Blossom: Determine mime type and supply extension in Blob Descriptor URL

Cargo.lock

@@ -1,6 +1,6 @@
 # This file is automatically @generated by Cargo.
 # It is not intended for manual editing.
-version = 3
+version = 4
 
 [[package]]
 name = "addr2line"
@@ -278,6 +278,8 @@ dependencies = [
  "hyper-util",
  "lazy_static",
  "log",
+ "mime-sniffer",
+ "new_mime_guess",
  "parking_lot",
  "pocket-db",
  "pocket-types",
@@ -1094,6 +1096,22 @@ dependencies = [
  "autocfg",
 ]
 
+[[package]]
+name = "mime"
+version = "0.3.17"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6877bb514081ee2a7ff5ef9de3281f14a4dd4bceac4c09388074a6b5df8a139a"
+
+[[package]]
+name = "mime-sniffer"
+version = "0.1.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6b8b2a64cd735f1d5f17ff6701ced3cc3c54851f9448caf454cd9c923d812408"
+dependencies = [
+ "mime",
+ "url",
+]
+
 [[package]]
 name = "minimal-lexical"
 version = "0.2.1"
@@ -1137,6 +1155,16 @@ dependencies = [
  "memmap2",
 ]
 
+[[package]]
+name = "new_mime_guess"
+version = "4.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "02a2dfb3559d53e90b709376af1c379462f7fb3085a0177deb73e6ea0d99eff4"
+dependencies = [
+ "mime",
+ "unicase",
+]
+
 [[package]]
 name = "nom"
 version = "7.1.3"
@@ -1931,6 +1959,12 @@ version = "1.17.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "42ff0bf0c66b8238c6f3b578df37d0b7848e55df8577b3f74f92a69acceeb825"
 
+[[package]]
+name = "unicase"
+version = "2.8.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7e51b68083f157f853b6379db119d1c1be0e6e4dec98101079dec41f6f5cf6df"
+
 [[package]]
 name = "unicode-ident"
 version = "1.0.13"

Cargo.toml

@@ -21,6 +21,8 @@ hyper-tungstenite = "0.15"
 hyper-util = "0.1"
 lazy_static = "1.5"
 log = "0.4"
+mime-sniffer = "0.1"
+new_mime_guess = "4.0"
 pocket-types = { git = "https://github.com/mikedilger/pocket", branch = "master" }
 pocket-db = { git = "https://github.com/mikedilger/pocket", branch = "master" }
 parking_lot = "0.12"

src/filestore/mod.rs

@@ -48,7 +48,7 @@ impl FileStore {
         &self,
         data: BoxBody<Bytes, Error>,
         expected_hash: Option<HashOutput>,
-    ) -> Result<(u64, HashOutput), Error> {
+    ) -> Result<(u64, HashOutput, Option<String>), Error> {
         use bitcoin_hashes::sha256;
         use std::io::Write; // for hash_engine.write_all()
 
@@ -78,6 +78,7 @@ impl FileStore {
 
         // Copy the data into the tempfile (hashing and counting as we go)
         let count = tokio::io::copy(&mut inspect_reader, &mut tempfile).await?;
+        drop(tempfile);
 
         // Verify our code was correct
         if count != size {
@@ -101,6 +102,16 @@ impl FileStore {
             }
         }
 
+        // Sniff the mime-type
+        let maybe_mime_string = {
+            use mime_sniffer::MimeTypeSniffer;
+            use tokio::io::AsyncReadExt;
+            let mut readtempfile = File::open(&temppathbuf).await?;
+            let mut buffer: Vec<u8> = vec![0; 128];
+            let _ = readtempfile.read(&mut buffer).await?;
+            buffer.sniff_mime_type().map(|s| s.to_string())
+        };
+
         // Compute the proper path
         let pathbuf = hash.to_pathbuf(&self.base);
 
@@ -109,7 +120,7 @@ impl FileStore {
             // Just clean up
             fs::remove_file(&temppathbuf).await?;
 
-            return Ok((size, hash));
+            return Ok((size, hash, maybe_mime_string));
         }
 
         // Make the parent directory
@@ -118,7 +129,7 @@ impl FileStore {
         // Move the file
         fs::rename(&temppathbuf, &pathbuf).await?;
 
-        Ok((size, hash))
+        Ok((size, hash, maybe_mime_string))
     }
 
     /// Retrieve a file from storage by its HashOutput, streamed to a hyper BoxBoxy

src/web/blossom/mod.rs

@@ -229,7 +229,15 @@ pub async fn handle_upload(
 
             let uri = request.uri().to_owned();
 
-            let (size, hash) = GLOBALS
+            let maybe_content_type = match request.headers().get(http::header::CONTENT_TYPE) {
+                Some(s) => match s.to_str() {
+                    Ok(s) => Some(s.to_owned()),
+                    Err(_) => None,
+                },
+                None => None,
+            };
+
+            let (size, hash, maybe_sniffed_mime_string) = GLOBALS
                 .filestore
                 .get()
                 .unwrap()
@@ -239,11 +247,26 @@ pub async fn handle_upload(
                 )
                 .await?;
 
+            let extension = {
+                let mut mime_string: String = "".to_owned();
+                if let Some(ms) = maybe_content_type {
+                    mime_string = ms.to_owned();
+                } else if let Some(ms) = maybe_sniffed_mime_string {
+                    mime_string = ms.to_owned();
+                }
+
+                if let Some(exts) = new_mime_guess::get_mime_extensions_str(&mime_string) {
+                    exts[0]
+                } else {
+                    "blob"
+                }
+            };
+
             let uri = {
                 let mut parts = GLOBALS.config.read().uri_parts(uri, true)?;
                 parts.path_and_query = Some(http::uri::PathAndQuery::from_maybe_shared(format!(
-                    "/{}",
+                    "/{}.{}",
-                    hash
+                    hash, extension
                 ))?);
                 http::Uri::from_parts(parts)?
             };