mailcat/src/parsing.rs
2021-07-18 16:05:10 +02:00

213 lines
6 KiB
Rust

use mailparse::{DispositionType, MailHeaderMap, ParsedMail};
#[derive(PartialEq, Eq, Debug, Default)]
pub struct Body {
pub text: Option<String>,
}
pub trait ParsedMailExt {
fn is_attachment(&self) -> bool;
fn subject(&self) -> Option<String>;
fn date(&self) -> Result<Option<chrono::DateTime<chrono::FixedOffset>>, anyhow::Error>;
fn body(&self) -> Result<Option<Body>, anyhow::Error>;
}
impl<'a> ParsedMailExt for ParsedMail<'a> {
fn is_attachment(&self) -> bool {
self.get_content_disposition().disposition == DispositionType::Attachment
}
fn subject(&self) -> Option<String> {
self.headers.get_first_value("Subject")
}
fn date(&self) -> Result<Option<chrono::DateTime<chrono::FixedOffset>>, anyhow::Error> {
let date = self
.headers
.get_first_value("Date")
.map(|s| chrono::DateTime::parse_from_rfc2822(&s))
.transpose()?;
Ok(date)
}
fn body(&self) -> Result<Option<Body>, anyhow::Error> {
let mimetype: mime::Mime = self.ctype.mimetype.parse()?;
if self.is_attachment() {
return Ok(None);
}
if mimetype == mime::TEXT_PLAIN {
return Ok(Some(Body {
text: Some(self.get_body()?),
}));
}
for subpart in &self.subparts {
if let Some(body) = subpart.body()? {
return Ok(Some(body));
}
}
Ok(None)
}
}
#[cfg(test)]
mod test {
use chrono::TimeZone;
use super::*;
const TEXT_PLAIN: &str = r#"From: Andrey Golovizin <ag@sologoc.com>
To: ag@sologoc.com
Subject: Plain text body
Date: Sat, 17 Jul 2021 18:04:10 +0200
Message-ID: <2148453.iZASKD2KPV@sakuragaoka>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="iso-8859-1"
Prost=FD text.
"#;
const MULTIPART_ALTERNATIVE: &str = r#"From: Andrey Golovizin <ag@sologoc.com>
Subject: Plain text with HTML
Date: Sat, 17 Jul 2021 18:13:26 +0200
Message-ID: <3609140.kQq0lBPeGt@sakuragaoka>
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="nextPart3377186.iIbC2pHGDl"
Content-Transfer-Encoding: 7Bit
This is a multi-part message in MIME format.
--nextPart3377186.iIbC2pHGDl
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="iso-8859-1"
Prost=FD text.
--nextPart3377186.iIbC2pHGDl
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html; charset="UTF-8"
<html>
<head>
<meta http-equiv=3D"content-type" content=3D"text/html; charset=3DUTF-8">
</head>
<body><p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0=
;"><strong>N=C4=9Bjaky HTML text.</strong></p>
</body>
</html>
--nextPart3377186.iIbC2pHGDl--
"#;
const MULTIPART_MIXED_ALTERNATIVE: &str = r#"From: Andrey Golovizin <ag@sologoc.com>
Subject: Plain text with HTML and attachment
Date: Sat, 17 Jul 2021 21:39:08 +0200
Message-ID: <3054314.5fSG56mABF@sakuragaoka>
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="nextPart1698715.VLH7GnMWUR"
Content-Transfer-Encoding: 7Bit
This is a multi-part message in MIME format.
--nextPart1698715.VLH7GnMWUR
Content-Type: multipart/alternative; boundary="nextPart5630828.MhkbZ0Pkbq"
Content-Transfer-Encoding: 7Bit
This is a multi-part message in MIME format.
--nextPart5630828.MhkbZ0Pkbq
Content-Transfer-Encoding: base64
Content-Type: text/plain; charset="UTF-8"
UHJvc3TDvSB0ZXh0Lg==
--nextPart5630828.MhkbZ0Pkbq
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html; charset="UTF-8"
<html>
<head>
<meta http-equiv=3D"content-type" content=3D"text/html; charset=3DUTF-8">
</head>
<body><p style=3D"margin-top:0;margin-bottom:0;margin-left:0;margin-right:0=
;"><strong>Tu=C4=8Dn=C3=BD HTML text</strong></p>
</body>
</html>
--nextPart5630828.MhkbZ0Pkbq--
--nextPart1698715.VLH7GnMWUR
Content-Disposition: attachment; filename="test.txt"
Content-Transfer-Encoding: base64
Content-Type: text/plain; charset="UTF-8"; name="test.txt"
TsSbamFrw6EgcMWZw61sb2hhLgo=
--nextPart1698715.VLH7GnMWUR--
"#;
#[test]
fn test_plain_text_body() -> Result<(), anyhow::Error> {
let message = mailparse::parse_mail(TEXT_PLAIN.as_bytes())?;
assert_eq!(message.subject(), Some("Plain text body".to_string()));
assert_eq!(
message.date()?,
Some(
chrono::FixedOffset::east(3600 * 2)
.ymd(2021, 7, 17)
.and_hms(18, 4, 10),
)
);
assert_eq!(
message.body()?,
Some(Body {
text: Some("Prostý text.".to_string())
}),
);
Ok(())
}
#[test]
fn test_multipart_alternative_text_body() -> Result<(), anyhow::Error> {
let message = mailparse::parse_mail(MULTIPART_ALTERNATIVE.as_bytes())?;
assert_eq!(message.subject(), Some("Plain text with HTML".to_string()));
assert_eq!(
message.date()?,
Some(
chrono::FixedOffset::east(3600 * 2)
.ymd(2021, 7, 17)
.and_hms(18, 13, 26),
)
);
assert_eq!(
message.body()?,
Some(Body {
text: Some("Prostý text.".to_string())
}),
);
Ok(())
}
#[test]
fn test_multipart_mixed_alternative_text_body() -> Result<(), anyhow::Error> {
let message = mailparse::parse_mail(MULTIPART_MIXED_ALTERNATIVE.as_bytes())?;
assert_eq!(
message.subject(),
Some("Plain text with HTML and attachment".to_string())
);
assert_eq!(
message.date()?,
Some(
chrono::FixedOffset::east(3600 * 2)
.ymd(2021, 7, 17)
.and_hms(21, 39, 8),
)
);
assert_eq!(
message.body()?,
Some(Body {
text: Some("Prostý text.".to_string())
}),
);
Ok(())
}
}