Prober: pipelined bodies, merge-first order, a probe cache across resets

- XoverClient.bodies: PROBE_PIPELINE (8) BODY commands per connection at
  once, replies read in order (one at a time it ran 3-10 articles/s on 6
  connections against a queue of 1M).
- Claims in tiers: lone files first, then par2 sets and archives of
  releases without a readable name (what the par2 merge needs), then those
  of readable releases.
- probe_cache: final answers by message-id (status, inspection with the
  16 KiB hash and par2 file list), PROBE_VERSION 2. A reset empties the
  probe queue, not the cache; recreated candidates of probed articles are
  answered from it without the news server.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

tests/unit/nntp/xover_client.spec.ts

@@ -79,6 +79,23 @@ test.group('nntp / XoverClient', () => {
     }
   })
 
+  test('reads several bodies pipelined, in order, a missing one as null', async ({ assert }) => {
+    const { port, close } = await server()
+    const client = await XoverClient.connect({ host: '127.0.0.1', port, user: 'u', password: 'p', compress: false })
+    try {
+      const bodies = await client.bodies(['one@test', '<gone@test>', '<one@test>'])
+      assert.equal(bodies.length, 3)
+      assert.equal(bodies[0]!.subarray(6).toString('latin1'), '..stuffed\r\nlast\r\n')
+      assert.isNull(bodies[1])
+      assert.equal(bodies[2]!.length, bodies[0]!.length)
+      // The connection is still in step afterwards.
+      assert.equal((await client.body('<one@test>')).length, bodies[0]!.length)
+    } finally {
+      client.close()
+      close()
+    }
+  })
+
   test('parses overview lines into the fields nntp-js gives', ({ assert }) => {
     const [[number, fields]] = parseOverviews(LINES[0] + '\r\n')
     assert.equal(number, 101)