From ce7fe5f019815aaebf0111d9c52e27bcb35ce84d Mon Sep 17 00:00:00 2001
From: Nathan Fisher <nfisher.sr@gmail.com>
Date: Fri, 7 Jul 2023 19:20:21 -0400
Subject: [PATCH] Add "magic" to file start

---
 Format.md     |  9 +++++++--
 src/error.rs  |  2 ++
 src/lib.rs    |  6 +++++-
 src/stream.rs | 26 ++++++++++++++++----------
 4 files changed, 30 insertions(+), 13 deletions(-)

diff --git a/Format.md b/Format.md
index 9e0fdc7..fcb51f4 100644
--- a/Format.md
+++ b/Format.md
@@ -68,10 +68,15 @@ by 8, the third byte by 16 and the fourth byte by 24, and the bits combined
 into a single 32-bit uint. This is dramatically more efficient than storing
 those numbers as ascii text, as is done by Tar.
 
+The first 7 bytes of a haggis file make up the "magic" identifier, consisting
+of the string "\x89haggis". To be clear, that is the integer `0x89` plus
+"haggis". The first file header immediately follows this string.
+
 | **bytes** | **meaning** |
 | ----- | ------- |
-| 0-8   | The length of the filename (64 bit unsigned int) |
-| 8 to the length specified above | The bytes making up the filename |
+| 0-6 | "\x89haggis" - the haggis "magic" identifier |
+| 7-15 | The **length** of the filename (64 bit unsigned int) |
+| the next **length** bytes | The bytes making up the filename |
 | the next 4 bytes | The files Unix permissions mode (32 bit unsigned int) |
 | the next 4 bytes | The uid of the file's owner (32 bit unsigned int) |
 | the next 4 bytes | the gid of the file's owner (32 bit unsigned int) |
diff --git a/src/error.rs b/src/error.rs
index 64e2d71..f753cca 100644
--- a/src/error.rs
+++ b/src/error.rs
@@ -10,6 +10,7 @@ pub enum Error {
     Slice(TryFromSliceError),
     InvalidChecksum,
     InvalidAlgorithm,
+    InvalidMagic,
     MissingData,
     MutexError,
     NulError,
@@ -28,6 +29,7 @@ impl fmt::Display for Error {
             Self::Io(e) => write!(f, "{e}"),
             Self::InvalidAlgorithm => write!(f, "invalid algorithm"),
             Self::InvalidChecksum => write!(f, "invalid checksum"),
+            Self::InvalidMagic => write!(f, "invalid magic"),
             Self::MissingData => write!(f, "missing data"),
             Self::MutexError => write!(f, "mutex error"),
             Self::NulError => write!(f, "nul error"),
diff --git a/src/lib.rs b/src/lib.rs
index 790b2a1..9b6f934 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -30,6 +30,8 @@ pub use {
     stream::Stream,
 };
 
+pub static MAGIC: [u8; 7] = [0x89, b'h', b'a', b'g', b'g', b'i', b's' ];
+
 /// Creates a haggis archive from a list of files
 pub fn create_archive(path: &str, files: Vec<String>, algorithm: Algorithm) -> Result<(), Error> {
     let fd = fs::OpenOptions::new()
@@ -48,6 +50,7 @@ pub fn stream_archive<W: Write>(
     algorithm: Algorithm,
 ) -> Result<(), Error> {
     let links = Mutex::new(HashMap::new());
+    writer.write_all(&MAGIC)?;
     for f in &files {
         let node = Node::from_path(f, algorithm, &links)?;
         node.write(&mut writer)?;
@@ -82,11 +85,12 @@ pub fn par_create_archive(
 /// Streams a Haggis archive from a list of files, processing each file in parallel
 #[cfg(feature = "parallel")]
 pub fn par_stream_archive<W: Write + Send>(
-    writer: W,
+    mut writer: W,
     files: Vec<String>,
     algorithm: Algorithm,
     sender: &Sender<Message>,
 ) -> Result<(), Error> {
+    writer.write_all(&MAGIC)?;
     let links = Mutex::new(HashMap::<u64, String>::new());
     let writer = Mutex::new(writer);
     let s = sender.clone();
diff --git a/src/stream.rs b/src/stream.rs
index e8f95d1..8a26769 100644
--- a/src/stream.rs
+++ b/src/stream.rs
@@ -1,3 +1,5 @@
+use crate::MAGIC;
+
 use {
     crate::{Error, Node},
     std::{
@@ -14,17 +16,11 @@ use {
 /// An iterator over a series of archive `Node`'s. This struct is generic over any
 /// type which implements `Read`, such as a file or a network stream.
 #[derive(Debug)]
-pub struct Stream<T: Read + Send> {
-    reader: T,
+pub struct Stream<R: Read + Send> {
+    reader: R,
 }
 
-impl<T: Read + Send> From<T> for Stream<T> {
-    fn from(value: T) -> Self {
-        Self { reader: value }
-    }
-}
-
-impl<T: Read + Send> Iterator for Stream<T> {
+impl<R: Read + Send> Iterator for Stream<R> {
     type Item = Result<Node, Error>;
 
     fn next(&mut self) -> Option<Self::Item> {
@@ -50,7 +46,17 @@ pub enum Message {
     Eof,
 }
 
-impl<T: Read + Send> Stream<T> {
+impl<R: Read + Send> Stream<R> {
+    pub fn new(mut reader: R) -> Result<Self, Error> {
+        let mut buf = [0; 7];
+        reader.read_exact(&mut buf)?;
+        if buf == MAGIC {
+            Ok(Self { reader })
+        } else {
+            Err(Error::InvalidMagic)
+        }
+    }
+
     pub fn extract(&mut self, prefix: Option<&str>) -> Result<(), Error> {
         for node in self {
             node?.extract(prefix)?;