@optimystic/db-p2p 0.24.0 → 0.24.2

package/src/logger.ts

@@ -1,27 +1,27 @@
-import debug from 'debug'
-
-const BASE_NAMESPACE = 'optimystic:db-p2p'
-
-/**
- * Build a `debug` logger under `optimystic:db-p2p:<subNamespace>`, optionally suffixed with the
- * owning node's peer id (`:<first 12 chars>`) so lines from several nodes sharing one process —
- * every integration test — are attributable. Omit `peerId` and the namespace is byte-for-byte the
- * un-suffixed one, so callers that don't know their peer id are unaffected.
- *
- * NOTE: 12 chars matches the truncation the rest of this package already uses in log payloads, but
- * every Ed25519 peer id starts with the constant `12D3KooW`, so only ~4 base58 characters actually
- * distinguish nodes (~11M combinations — ample for the handful of nodes a test process runs).
- * Widen this only if a run ever needs to match a namespace against a full peer id.
- *
- * NOTE: only the two classes the diagnosability ticket named pass a peer id today
- * (`Libp2pKeyPeerNetwork`, `CoordinatorRepo`); the package's other ~30 `createLogger` call sites
- * still log under a flat namespace. Thread a peer id through any of them if a future diagnosis
- * needs per-node attribution from that subsystem — the mechanism is already here.
- */
-export function createLogger(subNamespace: string, peerId?: string): debug.Debugger {
-	const suffix = peerId ? `:${peerId.substring(0, 12)}` : ''
-	return debug(`${BASE_NAMESPACE}:${subNamespace}${suffix}`)
-}
-
-export const verbose = typeof process !== 'undefined'
-	&& (process.env.OPTIMYSTIC_VERBOSE === '1' || process.env.OPTIMYSTIC_VERBOSE === 'true');
+import debug from 'debug'
+
+const BASE_NAMESPACE = 'optimystic:db-p2p'
+
+/**
+ * Build a `debug` logger under `optimystic:db-p2p:<subNamespace>`, optionally suffixed with the
+ * owning node's peer id (`:<first 12 chars>`) so lines from several nodes sharing one process —
+ * every integration test — are attributable. Omit `peerId` and the namespace is byte-for-byte the
+ * un-suffixed one, so callers that don't know their peer id are unaffected.
+ *
+ * NOTE: 12 chars matches the truncation the rest of this package already uses in log payloads, but
+ * every Ed25519 peer id starts with the constant `12D3KooW`, so only ~4 base58 characters actually
+ * distinguish nodes (~11M combinations — ample for the handful of nodes a test process runs).
+ * Widen this only if a run ever needs to match a namespace against a full peer id.
+ *
+ * NOTE: only a few call sites pass a peer id today (`Libp2pKeyPeerNetwork`, `CoordinatorRepo`, and
+ * the three `peer-address-book` sinks); the package's other ~30 `createLogger` call sites
+ * still log under a flat namespace. Thread a peer id through any of them if a future diagnosis
+ * needs per-node attribution from that subsystem — the mechanism is already here.
+ */
+export function createLogger(subNamespace: string, peerId?: string): debug.Debugger {
+	const suffix = peerId ? `:${peerId.substring(0, 12)}` : ''
+	return debug(`${BASE_NAMESPACE}:${subNamespace}${suffix}`)
+}
+
+export const verbose = typeof process !== 'undefined'
+	&& (process.env.OPTIMYSTIC_VERBOSE === '1' || process.env.OPTIMYSTIC_VERBOSE === 'true');