hygiene: track Cargo.lock, drop dead config + error variant, fix doc drift

- .gitignore no longer excludes Cargo.lock; track it (standard for a service crate)
- drop the unused GRINLWS_SCAN_BATCH_BLOCKS config field (the scanner pages by a
  fixed PAGE_MAX, never read this) and its .env.example line
- remove the unused Error::NotImplemented variant and its match arm
- replace the crate-level #![allow(dead_code)] with a targeted allow on
  mark_unspent (the single-commit un-spend counterpart to mark_spent; the reorg
  path un-spends in bulk via rollback_to)
- doc drift: ready() returns 502 when the node is down (not 503); main.rs status
  reflects the functional, in-production service

.env.example

@@ -16,9 +16,8 @@ GRIN_NODE_URL=http://127.0.0.1:3413/v2/foreign
 # Optional basic-auth secret for the node Foreign API (empty = none).
 GRIN_NODE_API_SECRET=
 
-# Background scanner tuning.
+# Background scanner tuning: how often it polls for new blocks.
 GRINLWS_SCAN_POLL_SECS=15
-GRINLWS_SCAN_BATCH_BLOCKS=100
 
 # Restore-depth bound: how many days behind the tip a newly registered account
 # may start scanning from. 0 = unbounded.

.gitignore

@@ -1,4 +1,3 @@
 /target
 .env
 *.log
-Cargo.lock

src/api.rs

@@ -235,7 +235,7 @@ pub async fn health() -> Json<HealthResponse> {
     })
 }
 
-/// Readiness — reaches the grin node (503 when the node is down).
+/// Readiness — reaches the grin node (502 when the node is down).
 pub async fn ready(State(state): State<Arc<AppState>>) -> Result<Json<HealthResponse>> {
     state.node.health_check().await?;
     Ok(Json(HealthResponse {

src/config.rs

@@ -25,9 +25,6 @@ pub struct Config {
 
     /// How often the background scanner polls for new blocks.
     pub scan_poll_secs: u64,
-    /// How many blocks the scanner processes per batch (bounds memory / a single
-    /// transaction's size).
-    pub scan_batch_blocks: u64,
     /// How far behind the tip a newly-registered account may start scanning from
     /// (restore-depth bound). `0` = unbounded.
     pub restore_max_depth_days: u32,
@@ -55,7 +52,6 @@ impl Config {
             ),
             node_foreign_api_secret: Secret::new(env_or("GRIN_NODE_API_SECRET", "")),
             scan_poll_secs: parse_or("GRINLWS_SCAN_POLL_SECS", 15),
-            scan_batch_blocks: parse_or("GRINLWS_SCAN_BATCH_BLOCKS", 100),
             restore_max_depth_days: parse_or("GRINLWS_RESTORE_MAX_DEPTH_DAYS", 0) as u32,
             admin_key: Secret::new(env_or("GRINLWS_ADMIN_KEY", "")),
         }

src/db.rs

@@ -18,12 +18,6 @@
 //!
 //! SECURITY: only view-only data is ever stored. No spend key, no private key.
 
-// Several write helpers (insert_output, mark_spent, set_cursor, rollback_to) are
-// exercised only by the scanner, which is a deferred milestone in this scaffold.
-// They are real and tested-shaped; allow them to be unused until the scanner
-// loop is wired up.
-#![allow(dead_code)]
-
 use sqlx::any::AnyPoolOptions;
 use sqlx::{AnyPool, Row};
 
@@ -197,7 +191,10 @@ pub async fn mark_spent(pool: &AnyPool, commit: &str, spent_height: i64) -> Resu
 }
 
 /// Un-mark a spend (a reorg brought a previously-spent output back into the
-/// unspent set). Self-heals an over-eager spend reconcile.
+/// unspent set). Self-heals an over-eager spend reconcile. Retained as the
+/// single-commit counterpart to `mark_spent`; the reorg path currently un-spends
+/// in bulk via `rollback_to`, so this is not yet called.
+#[allow(dead_code)]
 pub async fn mark_unspent(pool: &AnyPool, commit: &str) -> Result<()> {
     sqlx::query("UPDATE outputs SET spent = 0, spent_height = NULL WHERE \"commit\" = $1")
         .bind(commit)

src/error.rs

@@ -23,10 +23,6 @@ pub enum Error {
 
     #[error("database error")]
     Db(#[from] sqlx::Error),
-
-    /// A route that is deliberately stubbed in this scaffold.
-    #[error("{0} not implemented (scaffold)")]
-    NotImplemented(&'static str),
 }
 
 impl IntoResponse for Error {
@@ -41,7 +37,6 @@ impl IntoResponse for Error {
                 tracing::error!(error = %e, "database error");
                 StatusCode::INTERNAL_SERVER_ERROR
             }
-            Error::NotImplemented(_) => StatusCode::NOT_IMPLEMENTED,
         };
         let message = match &self {
             Error::Db(_) => "internal error".to_string(),

src/main.rs

@@ -29,10 +29,11 @@
 //! nothing; grin-lws holds the `rewind_hash` for scanning just as monero-lws
 //! holds the view key. NEVER a spend key.
 //!
-//! STATUS: FUNCTIONAL. The view-only rewind (`grin::rewind_output`), the
-//! PMMR-index node reads, and the background scanner (discover / spend-reconcile
-//! / reorg / backfill) are implemented + tested. Remaining before production: a
-//! smoke test against a live grin node. See README.md.
+//! STATUS: FUNCTIONAL and running in production. The view-only rewind
+//! (`grin::rewind_output`), the PMMR-index node reads, and the background scanner
+//! (discover / spend-reconcile / reorg / backfill) are implemented + tested, and
+//! the service has been validated against a live grin node with a funded wallet.
+//! See README.md.
 //!
 //! PUBLIC-CLEAN: all config is env-based with generic loopback defaults. No
 //! hostnames, operator IPs, secrets, or deploy specifics anywhere in this repo.
@@ -75,7 +76,7 @@ async fn main() {
         .await
         .expect("failed to build database pool");
     // In production, run `sqlx migrate run` (or embed `sqlx::migrate!()`) against
-    // DATABASE_URL before serving. Left explicit here so the scaffold starts
+    // DATABASE_URL before serving. Left explicit here so the service starts
     // without a live DB.
 
     let node = GrinNode::new(&config).expect("failed to build grin node client");

src/scanner.rs

@@ -357,7 +357,6 @@ mod integration {
             node_foreign_api_url: url,
             node_foreign_api_secret: Secret::new(String::new()),
             scan_poll_secs: 1,
-            scan_batch_blocks: 100,
             restore_max_depth_days: 0,
             admin_key: Secret::new(String::new()),
         }