feat(media): authenticated signed media gateway (C1, C2-sink, C3, H2, H10)

Replaces the DB-blind `/media` ServeDir with a signed, DB-aware gateway at
`GET /media/{kind}/{id}`. Every media byte now flows through an HMAC-SHA256
signature check (minted into feed/upload DTOs for authenticated members; <img>
can't carry a Bearer header) plus a DB lookup:

- C1: export ZIP/HTML have no upload row, so they are unreachable by path —
  download stays behind the authenticated /export endpoints.
- C2 (sink): responses carry X-Content-Type-Options: nosniff and a locked-down
  CSP (default-src 'none'; sandbox), neutralizing any active content.
- C3 / H2: find_by_id filters deleted_at and the handler rejects ban-hidden
  uploaders, so deleted and moderated artifacts 404 — and the unauthenticated
  get_original alias (the H2 hole) is removed entirely.
- H10: delete paths (owner + host) now unlink original/preview/thumbnail after
  commit; soft_delete returns the paths; an hourly reaper reclaims disk for
  rows soft-deleted past a 1-day grace and hard-deletes them (FKs cascade).

Signed URLs are bucketed to a 1h window so they stay stable across feed polls
(browser cache hits) while expiring within 24h. media_token sign/verify has a
unit test (roundtrip + tamper + expiry).

Frontend: FeedUpload/pickMediaUrl now use the backend-provided signed
original_url; no client constructs a media path anymore.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
MechaCat02
2026-06-27 15:30:30 +02:00
parent ab9f1d89b2
commit 8faf702208
16 changed files with 437 additions and 119 deletions

View File

@@ -0,0 +1,73 @@
//! Filesystem lifecycle for media artifacts.
//!
//! The DB-aware gateway hides deleted/hidden uploads, but the bytes still sit on
//! a fixed-size disk until something removes them. This module is that
//! something: best-effort unlinking on delete, plus a periodic reaper that
//! sweeps files belonging to soft-deleted rows (catching anything an in-process
//! unlink missed — e.g. a crash between the DB commit and the unlink).
use std::path::Path;
use sqlx::PgPool;
use crate::models::upload::DeletedPaths;
/// Best-effort removal of an upload's three on-disk artifacts. A missing file is
/// not an error (it may already be gone, or never existed for videos without a
/// preview); anything else is logged but never propagated — losing the bytes
/// must not fail the user's delete.
pub async fn unlink_media(media_path: &Path, paths: &DeletedPaths) {
let candidates = [
Some(&paths.original),
paths.preview.as_ref(),
paths.thumbnail.as_ref(),
];
for rel in candidates.into_iter().flatten() {
let abs = media_path.join(rel);
match tokio::fs::remove_file(&abs).await {
Ok(()) => {}
Err(e) if e.kind() == std::io::ErrorKind::NotFound => {}
Err(e) => tracing::warn!(path = %rel, error = ?e, "media unlink failed"),
}
}
}
/// Reaper: remove on-disk files for uploads that were soft-deleted more than
/// `grace` ago, then hard-delete those rows so they don't accumulate. Strictly
/// DB-row-driven — it never walks the filesystem, so it can never remove a file
/// belonging to a live upload.
pub async fn reap_deleted(pool: &PgPool, media_path: &Path) {
let rows: Vec<DeletedPaths> = match sqlx::query_as::<_, DeletedPaths>(
"SELECT original_path AS original, preview_path AS preview, thumbnail_path AS thumbnail
FROM upload
WHERE deleted_at IS NOT NULL AND deleted_at < NOW() - INTERVAL '1 day'",
)
.fetch_all(pool)
.await
{
Ok(rows) => rows,
Err(e) => {
tracing::warn!(error = ?e, "media reaper: query failed");
return;
}
};
if rows.is_empty() {
return;
}
for paths in &rows {
unlink_media(media_path, paths).await;
}
match sqlx::query(
"DELETE FROM upload
WHERE deleted_at IS NOT NULL AND deleted_at < NOW() - INTERVAL '1 day'",
)
.execute(pool)
.await
{
Ok(r) => tracing::info!("media reaper: removed {} deleted upload(s)", r.rows_affected()),
Err(e) => tracing::warn!(error = ?e, "media reaper: row delete failed"),
}
}