feat(paste): make duplicate uploads optional (#7)

Cargo.lock

@@ -980,6 +980,15 @@ version = "0.4.8"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b71991ff56294aa922b450139ee08b3bfc70982c6b2c7562771375cf73542dd4"
 
+[[package]]
+name = "js-sys"
+version = "0.3.55"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7cc9ffccd38c451a86bf13657df244e9c3f37493cce8e5e21e940963777acc84"
+dependencies = [
+ "wasm-bindgen",
+]
+
 [[package]]
 name = "kernel32-sys"
 version = "0.2.2"
@@ -1493,6 +1502,21 @@ dependencies = [
  "quick-error",
 ]
 
+[[package]]
+name = "ring"
+version = "0.16.20"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "3053cf52e236a3ed746dfc745aa9cacf1b791d846bdaf412f60a8d7d6e17c8fc"
+dependencies = [
+ "cc",
+ "libc",
+ "once_cell",
+ "spin",
+ "untrusted",
+ "web-sys",
+ "winapi 0.3.9",
+]
+
 [[package]]
 name = "rustc_version"
 version = "0.2.3"
@@ -1532,6 +1556,7 @@ dependencies = [
  "petname",
  "rand 0.8.4",
  "regex",
+ "ring",
  "serde",
  "serde_regex",
  "url",
@@ -1686,6 +1711,12 @@ dependencies = [
  "winapi 0.3.9",
 ]
 
+[[package]]
+name = "spin"
+version = "0.5.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6e63cff320ae2c57904679ba7cb63280a3dc4613885beafb148ee7bf9aa9042d"
+
 [[package]]
 name = "standback"
 version = "0.2.17"
@@ -2060,6 +2091,12 @@ version = "0.2.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "8ccb82d61f80a663efe1f787a51b16b5a51e3314d6ac365b08639f52387b33f3"
 
+[[package]]
+name = "untrusted"
+version = "0.7.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a156c684c91ea7d62626509bce3cb4e1d9ed5c4d978f7b4352658f96a4c26b4a"
+
 [[package]]
 name = "url"
 version = "2.2.2"
@@ -2136,9 +2173,9 @@ checksum = "fd6fbd9a79829dd1ad0cc20627bf1ed606756a7f77edff7b66b7064f9cb327c6"
 
 [[package]]
 name = "wasm-bindgen"
-version = "0.2.76"
+version = "0.2.78"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8ce9b1b516211d33767048e5d47fa2a381ed8b76fc48d2ce4aa39877f9f183e0"
+checksum = "632f73e236b219150ea279196e54e610f5dbafa5d61786303d4da54f84e47fce"
 dependencies = [
  "cfg-if 1.0.0",
  "wasm-bindgen-macro",
@@ -2146,9 +2183,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-backend"
-version = "0.2.76"
+version = "0.2.78"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "cfe8dc78e2326ba5f845f4b5bf548401604fa20b1dd1d365fb73b6c1d6364041"
+checksum = "a317bf8f9fba2476b4b2c85ef4c4af8ff39c3c7f0cdfeed4f82c34a880aa837b"
 dependencies = [
  "bumpalo",
  "lazy_static",
@@ -2161,9 +2198,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-macro"
-version = "0.2.76"
+version = "0.2.78"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "44468aa53335841d9d6b6c023eaab07c0cd4bddbcfdee3e2bb1e8d2cb8069fef"
+checksum = "d56146e7c495528bf6587663bea13a8eb588d39b36b679d83972e1a2dbbdacf9"
 dependencies = [
  "quote",
  "wasm-bindgen-macro-support",
@@ -2171,9 +2208,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-macro-support"
-version = "0.2.76"
+version = "0.2.78"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0195807922713af1e67dc66132c7328206ed9766af3858164fb583eedc25fbad"
+checksum = "7803e0eea25835f8abdc585cd3021b3deb11543c6fe226dcd30b228857c5c5ab"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -2184,9 +2221,19 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-shared"
-version = "0.2.76"
+version = "0.2.78"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "acdb075a845574a1fa5f09fd77e43f7747599301ea3417a9fbffdeedfc1f4a29"
+checksum = "0237232789cf037d5480773fe568aac745bfe2afbc11a863e97901780a6b47cc"
+
+[[package]]
+name = "web-sys"
+version = "0.3.55"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "38eb105f1c59d9eaa6b5cdc92b859d85b926e82cb2e0945cd0c9259faa6fe9fb"
+dependencies = [
+ "js-sys",
+ "wasm-bindgen",
+]
 
 [[package]]
 name = "widestring"

Cargo.toml

@@ -29,6 +29,7 @@ regex = "1.5.4"
 serde_regex = "1.1.0"
 humantime = "2.1.0"
 glob = "0.3.0"
+ring = "0.16.20"
 
 [dependencies.config]
 version = "0.11.0"

README.md

@@ -23,6 +23,7 @@ some text
   - supports one shot links (can only be viewed once)
   - guesses MIME types
     - supports overriding and blacklisting
+  - no duplicate uploads (optional)
 - Single binary
   - [binary releases](https://github.com/orhun/rustypaste/releases)
 - Easy to deploy

config.toml

@@ -22,3 +22,4 @@ mime_blacklist = [
     "application/java-archive",
     "application/java-vm"
 ]
+duplicate_files = false

src/config.rs

@@ -37,6 +37,8 @@ pub struct PasteConfig {
     pub mime_override: Vec<MimeMatcher>,
     /// Media type blacklist.
     pub mime_blacklist: Vec<String>,
+    /// Allow duplicate uploads
+    pub duplicate_files: Option<bool>,
 }
 
 impl Config {

src/file.rs

@@ -0,0 +1,75 @@
+use crate::util;
+use actix_web::{error, Error as ActixError};
+use glob::glob;
+use std::convert::TryFrom;
+use std::fs::File as OsFile;
+use std::path::{Path, PathBuf};
+
+/// [`PathBuf`] wrapper for storing checksums.
+#[derive(Debug)]
+pub struct File {
+    /// Path of the file.
+    pub path: PathBuf,
+    /// SHA256 checksum.
+    pub sha256sum: String,
+}
+
+/// Directory that contains [`File`]s.
+pub struct Directory {
+    /// Files in the directory.
+    pub files: Vec<File>,
+}
+
+impl<'a> TryFrom<&'a Path> for Directory {
+    type Error = ActixError;
+    fn try_from(directory: &'a Path) -> Result<Self, Self::Error> {
+        let files = glob(directory.join("**").join("*").to_str().ok_or_else(|| {
+            error::ErrorInternalServerError("directory contains invalid characters")
+        })?)
+        .map_err(error::ErrorInternalServerError)?
+        .filter_map(Result::ok)
+        .filter(|path| !path.is_dir())
+        .filter_map(|path| match OsFile::open(&path) {
+            Ok(file) => Some((path, file)),
+            _ => None,
+        })
+        .filter_map(|(path, file)| match util::sha256_digest(file) {
+            Ok(sha256sum) => Some(File { path, sha256sum }),
+            _ => None,
+        })
+        .collect();
+        Ok(Self { files })
+    }
+}
+
+impl Directory {
+    /// Returns the file that matches the given checksum.
+    pub fn get_file<S: AsRef<str>>(self, sha256sum: S) -> Option<File> {
+        self.files
+            .into_iter()
+            .find(|file| file.sha256sum == sha256sum.as_ref())
+    }
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_file_checksum() -> Result<(), ActixError> {
+        assert_eq!(
+            "rustypaste_logo.png",
+            Directory::try_from(
+                PathBuf::from(env!("CARGO_MANIFEST_DIR"))
+                    .join("img")
+                    .as_path()
+            )?
+            .get_file("2073f6f567dcba3b468c568d29cf8ed2e9d3f0f7305b9ab1b5a22861f5922e61")
+            .unwrap()
+            .path
+            .file_name()
+            .unwrap()
+        );
+        Ok(())
+    }
+}

src/lib.rs

@@ -19,6 +19,9 @@ pub mod auth;
 /// Storage handler.
 pub mod paste;
 
+/// File metadata handler.
+pub mod file;
+
 /// Media type handler.
 pub mod mime;
 

src/server.rs

@@ -1,5 +1,6 @@
 use crate::auth;
 use crate::config::Config;
+use crate::file::Directory;
 use crate::header::{self, ContentDisposition};
 use crate::mime;
 use crate::paste::{Paste, PasteType};
@@ -101,6 +102,24 @@ async fn upload(
                 log::warn!("{} sent zero bytes", host);
                 return Err(error::ErrorBadRequest("invalid file size"));
             }
+            if paste_type != PasteType::Oneshot && !config.paste.duplicate_files.unwrap_or(true) {
+                let bytes_checksum = util::sha256_digest(&*bytes)?;
+                if let Some(file) = Directory::try_from(config.server.upload_path.as_path())?
+                    .get_file(bytes_checksum)
+                {
+                    urls.push(format!(
+                        "{}://{}/{}\n",
+                        connection.scheme(),
+                        connection.host(),
+                        file.path
+                            .file_name()
+                            .map(|v| v.to_string_lossy())
+                            .unwrap_or_default()
+                            .to_string()
+                    ));
+                    continue;
+                }
+            }
             let bytes_unit = Byte::from_bytes(bytes.len() as u128).get_appropriate_unit(false);
             let paste = Paste {
                 data: bytes.to_vec(),

src/util.rs

@@ -1,5 +1,7 @@
 use actix_web::{error, Error as ActixError};
 use glob::glob;
+use ring::digest::{Context, SHA256};
+use std::io::{BufReader, Read};
 use std::path::PathBuf;
 use std::time::Duration;
 use std::time::{SystemTime, UNIX_EPOCH};
@@ -41,6 +43,29 @@ pub fn glob_match_file(mut path: PathBuf) -> Result<PathBuf, ActixError> {
     Ok(path)
 }
 
+/// Returns the SHA256 digest of the given input.
+pub fn sha256_digest<R: Read>(input: R) -> Result<String, ActixError> {
+    let mut reader = BufReader::new(input);
+    let mut context = Context::new(&SHA256);
+    let mut buffer = [0; 1024];
+    loop {
+        let bytes_read = reader.read(&mut buffer)?;
+        if bytes_read != 0 {
+            context.update(&buffer[..bytes_read]);
+        } else {
+            break;
+        }
+    }
+    Ok(context
+        .finish()
+        .as_ref()
+        .iter()
+        .collect::<Vec<&u8>>()
+        .iter()
+        .map(|byte| format!("{:02x}", byte))
+        .collect::<String>())
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -72,4 +97,17 @@ mod tests {
 
         Ok(())
     }
+
+    #[test]
+    fn test_sha256sum() -> Result<(), ActixError> {
+        assert_eq!(
+            "9f86d081884c7d659a2feaa0c55ad015a3bf4f1b2b0b822cd15d6c15b0f00a08",
+            sha256_digest(String::from("test").as_bytes())?
+        );
+        assert_eq!(
+            "2fc36f72540bb9145e95e67c41dccdc440c95173257032e32e111ebd7b6df960",
+            sha256_digest(env!("CARGO_PKG_NAME").as_bytes())?
+        );
+        Ok(())
+    }
 }