sync: batch the fallback getAlbum song fetches

The `search3`-less fallback emitted one batch per album, and
`SyncService.writeSongs` opens a transaction per batch, so syncing a large
library that way ran one transaction per album — thousands of fsyncs, and an
invalidation round for every list observing `songs` with each.

Accumulate what each `getAlbum` returns and flush at `PAGE_SIZE`, so the
fallback writes the same ~500-row batches every other fetch already does.
Albums are fetched concurrently and complete out of order, so the accumulator
sits behind a mutex and the send happens outside it.

Batches are in-flight memory, so the flush is per rows reached rather than
per album page. Measured per `Song` (strings, boxed longs and genre list
included): ~0.7 KiB, so a 500-row batch is ~334 KiB plus ~51 KiB of the
`names`/`genres` structures `writeSongs` derives, where a whole album page
(~5,500 rows) would be ~3.8 MiB. `albumSongs()` is also a bare `channelFlow`,
whose channel holds 64 elements whatever their size, so 500-row batches would
have made its worst case ~21 MiB; it now carries the same `.buffer(1)` that
`entityBatches()` already uses on itself. The rest of the in-flight rows
scale with `syncConcurrency`, so that leaves the default's batch in hand and
a few MiB at the highest setting.

`SubsonicSourceTest.albumSongsAreEmittedInBoundedBatches` asserts the emitted
sizes, [500, 100] for 600 songs spread over 200 three-song albums, which pins
the carried remainder as well as the flush: verified it fails on the
pre-change code (200 batches of 3) and against a mutant that drops the
remainder ([500, 99]). `docs/architecture.md`'s page cap was 20x out of date,
so it is corrected alongside. `ktlintCheck`, `testDebugUnitTest` and
`lintDebug` pass.

app/src/main/kotlin/com/subtracks/data/source/subsonic/SubsonicSource.kt

@@ -25,7 +25,9 @@ import kotlinx.coroutines.flow.flow
 import kotlinx.coroutines.flow.flowOn
 import kotlinx.coroutines.flow.map
 import kotlinx.coroutines.launch
+import kotlinx.coroutines.sync.Mutex
 import kotlinx.coroutines.sync.Semaphore
+import kotlinx.coroutines.sync.withLock
 import kotlinx.coroutines.sync.withPermit
 import kotlinx.coroutines.withContext
 import org.xml.sax.Attributes
@@ -224,6 +226,23 @@ class SubsonicSource(
     private fun albumSongs(): Flow<List<Song>> =
         channelFlow {
             val permits = Semaphore(maxConcurrentFetches)
+            val lock = Mutex()
+            val pending = ArrayList<Song>(PAGE_SIZE)
+
+            suspend fun accept(songs: List<Song>) {
+                lock
+                    .withLock {
+                        pending.addAll(songs)
+                        if (pending.size < PAGE_SIZE) {
+                            null
+                        } else {
+                            val batch = ArrayList(pending.subList(0, PAGE_SIZE))
+                            pending.subList(0, PAGE_SIZE).clear()
+                            batch
+                        }
+                    }?.let { send(it) }
+            }
+
             var offset = 0
             var pages = 0
             while (true) {
@@ -251,7 +270,7 @@ class SubsonicSource(
                                         onChild = { song, name, attrs -> if (name == "genres") song.addGenre(attrs) },
                                     ).collect { batch ->
                                         val accepted = batch.map { it.toSong() }.filter { it.id.isNotEmpty() }
-                                        if (accepted.isNotEmpty()) send(accepted)
+                                        accept(accepted)
                                     }
                                 }
                             }
@@ -265,7 +284,14 @@ class SubsonicSource(
                 if (pages > maxPages) throw pageCapExceeded()
                 offset += PAGE_SIZE
             }
-        }
+            val tail =
+                lock.withLock {
+                    val held = pending.toList()
+                    pending.clear()
+                    held
+                }
+            if (tail.isNotEmpty()) send(tail)
+        }.buffer(1)
 
     private fun supportsEmptyQuerySearch(): Boolean {
         emptyQuerySearchSupported?.let { return it }

app/src/test/kotlin/com/subtracks/data/source/subsonic/SubsonicSourceTest.kt

@@ -608,6 +608,40 @@ class SubsonicSourceTest {
             assertTrue(maxInFlight.get() in 2..DEFAULT_FETCH_CONCURRENCY)
         }
 
+    @Test
+    fun albumSongsAreEmittedInBoundedBatches() =
+        runBlocking {
+            val albumCount = 200
+            val songsPerAlbum = 3
+            server.dispatcher =
+                object : Dispatcher() {
+                    override fun dispatch(request: RecordedRequest): MockResponse {
+                        val url = request.requestUrl!!
+                        return when (url.encodedPath) {
+                            "/rest/search3.view" -> {
+                                MockResponse().setBody(failed(10, "Required parameter is missing"))
+                            }
+
+                            "/rest/getAlbumList2.view" -> {
+                                MockResponse().setBody(albumListOf(albumCount))
+                            }
+
+                            "/rest/getAlbum.view" -> {
+                                MockResponse().setBody(albumWithSongs(url.queryParameter("id")!!, songsPerAlbum))
+                            }
+
+                            else -> {
+                                MockResponse().setResponseCode(404)
+                            }
+                        }
+                    }
+                }
+
+            val batches = SubsonicSource(1, client()).songs().toList().map { it.size }
+
+            assertEquals(listOf(500, 100), batches)
+        }
+
     @Test
     fun playlistEntriesWithoutIdsAreSkipped() =
         runBlocking {
@@ -705,6 +739,13 @@ class SubsonicSourceTest {
             "<song id=\"$albumId-s1\" title=\"One\"/>" +
             "</album></subsonic-response>"
 
+    private fun albumWithSongs(
+        albumId: String,
+        count: Int,
+    ) = "<subsonic-response status=\"ok\"><album id=\"$albumId\">" +
+        (1..count).joinToString("") { "<song id=\"$albumId-s$it\" title=\"Title $it\"/>" } +
+        "</album></subsonic-response>"
+
     private fun failed(
         code: Int,
         message: String,

docs/architecture.md

@@ -120,7 +120,7 @@ Images: grids and lists request a 256 px `getCoverArt` thumbnail; the 120 dp det
 
 A run is triggered by the user (pull-to-refresh or a Sync action) or by `AutoSync`, a foreground-only timer. While the app has a visible Activity `AutoSync` re-checks at most every 15 minutes (the configured interval, capped) whether `syncIntervalMinutes` has elapsed since the last attempt and, if the mode allows (`Off`, `Wi-Fi only` or `Always`, default `Wi-Fi only`), the app is online and a server is configured, it asks `SyncManager` for a silent sync; nothing runs while the app is backgrounded, and a missed interval catches up on the next open. Only the time of the last attempt is persisted, so a server that keeps failing is retried once per interval rather than on every poll. A silent sync still shows the progress notification and the Settings spinner, but never a toast.
 
-Songs have a primary and a fallback fetch. The primary probes `search3` with an empty query once (memoized per source); if the server supports it, the library is paged through `search3` 500 songs at a time. Otherwise the fallback pages `getAlbumList2` and then issues one `getAlbum` per album — it works everywhere but costs a request per album instead of per 500 songs. Every paged loop stops on a short page and is capped at 1000 pages. Playlists are fetched whole, and their songs are stored by position so a playlist that got shorter (or was emptied) is trimmed at the end.
+Songs have a primary and a fallback fetch. The primary probes `search3` with an empty query once (memoized per source); if the server supports it, the library is paged through `search3` 500 songs at a time. Otherwise the fallback pages `getAlbumList2` and then issues one `getAlbum` per album — it works everywhere but costs a request per album instead of per 500 songs, though it still coalesces the songs into 500-row batches before writing them. Every paged loop stops on a short page and is capped at 20,000 pages. Playlists are fetched whole, and their songs are stored by position so a playlist that got shorter (or was emptied) is trimmed at the end.
 
 Cleanup is the "everything I had, minus everything the server just sent" step described above: artists, albums, songs and playlists delete the leftover IDs in 500-row chunks, disc titles are deleted per `(album, disc)` for albums that still exist, and playlist entries beyond the new count are removed along with entries for playlists that disappeared.