Refactoring libgitmail core to work on complete emails, with tests

6 files changed, 406 insertions, 0 deletions

diff --git a/libgitmail/README b/libgitmail/README
new file mode 100644
index 0000000..0d3ba07
--- /dev/null
+++ b/libgitmail/README
@@ -0,0 +1,18 @@
+libgitmail
+==========
+
+This library provides an abstraction over git sent email patches.  It
+parses incoming mail and provides a strongly typed interface for
+interacting with patches, patch trees, as well as building patch sets
+from a tree thread with many revisions.
+
+It then compiles a new thread of mail that can be piped directly into
+git am - by some user interface.
+
+
+License
+-------
+
+This library is part of dev-suite and as such licensed under the GNU
+General Public License 3.0 or, (at your choosing), any later version.
+See the LICENSE.md in the parent repository for more details.
diff --git a/libgitmail/email.txt b/libgitmail/email.txt
new file mode 100644
index 0000000..ac950ad
--- /dev/null
+++ b/libgitmail/email.txt
@@ -0,0 +1,86 @@
+Received: from wout5-smtp.messagingengine.com (wout5-smtp.messagingengine.com [64.147.123.21])
+	by mail.sr.ht (Postfix) with ESMTPS id 9C1CE402A6
+	for <~qaul/community@lists.sr.ht>; Sun, 29 Dec 2019 13:25:16 +0000 (UTC)
+Authentication-Results: mail.sr.ht;
+	dkim=pass (2048-bit key) header.d=alyssa.is header.i=@alyssa.is header.b=bK0bsSUp;
+	dkim=pass (2048-bit key) header.d=messagingengine.com header.i=@messagingengine.com header.b=HzRR6htT
+Received: from compute5.internal (compute5.nyi.internal [10.202.2.45])
+	by mailout.west.internal (Postfix) with ESMTP id 226173D1;
+	Sun, 29 Dec 2019 08:25:15 -0500 (EST)
+Received: from mailfrontend2 ([10.202.2.163])
+  by compute5.internal (MEProxy); Sun, 29 Dec 2019 08:25:15 -0500
+DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=alyssa.is; h=
+	from:to:cc:subject:date:message-id:mime-version
+	:content-transfer-encoding; s=fm2; bh=pcHi92xLPx2ZwvL3WD16XEuRb2
+	iD1LAkLhnN11wIKrE=; b=bK0bsSUpdzMvscbhhfMuAzxVdz/+/N9ceeiAXxJoBG
+	7Ley+qOF9lHXtXgJg8+cP32f51P871M+X9L5qzKTFKwYZE3PUl1sO0Xhj0FAHBqa
+	enMXkCvxZbSn3yz8Y/v+zk0GPIcvd99CkPELkkTusR19EME5hV3AymYZKiln42UV
+	Tc6NYsGzaEvdgH8Lq3xGz1mhcC5knHYelKZz2LZu2D6aQy8WZOBxZDRkwlfdfb85
+	yQ7faBRaxfiw6jO/8aspKr1DLiGbE+2OWOwj/kEcIyD1gbMwjJb8NGHHrYWxj32q
+	juSdIHdWZgmO0Y397wAi2CeBvch/03g9dkyy6LQEEvDQ==
+DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=
+	messagingengine.com; h=cc:content-transfer-encoding:date:from
+	:message-id:mime-version:subject:to:x-me-proxy:x-me-proxy
+	:x-me-sender:x-me-sender:x-sasl-enc; s=fm1; bh=pcHi92xLPx2ZwvL3W
+	D16XEuRb2iD1LAkLhnN11wIKrE=; b=HzRR6htTKCdXVTf44qHumeH7nU/wgi9Ic
+	NFdHpWpMW1eH24huf4K4bhRf5Zl+9UVcWLOhYgYwQcRUkBlOcKrGwmxeL+0/ZmZR
+	AJqONj6xNJAZatmKsPuk9ZWd/59YlDw6CwxCw7De5qPYS6xaYTKIlkC67OuftPIa
+	wykga/tK9r0cG6nYwiEKwNR2hhUi8VrfNAMhwGbaNAHvGy4UY8tn4fhzmJKc1hn1
+	s9LHzZUOriwFFE2PmCLEooNLosN47xp3T2g2lBCsT//SiJTv8Rm+Qq5kRMG2iUgp
+	Dh3Vx/OoYzmXUATFfG2BCWoiwQlCmC3/2N6ph4Pcjs17DtmOk8y4g==
+X-ME-Sender: <xms:OakIXj9KSSuHqRfGLe6tUc0G3UgzkdLMKEwQ09Mv9Mtug1erzbaCKA>
+X-ME-Proxy-Cause: gggruggvucftvghtrhhoucdtuddrgedufedrvdeffedgheduucetufdoteggodetrfdotf
+    fvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdfqfgfvpdfurfetoffkrfgpnffqhgen
+    uceurghilhhouhhtmecufedttdenucenucfjughrpefhvffufffkofgggfestdekredtre
+    dttdenucfhrhhomheptehlhihsshgrucftohhsshcuoehhihesrghlhihsshgrrdhisheq
+    necuffhomhgrihhnpegvmhgsvghrqdgtlhhirdgtohhmpdhgohhoghhlvgdrtghomhdpfh
+    hinhgvrdguohgtshdpnhhouggvjhhsrdhorhhgpdihrghrnhhpkhhgrdgtohhmpdhgihht
+    qdhstghmrdgtohhmnecukfhppeduhedurddvudejrddvudejrdduheegnecurfgrrhgrmh
+    epmhgrihhlfhhrohhmpehqhihlihhsshesgidvvddtrdhqhihlihhsshdrnhgvthenucev
+    lhhushhtvghrufhiiigvpedt
+X-ME-Proxy: <xmx:OakIXu9us3-tg5ttMt9F2aiP33qpkoccDHkUI6wgKBwjRBEOv1KQtA>
+    <xmx:OakIXj3qwzXAlA5-K5KvaNTRMDcPQ83Spc8FsTXreEkvZTkOg_oSog>
+    <xmx:OakIXm9h6ztAwRW9tHaBXehYrCR55Q-F9jD1NXrfwEB4iRuT_Vkjkg>
+    <xmx:OqkIXi95Uuou4hos4yeU2YHX4DogCTC9OISbfvyBX9mxhKulMTAuUg>
+Received: from x220.qyliss.net (unknown [151.217.217.154])
+	by mail.messagingengine.com (Postfix) with ESMTPA id A5F8B30608D7;
+	Sun, 29 Dec 2019 08:25:13 -0500 (EST)
+Received: by x220.qyliss.net (Postfix, from userid 1000)
+	id E0BA68D7; Sun, 29 Dec 2019 13:25:10 +0000 (UTC)
+From: Alyssa Ross <hi@alyssa.is>
+To: ~qaul/community@lists.sr.ht
+Cc: Katharina Fey <kookie@spacekookie.de>,
+	Mathias Jud <jud@open-communication.net>,
+	Lux <lux@lux.name>,
+	Alyssa Ross <hi@alyssa.is>
+Subject: [PATCH] docs: Node.js 10 works fine (now?)
+Date: Sun, 29 Dec 2019 13:24:05 +0000
+Message-Id: <20191229132404.14579-1-hi@alyssa.is>
+X-Mailer: git-send-email 2.23.0
+MIME-Version: 1.0
+Content-Transfer-Encoding: quoted-printable
+
+---
+I tested with Node 10 on master and feature/webgui/responsive, and
+both worked fine.
+
+ docs/contributors/src/technical/webgui/install.md | 2 +-
+ 1 file changed, 1 insertion(+), 1 deletion(-)
+
+diff --git a/docs/contributors/src/technical/webgui/install.md b/docs/con=
+tributors/src/technical/webgui/install.md
+index db197ff..6501d91 100644
+--- a/docs/contributors/src/technical/webgui/install.md
++++ b/docs/contributors/src/technical/webgui/install.md
+@@ -10,7 +10,7 @@ to get you up and running.
+ You will need the following things properly installed on your computer.
+=20
+ * [Git](https://git-scm.com/)
+-* [Node.js](https://nodejs.org/) *(you need node version <10 due depende=
+ncies..., 2018/05/07)*
++* [Node.js](https://nodejs.org/)
+ * [Yarn](https://yarnpkg.com/)
+ * [Ember CLI](https://ember-cli.com/)
+ * [Google Chrome](https://google.com/chrome/)
+--=20
+2.23.0
diff --git a/libgitmail/src/lib.rs b/libgitmail/src/lib.rs
index 81de51d..53bb6b8 100644
--- a/libgitmail/src/lib.rs
+++ b/libgitmail/src/lib.rs
@@ -10,6 +10,14 @@
 //! structure of a pathset with a way to select specific parts of the
 //! set to export.  Look at the `PatchSet` type for more information.
 
+mod patch;
+pub use patch::*;
+
+pub mod set;
+pub mod tree;
+
+pub type Result<T> = std::result::Result<T, Error>;
+
 /// A mail error type
 #[derive(Debug)]
 pub enum Error {
@@ -17,4 +25,6 @@ pub enum Error {
   FailedParsing,
   /// The provided email is not a valid get-sent mail
   NotAGitMail,
+  /// Generally a git email, but malformed
+  InvalidGitFormat,
 }
diff --git a/libgitmail/src/patch/mod.rs b/libgitmail/src/patch/mod.rs
new file mode 100644
index 0000000..b4b09a4
--- /dev/null
+++ b/libgitmail/src/patch/mod.rs
@@ -0,0 +1,103 @@
+use crate::{Error, Result};
+use std::collections::HashMap;
+
+mod parsers;
+use parsers::get_header;
+pub use parsers::{Segment, Subject};
+
+#[doc(hidden)]
+pub type HeaderMap = HashMap<String, String>;
+
+/// Message ID of an email in a thread
+#[derive(Clone, Debug, Eq, PartialEq)]
+pub struct Id(pub String);
+
+impl Id {
+  pub fn from(headers: &HeaderMap, key: &str) -> Result<Self> {
+    get_header(&headers, key).and_then(|h| match h {
+      Header::Single(h) => Ok(Self(h)),
+      _ => Err(Error::FailedParsing),
+    })
+  }
+}
+
+/// A semi typed header value for mail
+pub(crate) enum Header {
+  /// A single header value
+  Single(String),
+  /// A set of values that was separated by `,`
+  Multi(Vec<String>),
+}
+
+impl Header {
+  fn single(self) -> Result<String> {
+    match self {
+      Self::Single(s) => Ok(s),
+      Self::Multi(_) => Err(Error::FailedParsing),
+    }
+  }
+}
+
+/// A single patch, metadata and email body
+///
+/// This type is constructed from a single email in a thread, and can
+/// then be combined into a [PatchSet](struct.PatchSet.html) via the
+/// [PatchTree](struct.PatchTree.html) builder.
+#[derive(Clone, Debug, Eq, PartialEq)]
+pub struct Patch<'mail> {
+  pub id: Id,
+  pub reply_to: Option<Id>,
+  pub subject: Subject,
+  pub headers: HeaderMap,
+  pub raw: &'mail str,
+}
+
+impl<'mail> Patch<'mail> {
+  #[doc(hidden)]
+  pub fn preprocess(raw: &'mail str) -> HeaderMap {
+    let mail = mailparse::parse_mail(raw.as_bytes())
+      .map_err(|_| Error::FailedParsing)
+      .unwrap();
+    mail
+      .headers
+      .into_iter()
+      .fold(HashMap::new(), |mut acc, header| {
+        let key = header.get_key().unwrap();
+        let val = header.get_value().unwrap();
+        acc.insert(key, val);
+        acc
+      })
+  }
+
+  pub fn new(raw: &'mail str) -> Result<Self> {
+    let mail = mailparse::parse_mail(raw.as_bytes())
+      .map_err(|_| Error::FailedParsing)?;
+    let headers =
+      mail
+        .headers
+        .into_iter()
+        .fold(HashMap::new(), |mut acc, header| {
+          let key = header.get_key().unwrap();
+          let val = header.get_value().unwrap();
+          acc.insert(key, val);
+          acc
+        });
+
+    get_header(&headers, "X-Mailer").and_then(|h| match h {
+      Header::Single(s) if s.contains("git-send-email") => Ok(()),
+      _ => Err(Error::NotAGitMail),
+    })?;
+
+    Ok(Self {
+      id: Id::from(&headers, "Message-Id")?,
+      reply_to: Id::from(&headers, "In-Reply-To")
+        .map(|id| Some(id))
+        .unwrap_or(None),
+      subject: get_header(&headers, "Subject")
+        .and_then(|h| h.single())
+        .and_then(|s| Subject::from(s))?,
+      headers,
+      raw,
+    })
+  }
+}
diff --git a/libgitmail/src/patch/parsers.rs b/libgitmail/src/patch/parsers.rs
new file mode 100644
index 0000000..9779e2b
--- /dev/null
+++ b/libgitmail/src/patch/parsers.rs
@@ -0,0 +1,163 @@
+//! Contains additional parsers that are required for git patch emails
+//!
+//! The parsers are written in nom, and generate
+//! [Patch](../struct.Patch.html) objects, or an Error.  A parser
+//! should neven panic.
+//!
+//! There's also some unit and quick check tests that make sure that
+//! valid git patch emails can be parsed.  If you have emails you
+//! don't mind becoming part of the test set, feel free to email in a
+//! patch!
+
+use crate::{
+  patch::{Header, HeaderMap},
+  Error, Result,
+};
+use nom::{
+  bytes::complete::{is_not, tag, take_till},
+  character::complete::char,
+  combinator::opt,
+  sequence::{delimited, preceded},
+  IResult,
+};
+
+/// The segment data for a patch mail
+#[derive(Clone, Default, Debug, PartialEq, Eq)]
+pub struct Segment {
+  /// The number of a patch in it's series
+  pub this: u8,
+  /// The total number of patches in a seriesThe first
+  pub max: u8,
+}
+
+/// The typed subject line of a patch mail
+///
+/// A message might be: "[PATCH v2 1/5], which is parsed into this
+/// structure to be easier to work with
+#[derive(Clone, Default, Debug, PartialEq, Eq)]
+pub struct Subject {
+  /// The patch revision (e.g. `v2`)
+  pub version: u8,
+  /// Patch segment information (e.g. `1/5`)
+  pub segment: Option<Segment>,
+  /// Subject prefix (e.g. `PATCH`)
+  pub prefix: String,
+  /// The actual patch commit message
+  pub message: String,
+}
+
+impl Subject {
+  /// Create a typed subject from a subject string
+  pub fn from(s: String) -> Result<Self> {
+    Ok(Self::parse(s.as_str()).unwrap().1)
+  }
+
+  /// This is a terrible function that does terrible things
+  fn parse(input: &str) -> IResult<&str, Self> {
+    let input = opt(tag("Re:"))(input).map(|(i, _)| i.trim())?;
+    let (msg, input) = Self::parens(input).map(|(a, b)| (a.trim(), b))?;
+    let (input, prefix) =
+      take_till(|c| c == ' ')(input).map(|(a, b)| (a.trim(), b))?;
+
+    let (input, version) =
+      opt(preceded(tag("v"), take_till(|c| c == ' ')))(input)
+        .map(|(i, v)| (i.trim(), v.map_or(1, |v| v.parse::<u8>().unwrap())))?;
+
+    let (max_str, this_opt) = opt(take_till(|c| c == '/'))(input)?;
+
+    let segment = if max_str.len() > 1 {
+      let max = max_str[1..].parse::<u8>().unwrap_or(1);
+      let this = this_opt.unwrap().parse::<u8>().unwrap();
+      Some(Segment { max, this })
+    } else {
+      None
+    };
+
+    Ok((
+      input,
+      Self {
+        message: msg.into(),
+        prefix: prefix.into(),
+        version,
+        segment,
+      },
+    ))
+  }
+
+  /// Grab text between the [ ... ] section of a subject
+  fn parens(input: &str) -> IResult<&str, &str> {
+    delimited(char('['), is_not("]"), char(']'))(input)
+  }
+}
+
+/// A utility function that can get a header into a semi-typed field
+pub(crate) fn get_header(headers: &HeaderMap, key: &str) -> Result<Header> {
+  headers.get(key).map_or(Err(Error::FailedParsing), |h| {
+    match h.split(",").map(|s| s.trim()).collect::<Vec<_>>() {
+      vec if vec.len() == 0 => Err(Error::FailedParsing),
+      vec if vec.len() == 1 => {
+        Ok(Header::Single(String::from(*vec.get(0).unwrap())))
+      }
+      vec => Ok(Header::Multi(
+        vec.into_iter().map(|s| String::from(s)).collect(),
+      )),
+    }
+  })
+}
+
+#[test]
+fn full_test() {
+  let subject =
+    "Re: [PATCH v2 2/3] docs: fix spelling of \"contributors\"".into();
+  assert_eq!(
+    Subject::from(subject).unwrap(),
+    Subject {
+      message: "docs: fix spelling of \"contributors\"".into(),
+      prefix: "PATCH".into(),
+      version: 2,
+      segment: Some(Segment { max: 3, this: 2 })
+    }
+  );
+}
+
+#[test]
+fn minimal_test() {
+  let subject = "Re: [PATCH] docs: fix spelling of \"contributors\"".into();
+  assert_eq!(
+    Subject::from(subject).unwrap(),
+    Subject {
+      message: "docs: fix spelling of \"contributors\"".into(),
+      prefix: "PATCH".into(),
+      version: 1,
+      segment: None
+    }
+  );
+}
+
+#[test]
+fn version_test() {
+  let subject = "Re: [PATCH v4] docs: fix spelling of \"contributors\"".into();
+  assert_eq!(
+    Subject::from(subject).unwrap(),
+    Subject {
+      message: "docs: fix spelling of \"contributors\"".into(),
+      prefix: "PATCH".into(),
+      version: 4,
+      segment: None
+    }
+  );
+}
+
+#[test]
+fn double_digit_version_test() {
+  let subject = "Re: [PATCH v11] docs: fix spelling of \"contributors\"".into();
+  assert_eq!(
+    Subject::from(subject).unwrap(),
+    Subject {
+      message: "docs: fix spelling of \"contributors\"".into(),
+      prefix: "PATCH".into(),
+      version: 11,
+      segment: None
+    }
+  );
+}
diff --git a/libgitmail/tests/basic.rs b/libgitmail/tests/basic.rs
new file mode 100644
index 0000000..0271784
--- /dev/null
+++ b/libgitmail/tests/basic.rs
@@ -0,0 +1,26 @@
+#![allow(warnings, unused)]
+
+use libgitmail::{Id, Patch, Segment, Subject};
+
+const TEST_MAIL: &'static str = include_str!("../email.txt");
+
+/// Parse an e-mail that was generated by git-send-email in a real projectf
+#[test]
+fn parse_mail() {
+  let headers = Patch::preprocess(TEST_MAIL);
+  assert_eq!(
+    Patch::new(TEST_MAIL).unwrap(),
+    Patch {
+      id: Id("<20191229132404.14579-1-hi@alyssa.is>".into()),
+      headers,
+      raw: TEST_MAIL,
+      reply_to: None,
+      subject: Subject {
+        version: 1,
+        segment: None,
+        prefix: "PATCH".into(),
+        message: "docs: Node.js 10 works fine (now?)".into()
+      }
+    }
+  )
+}