tests

backend/Cargo.lock

@@ -1979,6 +1979,7 @@ version = "0.1.0"
 dependencies = [
  "anyhow",
  "argon2",
+ "async-trait",
  "aws-config",
  "aws-credential-types",
  "aws-sdk-s3",
@@ -1990,7 +1991,10 @@ dependencies = [
  "diesel_migrations",
  "dotenv",
  "hex",
+ "http-body-util",
+ "hyper 1.7.0",
  "jsonwebtoken",
+ "once_cell",
  "rand 0.8.5",
  "serde",
  "serde_json",
@@ -2054,6 +2058,26 @@ version = "2.3.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9b4f627cb1b25917193a259e49bdad08f671f8d9708acfd5fe0a8c1455d87220"
 
+[[package]]
+name = "pin-project"
+version = "1.1.10"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "677f1add503faace112b9f1373e43e9e054bfdd22ff1a63c1bc485eaec6a6a8a"
+dependencies = [
+ "pin-project-internal",
+]
+
+[[package]]
+name = "pin-project-internal"
+version = "1.1.10"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6e918e4ff8c4549eb882f14b3a4bc8c8bc93de829416eacf579f1207a8fbf861"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
 [[package]]
 name = "pin-project-lite"
 version = "0.2.16"
@@ -2931,7 +2955,9 @@ version = "0.4.13"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b8fa9be0de6cf49e536ce1851f987bd21a43b771b09473c3549a6c853db37c1c"
 dependencies = [
+ "futures-core",
  "futures-util",
+ "pin-project",
  "pin-project-lite",
  "tokio",
  "tower-layer",

backend/Cargo.toml

@@ -7,7 +7,7 @@ edition = "2021"
 # Web framework
 axum = { version = "0.7", features = ["multipart"] }
 tokio = { version = "1", features = ["full"] }
-tower = { version = "0.4", features = ["make"] }
+tower = { version = "0.4", features = ["make", "util"] }
 tower-http = { version = "0.5", features = ["cors", "trace"] }
 axum-extra = { version = "0.9", features = ["typed-header"] }
 
@@ -33,6 +33,7 @@ dotenv = "0.15"
 sha2 = "0.10"
 hex = "0.4"
 bytes = "1.5"
+async-trait = "0.1"
 
 # Error handling
 thiserror = "1.0"
@@ -44,3 +45,8 @@ jsonwebtoken = "9"
 
 # Misc
 rand = "0.8"
+
+[dev-dependencies]
+once_cell = "1.19"
+hyper = "1.2"
+http-body-util = "0.1"

backend/src/lib.rs

@@ -0,0 +1,10 @@
+pub mod auth;
+pub mod config;
+pub mod db;
+pub mod error;
+pub mod models;
+pub mod routes;
+pub mod s3;
+pub mod schema;
+pub mod state;
+pub mod storage;

backend/src/main.rs

@@ -1,23 +1,17 @@
-mod auth;
-mod config;
-mod db;
-mod error;
-mod models;
-mod routes;
-mod s3;
-mod schema;
-mod state;
-
 use std::net::SocketAddr;
+use std::sync::Arc;
 
 use tokio::net::TcpListener;
 use tower::make::Shared;
 use tracing_subscriber::EnvFilter;
 
-use crate::auth::jwt::JwtService;
+use paperless_backend::auth::jwt::JwtService;
-use crate::config::AppConfig;
+use paperless_backend::config::AppConfig;
-use crate::s3::build_client;
+use paperless_backend::db;
-use crate::state::AppState;
+use paperless_backend::routes;
+use paperless_backend::s3::build_client;
+use paperless_backend::state::AppState;
+use paperless_backend::storage::S3Storage;
 
 #[tokio::main]
 async fn main() -> anyhow::Result<()> {
@@ -27,9 +21,10 @@ async fn main() -> anyhow::Result<()> {
     let config = AppConfig::from_env()?;
     let pool = db::init_pool(&config.database_url)?;
     let s3_client = build_client(&config).await?;
+    let storage = Arc::new(S3Storage::new(s3_client, config.s3_bucket.clone()));
     let jwt = JwtService::from_config(&config)?;
 
-    let state = AppState::new(pool, config, s3_client, jwt);
+    let state = AppState::new(pool, config, storage, jwt);
 
     let router = routes::create_router(state.clone());
 

backend/src/routes/documents.rs

@@ -1,8 +1,6 @@
 use std::collections::HashMap;
 use std::time::Duration;
 
-use aws_sdk_s3::presigning::PresigningConfig;
-use aws_sdk_s3::primitives::ByteStream;
 use axum::extract::{Json, Multipart, Path, Query, State};
 use axum::http::StatusCode;
 use axum::response::IntoResponse;
@@ -267,21 +265,11 @@ pub async fn upload_document(
         }
     }
 
-    let mut put_request = state
+    state
-        .s3
+        .storage
-        .put_object()
+        .put_object(&s3_key, file_bytes.clone(), content_type.clone())
-        .bucket(&state.config.s3_bucket)
-        .key(&s3_key)
-        .body(ByteStream::from(file_bytes.clone()));
-
-    if let Some(ref ct) = content_type {
-        put_request = put_request.content_type(ct.clone());
-    }
-
-    put_request
-        .send()
         .await
-        .map_err(|err| AppError::internal(format!("failed to upload to s3: {err}")))?;
+        .map_err(|err| AppError::internal(format!("failed to store document: {err}")))?;
 
     let metadata_value = if metadata.is_null() {
         Value::Object(Default::default())
@@ -349,22 +337,17 @@ pub async fn download_document(
         .filter(document_versions::version_number.eq(doc.current_version))
         .first(&mut conn)?;
 
-    let presign_config = PresigningConfig::builder()
+    let presigned_url = state
-        .expires_in(Duration::from_secs(PRESIGNED_URL_EXPIRY_SECONDS))
+        .storage
-        .build()
+        .presign_get_object(
-        .map_err(|err| AppError::internal(format!("failed to build presigning config: {err}")))?;
+            &version.s3_key,
-
+            Duration::from_secs(PRESIGNED_URL_EXPIRY_SECONDS),
-    let presigned = state
+        )
-        .s3
-        .get_object()
-        .bucket(&state.config.s3_bucket)
-        .key(&version.s3_key)
-        .presigned(presign_config)
         .await
         .map_err(|err| AppError::internal(format!("failed to generate download URL: {err}")))?;
 
     Ok(Json(DocumentDownloadResponse {
-        url: presigned.uri().to_string(),
+        url: presigned_url,
         expires_in: PRESIGNED_URL_EXPIRY_SECONDS,
         filename: doc.original_name.clone(),
         content_type: doc.content_type.clone(),

backend/src/schema.rs

@@ -16,8 +16,6 @@ diesel::table! {
         version_number -> Int4,
         #[max_length = 500]
         s3_key -> Varchar,
-        #[max_length = 100]
-        s3_bucket -> Varchar,
         size_bytes -> Int8,
         #[max_length = 64]
         checksum -> Varchar,

backend/src/state.rs

@@ -1,6 +1,5 @@
 use std::sync::Arc;
 
-use aws_sdk_s3::Client as S3Client;
 use diesel::{
     pg::PgConnection,
     r2d2::{ConnectionManager, PooledConnection},
@@ -11,6 +10,7 @@ use crate::{
     config::AppConfig,
     db::PgPool,
     error::{AppError, AppResult},
+    storage::ObjectStorage,
 };
 
 type PgPooledConnection = PooledConnection<ConnectionManager<PgConnection>>;
@@ -19,16 +19,21 @@ type PgPooledConnection = PooledConnection<ConnectionManager<PgConnection>>;
 pub struct AppState {
     pub pool: PgPool,
     pub config: Arc<AppConfig>,
-    pub s3: S3Client,
+    pub storage: Arc<dyn ObjectStorage>,
     pub jwt: JwtService,
 }
 
 impl AppState {
-    pub fn new(pool: PgPool, config: AppConfig, s3: S3Client, jwt: JwtService) -> Self {
+    pub fn new(
+        pool: PgPool,
+        config: AppConfig,
+        storage: Arc<dyn ObjectStorage>,
+        jwt: JwtService,
+    ) -> Self {
         Self {
             pool,
             config: Arc::new(config),
-            s3,
+            storage,
             jwt,
         }
     }

backend/src/storage.rs

@@ -0,0 +1,79 @@
+use std::time::Duration;
+
+use anyhow::{Context, Result};
+use async_trait::async_trait;
+use aws_sdk_s3::presigning::PresigningConfig;
+use aws_sdk_s3::primitives::ByteStream;
+use aws_sdk_s3::Client as S3Client;
+
+#[async_trait]
+pub trait ObjectStorage: Send + Sync + 'static {
+    async fn put_object(
+        &self,
+        key: &str,
+        bytes: Vec<u8>,
+        content_type: Option<String>,
+    ) -> Result<()>;
+
+    async fn presign_get_object(&self, key: &str, expires_in: Duration) -> Result<String>;
+}
+
+pub struct S3Storage {
+    client: S3Client,
+    bucket: String,
+}
+
+impl S3Storage {
+    pub fn new(client: S3Client, bucket: impl Into<String>) -> Self {
+        Self {
+            client,
+            bucket: bucket.into(),
+        }
+    }
+}
+
+#[async_trait]
+impl ObjectStorage for S3Storage {
+    async fn put_object(
+        &self,
+        key: &str,
+        bytes: Vec<u8>,
+        content_type: Option<String>,
+    ) -> Result<()> {
+        let mut request = self
+            .client
+            .put_object()
+            .bucket(&self.bucket)
+            .key(key)
+            .body(ByteStream::from(bytes));
+
+        if let Some(content_type) = content_type {
+            request = request.content_type(content_type);
+        }
+
+        request
+            .send()
+            .await
+            .context("failed to upload object to S3")?;
+
+        Ok(())
+    }
+
+    async fn presign_get_object(&self, key: &str, expires_in: Duration) -> Result<String> {
+        let presign_config = PresigningConfig::builder()
+            .expires_in(expires_in)
+            .build()
+            .context("failed to build S3 presigning config")?;
+
+        let presigned = self
+            .client
+            .get_object()
+            .bucket(&self.bucket)
+            .key(key)
+            .presigned(presign_config)
+            .await
+            .context("failed to generate presigned download URL")?;
+
+        Ok(presigned.uri().to_string())
+    }
+}

backend/tests/auth_flow.rs

@@ -0,0 +1,37 @@
+mod common;
+
+use anyhow::Result;
+use axum::http::StatusCode;
+use common::{acquire_db_lock, body_to_vec, TestApp};
+use serde::Deserialize;
+
+#[derive(Deserialize)]
+struct AuthenticatedUser {
+    username: String,
+    role: String,
+}
+
+#[tokio::test]
+async fn login_and_me_roundtrip() -> Result<()> {
+    let _lock = acquire_db_lock().await;
+    let Some(app) = TestApp::new().await? else {
+        eprintln!("skipping test: TEST_DATABASE_URL not set");
+        return Ok(());
+    };
+
+    let password = "s3cret";
+    app.insert_user("alice", password, "admin").await?;
+
+    let token = app.login_token("alice", password).await?;
+
+    let response = app.get("/api/auth/me", Some(&token)).await?;
+    assert_eq!(response.status(), StatusCode::OK);
+    let body = body_to_vec(response.into_body()).await?;
+    let user: AuthenticatedUser = serde_json::from_slice(&body)?;
+
+    assert_eq!(user.username, "alice");
+    assert_eq!(user.role, "admin");
+
+    app.cleanup().await?;
+    Ok(())
+}

backend/tests/common/mod.rs

@@ -0,0 +1,392 @@
+use std::collections::HashMap;
+use std::env;
+use std::sync::Arc;
+use std::time::Duration;
+
+use anyhow::{anyhow, ensure, Context, Result};
+use async_trait::async_trait;
+use axum::body::Body;
+use axum::http::{Method, Request, StatusCode};
+use axum::Router;
+use diesel::connection::SimpleConnection;
+use diesel::prelude::*;
+use diesel::PgConnection;
+use diesel_migrations::{embed_migrations, EmbeddedMigrations, MigrationHarness};
+use http_body_util::BodyExt;
+use once_cell::sync::Lazy;
+use paperless_backend::auth::jwt::JwtService;
+use paperless_backend::config::AppConfig;
+use paperless_backend::db::{self, PgPool};
+use paperless_backend::models::NewUser;
+use paperless_backend::routes;
+use paperless_backend::state::AppState;
+use paperless_backend::storage::ObjectStorage;
+use rand::rngs::OsRng;
+use serde::Serialize;
+use tokio::sync::Mutex;
+use tower::util::ServiceExt;
+use uuid::Uuid;
+
+const MIGRATIONS: EmbeddedMigrations = embed_migrations!("migrations");
+
+static DB_LOCK: Lazy<Mutex<()>> = Lazy::new(|| Mutex::new(()));
+
+#[allow(dead_code)]
+#[derive(Clone)]
+pub struct StoredObject {
+    pub key: String,
+    pub bytes: Vec<u8>,
+    pub content_type: Option<String>,
+}
+
+#[derive(Default)]
+pub struct FakeStorage {
+    objects: Mutex<HashMap<String, StoredObject>>,
+}
+
+#[async_trait]
+impl ObjectStorage for FakeStorage {
+    async fn put_object(
+        &self,
+        key: &str,
+        bytes: Vec<u8>,
+        content_type: Option<String>,
+    ) -> Result<()> {
+        let stored = StoredObject {
+            key: key.to_string(),
+            bytes,
+            content_type,
+        };
+        let mut guard = self.objects.lock().await;
+        guard.insert(stored.key.clone(), stored);
+        Ok(())
+    }
+
+    async fn presign_get_object(&self, key: &str, expires_in: Duration) -> Result<String> {
+        let guard = self.objects.lock().await;
+        ensure!(guard.contains_key(key), "object {key} missing");
+        Ok(format!(
+            "https://fake-storage/{key}?expires_in={}",
+            expires_in.as_secs()
+        ))
+    }
+}
+
+impl FakeStorage {
+    #[allow(dead_code)]
+    pub async fn get(&self, key: &str) -> Option<StoredObject> {
+        let guard = self.objects.lock().await;
+        guard.get(key).cloned()
+    }
+
+    #[allow(dead_code)]
+    pub async fn object_count(&self) -> usize {
+        let guard = self.objects.lock().await;
+        guard.len()
+    }
+}
+
+pub struct TestApp {
+    pub state: AppState,
+    router: Router,
+    storage: Arc<FakeStorage>,
+}
+
+impl TestApp {
+    pub async fn new() -> Result<Option<Self>> {
+        let database_url = match env::var("TEST_DATABASE_URL") {
+            Ok(url) => url,
+            Err(_) => return Ok(None),
+        };
+
+        let config = AppConfig {
+            database_url: database_url.clone(),
+            server_host: "127.0.0.1".to_string(),
+            server_port: 0,
+            jwt_secret: "test-secret".to_string(),
+            jwt_issuer: "test-issuer".to_string(),
+            jwt_audience: "test-audience".to_string(),
+            jwt_expiry_minutes: 60,
+            aws_endpoint_url: None,
+            aws_access_key_id: None,
+            aws_secret_access_key: None,
+            aws_region: "us-east-1".to_string(),
+            s3_bucket: "test-bucket".to_string(),
+        };
+
+        let pool = db::init_pool(&config.database_url)?;
+        prepare_database(&pool).await?;
+
+        let storage = Arc::new(FakeStorage::default());
+        let storage_for_state: Arc<dyn ObjectStorage> = storage.clone();
+        let jwt = JwtService::from_config(&config)?;
+        let state = AppState::new(pool.clone(), config, storage_for_state, jwt);
+        let router = routes::create_router(state.clone());
+
+        Ok(Some(Self {
+            state,
+            router,
+            storage,
+        }))
+    }
+
+    pub async fn cleanup(&self) -> Result<()> {
+        let pool = self.state.pool.clone();
+        tokio::task::spawn_blocking(move || -> Result<()> {
+            let mut conn = pool
+                .get()
+                .map_err(|err| anyhow!("failed to get cleanup connection: {err}"))?;
+            truncate_all(&mut conn)?;
+            Ok(())
+        })
+        .await
+        .context("cleanup task panicked")?
+    }
+
+    #[allow(dead_code)]
+    pub fn storage(&self) -> Arc<FakeStorage> {
+        self.storage.clone()
+    }
+
+    pub async fn insert_user(&self, username: &str, password: &str, role: &str) -> Result<Uuid> {
+        let username = username.to_string();
+        let password = password.to_string();
+        let role = role.to_string();
+        self.with_conn(move |conn| {
+            let password_hash = hash_password(&password)?;
+            let user = NewUser {
+                id: Uuid::new_v4(),
+                username,
+                password_hash,
+                role,
+            };
+            diesel::insert_into(paperless_backend::schema::users::table)
+                .values(&user)
+                .execute(conn)
+                .context("failed to insert user")?;
+            Ok(user.id)
+        })
+        .await
+    }
+
+    pub async fn login_token(&self, username: &str, password: &str) -> Result<String> {
+        #[derive(Serialize)]
+        struct LoginPayload<'a> {
+            username: &'a str,
+            password: &'a str,
+        }
+
+        let response = self
+            .post_json(
+                "/api/auth/login",
+                &LoginPayload { username, password },
+                None,
+            )
+            .await?;
+
+        ensure!(
+            response.status() == StatusCode::OK,
+            "login failed with status {}",
+            response.status()
+        );
+
+        let body = body_to_vec(response.into_body()).await?;
+        #[derive(serde::Deserialize)]
+        struct LoginResponse {
+            access_token: String,
+        }
+        let parsed: LoginResponse = serde_json::from_slice(&body)?;
+        Ok(parsed.access_token)
+    }
+
+    pub async fn post_json<T: Serialize + ?Sized>(
+        &self,
+        path: &str,
+        payload: &T,
+        token: Option<&str>,
+    ) -> Result<hyper::Response<Body>> {
+        let body = serde_json::to_vec(payload)?;
+        let mut builder = Request::builder()
+            .method(Method::POST)
+            .uri(path)
+            .header("content-type", "application/json");
+        if let Some(token) = token {
+            builder = builder.header("authorization", format!("Bearer {token}"));
+        }
+        let request = builder.body(Body::from(body))?;
+        Ok(self
+            .router
+            .clone()
+            .oneshot(request)
+            .await
+            .expect("infallible response"))
+    }
+
+    #[allow(dead_code)]
+    pub async fn patch_json<T: Serialize + ?Sized>(
+        &self,
+        path: &str,
+        payload: &T,
+        token: Option<&str>,
+    ) -> Result<hyper::Response<Body>> {
+        let body = serde_json::to_vec(payload)?;
+        let mut builder = Request::builder()
+            .method(Method::PATCH)
+            .uri(path)
+            .header("content-type", "application/json");
+        if let Some(token) = token {
+            builder = builder.header("authorization", format!("Bearer {token}"));
+        }
+        let request = builder.body(Body::from(body))?;
+        Ok(self
+            .router
+            .clone()
+            .oneshot(request)
+            .await
+            .expect("infallible response"))
+    }
+
+    pub async fn get(&self, path: &str, token: Option<&str>) -> Result<hyper::Response<Body>> {
+        let mut builder = Request::builder().method(Method::GET).uri(path);
+        if let Some(token) = token {
+            builder = builder.header("authorization", format!("Bearer {token}"));
+        }
+        let request = builder.body(Body::empty())?;
+        Ok(self
+            .router
+            .clone()
+            .oneshot(request)
+            .await
+            .expect("infallible response"))
+    }
+
+    #[allow(dead_code)]
+    pub async fn delete(&self, path: &str, token: Option<&str>) -> Result<hyper::Response<Body>> {
+        let builder = Request::builder().method(Method::DELETE).uri(path);
+        let builder = if let Some(token) = token {
+            builder.header("authorization", format!("Bearer {token}"))
+        } else {
+            builder
+        };
+        let request = builder.body(Body::empty())?;
+        Ok(self
+            .router
+            .clone()
+            .oneshot(request)
+            .await
+            .expect("infallible response"))
+    }
+
+    #[allow(dead_code)]
+    pub async fn upload_document(
+        &self,
+        path: &str,
+        filename: &str,
+        content_type: &str,
+        data: &[u8],
+        folder_id: Option<Uuid>,
+        token: &str,
+    ) -> Result<hyper::Response<Body>> {
+        let boundary = format!("boundary-{}", Uuid::new_v4());
+        let mut body = Vec::new();
+        body.extend(format!("--{boundary}\r\n").as_bytes());
+        body.extend(
+            format!(
+                "Content-Disposition: form-data; name=\"file\"; filename=\"{}\"\r\n",
+                filename
+            )
+            .as_bytes(),
+        );
+        body.extend(format!("Content-Type: {}\r\n\r\n", content_type).as_bytes());
+        body.extend(data);
+        body.extend(b"\r\n");
+
+        if let Some(folder) = folder_id {
+            body.extend(format!("--{boundary}\r\n").as_bytes());
+            body.extend(b"Content-Disposition: form-data; name=\"folder_id\"\r\n\r\n");
+            body.extend(folder.to_string().as_bytes());
+            body.extend(b"\r\n");
+        }
+
+        body.extend(format!("--{boundary}--\r\n").as_bytes());
+
+        let builder = Request::builder()
+            .method(Method::POST)
+            .uri(path)
+            .header(
+                "content-type",
+                format!("multipart/form-data; boundary={boundary}"),
+            )
+            .header("authorization", format!("Bearer {token}"));
+
+        let request = builder.body(Body::from(body))?;
+        Ok(self
+            .router
+            .clone()
+            .oneshot(request)
+            .await
+            .expect("infallible response"))
+    }
+
+    async fn with_conn<F, T>(&self, f: F) -> Result<T>
+    where
+        F: FnOnce(&mut PgConnection) -> Result<T> + Send + 'static,
+        T: Send + 'static,
+    {
+        let pool = self.state.pool.clone();
+        tokio::task::spawn_blocking(move || {
+            let mut conn = pool
+                .get()
+                .map_err(|err| anyhow!("failed to get database connection: {err}"))?;
+            f(&mut conn)
+        })
+        .await
+        .context("connection task panicked")?
+    }
+}
+
+pub async fn acquire_db_lock() -> tokio::sync::MutexGuard<'static, ()> {
+    DB_LOCK.lock().await
+}
+
+pub async fn body_to_vec(body: Body) -> Result<Vec<u8>> {
+    let collected = body
+        .collect()
+        .await
+        .map_err(|err| anyhow!("failed to read response body: {err}"))?;
+    Ok(collected.to_bytes().to_vec())
+}
+
+async fn prepare_database(pool: &PgPool) -> Result<()> {
+    let pool = pool.clone();
+    tokio::task::spawn_blocking(move || -> Result<()> {
+        let mut conn = pool
+            .get()
+            .map_err(|err| anyhow!("failed to acquire connection: {err}"))?;
+        conn.run_pending_migrations(MIGRATIONS)
+            .map_err(|err| anyhow!("failed to run migrations: {err}"))?;
+        truncate_all(&mut conn)?;
+        Ok(())
+    })
+    .await
+    .context("migration task panicked")?
+}
+
+fn truncate_all(conn: &mut PgConnection) -> Result<()> {
+    conn.batch_execute(
+        "TRUNCATE TABLE document_tags, document_versions, documents, folders, tags, users RESTART IDENTITY CASCADE;",
+    )
+    .context("failed to truncate tables")?;
+    Ok(())
+}
+
+fn hash_password(password: &str) -> Result<String> {
+    use argon2::password_hash::{PasswordHasher, SaltString};
+    use argon2::Argon2;
+
+    let salt = SaltString::generate(&mut OsRng);
+    Ok(Argon2::default()
+        .hash_password(password.as_bytes(), &salt)
+        .map_err(|err| anyhow!("failed to hash password: {err}"))?
+        .to_string())
+}

backend/tests/documents_flow.rs

@@ -0,0 +1,177 @@
+mod common;
+
+use anyhow::Result;
+use axum::http::StatusCode;
+use common::{acquire_db_lock, body_to_vec, TestApp};
+use serde::Deserialize;
+use serde_json::Value;
+use uuid::Uuid;
+
+#[derive(Deserialize)]
+struct DocumentDetail {
+    document: DocumentInfo,
+    current_version: DocumentVersion,
+}
+
+#[derive(Deserialize)]
+struct DocumentInfo {
+    id: Uuid,
+    original_name: String,
+    current_version: i32,
+    deleted_at: Option<String>,
+    tags: Vec<Value>,
+}
+
+#[derive(Deserialize)]
+struct DocumentVersion {
+    s3_key: String,
+    size_bytes: i64,
+}
+
+#[derive(Deserialize)]
+struct DocumentListItem {
+    id: Uuid,
+    current_version: i32,
+}
+
+#[derive(Deserialize)]
+struct DocumentDownload {
+    url: String,
+    filename: String,
+}
+
+#[tokio::test]
+async fn upload_and_list_document() -> Result<()> {
+    let _lock = acquire_db_lock().await;
+    let Some(app) = TestApp::new().await? else {
+        eprintln!("skipping test: TEST_DATABASE_URL not set");
+        return Ok(());
+    };
+
+    let password = "passw0rd";
+    app.insert_user("dana", password, "admin").await?;
+    let token = app.login_token("dana", password).await?;
+
+    let file_bytes = b"example document body".to_vec();
+    let upload = app
+        .upload_document(
+            "/api/documents",
+            "doc.txt",
+            "text/plain",
+            &file_bytes,
+            None,
+            &token,
+        )
+        .await?;
+    assert_eq!(upload.status(), StatusCode::OK);
+    let body = body_to_vec(upload.into_body()).await?;
+    let detail: DocumentDetail = serde_json::from_slice(&body)?;
+
+    assert_eq!(detail.document.original_name, "doc.txt");
+    assert_eq!(detail.document.current_version, 1);
+    assert_eq!(detail.document.deleted_at, None);
+    assert!(detail.document.tags.is_empty());
+    assert_eq!(detail.current_version.size_bytes, file_bytes.len() as i64);
+
+    let stored = app
+        .storage()
+        .get(&detail.current_version.s3_key)
+        .await
+        .expect("object stored");
+    assert_eq!(stored.bytes, file_bytes);
+
+    let response = app.get("/api/documents", Some(&token)).await?;
+    assert_eq!(response.status(), StatusCode::OK);
+    let body = body_to_vec(response.into_body()).await?;
+    let mut list: Vec<DocumentListItem> = serde_json::from_slice(&body)?;
+    assert_eq!(list.len(), 1);
+    let item = list.pop().unwrap();
+    assert_eq!(item.id, detail.document.id);
+    assert_eq!(item.current_version, 1);
+
+    let download = app
+        .get(
+            &format!("/api/documents/{}/download", detail.document.id),
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(download.status(), StatusCode::OK);
+    let body = body_to_vec(download.into_body()).await?;
+    let download_info: DocumentDownload = serde_json::from_slice(&body)?;
+    assert!(download_info.url.contains(&detail.current_version.s3_key));
+    assert_eq!(download_info.filename, "doc.txt");
+
+    app.cleanup().await?;
+    Ok(())
+}
+
+#[tokio::test]
+async fn duplicate_and_restore_document() -> Result<()> {
+    let _lock = acquire_db_lock().await;
+    let Some(app) = TestApp::new().await? else {
+        eprintln!("skipping test: TEST_DATABASE_URL not set");
+        return Ok(());
+    };
+
+    let password = "pass1234";
+    app.insert_user("sam", password, "admin").await?;
+    let token = app.login_token("sam", password).await?;
+
+    let payload = b"same bytes".to_vec();
+    let first = app
+        .upload_document(
+            "/api/documents",
+            "dup.bin",
+            "application/octet-stream",
+            &payload,
+            None,
+            &token,
+        )
+        .await?;
+    let first_body = body_to_vec(first.into_body()).await?;
+    let first_detail: DocumentDetail = serde_json::from_slice(&first_body)?;
+
+    let second = app
+        .upload_document(
+            "/api/documents",
+            "dup.bin",
+            "application/octet-stream",
+            &payload,
+            None,
+            &token,
+        )
+        .await?;
+    assert_eq!(second.status(), StatusCode::OK);
+    let second_body = body_to_vec(second.into_body()).await?;
+    let second_detail: DocumentDetail = serde_json::from_slice(&second_body)?;
+
+    assert_eq!(first_detail.document.id, second_detail.document.id);
+    assert_eq!(second_detail.document.deleted_at, None);
+
+    let delete = app
+        .delete(
+            &format!("/api/documents/{}", first_detail.document.id),
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(delete.status(), StatusCode::NO_CONTENT);
+
+    let third = app
+        .upload_document(
+            "/api/documents",
+            "dup.bin",
+            "application/octet-stream",
+            &payload,
+            None,
+            &token,
+        )
+        .await?;
+    let third_body = body_to_vec(third.into_body()).await?;
+    let third_detail: DocumentDetail = serde_json::from_slice(&third_body)?;
+
+    assert_eq!(third_detail.document.id, first_detail.document.id);
+    assert_eq!(third_detail.document.deleted_at, None);
+
+    app.cleanup().await?;
+    Ok(())
+}

backend/tests/folders_flow.rs

@@ -0,0 +1,137 @@
+mod common;
+
+use anyhow::Result;
+use axum::http::StatusCode;
+use common::{acquire_db_lock, body_to_vec, TestApp};
+use serde::Deserialize;
+use serde::Serialize;
+use uuid::Uuid;
+
+#[derive(Deserialize)]
+struct FolderResponse {
+    folder: FolderInfo,
+}
+
+#[derive(Deserialize)]
+struct FolderInfo {
+    id: Uuid,
+}
+
+#[derive(Deserialize)]
+struct FolderContents {
+    documents: Vec<DocSummary>,
+}
+
+#[derive(Deserialize)]
+struct DocSummary {
+    id: Uuid,
+}
+
+#[derive(Serialize)]
+struct CreateFolder<'a> {
+    name: &'a str,
+    parent_id: Option<Uuid>,
+}
+
+#[derive(Serialize)]
+struct MoveDocumentRequest {
+    folder_id: Option<Uuid>,
+}
+
+#[derive(Deserialize)]
+struct DocumentDetail {
+    document: DocSummary,
+}
+
+#[tokio::test]
+async fn folder_move_and_delete_flow() -> Result<()> {
+    let _lock = acquire_db_lock().await;
+    let Some(app) = TestApp::new().await? else {
+        eprintln!("skipping test: TEST_DATABASE_URL not set");
+        return Ok(());
+    };
+
+    let password = "folderpass";
+    app.insert_user("folder-admin", password, "admin").await?;
+    let token = app.login_token("folder-admin", password).await?;
+
+    let folder_resp = app
+        .post_json(
+            "/api/folders",
+            &CreateFolder {
+                name: "Projects",
+                parent_id: None,
+            },
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(folder_resp.status(), StatusCode::OK);
+    let folder_body = body_to_vec(folder_resp.into_body()).await?;
+    let folder: FolderResponse = serde_json::from_slice(&folder_body)?;
+
+    let upload = app
+        .upload_document(
+            "/api/documents",
+            "plan.pdf",
+            "application/pdf",
+            b"dummy",
+            None,
+            &token,
+        )
+        .await?;
+    let upload_body = body_to_vec(upload.into_body()).await?;
+    let detail: DocumentDetail = serde_json::from_slice(&upload_body)?;
+
+    let move_resp = app
+        .patch_json(
+            &format!("/api/documents/{}/folder", detail.document.id),
+            &MoveDocumentRequest {
+                folder_id: Some(folder.folder.id),
+            },
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(move_resp.status(), StatusCode::NO_CONTENT);
+
+    let contents = app
+        .get(
+            &format!("/api/folders/{}/contents", folder.folder.id),
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(contents.status(), StatusCode::OK);
+    let contents_body = body_to_vec(contents.into_body()).await?;
+    let contents: FolderContents = serde_json::from_slice(&contents_body)?;
+    assert_eq!(contents.documents.len(), 1);
+    assert_eq!(contents.documents[0].id, detail.document.id);
+
+    let root_contents = app.get("/api/folders/root/contents", Some(&token)).await?;
+    let root_body = body_to_vec(root_contents.into_body()).await?;
+    let root: FolderContents = serde_json::from_slice(&root_body)?;
+    assert!(root
+        .documents
+        .iter()
+        .all(|doc| doc.id != detail.document.id));
+
+    let delete_attempt = app
+        .delete(&format!("/api/folders/{}", folder.folder.id), Some(&token))
+        .await?;
+    assert_eq!(delete_attempt.status(), StatusCode::BAD_REQUEST);
+
+    let move_back = app
+        .patch_json(
+            &format!("/api/documents/{}/folder", detail.document.id),
+            &MoveDocumentRequest { folder_id: None },
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(move_back.status(), StatusCode::NO_CONTENT);
+
+    let delete = app
+        .delete(&format!("/api/folders/{}", folder.folder.id), Some(&token))
+        .await?;
+    assert_eq!(delete.status(), StatusCode::NO_CONTENT);
+
+    app.cleanup().await?;
+    Ok(())
+}

backend/tests/tags_flow.rs

@@ -0,0 +1,124 @@
+mod common;
+
+use anyhow::Result;
+use axum::http::StatusCode;
+use common::{acquire_db_lock, body_to_vec, TestApp};
+use serde::Deserialize;
+use serde::Serialize;
+use uuid::Uuid;
+
+#[derive(Deserialize)]
+struct DocumentDetail {
+    document: DocumentInfo,
+}
+
+#[derive(Deserialize)]
+struct DocumentInfo {
+    id: Uuid,
+    tags: Vec<TagInfo>,
+}
+
+#[derive(Deserialize)]
+struct TagInfo {
+    label: String,
+}
+
+#[derive(Deserialize)]
+struct TagResponse {
+    id: Uuid,
+}
+
+#[derive(Serialize)]
+struct AssignTagsRequest {
+    tag_ids: Vec<Uuid>,
+}
+
+#[tokio::test]
+async fn tag_assignment_flow() -> Result<()> {
+    let _lock = acquire_db_lock().await;
+    let Some(app) = TestApp::new().await? else {
+        eprintln!("skipping test: TEST_DATABASE_URL not set");
+        return Ok(());
+    };
+
+    let password = "tagpass";
+    app.insert_user("tagger", password, "admin").await?;
+    let token = app.login_token("tagger", password).await?;
+
+    let upload = app
+        .upload_document(
+            "/api/documents",
+            "tagged.txt",
+            "text/plain",
+            b"tag me",
+            None,
+            &token,
+        )
+        .await?;
+    let upload_body = body_to_vec(upload.into_body()).await?;
+    let detail: DocumentDetail = serde_json::from_slice(&upload_body)?;
+
+    #[derive(Serialize)]
+    struct CreateTagPayload<'a> {
+        label: &'a str,
+        color: Option<&'a str>,
+    }
+
+    let create_tag = app
+        .post_json(
+            "/api/tags",
+            &CreateTagPayload {
+                label: "Important",
+                color: Some("#FF0000"),
+            },
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(create_tag.status(), StatusCode::OK);
+    let body = body_to_vec(create_tag.into_body()).await?;
+    let tag: TagResponse = serde_json::from_slice(&body)?;
+
+    let assign = app
+        .post_json(
+            &format!("/api/documents/{}/tags", detail.document.id),
+            &AssignTagsRequest {
+                tag_ids: vec![tag.id],
+            },
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(assign.status(), StatusCode::NO_CONTENT);
+
+    let refreshed = app
+        .get(
+            &format!("/api/documents/{}", detail.document.id),
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(refreshed.status(), StatusCode::OK);
+    let refreshed_body = body_to_vec(refreshed.into_body()).await?;
+    let refreshed_detail: DocumentDetail = serde_json::from_slice(&refreshed_body)?;
+    assert_eq!(refreshed_detail.document.tags.len(), 1);
+    assert_eq!(refreshed_detail.document.tags[0].label, "Important");
+
+    let remove = app
+        .delete(
+            &format!("/api/documents/{}/tags/{}", detail.document.id, tag.id),
+            Some(&token),
+        )
+        .await?;
+    assert_eq!(remove.status(), StatusCode::NO_CONTENT);
+
+    let final_check = app
+        .get(
+            &format!("/api/documents/{}", detail.document.id),
+            Some(&token),
+        )
+        .await?;
+    let final_body = body_to_vec(final_check.into_body()).await?;
+    let final_detail: DocumentDetail = serde_json::from_slice(&final_body)?;
+    assert!(final_detail.document.tags.is_empty());
+
+    app.cleanup().await?;
+    Ok(())
+}