Files
snake-python/server/database/backend/SqliteGameplayBackend.py
T
daniel156161 39a7a53873
Build and Push Docker Container / build-and-push (push) Successful in 3m52s
feat(database): store gameplay replays in compact form
- Add compact storage package: codec, records, hydrate, replay blobs, cache.
- Running games write game_turns and snake_states keyed by integer ids.
- Encode moves as 1-byte codes and bodies as head plus step codes.
- Store food and hazards only when they changed since the previous turn.
- Strip reasoning fields that duplicate stored values and restore on read.
- Pack finished games into one zstd blob per game in game_replays.
- Merge late turns into packed blobs during the stale-game sweep.
- Extract SQLite and PostgreSQL schemas into dedicated modules.
- Add read/import game bundle and list_game_ids to the backend template.
- Add legacy readers and converter for the old turns/snake_turns layout.
- Add migrate_gameplay_to_compact.py with in-place or cross-DB migration.
- Verify every migrated game and keep legacy rows on failure.
- Support --since, --minimum-quality and guarded --drop-legacy.
- Add just recipe migrate-gameplay-postgres.
- Rewrite merge script to copy game by game via backend bundles.
- Move benchmark state loading into benchmark_states for both backends.
- Update PostgreSQL cleanup script to the new replay tables.
- Drop sqlite-zstd build stages from the Dockerfile.
- Remove old migration scripts, normalized_turn and their tests.
- Add tests for compact storage, schema, legacy migration and PostgreSQL.
- Document the compact layout and migration workflow in the README.
2026-11-05 00:02:56 +01:00

516 lines
21 KiB
Python

"""SQLite gameplay backend with the same compact layout as PostgreSQL.
Running games use ``game_turns``/``snake_states``; finished games are packed
into one zstd blob per game in ``game_replays``.
"""
import asyncio, json, sqlite3, logging, sys
from datetime import datetime, timezone
from pathlib import Path
from server.database.backend.Template import GameplayBackendTemplate
from server.database.backend.sqlite_schema import create_schema
from server.database.compact.cache import RunningGame, RunningGameCache
from server.database.compact.codec import (
decode_body, decode_move, decode_points, encode_body, encode_move, encode_points,
)
from server.database.compact.games import GAME_COLUMNS, normalize_game
from server.database.compact.records import (
GameBundle, SnakeIdentity, SnakeState, TurnRecord, merge_records, records_from_game_state,
)
from server.database.compact.replay_blob import REPLAY_FORMAT, pack_replay, unpack_replay
logger = logging.getLogger(__name__)
if not logger.handlers:
_handler = logging.StreamHandler(stream=sys.stdout)
_handler.setFormatter(logging.Formatter(fmt="%(levelname)s %(module)s: %(message)s"))
logger.addHandler(_handler)
logger.propagate = False
_GAME_SELECT = """
SELECT gid, game_id, started_at, ended_at, width, height, source, map_name,
ruleset_name, ruleset_version, game_type, your_snake_id, your_snake_name,
your_snake_type, your_snake_version, winner_name, winner_you, final_turn,
status, has_replay, quality_status, quality_score, quality_tier, quality_reasons_json
FROM games
"""
def _game_from_row(row:sqlite3.Row) -> dict:
data = dict(row)
reasons = data.pop("quality_reasons_json", None)
data["quality_reasons"] = json.loads(reasons) if reasons else None
return normalize_game(data)
class SqliteGameplayBackend(GameplayBackendTemplate):
def __init__(self, db_path:str, busy_timeout_ms:int=5000, journal_mode:str="WAL", read_only:bool=False):
self.db_path = db_path
self.busy_timeout_ms = max(1000, int(busy_timeout_ms))
self._journal_mode = journal_mode
self._read_only = read_only
self._running = RunningGameCache()
if not read_only:
self._initialize_database()
# ── connection ─────────────────────────────────────────────────────────────
def _connect(self) -> sqlite3.Connection:
if self._read_only:
connection = sqlite3.connect(f"file:{self.db_path}?mode=ro", uri=True, isolation_level=None)
connection.row_factory = sqlite3.Row
connection.execute("PRAGMA query_only = ON")
return connection
connection = sqlite3.connect(
self.db_path,
timeout=max(1, self.busy_timeout_ms // 1000),
isolation_level=None,
)
connection.row_factory = sqlite3.Row
connection.execute("PRAGMA foreign_keys = ON")
connection.execute(f"PRAGMA journal_mode = {self._journal_mode}")
connection.execute("PRAGMA synchronous = NORMAL")
connection.execute("PRAGMA temp_store = MEMORY")
connection.execute("PRAGMA journal_size_limit = 1048576")
connection.execute(f"PRAGMA busy_timeout = {self.busy_timeout_ms}")
return connection
def _initialize_database(self) -> None:
Path(self.db_path).parent.mkdir(parents=True, exist_ok=True)
connection = self._connect()
try:
if connection.execute("PRAGMA auto_vacuum").fetchone()[0] != 1:
connection.execute("PRAGMA auto_vacuum = FULL")
connection.execute("VACUUM")
create_schema(connection)
connection.execute("PRAGMA optimize")
finally:
connection.close()
# ── record helpers ─────────────────────────────────────────────────────────
def _running_game(self, connection:sqlite3.Connection, game_id:str) -> RunningGame:
game = self._running.get(game_id)
if game is None:
row = connection.execute("SELECT gid FROM games WHERE game_id = ?", (game_id,)).fetchone()
if row is None:
raise LookupError(f"Turn for unknown game {game_id!r}")
game = RunningGame(gid=row["gid"])
return game
def _upsert_identities(self, connection:sqlite3.Connection, game_id:str, identities:list[SnakeIdentity]) -> dict[str, int]:
gsids = {}
for identity in identities:
row = connection.execute("""
INSERT INTO game_snakes (game_id, snake_id, snake_name, is_you, customizations_json, gsid)
VALUES (?, ?, ?, ?, ?, (SELECT COALESCE(MAX(gsid), 0) + 1 FROM game_snakes))
ON CONFLICT(game_id, snake_id) DO UPDATE SET
snake_name = excluded.snake_name,
is_you = excluded.is_you,
customizations_json = excluded.customizations_json
RETURNING gsid
""",
(
game_id, identity.snake_id, identity.name, 1 if identity.is_you else 0,
self._to_json(identity.customizations or {}),
),
).fetchone()
gsids[identity.snake_id] = row["gsid"]
return gsids
def _write_rows(self, connection:sqlite3.Connection, gid:int, gsids:dict[str, int], turns:list[TurnRecord], states:list[SnakeState]) -> None:
connection.executemany("""
INSERT INTO game_turns (gid, turn, observed_at, my_move, thinking_json, food, hazards)
VALUES (?, ?, ?, ?, ?, ?, ?)
ON CONFLICT(gid, turn) DO UPDATE SET
observed_at = excluded.observed_at,
my_move = excluded.my_move,
thinking_json = excluded.thinking_json,
food = excluded.food,
hazards = excluded.hazards
""",
[
(
gid, record.turn, self._ts_to_str(record.observed_at) or self._utc_now(),
encode_move(record.move),
self._to_json(record.thinking) if record.thinking is not None else None,
encode_points(record.food), encode_points(record.hazards),
)
for record in turns
],
)
connection.executemany("""
INSERT INTO snake_states (gsid, turn, health, latency, body) VALUES (?, ?, ?, ?, ?)
ON CONFLICT(gsid, turn) DO UPDATE SET
health = excluded.health,
latency = excluded.latency,
body = excluded.body
""",
[
(gsids[state.snake_id], state.turn, state.health, state.latency, encode_body(state.body))
for state in states if state.snake_id in gsids
],
)
def _load_records(self, connection:sqlite3.Connection, game_id:str, gid:int) -> tuple[list[SnakeIdentity], list[TurnRecord], list[SnakeState]]:
identities = [
SnakeIdentity(row["snake_id"], row["snake_name"], bool(row["is_you"]), self._from_json(row["customizations_json"]) or {})
for row in connection.execute("""
SELECT snake_id, snake_name, is_you, customizations_json
FROM game_snakes WHERE game_id = ?
""", (game_id,))
]
turns = [
TurnRecord(
turn=row["turn"], observed_at=row["observed_at"], move=decode_move(row["my_move"]),
thinking=self._from_json(row["thinking_json"]),
food=decode_points(row["food"]), hazards=decode_points(row["hazards"]),
)
for row in connection.execute("""
SELECT turn, observed_at, my_move, thinking_json, food, hazards
FROM game_turns WHERE gid = ? ORDER BY turn
""", (gid,))
]
states = [
SnakeState(row["snake_id"], row["turn"], row["health"], row["latency"], decode_body(row["body"]))
for row in connection.execute("""
SELECT gs.snake_id, s.turn, s.health, s.latency, s.body
FROM game_snakes AS gs
JOIN snake_states AS s ON s.gsid = gs.gsid
WHERE gs.game_id = ?
""", (game_id,))
]
blob = connection.execute("SELECT format, data FROM game_replays WHERE gid = ?", (gid,)).fetchone()
if blob is not None:
packed_turns, packed_states = unpack_replay(blob["data"], blob["format"])
turns, states = merge_records(packed_turns, packed_states, turns, states)
return identities, turns, states
def _store_replay(self, connection:sqlite3.Connection, game_id:str, gid:int, turns:list[TurnRecord], states:list[SnakeState]) -> None:
connection.execute("""
INSERT INTO game_replays (gid, format, turn_count, data) VALUES (?, ?, ?, ?)
ON CONFLICT(gid) DO UPDATE SET
format = excluded.format, turn_count = excluded.turn_count, data = excluded.data
""", (gid, REPLAY_FORMAT, len(turns), pack_replay(turns, states)))
connection.execute("DELETE FROM game_turns WHERE gid = ?", (gid,))
connection.execute(
"DELETE FROM snake_states WHERE gsid IN (SELECT gsid FROM game_snakes WHERE game_id = ?)",
(game_id,),
)
def _pack_game(self, connection:sqlite3.Connection, game_id:str) -> None:
connection.execute("BEGIN IMMEDIATE")
try:
row = connection.execute("SELECT gid FROM games WHERE game_id = ?", (game_id,)).fetchone()
if row is not None and connection.execute(
"SELECT 1 FROM game_turns WHERE gid = ? LIMIT 1", (row["gid"],),
).fetchone():
_, turns, states = self._load_records(connection, game_id, row["gid"])
self._store_replay(connection, game_id, row["gid"], turns, states)
connection.execute("COMMIT")
except Exception:
connection.execute("ROLLBACK")
raise
# ── sync write methods ─────────────────────────────────────────────────────
def _record_game_start_sync(self, game_state:dict, snake_type:str|None=None, snake_version:str|None=None) -> None:
game = game_state.get("game", {})
board = game_state.get("board", {})
you = self._extract_you(game_state)
ruleset = game.get("ruleset", {})
game_type = self._derive_game_type(board, ruleset)
connection = self._connect()
try:
connection.execute("""
INSERT INTO games (
game_id, started_at, width, height, source, map_name,
ruleset_name, ruleset_version, your_snake_id, your_snake_name,
your_snake_type, your_snake_version, game_type, status, gid
) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, 'running',
(SELECT COALESCE(MAX(gid), 0) + 1 FROM games))
ON CONFLICT(game_id) DO UPDATE SET
width = excluded.width,
height = excluded.height,
source = excluded.source,
map_name = excluded.map_name,
ruleset_name = excluded.ruleset_name,
ruleset_version = excluded.ruleset_version,
your_snake_id = excluded.your_snake_id,
your_snake_name = excluded.your_snake_name,
your_snake_type = excluded.your_snake_type,
your_snake_version = excluded.your_snake_version,
game_type = excluded.game_type,
status = 'running'
""",
(
game.get("id"), self._utc_now(), board.get("width"), board.get("height"),
game.get("source"), game.get("map"), ruleset.get("name"), ruleset.get("version"),
you.get("id"), you.get("name"), snake_type, snake_version, game_type,
),
)
connection.execute("PRAGMA wal_checkpoint(PASSIVE)")
finally:
connection.close()
def _record_turn_sync(self, game_state:dict, my_move:str|None, my_thinking:dict|None) -> None:
game_id = game_state.get("game", {}).get("id")
connection = self._connect()
try:
connection.execute("BEGIN IMMEDIATE")
try:
running = self._running_game(connection, game_id)
record, states, identities, points = records_from_game_state(
game_state, my_move, my_thinking, running.points,
)
if any(identity.snake_id not in running.snake_ids for identity in identities):
running.snake_ids.update(self._upsert_identities(connection, game_id, identities))
self._write_rows(connection, running.gid, running.snake_ids, [record], states)
connection.execute(
"UPDATE games SET final_turn = MAX(final_turn, ?) WHERE gid = ?",
(record.turn, running.gid),
)
connection.execute("COMMIT")
except Exception:
connection.execute("ROLLBACK")
raise
finally:
connection.close()
running.points = points
self._running.put(game_id, running)
def _record_game_end_sync(self, game_state:dict) -> None:
game_id, winner_name, winner_you, turn = self._extract_game_end_params(game_state)
connection = self._connect()
try:
connection.execute("""
UPDATE games
SET ended_at = ?,
winner_name = ?,
winner_you = ?,
final_turn = MAX(final_turn, ?),
status = 'finished'
WHERE game_id = ?
""",
(self._utc_now(), winner_name, 1 if winner_you else 0, turn, game_id),
)
self._pack_game(connection, game_id)
finally:
connection.close()
self._running.drop(game_id)
def _finalize_stale_running_games_sync(self, stale_after_seconds:int=600) -> int:
threshold = max(60, int(stale_after_seconds))
now_utc = datetime.now(timezone.utc)
finalized = 0
connection = self._connect()
try:
rows = connection.execute("""
SELECT gid, game_id, started_at, final_turn, your_snake_id
FROM games
WHERE status = 'running'
ORDER BY started_at ASC
""").fetchall()
for row in rows:
started_at = self._parse_utc_timestamp(row["started_at"])
if started_at is None or (now_utc - started_at).total_seconds() < threshold:
continue
game_id = row["game_id"]
identities, _, states = self._load_records(connection, game_id, row["gid"])
final_turn, survivors = self._survivors(identities, states, int(row["final_turn"] or 0))
winner_you, survivor_name = self._calculate_survivor(survivors, row["your_snake_id"])
result = connection.execute("""
UPDATE games
SET ended_at = ?,
winner_name = ?,
winner_you = ?,
final_turn = MAX(final_turn, ?),
status = 'finished'
WHERE game_id = ? AND status = 'running'
""",
(self._utc_now(), survivor_name, 1 if winner_you else 0, final_turn, game_id),
)
if result.rowcount > 0:
finalized += 1
self._pack_game(connection, game_id)
self._running.drop(game_id)
late = connection.execute("""
SELECT DISTINCT g.game_id
FROM game_turns AS t JOIN games AS g ON g.gid = t.gid
WHERE g.status = 'finished'
""").fetchall()
for late_row in late:
self._pack_game(connection, late_row["game_id"])
finally:
connection.close()
return finalized
# ── sync read methods ──────────────────────────────────────────────────────
def _get_summary_sync(self, recent_limit:int=15) -> dict:
connection = self._connect()
try:
totals = connection.execute("""
SELECT
COUNT(*) AS total_games,
SUM(CASE WHEN status = 'running' THEN 1 ELSE 0 END) AS running_games,
SUM(CASE WHEN status = 'finished' THEN 1 ELSE 0 END) AS finished_games,
SUM(CASE WHEN status = 'finished' AND winner_you = 1 THEN 1 ELSE 0 END) AS wins,
SUM(CASE WHEN status = 'finished' AND winner_you = 0 THEN 1 ELSE 0 END) AS losses,
AVG(CASE WHEN status = 'finished' THEN final_turn ELSE NULL END) AS avg_turns
FROM games
""").fetchone()
by_type = connection.execute("""
SELECT
COALESCE(game_type, ruleset_name, 'unknown') AS type_label,
COUNT(*) AS total,
SUM(CASE WHEN status = 'finished' AND winner_you = 1 THEN 1 ELSE 0 END) AS wins,
SUM(CASE WHEN status = 'finished' AND winner_you = 0 THEN 1 ELSE 0 END) AS losses
FROM games
WHERE status = 'finished'
GROUP BY type_label
ORDER BY total DESC
""").fetchall()
recent = connection.execute("""
SELECT game_id, started_at, ended_at, map_name, ruleset_name, game_type,
your_snake_name, your_snake_type, your_snake_version, winner_you, final_turn, status
FROM games
WHERE has_replay = 1
ORDER BY started_at DESC
LIMIT ?
""",
(max(1, int(recent_limit)),),
).fetchall()
finally:
connection.close()
return self._build_summary_output(totals, by_type, recent)
def _list_games_sync(self, limit:int=50) -> list[dict]:
connection = self._connect()
try:
rows = connection.execute("""
SELECT game_id, started_at, ended_at, map_name, source, ruleset_name, game_type,
your_snake_name, your_snake_type, your_snake_version,
winner_you, winner_name, final_turn, status
FROM games
WHERE has_replay = 1
ORDER BY started_at DESC
LIMIT ?
""",
(max(1, int(limit)),),
).fetchall()
finally:
connection.close()
return self._build_game_list_output(rows)
def _get_game_replay_sync(self, game_id:str) -> dict|None:
connection = self._connect()
try:
game_row = connection.execute(_GAME_SELECT + " WHERE game_id = ? AND has_replay = 1", (game_id,)).fetchone()
if game_row is None:
return None
identities, turns, states = self._load_records(connection, game_id, game_row["gid"])
finally:
connection.close()
return self._build_game_replay_output(game_row, identities, turns, states)
# ── bundles (migration / merge) ────────────────────────────────────────────
def list_game_ids_sync(self) -> list[str]:
connection = self._connect()
try:
return [row[0] for row in connection.execute("SELECT game_id FROM games ORDER BY started_at, game_id")]
finally:
connection.close()
def read_game_bundle_sync(self, game_id:str) -> GameBundle|None:
connection = self._connect()
try:
row = connection.execute(_GAME_SELECT + " WHERE game_id = ?", (game_id,)).fetchone()
if row is None:
return None
identities, turns, states = self._load_records(connection, game_id, row["gid"])
finally:
connection.close()
return GameBundle(_game_from_row(row), identities, turns, states)
def import_game_bundle_sync(self, bundle:GameBundle) -> None:
game = normalize_game(bundle.game)
columns = [column for column in GAME_COLUMNS if column != "quality_reasons"] + ["quality_reasons_json"]
values = [game[column] for column in GAME_COLUMNS if column != "quality_reasons"]
values.append(self._to_json(game["quality_reasons"]) if game["quality_reasons"] is not None else None)
values = [self._ts_to_str(value) if isinstance(value, datetime) else value for value in values]
updates = ", ".join(f"{column} = excluded.{column}" for column in columns[1:])
connection = self._connect()
try:
connection.execute("BEGIN IMMEDIATE")
try:
connection.execute(
f"INSERT INTO games ({', '.join(columns)}, gid) "
f"VALUES ({', '.join('?' for _ in columns)}, (SELECT COALESCE(MAX(gid), 0) + 1 FROM games)) "
f"ON CONFLICT(game_id) DO UPDATE SET {updates}",
values,
)
gid = connection.execute("SELECT gid FROM games WHERE game_id = ?", (game["game_id"],)).fetchone()["gid"]
gsids = self._upsert_identities(connection, game["game_id"], bundle.snakes)
if bundle.turns and game["has_replay"]:
if game["status"] == "finished":
self._store_replay(connection, game["game_id"], gid, bundle.turns, bundle.states)
else:
self._write_rows(connection, gid, gsids, bundle.turns, bundle.states)
connection.execute("COMMIT")
except Exception:
connection.execute("ROLLBACK")
raise
finally:
connection.close()
self._running.drop(game["game_id"])
# ── public async interface ─────────────────────────────────────────────────
async def record_game_start(self, game_state:dict, snake_type:str|None=None, snake_version:str|None=None) -> None:
await asyncio.to_thread(self._record_game_start_sync, game_state, snake_type, snake_version)
async def record_turn(self, game_state:dict, my_move:str|None, my_thinking:dict|None=None) -> None:
await asyncio.to_thread(self._record_turn_sync, game_state, my_move, my_thinking)
async def record_game_end(self, game_state:dict) -> None:
await asyncio.to_thread(self._record_game_end_sync, game_state)
async def get_summary(self, recent_limit:int=15) -> dict:
return await asyncio.to_thread(self._get_summary_sync, recent_limit)
async def list_games(self, limit:int=50) -> list[dict]:
return await asyncio.to_thread(self._list_games_sync, limit)
async def finalize_stale_running_games(self, stale_after_seconds:int=600) -> int:
return await asyncio.to_thread(self._finalize_stale_running_games_sync, stale_after_seconds)
async def get_game_replay(self, game_id:str) -> dict|None:
return await asyncio.to_thread(self._get_game_replay_sync, game_id)
async def list_game_ids(self) -> list[str]:
return await asyncio.to_thread(self.list_game_ids_sync)
async def read_game_bundle(self, game_id:str) -> GameBundle|None:
return await asyncio.to_thread(self.read_game_bundle_sync, game_id)
async def import_game_bundle(self, bundle:GameBundle) -> None:
await asyncio.to_thread(self.import_game_bundle_sync, bundle)
async def close(self) -> None:
return None