support using either infer or xdg_mime for mime detection

i guess we can build for windows now

Cargo.lock

@@ -174,6 +174,7 @@ dependencies = [
  "cached",
  "clap",
  "env_logger",
+ "infer",
  "log",
  "mime_guess",
  "once_cell",
@@ -235,6 +236,12 @@ dependencies = [
  "hashbrown",
 ]
 
+[[package]]
+name = "infer"
+version = "0.3.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8906512588cd815b8f759fd0ac11de2a84c985c0f792f70df611e9325c270c1f"
+
 [[package]]
 name = "lazy_static"
 version = "1.4.0"

Cargo.toml

@@ -10,20 +10,23 @@ rust-version = "1.43.0" # cached requires 1.42.0
 #license-file = "LICENSE"
 
 [features]
-default = ["mini-buffer", "multi-threaded"]
+default = ["mini-buffer", "multi-threaded", "infer-backend"]
 mini-buffer = []
 multi-threaded = []
+infer-backend = ["infer"]
+xdg-mime-backend = ["xdg-mime"]
 
 [dependencies]
 walkdir = "2.3.1"
 log = "0.4.14"
 smartstring = "0.2.6"
-# use git version while waiting on a release incorporating https://github.com/ebassi/xdg-mime-rs/commit/de5a6dd
-xdg-mime = {git = "https://github.com/ebassi/xdg-mime-rs", version = "0.3", rev = "de5a6dd"}
 mime_guess = "2.0.3"
 rayon = "1.5.0"
 snailquote = "0.3.0"
 once_cell = "1.5.2"
+infer = { version = "0.3.4", optional = true }
+# use git version while waiting on a release incorporating https://github.com/ebassi/xdg-mime-rs/commit/de5a6dd
+xdg-mime = {git = "https://github.com/ebassi/xdg-mime-rs", version = "0.3", rev = "de5a6dd", optional = true }
 
 [dependencies.clap]
 version = "3.0.0-beta.2"

src/inspectors.rs

@@ -6,7 +6,8 @@ use std::path::Path;
 use cached::cached;
 use mime_guess::Mime;
 use smartstring::alias::String;
-use xdg_mime::SharedMimeInfo;
+
+use crate::mimedb::MimeDb;
 
 // use log::{debug, warn};
 
@@ -20,7 +21,7 @@ use xdg_mime::SharedMimeInfo;
 // at least 265 bytes to identify a tar file.
 const BUF_SIZE: usize = 512;
 
-pub fn mime_type(db: &SharedMimeInfo, path: &Path) -> io::Result<Option<Mime>, > {
+pub fn mime_type<T: MimeDb>(db: &T, path: &Path) -> io::Result<Option<Mime>, > {
 	// attempt to read up to the BUF_SIZE bytes of the file
 	let mut buffer = [0; 64];
 	let mut file = File::open(path)?;
@@ -31,7 +32,7 @@ pub fn mime_type(db: &SharedMimeInfo, path: &Path) -> io::Result<Option<Mime>, >
 		file.read(&mut buffer)?;
 
 
-	let r = db.get_mime_type_for_data(&buffer).map(|m| m.0);
+	let r = db.get_type(&buffer);
 	if r.is_some() {
 		return Ok(r);
 	}
@@ -40,7 +41,7 @@ pub fn mime_type(db: &SharedMimeInfo, path: &Path) -> io::Result<Option<Mime>, >
 	file.seek(SeekFrom::Start(0))?;
 	file.read(&mut buffer)?;
 	// warn!("dang");
-	Ok(db.get_mime_type_for_data(&buffer).map(|m| m.0))
+	Ok(db.get_type(&buffer))
 }
 
 // TODO: avoid cloning mime if possible, although i don't really see how it would be - maybe instead of passing the mime

src/main.rs

@@ -26,6 +26,7 @@ use walkdir::{DirEntry, WalkDir};
 
 use crate::findings::Findings;
 use crate::formats::{Format, Script};
+use crate::mimedb::MimeDb;
 use crate::parameters::OutputFormat;
 use crate::scanerror::ScanError;
 
@@ -34,11 +35,15 @@ mod inspectors;
 mod formats;
 mod scanerror;
 mod findings;
+mod mimedb;
 
-static MIMEDB: OnceCell<xdg_mime::SharedMimeInfo> = OnceCell::new();
+#[cfg(feature = "infer-backend")]
+	static MIMEDB: OnceCell<mimedb::InferDb> = OnceCell::new();
 
-// TODO: test if this actually works on a windows machine - not there's much of a point right now, considering
+#[cfg(feature = "xdg-mime-backend")]
-// xdg-mime-rs doesn't support windows
+	static MIMEDB: OnceCell<mimedb::XdgDb> = OnceCell::new();
+
+// TODO: test if this actually works on a windows machine
 #[cfg(windows)]
 fn is_hidden(entry: &DirEntry) -> bool {
 	use std::os::windows::prelude::*;
@@ -154,7 +159,12 @@ fn main() {
 		// .target(env_logger::Target::Stdout) // log to stdout rather than stderr
 		.init();
 
-	MIMEDB.set(xdg_mime::SharedMimeInfo::new()).or(Err("Failed to initialise MIMEDB")).unwrap();
+	#[cfg(feature = "infer-backend")]
+		MIMEDB.set(mimedb::InferDb::init()).or(Err("Failed to initialise MIMEDB")).unwrap();
+
+	#[cfg(feature = "xdg-mime-backend")]
+		MIMEDB.set(mimedb::XdgDb::init()).or(Err("Failed to initialise MIMEDB")).unwrap();
+
 	debug!("Iterating directory: {:?}", args.dirs);
 
 	let stepper = WalkDir::new(&args.dirs).into_iter();

src/mimedb.rs

@@ -0,0 +1,73 @@
+use std::str::FromStr;
+
+use mime_guess::Mime;
+
+pub trait MimeDb {
+	fn init() -> Self;
+	fn get_type(&self, data: &[u8]) -> Option<Mime>;
+}
+
+#[cfg(feature = "infer-backend")]
+pub struct InferDb {
+	db: infer::Infer
+}
+
+#[cfg(feature = "infer-backend")]
+impl MimeDb for InferDb {
+	fn init() -> Self {
+		let mut info = infer::Infer::new();
+		// add a random file type just to make sure adding works and such
+		info.add(
+			"image/jpeg2000",
+			".jp2",
+			|buf| {
+				buf.len() > 23
+				&& buf[0] == 0x00
+				&& buf[1] == 0x00
+				&& buf[2] == 0x00
+				&& buf[3] == 0x0C
+				&& buf[4] == 0x6A
+				&& buf[5] == 0x50
+				&& buf[6] == 0x20
+				&& buf[7] == 0x20
+				&& buf[8] == 0x0D
+				&& buf[9] == 0x0A
+				&& buf[10] == 0x87
+				&& buf[11] == 0x0A
+				&& buf[20] == 0x6A
+				&& buf[21] == 0x70
+				&& buf[22] == 0x32
+				&& buf[23] == 0x20
+			}
+		);
+
+		// unmut
+		let info = info;
+
+		Self { db: info }
+	}
+
+	fn get_type(&self, data: &[u8]) -> Option<Mime> {
+		self.db
+			.get(data)
+			.map(|f| Mime::from_str(f.mime_type()).unwrap())
+	}
+}
+
+#[cfg(feature = "xdg-mime-backend")]
+pub struct XdgDb {
+	db: xdg_mime::SharedMimeInfo
+}
+
+#[cfg(feature = "xdg-mime-backend")]
+impl MimeDb for XdgDb {
+	fn init() -> Self {
+		Self {
+			db: xdg_mime::SharedMimeInfo::new()
+		}
+	}
+
+	fn get_type(&self, data: &[u8]) -> Option<Mime> {
+		self.db.get_mime_type_for_data(&data).map(|m| m.0)
+	}
+}