feat: write chunks directly to temp upload files

Completed uploads used to copy every staged chunk into a second file before
renaming the result into data/complete. That doubled write volume and required
peak disk space for both the chunk set and the final file.

Write each chunk directly into one private temp upload file at its final offset
instead. After a chunk write succeeds, record a tiny durable completion marker
for progress and resume scans. Completion now verifies the temp file length and
all markers, then renames the temp file into the completed upload directory.

Add UPL_TEMP_DIR and --temp-dir so operators can choose where upload metadata,
markers, and temp files live. The default remains data/staging, and docs call
out that the temp directory must be on the same filesystem as data/complete for
atomic promotion. The nginx example now aliases only the completed upload
directory, and the smoke test verifies that final-file alias.

This keeps the existing length-based validation model; it does not add per-chunk
hashing.

Test Plan:
- just check
- just nginx-smoke
- cargo clippy && cargo clippy --benches && cargo clippy --tests
- cargo +nightly fmt --all
- cargo clippy && cargo clippy --benches && cargo clippy --tests

Refs: none

scripts/nginx-smoke.sh

@@ -9,10 +9,13 @@ workspace_dir="$(pwd)"
 mkdir -p "$workspace_dir/target/nginx-smoke"
 tmp_dir="$(mktemp -d "$workspace_dir/target/nginx-smoke/run.XXXXXXXX")"
 data_dir="$tmp_dir/data"
+complete_dir="$data_dir/complete"
+temp_dir="$tmp_dir/upload-temp"
 nginx_conf_dir="$tmp_dir/nginx-conf.d"
 nginx_conf="$nginx_conf_dir/default.conf"
 backend_log="$tmp_dir/backend.log"
 source_file="$tmp_dir/source.bin"
+served_file="$tmp_dir/served.bin"
 chunk0="$tmp_dir/chunk0.part"
 chunk1="$tmp_dir/chunk1.part"
 backend_pid=""
@@ -29,7 +32,7 @@ cleanup() {
 trap cleanup EXIT
 
 start_backend() {
-    UPL_BIND="0.0.0.0:$backend_port" UPL_DATA_DIR="$data_dir" \
+    UPL_BIND="0.0.0.0:$backend_port" UPL_DATA_DIR="$data_dir" UPL_TEMP_DIR="$temp_dir" \
         cargo run --quiet >"$backend_log" 2>&1 &
     backend_pid="$!"
     wait_for "http://127.0.0.1:$backend_port/healthz"
@@ -66,13 +69,19 @@ process.stdin.on("end", () => {
 ' "$field"
 }
 
-mkdir -p "$data_dir" "$nginx_conf_dir"
+mkdir -p "$complete_dir" "$temp_dir" "$nginx_conf_dir"
 
 cat >"$nginx_conf" <<EOF
 server {
     listen $proxy_port;
     client_max_body_size 64m;
 
+    location /files/ {
+        alias /upl-complete/;
+        autoindex off;
+        try_files \$uri =404;
+    }
+
     location / {
         proxy_pass http://host.docker.internal:$backend_port;
         proxy_http_version 1.1;
@@ -95,6 +104,7 @@ docker run -d --rm \
     --add-host host.docker.internal:host-gateway \
     -p "127.0.0.1:$proxy_port:$proxy_port" \
     -v "$nginx_conf_dir:/etc/nginx/conf.d:ro" \
+    -v "$complete_dir:/upl-complete:ro" \
     "$nginx_image" >/dev/null
 wait_for "http://127.0.0.1:$proxy_port/healthz"
 
@@ -143,10 +153,17 @@ complete_path="$(printf '%s' "$complete_response" | json_field file_path)"
 
 source_hash="$(sha256sum "$source_file" | awk '{print $1}')"
 complete_hash="$(sha256sum "$complete_path" | awk '{print $1}')"
+curl -fsS "http://127.0.0.1:$proxy_port/files/source.bin" -o "$served_file"
+served_hash="$(sha256sum "$served_file" | awk '{print $1}')"
 
 if [[ "$source_hash" != "$complete_hash" ]]; then
     echo "Checksum mismatch after nginx-proxied resume" >&2
     exit 1
 fi
 
+if [[ "$source_hash" != "$served_hash" ]]; then
+    echo "Checksum mismatch through nginx completed-file alias" >&2
+    exit 1
+fi
+
 echo "nginx smoke ok: $upload_id"