initial s3 support

probably incomplete

Cargo.lock

@@ -115,6 +115,12 @@ dependencies = [
  "const-random",
 ]
 
+[[package]]
+name = "ahash"
+version = "0.4.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "739f4a8db6605981345c5654f3a85b056ce52f37a39d34da03f25bf2151ea16e"
+
 [[package]]
 name = "ahash"
 version = "0.7.6"
@@ -166,6 +172,12 @@ dependencies = [
  "winapi 0.3.9",
 ]
 
+[[package]]
+name = "anyhow"
+version = "1.0.71"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9c7d0618f0e0b7e8ff11427422b64564d5fb0be1940354bfe2e0529b18a9d9b8"
+
 [[package]]
 name = "arc-swap"
 version = "1.6.0"
@@ -227,7 +239,7 @@ dependencies = [
  "derive_builder",
  "diligent-date-parser",
  "never",
- "quick-xml",
+ "quick-xml 0.27.1",
 ]
 
 [[package]]
@@ -241,6 +253,22 @@ dependencies = [
  "winapi 0.3.9",
 ]
 
+[[package]]
+name = "attohttpc"
+version = "0.18.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e69e13a99a7e6e070bb114f7ff381e58c7ccc188630121fc4c2fe4bcf24cd072"
+dependencies = [
+ "http 0.2.8",
+ "log 0.4.17",
+ "native-tls",
+ "openssl",
+ "serde 1.0.152",
+ "serde_json",
+ "url 2.3.1",
+ "wildmatch",
+]
+
 [[package]]
 name = "atty"
 version = "0.2.14"
@@ -267,6 +295,31 @@ version = "1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
 
+[[package]]
+name = "aws-creds"
+version = "0.27.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "460a75eac8f3cb7683e0a9a588a83c3ff039331ea7bfbfbfcecf1dacab276e11"
+dependencies = [
+ "anyhow",
+ "attohttpc",
+ "dirs",
+ "rust-ini 0.17.0",
+ "serde 1.0.152",
+ "serde-xml-rs",
+ "serde_derive",
+ "url 2.3.1",
+]
+
+[[package]]
+name = "aws-region"
+version = "0.23.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "10110ddbd800fb47e6bef95e88fc13495795d252f585272a4fa3ac4f5b2e0a4d"
+dependencies = [
+ "anyhow",
+]
+
 [[package]]
 name = "backtrace"
 version = "0.1.8"
@@ -389,6 +442,16 @@ dependencies = [
  "generic-array",
 ]
 
+[[package]]
+name = "block_on_proc"
+version = "0.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b872f3528eeeb4370ee73b51194dc1cd93680c2d0eb6c7a223889038d2c1a167"
+dependencies = [
+ "quote 1.0.23",
+ "syn 1.0.107",
+]
+
 [[package]]
 name = "blowfish"
 version = "0.9.1"
@@ -578,7 +641,7 @@ checksum = "19b076e143e1d9538dde65da30f8481c2a6c44040edb8e02b9bf1351edb92ce3"
 dependencies = [
  "lazy_static",
  "nom 5.1.2",
- "rust-ini",
+ "rust-ini 0.13.0",
  "serde 1.0.152",
  "serde-hjson",
  "serde_json",
@@ -636,7 +699,7 @@ dependencies = [
  "aes-gcm",
  "base64 0.13.1",
  "hkdf",
- "hmac",
+ "hmac 0.10.1",
  "percent-encoding 2.2.0",
  "rand 0.8.5",
  "sha2",
@@ -876,6 +939,16 @@ dependencies = [
  "subtle",
 ]
 
+[[package]]
+name = "crypto-mac"
+version = "0.11.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b1d1a86f49236c215f271d40892d5fc950490551400b02ef360692c29815c714"
+dependencies = [
+ "generic-array",
+ "subtle",
+]
+
 [[package]]
 name = "ctr"
 version = "0.6.0"
@@ -1150,6 +1223,35 @@ dependencies = [
  "chrono",
 ]
 
+[[package]]
+name = "dirs"
+version = "4.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ca3aa72a6f96ea37bbc5aa912f6788242832f75369bdfdadcb0e38423f100059"
+dependencies = [
+ "dirs-sys",
+]
+
+[[package]]
+name = "dirs-sys"
+version = "0.3.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1b1d1d91c932ef41c0f2663aa8b0ca0342d444d842c06914aa0a7e352d0bada6"
+dependencies = [
+ "libc",
+ "redox_users",
+ "winapi 0.3.9",
+]
+
+[[package]]
+name = "dlv-list"
+version = "0.2.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "68df3f2b690c1b86e65ef7830956aededf3cb0a16f898f79b9a6f421a7b6211b"
+dependencies = [
+ "rand 0.8.5",
+]
+
 [[package]]
 name = "dotenv"
 version = "0.15.0"
@@ -1745,6 +1847,15 @@ dependencies = [
  "tracing",
 ]
 
+[[package]]
+name = "hashbrown"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d7afe4a420e3fe79967a00898cc1f4db7c8a49a9333a29f8a4bd76a253d5cd04"
+dependencies = [
+ "ahash 0.4.7",
+]
+
 [[package]]
 name = "hashbrown"
 version = "0.11.2"
@@ -1794,7 +1905,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "51ab2f639c231793c5f6114bdb9bbe50a7dbbfcd7c7c6bd8475dec2d991e964f"
 dependencies = [
  "digest",
- "hmac",
+ "hmac 0.10.1",
 ]
 
 [[package]]
@@ -1803,7 +1914,17 @@ version = "0.10.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c1441c6b1e930e2817404b5046f1f989899143a12bf92de603b69f4e0aee1e15"
 dependencies = [
- "crypto-mac",
+ "crypto-mac 0.10.1",
+ "digest",
+]
+
+[[package]]
+name = "hmac"
+version = "0.11.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2a2a2320eb7ec0ebe8da8f744d7812d9fc4cb4d09344ac01898dbcb6a20ae69b"
+dependencies = [
+ "crypto-mac 0.11.1",
  "digest",
 ]
 
@@ -2549,6 +2670,17 @@ version = "0.1.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a3e378b66a060d48947b590737b30a1be76706c8dd7b8ba0f2fe3989c68a853f"
 
+[[package]]
+name = "maybe-async"
+version = "0.2.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0f1b8c13cb1f814b634a96b2c725449fe7ed464a7b8781de8688be5ffbd3f305"
+dependencies = [
+ "proc-macro2 1.0.49",
+ "quote 1.0.23",
+ "syn 1.0.107",
+]
+
 [[package]]
 name = "maybe-uninit"
 version = "2.0.0"
@@ -2641,6 +2773,15 @@ dependencies = [
  "unicase 2.6.0",
 ]
 
+[[package]]
+name = "minidom"
+version = "0.13.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "332592c2149fc7dd40a64fc9ef6f0d65607284b474cef9817d1fc8c7e7b3608e"
+dependencies = [
+ "quick-xml 0.20.0",
+]
+
 [[package]]
 name = "minimal-lexical"
 version = "0.2.1"
@@ -3063,6 +3204,16 @@ dependencies = [
  "vcpkg",
 ]
 
+[[package]]
+name = "ordered-multimap"
+version = "0.3.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1c672c7ad9ec066e428c00eb917124a06f08db19e2584de982cc34b1f4c12485"
+dependencies = [
+ "dlv-list",
+ "hashbrown 0.9.1",
+]
+
 [[package]]
 name = "overload"
 version = "0.1.1"
@@ -3294,6 +3445,7 @@ dependencies = [
  "rocket_i18n",
  "rsass",
  "ructe",
+ "rust-s3",
  "scheduled-thread-pool",
  "serde 1.0.152",
  "serde_json",
@@ -3409,6 +3561,7 @@ dependencies = [
  "riker",
  "rocket",
  "rocket_i18n",
+ "rust-s3",
  "scheduled-thread-pool",
  "serde 1.0.152",
  "serde_derive",
@@ -3552,6 +3705,15 @@ version = "1.2.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "a1d01941d82fa2ab50be1e79e6714289dd7cde78eba4c074bc5a4374f650dfe0"
 
+[[package]]
+name = "quick-xml"
+version = "0.20.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "26aab6b48e2590e4a64d1ed808749ba06257882b461d01ca71baeb747074a6dd"
+dependencies = [
+ "memchr",
+]
+
 [[package]]
 name = "quick-xml"
 version = "0.27.1"
@@ -3836,6 +3998,17 @@ dependencies = [
  "bitflags 1.3.2",
 ]
 
+[[package]]
+name = "redox_users"
+version = "0.4.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b033d837a7cf162d7993aded9304e30a83213c648b6e389db233191f891e5c2b"
+dependencies = [
+ "getrandom 0.2.8",
+ "redox_syscall 0.2.16",
+ "thiserror",
+]
+
 [[package]]
 name = "regex"
 version = "1.7.0"
@@ -3967,6 +4140,7 @@ dependencies = [
  "tokio 1.24.1",
  "tokio-native-tls",
  "tokio-socks",
+ "tokio-util 0.7.4",
  "tower-service",
  "url 2.3.1",
  "wasm-bindgen",
@@ -4159,6 +4333,48 @@ version = "0.13.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3e52c148ef37f8c375d49d5a73aa70713125b7f19095948a923f80afdeb22ec2"
 
+[[package]]
+name = "rust-ini"
+version = "0.17.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "63471c4aa97a1cf8332a5f97709a79a4234698de6a1f5087faf66f2dae810e22"
+dependencies = [
+ "cfg-if 1.0.0",
+ "ordered-multimap",
+]
+
+[[package]]
+name = "rust-s3"
+version = "0.29.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6a4e82923ed07143871571852a390742200607e5058ce633afec89752f9c3f82"
+dependencies = [
+ "anyhow",
+ "async-trait",
+ "aws-creds",
+ "aws-region",
+ "base64 0.13.1",
+ "block_on_proc",
+ "cfg-if 1.0.0",
+ "hex",
+ "hmac 0.11.0",
+ "http 0.2.8",
+ "log 0.4.17",
+ "maybe-async",
+ "md5",
+ "minidom",
+ "percent-encoding 2.2.0",
+ "reqwest 0.11.13",
+ "serde 1.0.152",
+ "serde-xml-rs",
+ "serde_derive",
+ "sha2",
+ "time 0.3.17",
+ "tokio 1.24.1",
+ "tokio-stream",
+ "url 2.3.1",
+]
+
 [[package]]
 name = "rust-stemmers"
 version = "1.2.0"
@@ -4302,6 +4518,18 @@ dependencies = [
  "serde 0.8.23",
 ]
 
+[[package]]
+name = "serde-xml-rs"
+version = "0.5.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "65162e9059be2f6a3421ebbb4fef3e74b7d9e7c60c50a0e292c6239f19f1edfa"
+dependencies = [
+ "log 0.4.17",
+ "serde 1.0.152",
+ "thiserror",
+ "xml-rs",
+]
+
 [[package]]
 name = "serde_derive"
 version = "1.0.152"
@@ -5679,6 +5907,12 @@ dependencies = [
  "once_cell",
 ]
 
+[[package]]
+name = "wildmatch"
+version = "2.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ee583bdc5ff1cf9db20e9db5bb3ff4c3089a8f6b8b31aff265c9aba85812db86"
+
 [[package]]
 name = "winapi"
 version = "0.2.8"

Cargo.toml

@@ -19,6 +19,7 @@ rocket = "0.4.11"
 rocket_contrib = { version = "0.4.11", features = ["json"] }
 rocket_i18n = "0.4.1"
 scheduled-thread-pool = "0.2.6"
+rust-s3 = { version = "0.29.0", no-default-features = true, features = ["blocking"], optional = true}
 serde = "1.0.137"
 serde_json = "1.0.81"
 shrinkwraprs = "0.3.0"
@@ -74,6 +75,7 @@ sqlite = ["plume-models/sqlite", "diesel/sqlite"]
 debug-mailer = []
 test = []
 search-lindera = ["plume-models/search-lindera"]
+s3 = ["rust-s3"]
 
 [workspace]
 members = ["plume-api", "plume-cli", "plume-models", "plume-common", "plume-front", "plume-macro"]

plume-models/Cargo.toml

@@ -18,6 +18,7 @@ rocket_i18n = "0.4.1"
 reqwest = "0.11.11"
 scheduled-thread-pool = "0.2.6"
 serde = "1.0.137"
+rust-s3 = { version = "0.29.0", no-default-features = true, features = ["blocking"] }
 serde_derive = "1.0"
 serde_json = "1.0.81"
 tantivy = "0.13.3"

plume-models/src/config.rs

@@ -6,6 +6,10 @@ use rocket::Config as RocketConfig;
 use std::collections::HashSet;
 use std::env::{self, var};
 
+use s3::{Bucket, Region};
+use s3::creds::Credentials;
+
+
 #[cfg(not(test))]
 const DB_NAME: &str = "plume";
 #[cfg(test)]
@@ -27,13 +31,23 @@ pub struct Config {
     pub mail: Option<MailConfig>,
     pub ldap: Option<LdapConfig>,
     pub proxy: Option<ProxyConfig>,
+    pub s3: Option<S3Config>,
 }
+
 impl Config {
     pub fn proxy(&self) -> Option<&reqwest::Proxy> {
         self.proxy.as_ref().map(|p| &p.proxy)
     }
 }
 
+fn string_to_bool(val: &str, name: &str) -> bool {
+    match val {
+        "1" | "true" | "TRUE" => true,
+        "0" | "false" | "FALSE" => false,
+        _ => panic!("Invalid configuration: {} is not boolean", name),
+    }
+}
+
 #[derive(Debug, Clone)]
 pub enum InvalidRocketConfig {
     Env,
@@ -288,11 +302,7 @@ fn get_ldap_config() -> Option<LdapConfig> {
     match (addr, base_dn) {
         (Some(addr), Some(base_dn)) => {
             let tls = var("LDAP_TLS").unwrap_or_else(|_| "false".to_owned());
-            let tls = match tls.as_ref() {
+            let tls = string_to_bool(&tls, "LDAP_TLS");
-                "1" | "true" | "TRUE" => true,
-                "0" | "false" | "FALSE" => false,
-                _ => panic!("Invalid LDAP configuration : tls"),
-            };
             let user_name_attr = var("LDAP_USER_NAME_ATTR").unwrap_or_else(|_| "cn".to_owned());
             let mail_attr = var("LDAP_USER_MAIL_ATTR").unwrap_or_else(|_| "mail".to_owned());
             Some(LdapConfig {
@@ -349,6 +359,93 @@ fn get_proxy_config() -> Option<ProxyConfig> {
     })
 }
 
+pub struct S3Config {
+    pub bucket: String,
+    pub access_key_id: String,
+    pub access_key_secret: String,
+
+    // region? If not set, default to us-east-1
+    pub region: String,
+    // hostname for s3. If not set, default to $region.amazonaws.com
+    pub hostname: String,
+    // may be useful when using self hosted s3. Won't work with recent AWS buckets
+    pub path_style: bool,
+    pub protocol: String,
+
+    // options below this comment are not used yet
+    // upload directly from user to S3, without going through Plume. Uses PostObject endpoint
+    pub direct_upload: bool,
+    // download directly from s3 to user, wihout going through Plume. Require public read on bucket
+    pub direct_download: bool,
+    // use this hostname for downloads, can be used with caching proxy in front of s3
+    pub alias: Option<String>,
+}
+
+impl S3Config {
+    pub fn get_bucket(&self) -> Bucket {
+        let region = Region::Custom {
+            region: self.region.clone(),
+            endpoint: format!("{}://{}", self.protocol, self.hostname),
+        };
+        let credentials = Credentials {
+            access_key: Some(self.access_key_id.clone()),
+            secret_key: Some(self.access_key_secret.clone()),
+            security_token: None,
+            session_token: None,
+        };
+
+        if self.path_style {
+            Bucket::new_with_path_style(&self.bucket, region, credentials)
+        } else {
+            Bucket::new(&self.bucket, region, credentials)
+        }.unwrap()
+    }
+}
+
+fn get_s3_config() -> Option<S3Config> {
+    let bucket = var("S3_BUCKET").ok();
+    let access_key_id = var("AWS_ACCESS_KEY_ID").ok();
+    let access_key_secret = var("AWS_SECRET_ACCESS_KEY").ok();
+    if bucket.is_none() && access_key_id.is_none() && access_key_secret.is_none() {
+        return None;
+    }
+    if bucket.is_none() || access_key_id.is_none() || access_key_secret.is_none() {
+        panic!("Invalid S3 configuration: some required values are set, but not others");
+    }
+    let bucket = bucket.unwrap();
+    let access_key_id = access_key_id.unwrap();
+    let access_key_secret = access_key_secret.unwrap();
+
+    let region = var("S3_REGION").unwrap_or_else(|_| "us-east-1".to_owned());
+    let hostname = var("S3_HOSTNAME").unwrap_or_else(|_| format!("{}.amazonaws.com", region));
+
+    let protocol = var("S3_PROTOCOL").unwrap_or_else(|_| "https".to_owned());
+    if protocol != "http" && protocol != "https" {
+        panic!("Invalid S3 configuration: invalid protocol {}", protocol);
+    }
+
+    let path_style = var("S3_PATH_STYLE").unwrap_or_else(|_| "false".to_owned());
+    let path_style = string_to_bool(&path_style, "S3_PATH_STYLE");
+    let direct_upload = var("S3_DIRECT_UPLOAD").unwrap_or_else(|_| "false".to_owned());
+    let direct_upload = string_to_bool(&direct_upload, "S3_DIRECT_UPLOAD");
+    let direct_download = var("S3_DIRECT_DOWNLOAD").unwrap_or_else(|_| "false".to_owned());
+    let direct_download = string_to_bool(&direct_download, "S3_DIRECT_DOWNLOAD");
+
+    let alias = var("S3_ALIAS_HOST").ok();
+    Some(S3Config {
+        bucket,
+        access_key_id,
+        access_key_secret,
+        region,
+        hostname,
+        protocol,
+        path_style,
+        direct_upload,
+        direct_download,
+        alias,
+    })
+}
+
 lazy_static! {
     pub static ref CONFIG: Config = Config {
         base_url: var("BASE_URL").unwrap_or_else(|_| format!(
@@ -380,5 +477,6 @@ lazy_static! {
         mail: get_mail_config(),
         ldap: get_ldap_config(),
         proxy: get_proxy_config(),
+        s3: get_s3_config(),
     };
 }

plume-models/src/lib.rs

@@ -69,6 +69,7 @@ pub enum Error {
     Webfinger,
     Expired,
     UserAlreadyExists,
+    Anyhow(anyhow::Error),
 }
 
 impl From<bcrypt::BcryptError> for Error {
@@ -170,6 +171,12 @@ impl From<request::Error> for Error {
     }
 }
 
+impl From<anyhow::Error> for Error {
+    fn from(err: anyhow::Error) -> Error {
+        Error::Anyhow(err)
+    }
+}
+
 pub type Result<T> = std::result::Result<T, Error>;
 
 /// Adds a function to a model, that returns the first

plume-models/src/medias.rs

@@ -170,8 +170,13 @@ impl Media {
 
     pub fn delete(&self, conn: &Connection) -> Result<()> {
         if !self.is_remote {
+            if let Some(config) = &CONFIG.s3 {
+                config.get_bucket()
+                    .delete_object_blocking(&self.file_path)?;
+            } else {
                 fs::remove_file(self.file_path.as_str())?;
             }
+        }
         diesel::delete(self)
             .execute(conn)
             .map(|_| ())

src/routes/mod.rs

@@ -9,7 +9,7 @@ use rocket::{
     http::{
         hyper::header::{CacheControl, CacheDirective, ETag, EntityTag},
         uri::{FromUriParam, Query},
-        RawStr, Status,
+        ContentType, RawStr, Status,
     },
     request::{self, FromFormValue, FromRequest, Request},
     response::{self, Flash, NamedFile, Redirect, Responder, Response},
@@ -204,10 +204,16 @@ pub mod timelines;
 pub mod user;
 pub mod well_known;
 
+#[derive(Responder)]
+enum FileKind {
+    Local(NamedFile),
+    S3(Vec<u8>, ContentType),
+}
+
 #[derive(Responder)]
 #[response()]
 pub struct CachedFile {
-    inner: NamedFile,
+    inner: FileKind,
     cache_control: CacheControl,
 }
 
@@ -253,19 +259,36 @@ pub fn plume_static_files(file: PathBuf, build_id: &RawStr) -> Option<CachedFile
 }
 #[get("/static/media/<file..>")]
 pub fn plume_media_files(file: PathBuf) -> Option<CachedFile> {
+    if let Some(config) = &CONFIG.s3 {
+        let ct = file.extension()
+            .and_then(|ext| ContentType::from_extension(&ext.to_string_lossy()))
+            .unwrap_or(ContentType::Binary);
+
+        let (data, code) = config.get_bucket()
+            .get_object_blocking(format!("plume-media/{}", file.to_string_lossy())).ok()?;
+        if code != 200 {
+            return None;
+        }
+
+        Some(CachedFile {
+            inner: FileKind::S3 ( data, ct),
+            cache_control: CacheControl(vec![CacheDirective::MaxAge(60 * 60 * 24 * 30)]),
+        })
+    } else {
         NamedFile::open(Path::new(&CONFIG.media_directory).join(file))
             .ok()
             .map(|f| CachedFile {
-            inner: f,
+                inner: FileKind::Local(f),
                 cache_control: CacheControl(vec![CacheDirective::MaxAge(60 * 60 * 24 * 30)]),
             })
     }
+}
 #[get("/static/<file..>", rank = 3)]
 pub fn static_files(file: PathBuf) -> Option<CachedFile> {
     NamedFile::open(Path::new("static/").join(file))
         .ok()
         .map(|f| CachedFile {
-            inner: f,
+            inner: FileKind::Local(f),
             cache_control: CacheControl(vec![CacheDirective::MaxAge(60 * 60 * 24 * 30)]),
         })
 }