_interpreted<T: WithDefaultType>(self) -> AnyMIME<'a> {
self.ctype
.as_ref()
.map(|c| c.to_type())
.unwrap_or(T::default_type())
.to_mime(self)
.into()
self.ctype.as_ref().map(|c| c.to_type()).unwrap_or(T::default_type()).to_mime(self).into()
}
}
pub trait WithDefaultType {
fn default_type() -> AnyType;
}
@ -129,13 +103,13 @@ pub trait WithDefaultType {
pub struct WithGenericDefault {}
impl WithDefaultType for WithGenericDefault {
fn default_type() -> AnyType {
AnyType::Text(r#type::DeductibleText::default())
AnyType::Text(r#type::Text::default())
}
}
pub struct WithDigestDefault {}
impl WithDefaultType for WithDigestDefault {
fn default_type() -> AnyType {
AnyType::Message(r#type::DeductibleMessage::default())
AnyType::Message(r#type::Message::default())
}
}

169
src/mime/type.rs
View file

@ -5,29 +5,19 @@ use nom::{
sequence::{preceded, terminated, tuple},
IResult,
};
use std::fmt;
use crate::mime::charset::EmailCharset;
use crate::mime::{AnyMIME, NaiveMIME, MIME};
use crate::text::misc_token::{mime_word, MIMEWord};
use crate::text::words::mime_atom;
use crate::mime::{AnyMIME, MIME, NaiveMIME};
// --------- NAIVE TYPE
#[derive(PartialEq, Clone)]
#[derive(Debug, PartialEq, Clone)]
pub struct NaiveType<'a> {
pub main: &'a [u8],
pub sub: &'a [u8],
pub params: Vec<Parameter<'a>>,
}
impl<'a> fmt::Debug for NaiveType<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("mime::NaiveType")
.field("main", &String::from_utf8_lossy(self.main))
.field("sub", &String::from_utf8_lossy(self.sub))
.field("params", &self.params)
.finish()
}
}
impl<'a> NaiveType<'a> {
pub fn to_type(&self) -> AnyType {
self.into()
@ -40,20 +30,11 @@ pub fn naive_type(input: &[u8]) -> IResult<&[u8], NaiveType> {
)(input)
}
#[derive(PartialEq, Clone)]
#[derive(Debug, PartialEq, Clone)]
pub struct Parameter<'a> {
pub name: &'a [u8],
pub value: MIMEWord<'a>,
}
impl<'a> fmt::Debug for Parameter<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("mime::Parameter")
.field("name", &String::from_utf8_lossy(self.name))
.field("value", &self.value)
.finish()
}
}
pub fn parameter(input: &[u8]) -> IResult<&[u8], Parameter> {
map(
tuple((mime_atom, tag(b"="), mime_word)),
@ -70,10 +51,10 @@ pub fn parameter_list(input: &[u8]) -> IResult<&[u8], Vec<Parameter>> {
pub enum AnyType {
// Composite types
Multipart(Multipart),
Message(Deductible<Message>),
Message(Message),
// Discrete types
Text(Deductible<Text>),
Text(Text),
Binary(Binary),
}
@ -82,60 +63,30 @@ impl<'a> From<&'a NaiveType<'a>> for AnyType {
match nt.main.to_ascii_lowercase().as_slice() {
b"multipart" => Multipart::try_from(nt)
.map(Self::Multipart)
.unwrap_or(Self::Text(DeductibleText::default())),
b"message" => Self::Message(DeductibleMessage::Explicit(Message::from(nt))),
b"text" => Self::Text(DeductibleText::Explicit(Text::from(nt))),
.unwrap_or(Self::Text(Text::default())),
b"message" => Self::Message(Message::from(nt)),
b"text" => Self::Text(Text::from(nt)),
_ => Self::Binary(Binary::default()),
}
}
}
impl<'a> AnyType {
pub fn to_mime(self, fields: NaiveMIME<'a>) -> AnyMIME<'a> {
match self {
Self::Multipart(interpreted_type) => AnyMIME::Mult(MIME::<Multipart> {
interpreted_type,
fields,
}),
Self::Message(interpreted_type) => AnyMIME::Msg(MIME::<DeductibleMessage> {
interpreted_type,
fields,
}),
Self::Text(interpreted_type) => AnyMIME::Txt(MIME::<DeductibleText> {
interpreted_type,
fields,
}),
Self::Binary(interpreted_type) => AnyMIME::Bin(MIME::<Binary> {
interpreted_type,
fields,
}),
pub fn to_mime(self, parsed: NaiveMIME<'a>) -> AnyMIME<'a> {
match self {
Self::Multipart(interpreted) => AnyMIME::Mult(MIME::<Multipart> { interpreted, parsed }),
Self::Message(interpreted) => AnyMIME::Msg(MIME::<Message> { interpreted, parsed }),
Self::Text(interpreted) => AnyMIME::Txt(MIME::<Text> { interpreted, parsed }),
Self::Binary(interpreted) => AnyMIME::Bin(MIME::<Binary> { interpreted, parsed }),
}
}
}
#[derive(Debug, PartialEq, Clone)]
pub enum Deductible<T: Default> {
Inferred(T),
Explicit(T),
}
impl<T: Default> Default for Deductible<T> {
fn default() -> Self {
Self::Inferred(T::default())
}
}
// REAL PARTS
#[derive(Debug, PartialEq, Clone)]
pub struct Multipart {
pub subtype: MultipartSubtype,
pub boundary: String,
}
impl Multipart {
pub fn main_type(&self) -> String {
"multipart".into()
}
}
impl<'a> TryFrom<&'a NaiveType<'a>> for Multipart {
type Error = ();
@ -160,19 +111,6 @@ pub enum MultipartSubtype {
Report,
Unknown,
}
impl ToString for MultipartSubtype {
fn to_string(&self) -> String {
match self {
Self::Alternative => "alternative",
Self::Mixed => "mixed",
Self::Digest => "digest",
Self::Parallel => "parallel",
Self::Report => "report",
Self::Unknown => "mixed",
}
.into()
}
}
impl<'a> From<&NaiveType<'a>> for MultipartSubtype {
fn from(nt: &NaiveType<'a>) -> Self {
match nt.sub.to_ascii_lowercase().as_slice() {
@ -187,61 +125,28 @@ impl<'a> From<&NaiveType<'a>> for MultipartSubtype {
}
#[derive(Debug, PartialEq, Default, Clone)]
pub enum MessageSubtype {
pub enum Message {
#[default]
RFC822,
Partial,
External,
Unknown,
}
impl ToString for MessageSubtype {
fn to_string(&self) -> String {
match self {
Self::RFC822 => "rfc822",
Self::Partial => "partial",
Self::External => "external",
Self::Unknown => "rfc822",
}
.into()
}
}
pub type DeductibleMessage = Deductible<Message>;
#[derive(Debug, PartialEq, Default, Clone)]
pub struct Message {
pub subtype: MessageSubtype,
}
impl<'a> From<&NaiveType<'a>> for Message {
fn from(nt: &NaiveType<'a>) -> Self {
match nt.sub.to_ascii_lowercase().as_slice() {
b"rfc822" => Self {
subtype: MessageSubtype::RFC822,
},
b"partial" => Self {
subtype: MessageSubtype::Partial,
},
b"external" => Self {
subtype: MessageSubtype::External,
},
_ => Self {
subtype: MessageSubtype::Unknown,
},
}
}
}
impl From<Deductible<Message>> for Message {
fn from(d: Deductible<Message>) -> Self {
match d {
Deductible::Inferred(t) | Deductible::Explicit(t) => t,
b"rfc822" => Self::RFC822,
b"partial" => Self::Partial,
b"external" => Self::External,
_ => Self::Unknown,
}
}
}
pub type DeductibleText = Deductible<Text>;
#[derive(Debug, PartialEq, Default, Clone)]
pub struct Text {
pub subtype: TextSubtype,
pub charset: Deductible<EmailCharset>,
pub charset: EmailCharset,
}
impl<'a> From<&NaiveType<'a>> for Text {
fn from(nt: &NaiveType<'a>) -> Self {
@ -251,15 +156,8 @@ impl<'a> From<&NaiveType<'a>> for Text {
.params
.iter()
.find(|x| x.name.to_ascii_lowercase().as_slice() == b"charset")
.map(|x| Deductible::Explicit(EmailCharset::from(x.value.to_string().as_bytes())))
.unwrap_or(Deductible::Inferred(EmailCharset::US_ASCII)),
}
}
}
impl From<Deductible<Text>> for Text {
fn from(d: Deductible<Text>) -> Self {
match d {
Deductible::Inferred(t) | Deductible::Explicit(t) => t,
.map(|x| EmailCharset::from(x.value.to_string().as_bytes()))
.unwrap_or(EmailCharset::US_ASCII),
}
}
}
@ -271,15 +169,6 @@ pub enum TextSubtype {
Html,
Unknown,
}
impl ToString for TextSubtype {
fn to_string(&self) -> String {
match self {
Self::Plain | Self::Unknown => "plain",
Self::Html => "html",
}
.into()
}
}
impl<'a> From<&NaiveType<'a>> for TextSubtype {
fn from(nt: &NaiveType<'a>) -> Self {
match nt.sub.to_ascii_lowercase().as_slice() {
@ -297,7 +186,6 @@ pub struct Binary {}
mod tests {
use super::*;
use crate::mime::charset::EmailCharset;
use crate::mime::r#type::Deductible;
use crate::text::quoted::QuotedString;
#[test]
@ -331,10 +219,10 @@ mod tests {
assert_eq!(
nt.to_type(),
AnyType::Text(Deductible::Explicit(Text {
charset: Deductible::Explicit(EmailCharset::UTF_8),
AnyType::Text(Text {
charset: EmailCharset::UTF_8,
subtype: TextSubtype::Plain,
}))
})
);
}
@ -356,12 +244,7 @@ mod tests {
let (rest, nt) = naive_type(b"message/rfc822").unwrap();
assert_eq!(rest, &[]);
assert_eq!(
nt.to_type(),
AnyType::Message(Deductible::Explicit(Message {
subtype: MessageSubtype::RFC822
}))
);
assert_eq!(nt.to_type(), AnyType::Message(Message::RFC822),);
}
#[test]

2
src/parse.rs
View file

@ -8,7 +8,7 @@ fn main() {
let mut rawmail = Vec::new();
io::stdin().lock().read_to_end(&mut rawmail).unwrap();
let (_, eml) = eml_codec::parse_message(&rawmail).unwrap();
let (_, eml) = eml_codec::email(&rawmail).unwrap();
println!("{:#?}", eml);
assert!(eml.imf.date.is_some());
assert!(!eml.imf.from.is_empty());

368
src/part/composite.rs
View file

@ -1,49 +1,23 @@
use nom::IResult;
use std::fmt;
use crate::header;
use crate::header::{header, self};
use crate::imf;
use crate::mime;
use crate::part::{self, AnyPart};
use crate::pointers;
use crate::part::{self, AnyPart, field::MixedField};
use crate::text::boundary::{boundary, Delimiter};
//--- Multipart
#[derive(PartialEq)]
#[derive(Debug, PartialEq)]
pub struct Multipart<'a> {
pub mime: mime::MIME<'a, mime::r#type::Multipart>,
pub interpreted: mime::MIME<'a, mime::r#type::Multipart>,
pub children: Vec<AnyPart<'a>>,
pub raw_part_inner: &'a [u8],
pub raw_part_outer: &'a [u8],
}
impl<'a> fmt::Debug for Multipart<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("part::Multipart")
.field("mime", &self.mime)
.field("children", &self.children)
.field(
"raw_part_inner",
&String::from_utf8_lossy(self.raw_part_inner),
)
.field(
"raw_part_outer",
&String::from_utf8_lossy(self.raw_part_outer),
)
.finish()
}
pub preamble: &'a [u8],
pub epilogue: &'a [u8],
}
impl<'a> Multipart<'a> {
pub fn preamble(&self) -> &'a [u8] {
pointers::parsed(self.raw_part_outer, self.raw_part_inner)
}
pub fn epilogue(&self) -> &'a [u8] {
pointers::rest(self.raw_part_outer, self.raw_part_inner)
}
pub fn preamble_and_body(&self) -> &'a [u8] {
pointers::with_preamble(self.raw_part_outer, self.raw_part_inner)
}
pub fn body_and_epilogue(&self) -> &'a [u8] {
pointers::with_epilogue(self.raw_part_outer, self.raw_part_inner)
pub fn with_epilogue(mut self, e: &'a [u8]) -> Self {
self.epilogue = e;
self
}
}
@ -53,25 +27,19 @@ pub fn multipart<'a>(
let m = m.clone();
move |input| {
// init
let outer_orig = input;
let bound = m.interpreted_type.boundary.as_bytes();
let bound = m.interpreted.boundary.as_bytes();
let (mut input_loop, preamble) = part::part_raw(bound)(input)?;
let mut mparts: Vec<AnyPart> = vec![];
// skip preamble
let (mut input_loop, _) = part::part_raw(bound)(input)?;
let inner_orig = input_loop;
loop {
let input = match boundary(bound)(input_loop) {
Err(_) => {
return Ok((
input_loop,
Multipart {
mime: m.clone(),
interpreted: m.clone(),
children: mparts,
raw_part_inner: pointers::parsed(inner_orig, input_loop),
raw_part_outer: pointers::parsed(outer_orig, input_loop),
preamble,
epilogue: &[],
},
))
}
@ -79,13 +47,10 @@ pub fn multipart<'a>(
return Ok((
inp,
Multipart {
mime: m.clone(),
interpreted: m.clone(),
children: mparts,
raw_part_inner: pointers::parsed(inner_orig, inp),
raw_part_outer: pointers::parsed(
outer_orig,
&outer_orig[outer_orig.len()..],
),
preamble,
epilogue: &[],
},
))
}
@ -93,40 +58,23 @@ pub fn multipart<'a>(
};
// parse mime headers, otherwise pick default mime
let (input, naive_mime) = match header::header_kv(input) {
Ok((input_eom, fields)) => {
let raw_hdrs = pointers::parsed(input, input_eom);
let mime = fields
.iter()
.flat_map(mime::field::Content::try_from)
.into_iter()
.collect::<mime::NaiveMIME>();
let mime = mime.with_kv(fields).with_raw(raw_hdrs);
(input_eom, mime)
}
let (input, naive_mime) = match header(mime::field::content)(input) {
Ok((input, (known, unknown, bad))) => (input, known.into_iter().collect::<mime::NaiveMIME>().with_opt(unknown).with_bad(bad)),
Err(_) => (input, mime::NaiveMIME::default()),
};
// interpret mime according to context
let mime = match m.interpreted_type.subtype {
mime::r#type::MultipartSubtype::Digest => naive_mime
.to_interpreted::<mime::WithDigestDefault>()
.into(),
_ => naive_mime
.to_interpreted::<mime::WithGenericDefault>()
.into(),
let mime = match m.interpreted.subtype {
mime::r#type::MultipartSubtype::Digest => naive_mime.to_interpreted::<mime::WithDigestDefault>().into(),
_ => naive_mime.to_interpreted::<mime::WithGenericDefault>().into(),
};
// parse raw part
let (input, rpart) = part::part_raw(bound)(input)?;
// parse mime body
// -- we do not keep the input as we are using the
// part_raw function as our cursor here.
let (_, part) = part::anypart(mime)(rpart)?;
mparts.push(part);
mparts.push(part::to_anypart(mime, rpart));
input_loop = input;
}
@ -135,72 +83,47 @@ pub fn multipart<'a>(
//--- Message
#[derive(PartialEq)]
#[derive(Debug, PartialEq)]
pub struct Message<'a> {
pub mime: mime::MIME<'a, mime::r#type::DeductibleMessage>,
pub interpreted: mime::MIME<'a, mime::r#type::Message>,
pub imf: imf::Imf<'a>,
pub child: Box<AnyPart<'a>>,
pub raw_part: &'a [u8],
pub raw_headers: &'a [u8],
pub raw_body: &'a [u8],
pub epilogue: &'a [u8],
}
impl<'a> fmt::Debug for Message<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("part::Message")
.field("mime", &self.mime)
.field("imf", &self.imf)
.field("child", &self.child)
.field("raw_part", &String::from_utf8_lossy(self.raw_part))
.field("raw_headers", &String::from_utf8_lossy(self.raw_headers))
.field("raw_body", &String::from_utf8_lossy(self.raw_body))
.finish()
impl<'a> Message<'a> {
pub fn with_epilogue(mut self, e: &'a [u8]) -> Self {
self.epilogue = e;
self
}
}
pub fn message<'a>(
m: mime::MIME<'a, mime::r#type::DeductibleMessage>,
m: mime::MIME<'a, mime::r#type::Message>,
) -> impl Fn(&'a [u8]) -> IResult<&'a [u8], Message<'a>> {
move |input: &[u8]| {
let orig = input;
// parse header fields
let (input, headers) = header::header_kv(input)?;
let (input, (known, unknown, bad)): (_, (Vec::<MixedField>, Vec<header::Kv>, Vec<&[u8]>)) =
header(part::field::mixed_field)(input)?;
// extract raw parts 1/2
let raw_headers = pointers::parsed(orig, input);
let body_orig = input;
//---------------
// aggregate header fields
let (naive_mime, imf) = part::field::split_and_build(&headers);
let (naive_mime, imf) = part::field::sections(known);
// Bind headers to mime
let naive_mime = naive_mime.with_kv(headers);
// attach bad headers to imf
let imf = imf.with_opt(unknown).with_bad(bad);
// interpret headers to choose the child mime type
let in_mime = naive_mime
.with_raw(raw_headers)
.to_interpreted::<mime::WithGenericDefault>()
.into();
//---------------
// interpret headers to choose a mime type
let in_mime = naive_mime.to_interpreted::<mime::WithGenericDefault>().into();
// parse a part following this mime specification
let (input, part) = part::anypart(in_mime)(input)?;
// extract raw parts 2/2
let raw_body = pointers::parsed(body_orig, input);
let raw_part = pointers::parsed(orig, input);
// parse this mimetype
let part = part::to_anypart(in_mime, input);
Ok((
input,
&[],
Message {
mime: m.clone(),
interpreted: m.clone(),
imf,
raw_part,
raw_headers,
raw_body,
child: Box::new(part),
epilogue: &[],
},
))
}
@ -212,21 +135,22 @@ mod tests {
use crate::part::discrete::Text;
use crate::part::AnyPart;
use crate::text::encoding::{Base64Word, EncodedWord, QuotedChunk, QuotedWord};
use crate::text::misc_token::{MIMEWord, Phrase, UnstrToken, Unstructured, Word};
use crate::text::misc_token::{Phrase, UnstrToken, Unstructured, Word, MIMEWord};
use crate::text::quoted::QuotedString;
use chrono::{FixedOffset, TimeZone};
#[test]
fn test_multipart() {
let base_mime = mime::MIME {
interpreted_type: mime::r#type::Multipart {
interpreted: mime::r#type::Multipart {
subtype: mime::r#type::MultipartSubtype::Alternative,
boundary: "simple boundary".to_string(),
},
fields: mime::NaiveMIME::default(),
parsed: mime::NaiveMIME::default(),
};
let input = b"This is the preamble. It is to be ignored, though it
assert_eq!(
multipart(base_mime.clone())(b"This is the preamble. It is to be ignored, though it
is a handy place for composition agents to include an
explanatory note to non-MIME conformant readers.
@ -243,50 +167,30 @@ It DOES end with a linebreak.
--simple boundary--
This is the epilogue. It is also to be ignored.
";
let inner = b"
--simple boundary
This is implicitly typed plain US-ASCII text.
It does NOT end with a linebreak.
--simple boundary
Content-type: text/plain; charset=us-ascii
This is explicitly typed plain US-ASCII text.
It DOES end with a linebreak.
--simple boundary--
";
assert_eq!(
multipart(base_mime.clone())(input),
"),
Ok((&b"\nThis is the epilogue. It is also to be ignored.\n"[..],
Multipart {
mime: base_mime,
raw_part_outer: input,
raw_part_inner: inner,
interpreted: base_mime,
preamble: &b"This is the preamble. It is to be ignored, though it\nis a handy place for composition agents to include an\nexplanatory note to non-MIME conformant readers.\n"[..],
epilogue: &b""[..],
children: vec![
AnyPart::Txt(Text {
mime: mime::MIME {
interpreted_type: mime::r#type::Deductible::Inferred(mime::r#type::Text {
interpreted: mime::MIME {
interpreted: mime::r#type::Text {
subtype: mime::r#type::TextSubtype::Plain,
charset: mime::r#type::Deductible::Inferred(mime::charset::EmailCharset::US_ASCII),
}),
fields: mime::NaiveMIME {
raw: &b"\n"[..],
..mime::NaiveMIME::default()
charset: mime::charset::EmailCharset::US_ASCII,
},
parsed: mime::NaiveMIME::default(),
},
body: &b"This is implicitly typed plain US-ASCII text.\nIt does NOT end with a linebreak."[..],
}),
AnyPart::Txt(Text {
mime: mime::MIME {
interpreted_type: mime::r#type::Deductible::Explicit(mime::r#type::Text {
interpreted: mime::MIME {
interpreted: mime::r#type::Text {
subtype: mime::r#type::TextSubtype::Plain,
charset: mime::r#type::Deductible::Explicit(mime::charset::EmailCharset::US_ASCII),
}),
fields: mime::NaiveMIME {
charset: mime::charset::EmailCharset::US_ASCII,
},
parsed: mime::NaiveMIME {
ctype: Some(mime::r#type::NaiveType {
main: &b"text"[..],
sub: &b"plain"[..],
@ -297,10 +201,6 @@ It DOES end with a linebreak.
}
]
}),
raw: &b"Content-type: text/plain; charset=us-ascii\n\n"[..],
kv: vec![
header::Field::Good(header::Kv2(&b"Content-type"[..], &b"text/plain; charset=us-ascii"[..]))
],
..mime::NaiveMIME::default()
},
},
@ -359,90 +259,14 @@ OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
"#
.as_bytes();
let hdrs = br#"Date: Sat, 8 Jul 2023 07:14:29 +0200
From: Grrrnd Zero <grrrndzero@example.org>
To: John Doe <jdoe@machine.example>
CC: =?ISO-8859-1?Q?Andr=E9?= Pirard <PIRARD@vm1.ulg.ac.be>
Subject: =?ISO-8859-1?B?SWYgeW91IGNhbiByZWFkIHRoaXMgeW8=?=
=?ISO-8859-2?B?dSB1bmRlcnN0YW5kIHRoZSBleGFtcGxlLg==?=
X-Unknown: something something
Bad entry
on multiple lines
Message-ID: <NTAxNzA2AC47634Y366BAMTY4ODc5MzQyODY0ODY5@www.grrrndzero.org>
MIME-Version: 1.0
Content-Type: multipart/alternative;
boundary="b1_e376dc71bafc953c0b0fdeb9983a9956"
Content-Transfer-Encoding: 7bit
"#;
let body = br#"This is a multi-part message in MIME format.
--b1_e376dc71bafc953c0b0fdeb9983a9956
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
GZ
OoOoO
oOoOoOoOo
oOoOoOoOoOoOoOoOo
oOoOoOoOoOoOoOoOoOoOoOo
oOoOoOoOoOoOoOoOoOoOoOoOoOoOo
OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO
--b1_e376dc71bafc953c0b0fdeb9983a9956
Content-Type: text/html; charset=us-ascii
<div style="text-align: center;"><strong>GZ</strong><br />
OoOoO<br />
oOoOoOoOo<br />
oOoOoOoOoOoOoOoOo<br />
oOoOoOoOoOoOoOoOoOoOoOo<br />
oOoOoOoOoOoOoOoOoOoOoOoOoOoOo<br />
OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
</div>
--b1_e376dc71bafc953c0b0fdeb9983a9956--
"#;
let inner = br#"
--b1_e376dc71bafc953c0b0fdeb9983a9956
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
GZ
OoOoO
oOoOoOoOo
oOoOoOoOoOoOoOoOo
oOoOoOoOoOoOoOoOoOoOoOo
oOoOoOoOoOoOoOoOoOoOoOoOoOoOo
OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO
--b1_e376dc71bafc953c0b0fdeb9983a9956
Content-Type: text/html; charset=us-ascii
<div style="text-align: center;"><strong>GZ</strong><br />
OoOoO<br />
oOoOoOoOo<br />
oOoOoOoOoOoOoOoOo<br />
oOoOoOoOoOoOoOoOoOoOoOo<br />
oOoOoOoOoOoOoOoOoOoOoOoOoOoOo<br />
OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
</div>
--b1_e376dc71bafc953c0b0fdeb9983a9956--
"#;
let base_mime = mime::MIME::<mime::r#type::DeductibleMessage>::default();
let base_mime = mime::MIME::<mime::r#type::Message>::default();
assert_eq!(
message(base_mime.clone())(fullmail),
Ok((
&[][..],
Message {
mime: base_mime,
raw_part: fullmail,
raw_headers: hdrs,
raw_body: body,
interpreted: base_mime,
epilogue: &b""[..],
imf: imf::Imf {
date: Some(FixedOffset::east_opt(2 * 3600)
.unwrap()
@ -506,15 +330,24 @@ OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
right: &b"www.grrrndzero.org"[..],
}),
mime_version: Some(imf::mime::Version { major: 1, minor: 0}),
header_ext: vec![
header::Kv(&b"X-Unknown"[..], Unstructured(vec![
UnstrToken::Plain(&b"something"[..]),
UnstrToken::Plain(&b"something"[..]),
]))
],
header_bad: vec![
&b"Bad entry\n on multiple lines\n"[..],
],
..imf::Imf::default()
},
child: Box::new(AnyPart::Mult(Multipart {
mime: mime::MIME {
interpreted_type: mime::r#type::Multipart {
interpreted: mime::MIME {
interpreted: mime::r#type::Multipart {
subtype: mime::r#type::MultipartSubtype::Alternative,
boundary: "b1_e376dc71bafc953c0b0fdeb9983a9956".to_string(),
},
fields: mime::NaiveMIME {
parsed: mime::NaiveMIME {
ctype: Some(mime::r#type::NaiveType {
main: &b"multipart"[..],
sub: &b"alternative"[..],
@ -525,33 +358,19 @@ OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
}
]
}),
raw: hdrs,
kv: vec![
header::Field::Good(header::Kv2(&b"Date"[..], &b"Sat, 8 Jul 2023 07:14:29 +0200"[..])),
header::Field::Good(header::Kv2(&b"From"[..], &b"Grrrnd Zero <grrrndzero@example.org>"[..])),
header::Field::Good(header::Kv2(&b"To"[..], &b"John Doe <jdoe@machine.example>"[..])),
header::Field::Good(header::Kv2(&b"CC"[..], &b"=?ISO-8859-1?Q?Andr=E9?= Pirard <PIRARD@vm1.ulg.ac.be>"[..])),
header::Field::Good(header::Kv2(&b"Subject"[..], &b"=?ISO-8859-1?B?SWYgeW91IGNhbiByZWFkIHRoaXMgeW8=?=\n =?ISO-8859-2?B?dSB1bmRlcnN0YW5kIHRoZSBleGFtcGxlLg==?="[..])),
header::Field::Good(header::Kv2(&b"X-Unknown"[..], &b"something something"[..])),
header::Field::Bad(&b"Bad entry\n on multiple lines\n"[..]),
header::Field::Good(header::Kv2(&b"Message-ID"[..], &b"<NTAxNzA2AC47634Y366BAMTY4ODc5MzQyODY0ODY5@www.grrrndzero.org>"[..])),
header::Field::Good(header::Kv2(&b"MIME-Version"[..], &b"1.0"[..])),
header::Field::Good(header::Kv2(&b"Content-Type"[..], &b"multipart/alternative;\n boundary=\"b1_e376dc71bafc953c0b0fdeb9983a9956\""[..])),
header::Field::Good(header::Kv2(&b"Content-Transfer-Encoding"[..], &b"7bit"[..])),
],
..mime::NaiveMIME::default()
},
},
raw_part_inner: inner,
raw_part_outer: body,
preamble: &b"This is a multi-part message in MIME format.\n"[..],
epilogue: &b""[..],
children: vec![
AnyPart::Txt(Text {
mime: mime::MIME {
interpreted_type: mime::r#type::Deductible::Explicit(mime::r#type::Text {
interpreted: mime::MIME {
interpreted: mime::r#type::Text {
subtype: mime::r#type::TextSubtype::Plain,
charset: mime::r#type::Deductible::Explicit(mime::charset::EmailCharset::UTF_8),
}),
fields: mime::NaiveMIME {
charset: mime::charset::EmailCharset::UTF_8,
},
parsed: mime::NaiveMIME {
ctype: Some(mime::r#type::NaiveType {
main: &b"text"[..],
sub: &b"plain"[..],
@ -563,24 +382,19 @@ OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
]
}),
transfer_encoding: mime::mechanism::Mechanism::QuotedPrintable,
kv: vec![
header::Field::Good(header::Kv2(&b"Content-Type"[..], &b"text/plain; charset=utf-8"[..])),
header::Field::Good(header::Kv2(&b"Content-Transfer-Encoding"[..], &b"quoted-printable"[..])),
],
raw: &b"Content-Type: text/plain; charset=utf-8\nContent-Transfer-Encoding: quoted-printable\n\n"[..],
..mime::NaiveMIME::default()
}
},
body: &b"GZ\nOoOoO\noOoOoOoOo\noOoOoOoOoOoOoOoOo\noOoOoOoOoOoOoOoOoOoOoOo\noOoOoOoOoOoOoOoOoOoOoOoOoOoOo\nOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO\n"[..],
}),
AnyPart::Txt(Text {
mime: mime::MIME {
interpreted_type: mime::r#type::Deductible::Explicit(mime::r#type::Text {
interpreted: mime::MIME {
interpreted: mime::r#type::Text {
subtype: mime::r#type::TextSubtype::Html,
charset: mime::r#type::Deductible::Explicit(mime::charset::EmailCharset::US_ASCII),
}),
charset: mime::charset::EmailCharset::US_ASCII,
},
fields: mime::NaiveMIME {
parsed: mime::NaiveMIME {
ctype: Some(mime::r#type::NaiveType {
main: &b"text"[..],
sub: &b"html"[..],
@ -591,10 +405,6 @@ OoOoOoOoOoOoOoOoOoOoOoOoOoOoOoOoO<br />
}
]
}),
kv: vec![
header::Field::Good(header::Kv2(&b"Content-Type"[..], &b"text/html; charset=us-ascii"[..])),
],
raw: &b"Content-Type: text/html; charset=us-ascii\n\n"[..],
..mime::NaiveMIME::default()
},
},

18
src/part/discrete.rs
View file

@ -4,30 +4,36 @@ use crate::mime;
#[derive(PartialEq)]
pub struct Text<'a> {
pub mime: mime::MIME<'a, mime::r#type::DeductibleText>,
pub interpreted: mime::MIME<'a, mime::r#type::Text>,
pub body: &'a [u8],
}
impl<'a> fmt::Debug for Text<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("part::Text")
.field("mime", &self.mime)
.field("body", &String::from_utf8_lossy(self.body))
.field("mime", &self.interpreted)
.field(
"body",
&format_args!("\"{}\"", String::from_utf8_lossy(self.body)),
)
.finish()
}
}
#[derive(PartialEq)]
pub struct Binary<'a> {
pub mime: mime::MIME<'a, mime::r#type::Binary>,
pub interpreted: mime::MIME<'a, mime::r#type::Binary>,
pub body: &'a [u8],
}
impl<'a> fmt::Debug for Binary<'a> {
fn fmt(&self, fmt: &mut fmt::Formatter<'_>) -> fmt::Result {
fmt.debug_struct("part::Binary")
.field("mime", &self.mime)
.field("body", &String::from_utf8_lossy(self.body))
.field("mime", &self.interpreted)
.field(
"body",
&format_args!("\"{}\"", String::from_utf8_lossy(self.body)),
)
.finish()
}
}

66
src/part/field.rs
View file

@ -1,24 +1,50 @@
use crate::header;
use nom::{branch::alt, combinator::map, IResult};
use crate::imf;
use crate::mime;
pub fn split_and_build<'a>(v: &Vec<header::Field<'a>>) -> (mime::NaiveMIME<'a>, imf::Imf<'a>) {
let (mimev, imfv) = v.iter().fold(
(
Vec::<mime::field::Content>::new(),
Vec::<imf::field::Field>::new(),
),
|(mut mime, mut imf), f| {
if let Ok(m) = mime::field::Content::try_from(f) {
mime.push(m);
} else if let Ok(i) = imf::field::Field::try_from(f) {
imf.push(i);
}
(mime, imf)
},
);
let fmime = mimev.into_iter().collect::<mime::NaiveMIME>();
let fimf = imfv.into_iter().collect::<imf::Imf>();
(fmime, fimf)
pub enum MixedField<'a> {
MIME(mime::field::Content<'a>),
IMF(imf::field::Field<'a>),
}
#[allow(dead_code)]
impl<'a> MixedField<'a> {
pub fn mime(&self) -> Option<&mime::field::Content<'a>> {
match self {
Self::MIME(v) => Some(v),
_ => None,
}
}
pub fn to_mime(self) -> Option<mime::field::Content<'a>> {
match self {
Self::MIME(v) => Some(v),
_ => None,
}
}
pub fn imf(&self) -> Option<&imf::field::Field<'a>> {
match self {
Self::IMF(v) => Some(v),
_ => None,
}
}
pub fn to_imf(self) -> Option<imf::field::Field<'a>> {
match self {
Self::IMF(v) => Some(v),
_ => None,
}
}
}
pub fn sections<'a>(list: Vec<MixedField<'a>>) -> (mime::NaiveMIME<'a>, imf::Imf<'a>) {
let (v1, v2): (Vec<MixedField>, Vec<_>) = list.into_iter().partition(|v| v.mime().is_some());
let mime = v1.into_iter().flat_map(MixedField::to_mime).collect::<mime::NaiveMIME>();
let imf = v2.into_iter().flat_map(MixedField::to_imf).collect::<imf::Imf>();
(mime, imf)
}
pub fn mixed_field(input: &[u8]) -> IResult<&[u8], MixedField> {
alt((
map(mime::field::content, MixedField::MIME),
map(imf::field::field, MixedField::IMF),
))(input)
}

77
src/part/mod.rs
View file

@ -17,7 +17,7 @@ use nom::{
};
use crate::mime;
use crate::mime::{AnyMIME, NaiveMIME};
use crate::mime::AnyMIME;
use crate::part::{
composite::{message, multipart, Message, Multipart},
discrete::{Binary, Text},
@ -58,61 +58,30 @@ impl<'a> AnyPart<'a> {
_ => None,
}
}
pub fn mime(&self) -> &NaiveMIME<'a> {
match self {
Self::Mult(v) => &v.mime.fields,
Self::Msg(v) => &v.mime.fields,
Self::Txt(v) => &v.mime.fields,
Self::Bin(v) => &v.mime.fields,
}
}
}
impl<'a> From<Multipart<'a>> for AnyPart<'a> {
fn from(m: Multipart<'a>) -> Self {
Self::Mult(m)
}
}
impl<'a> From<Message<'a>> for AnyPart<'a> {
fn from(m: Message<'a>) -> Self {
Self::Msg(m)
}
}
/// Parse any type of part
///
/// ## Note
///
/// Multiparts are a bit special as they have a clearly delimited beginning
/// and end contrary to all the other parts that are going up to the end of the buffer
pub fn anypart<'a>(m: AnyMIME<'a>) -> impl FnOnce(&'a [u8]) -> IResult<&'a [u8], AnyPart<'a>> {
move |input| {
let part = match m {
AnyMIME::Mult(a) => multipart(a)(input)
.map(|(_, multi)| multi.into())
.unwrap_or(AnyPart::Txt(Text {
mime: mime::MIME::<mime::r#type::DeductibleText>::default(),
body: input,
})),
AnyMIME::Msg(a) => {
message(a)(input)
.map(|(_, msg)| msg.into())
.unwrap_or(AnyPart::Txt(Text {
mime: mime::MIME::<mime::r#type::DeductibleText>::default(),
body: input,
}))
}
AnyMIME::Txt(a) => AnyPart::Txt(Text {
mime: a,
body: input,
}),
AnyMIME::Bin(a) => AnyPart::Bin(Binary {
mime: a,
body: input,
}),
};
// This function always consumes the whole input
Ok((&input[input.len()..], part))
pub fn to_anypart<'a>(m: AnyMIME<'a>, rpart: &'a [u8]) -> AnyPart<'a> {
match m {
AnyMIME::Mult(a) => multipart(a)(rpart)
.map(|(rest, multi)| AnyPart::Mult(multi.with_epilogue(rest)))
.unwrap_or(AnyPart::Txt(Text {
interpreted: mime::MIME::<mime::r#type::Text>::default(),
body: rpart,
})),
AnyMIME::Msg(a) => message(a)(rpart)
.map(|(rest, msg)| AnyPart::Msg(msg.with_epilogue(rest)))
.unwrap_or(AnyPart::Txt(Text {
interpreted: mime::MIME::<mime::r#type::Text>::default(),
body: rpart,
})),
AnyMIME::Txt(a) => AnyPart::Txt(Text {
interpreted: a,
body: rpart,
}),
AnyMIME::Bin(a) => AnyPart::Bin(Binary {
interpreted: a,
body: rpart,
}),
}
}

55
src/pointers.rs
View file

@ -1,55 +0,0 @@
pub fn parsed<'a>(input: &'a [u8], rest: &'a [u8]) -> &'a [u8] {
let start = input.as_ptr();
let offset = rest.as_ptr();
let idx = (offset as usize - start as usize) / std::mem::size_of::<u8>();
assert!(idx <= input.len());
&input[..idx]
}
pub fn rest<'a>(input: &'a [u8], parsed: &'a [u8]) -> &'a [u8] {
let start = input.as_ptr();
let offset = (&parsed[parsed.len()..]).as_ptr();
let idx = (offset as usize - start as usize) / std::mem::size_of::<u8>();
assert!(idx <= input.len());
&input[idx..]
}
pub fn with_preamble<'a>(input: &'a [u8], parsed: &'a [u8]) -> &'a [u8] {
let start = input.as_ptr();
let offset = (&parsed[parsed.len()..]).as_ptr();
let idx = (offset as usize - start as usize) / std::mem::size_of::<u8>();
assert!(idx <= input.len());
&input[..idx]
}
pub fn with_epilogue<'a>(input: &'a [u8], rest: &'a [u8]) -> &'a [u8] {
let start = input.as_ptr();
let offset = rest.as_ptr();
let idx = (offset as usize - start as usize) / std::mem::size_of::<u8>();
assert!(idx <= input.len());
&input[idx..]
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_all() {
let outer = b"aa bb cc";
let inner = &outer[3..5];
assert_eq!(inner, b"bb");
let p = parsed(outer, inner);
assert_eq!(p, b"aa ");
let r = rest(outer, inner);
assert_eq!(r, b" cc");
let wp = with_preamble(outer, inner);
assert_eq!(wp, b"aa bb");
let we = with_epilogue(outer, inner);
assert_eq!(we, b"bb cc");
}
}

5
src/text/ascii.rs
View file

@ -136,11 +136,6 @@ pub const TILDE: u8 = 0x7E; // ~
// GROUP OF CHARACTERS
// -- CRLF
pub const CRLF: &[u8] = &[CR, LF];
// -- CRCRLF
// Sometimes portable libraries replace transparently
// the "\n" with "\r\n" on Windows. When developpers
// explicitly write "\r\n", the library generates "\r\r\n".
pub const CRCRLF: &[u8] = &[CR, CR, LF];
// -- WHITESPACE
pub const WS: &[u8] = &[HT, SP];

7
src/text/whitespace.rs
View file

@ -22,12 +22,7 @@ use nom::{
/// \r or \n is allowed nowhere else, so we also add this support.
pub fn obs_crlf(input: &[u8]) -> IResult<&[u8], &[u8]> {
alt((
tag(ascii::CRLF),
tag(ascii::CRCRLF),
tag(&[ascii::CR]),
tag(&[ascii::LF]),
))(input)
alt((tag(ascii::CRLF), tag(&[ascii::CR]), tag(&[ascii::LF])))(input)
}
/// ```abnf