rubydb 0.1.4 → 0.1.5

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (420) hide show
  1. checksums.yaml +4 -4
  2. data/.github/PULL_REQUEST_TEMPLATE.md +15 -15
  3. data/.github/workflows/benchmark.yml +26 -26
  4. data/.github/workflows/compatibility.yml +63 -63
  5. data/.github/workflows/fuzz.yml +33 -33
  6. data/.github/workflows/lint.yml +21 -21
  7. data/.github/workflows/operations.yml +24 -24
  8. data/.github/workflows/production-validation.yml +111 -111
  9. data/.github/workflows/release.yml +77 -77
  10. data/.github/workflows/security.yml +39 -37
  11. data/.github/workflows/test.yml +26 -26
  12. data/.github/workflows/workload.yml +58 -58
  13. data/.gitignore +12 -5
  14. data/.rubocop.yml +50 -44
  15. data/.standard.yml +9 -14
  16. data/ARCHITECTURE.md +21 -21
  17. data/CHANGELOG.md +43 -27
  18. data/CODE_OF_CONDUCT.md +13 -13
  19. data/CONTRIBUTING.md +29 -29
  20. data/GOVERNANCE.md +16 -16
  21. data/Gemfile +18 -17
  22. data/Gemfile.lock +125 -71
  23. data/README.md +249 -162
  24. data/ROADMAP.md +27 -27
  25. data/Rakefile +71 -71
  26. data/SECURITY.md +54 -54
  27. data/SUPPORT.md +14 -14
  28. data/adapters/activerecord/Gemfile +11 -11
  29. data/adapters/activerecord/lib/active_record/connection_adapters/rubydb_adapter.rb +873 -879
  30. data/adapters/activerecord/rubydb-activerecord.gemspec +20 -20
  31. data/adapters/activerecord/spec/rubydb_adapter_integration_spec.rb +143 -143
  32. data/adapters/ruby/README.md +18 -18
  33. data/adapters/sequel/README.md +11 -11
  34. data/config/monitoring/prometheus-alerts.yml +39 -39
  35. data/config/production.yml +36 -36
  36. data/docs/README.md +75 -71
  37. data/docs/architecture/concurrency.md +14 -14
  38. data/docs/architecture/current-state.md +125 -125
  39. data/docs/architecture/execution-engine.md +25 -25
  40. data/docs/architecture/indexes.md +19 -19
  41. data/docs/architecture/mvcc.md +19 -19
  42. data/docs/architecture/overview.md +13 -13
  43. data/docs/architecture/pages.md +11 -11
  44. data/docs/architecture/production-roadmap.md +82 -82
  45. data/docs/architecture/query-planner.md +20 -20
  46. data/docs/architecture/recovery.md +18 -18
  47. data/docs/architecture/sql-engine.md +12 -12
  48. data/docs/architecture/storage-engine.md +14 -14
  49. data/docs/architecture/transactions.md +10 -10
  50. data/docs/architecture/wal.md +28 -28
  51. data/docs/cli-cheatsheet.md +98 -98
  52. data/docs/cli.md +275 -275
  53. data/docs/contributing/architecture.md +9 -9
  54. data/docs/contributing/benchmarking.md +14 -14
  55. data/docs/contributing/development.md +16 -16
  56. data/docs/contributing/release-process.md +49 -49
  57. data/docs/contributing/testing.md +16 -16
  58. data/docs/debugging.md +229 -229
  59. data/docs/developer/branching.md +10 -10
  60. data/docs/developer/database-diff.md +10 -10
  61. data/docs/developer/local-development.md +17 -17
  62. data/docs/developer/snapshots.md +9 -9
  63. data/docs/developer/temporal-data.md +10 -10
  64. data/docs/developer-guide.md +297 -297
  65. data/docs/getting-started/first-database.md +16 -16
  66. data/docs/getting-started/first-query.md +13 -13
  67. data/docs/getting-started/installation.md +19 -19
  68. data/docs/getting-started/local-to-production.md +300 -300
  69. data/docs/getting-started/quickstart.md +17 -17
  70. data/docs/getting-started/rails.md +16 -16
  71. data/docs/hardening_backlog.md +93 -93
  72. data/docs/lessons-learned.md +112 -112
  73. data/docs/operations/backups.md +33 -33
  74. data/docs/operations/disaster-recovery.md +31 -31
  75. data/docs/operations/failover.md +30 -30
  76. data/docs/operations/monitoring.md +25 -25
  77. data/docs/operations/production-guide.md +295 -295
  78. data/docs/operations/production-runbook.md +45 -45
  79. data/docs/operations/replication.md +33 -33
  80. data/docs/operations/restore.md +6 -6
  81. data/docs/operations/runbook.md +34 -34
  82. data/docs/operations/upgrades.md +14 -14
  83. data/docs/operations/workload-testing.md +17 -17
  84. data/docs/production-readiness.md +118 -118
  85. data/docs/production_validation.md +150 -150
  86. data/docs/rails/active-record.md +11 -11
  87. data/docs/rails/compatibility-guide.md +90 -90
  88. data/docs/rails/database-yml.md +92 -92
  89. data/docs/rails/installation.md +17 -17
  90. data/docs/rails/migrations.md +17 -17
  91. data/docs/rails/production.md +82 -82
  92. data/docs/rails/troubleshooting.md +18 -18
  93. data/docs/release.md +25 -25
  94. data/docs/server/architecture.md +10 -10
  95. data/docs/server/authentication.md +10 -10
  96. data/docs/server/configuration.md +16 -16
  97. data/docs/server/connection-pooling.md +10 -10
  98. data/docs/server/deployment.md +10 -10
  99. data/docs/server/protocol.md +12 -12
  100. data/docs/sql/compatibility-guide.md +82 -82
  101. data/docs/sql/compatibility.md +39 -39
  102. data/docs/sql/data-types.md +10 -10
  103. data/docs/sql/functions.md +9 -9
  104. data/docs/sql/joins.md +9 -9
  105. data/docs/sql/operators.md +9 -9
  106. data/docs/sql/sqlite-compatibility.md +21 -21
  107. data/docs/sql/syntax.md +10 -10
  108. data/docs/sql/transactions.md +10 -10
  109. data/docs/troubleshooting.md +244 -244
  110. data/lessons/01-foundations.md +73 -0
  111. data/lessons/02-local-development.md +121 -0
  112. data/lessons/03-embedded-rubydb.md +99 -0
  113. data/lessons/04-rails-complex-apps.md +138 -0
  114. data/lessons/05-rubydb-production-server.md +237 -0
  115. data/lessons/06-postgresql-massive-apps.md +96 -0
  116. data/lessons/07-hybrid-microservices.md +94 -0
  117. data/lessons/08-migrations-backups-recovery.md +86 -0
  118. data/lessons/09-observability-security-scale.md +87 -0
  119. data/lessons/10-release-readiness.md +125 -0
  120. data/lib/rubydb/backup/archive.rb +332 -334
  121. data/lib/rubydb/backup/backup.rb +400 -401
  122. data/lib/rubydb/backup/incremental.rb +349 -353
  123. data/lib/rubydb/backup/restore.rb +289 -290
  124. data/lib/rubydb/backup/snapshot.rb +265 -267
  125. data/lib/rubydb/backup/verification.rb +276 -279
  126. data/lib/rubydb/branching/branch.rb +181 -181
  127. data/lib/rubydb/branching/branch_manager.rb +307 -311
  128. data/lib/rubydb/branching/branch_metadata.rb +140 -140
  129. data/lib/rubydb/branching/checkout.rb +165 -166
  130. data/lib/rubydb/branching/copy_on_write.rb +272 -272
  131. data/lib/rubydb/branching/diff.rb +137 -138
  132. data/lib/rubydb/branching/merge.rb +282 -285
  133. data/lib/rubydb/build_info.rb +15 -15
  134. data/lib/rubydb/catalog/catalog.rb +391 -391
  135. data/lib/rubydb/catalog/column.rb +112 -112
  136. data/lib/rubydb/catalog/constraint.rb +180 -180
  137. data/lib/rubydb/catalog/database.rb +184 -184
  138. data/lib/rubydb/catalog/index.rb +97 -97
  139. data/lib/rubydb/catalog/schema.rb +103 -103
  140. data/lib/rubydb/catalog/sequence.rb +90 -90
  141. data/lib/rubydb/catalog/system_catalog.rb +698 -698
  142. data/lib/rubydb/catalog/table.rb +178 -178
  143. data/lib/rubydb/catalog/trigger.rb +102 -102
  144. data/lib/rubydb/catalog/view.rb +66 -66
  145. data/lib/rubydb/cli/application.rb +163 -163
  146. data/lib/rubydb/cli/commands/backup.rb +80 -81
  147. data/lib/rubydb/cli/commands/branch.rb +72 -72
  148. data/lib/rubydb/cli/commands/checkout.rb +54 -54
  149. data/lib/rubydb/cli/commands/create.rb +58 -58
  150. data/lib/rubydb/cli/commands/diff.rb +76 -77
  151. data/lib/rubydb/cli/commands/doctor.rb +74 -74
  152. data/lib/rubydb/cli/commands/drop.rb +57 -57
  153. data/lib/rubydb/cli/commands/init.rb +101 -102
  154. data/lib/rubydb/cli/commands/inspect.rb +95 -95
  155. data/lib/rubydb/cli/commands/merge.rb +63 -63
  156. data/lib/rubydb/cli/commands/migrate.rb +62 -62
  157. data/lib/rubydb/cli/commands/restart.rb +42 -39
  158. data/lib/rubydb/cli/commands/restore.rb +121 -121
  159. data/lib/rubydb/cli/commands/shell.rb +365 -365
  160. data/lib/rubydb/cli/commands/snapshot.rb +79 -79
  161. data/lib/rubydb/cli/commands/start.rb +88 -82
  162. data/lib/rubydb/cli/commands/status.rb +96 -92
  163. data/lib/rubydb/cli/commands/stop.rb +47 -47
  164. data/lib/rubydb/cli/commands/vacuum.rb +58 -58
  165. data/lib/rubydb/cli/formatter.rb +221 -221
  166. data/lib/rubydb/cli/output.rb +168 -168
  167. data/lib/rubydb/client/client.rb +302 -304
  168. data/lib/rubydb/client/connection.rb +414 -415
  169. data/lib/rubydb/client/connection_pool.rb +168 -168
  170. data/lib/rubydb/client/connection_url.rb +96 -96
  171. data/lib/rubydb/client/prepared_statement.rb +60 -60
  172. data/lib/rubydb/client/result.rb +123 -123
  173. data/lib/rubydb/client/statement.rb +52 -52
  174. data/lib/rubydb/client/transaction.rb +130 -130
  175. data/lib/rubydb/concurrency/concurrency.rb +19 -19
  176. data/lib/rubydb/concurrency/deadlock_detector.rb +148 -150
  177. data/lib/rubydb/concurrency/latch.rb +101 -101
  178. data/lib/rubydb/concurrency/lock_graph.rb +163 -165
  179. data/lib/rubydb/concurrency/mutex.rb +181 -183
  180. data/lib/rubydb/concurrency/rw_lock.rb +180 -180
  181. data/lib/rubydb/concurrency/scheduler.rb +248 -250
  182. data/lib/rubydb/concurrency/worker_pool.rb +145 -143
  183. data/lib/rubydb/configuration/config.rb +170 -170
  184. data/lib/rubydb/configuration/defaults.rb +179 -179
  185. data/lib/rubydb/configuration/environment.rb +152 -152
  186. data/lib/rubydb/configuration/parser.rb +185 -185
  187. data/lib/rubydb/configuration/validation.rb +221 -221
  188. data/lib/rubydb/constants.rb +74 -74
  189. data/lib/rubydb/constraints/check.rb +181 -181
  190. data/lib/rubydb/constraints/constraint.rb +101 -101
  191. data/lib/rubydb/constraints/foreign_key.rb +130 -130
  192. data/lib/rubydb/constraints/not_null.rb +64 -64
  193. data/lib/rubydb/constraints/primary_key.rb +99 -99
  194. data/lib/rubydb/constraints/unique.rb +106 -108
  195. data/lib/rubydb/constraints/validator.rb +349 -350
  196. data/lib/rubydb/errors/authentication_error.rb +10 -10
  197. data/lib/rubydb/errors/authorization_error.rb +23 -23
  198. data/lib/rubydb/errors/client_error.rb +10 -10
  199. data/lib/rubydb/errors/configuration_error.rb +10 -10
  200. data/lib/rubydb/errors/connection_error.rb +10 -10
  201. data/lib/rubydb/errors/constraint_error.rb +23 -23
  202. data/lib/rubydb/errors/corruption_error.rb +10 -10
  203. data/lib/rubydb/errors/database_error.rb +10 -10
  204. data/lib/rubydb/errors/error.rb +20 -20
  205. data/lib/rubydb/errors/execution_error.rb +10 -10
  206. data/lib/rubydb/errors/parser_error.rb +10 -10
  207. data/lib/rubydb/errors/recovery_error.rb +10 -10
  208. data/lib/rubydb/errors/replication_error.rb +10 -10
  209. data/lib/rubydb/errors/server_error.rb +6 -6
  210. data/lib/rubydb/errors/storage_error.rb +10 -10
  211. data/lib/rubydb/errors/transaction_error.rb +10 -10
  212. data/lib/rubydb/execution/aggregate_executor.rb +134 -138
  213. data/lib/rubydb/execution/delete_executor.rb +110 -112
  214. data/lib/rubydb/execution/distinct_executor.rb +131 -135
  215. data/lib/rubydb/execution/executor.rb +1182 -1188
  216. data/lib/rubydb/execution/expression.rb +191 -193
  217. data/lib/rubydb/execution/index_scan.rb +142 -142
  218. data/lib/rubydb/execution/insert_executor.rb +215 -217
  219. data/lib/rubydb/execution/join_executor.rb +243 -249
  220. data/lib/rubydb/execution/limit_executor.rb +83 -85
  221. data/lib/rubydb/execution/optimizer.rb +227 -215
  222. data/lib/rubydb/execution/plan.rb +355 -353
  223. data/lib/rubydb/execution/planner.rb +542 -536
  224. data/lib/rubydb/execution/predicate.rb +235 -235
  225. data/lib/rubydb/execution/scan.rb +49 -49
  226. data/lib/rubydb/execution/sequential_scan.rb +63 -63
  227. data/lib/rubydb/execution/sort_executor.rb +180 -185
  228. data/lib/rubydb/execution/update_executor.rb +160 -162
  229. data/lib/rubydb/functions/aggregate.rb +70 -70
  230. data/lib/rubydb/functions/date_functions.rb +274 -278
  231. data/lib/rubydb/functions/function.rb +85 -85
  232. data/lib/rubydb/functions/json_functions.rb +231 -215
  233. data/lib/rubydb/functions/numeric_functions.rb +346 -346
  234. data/lib/rubydb/functions/scalar.rb +52 -52
  235. data/lib/rubydb/functions/string_functions.rb +383 -383
  236. data/lib/rubydb/functions/system_functions.rb +258 -246
  237. data/lib/rubydb/history/as_of.rb +238 -238
  238. data/lib/rubydb/history/change.rb +105 -105
  239. data/lib/rubydb/history/history.rb +131 -131
  240. data/lib/rubydb/history/history_manager.rb +228 -229
  241. data/lib/rubydb/history/temporal_query.rb +202 -202
  242. data/lib/rubydb/history/timeline.rb +144 -144
  243. data/lib/rubydb/indexes/btree.rb +186 -186
  244. data/lib/rubydb/indexes/btree_cursor.rb +258 -258
  245. data/lib/rubydb/indexes/btree_node.rb +384 -385
  246. data/lib/rubydb/indexes/hash_index.rb +150 -150
  247. data/lib/rubydb/indexes/index.rb +71 -71
  248. data/lib/rubydb/indexes/index_manager.rb +408 -406
  249. data/lib/rubydb/indexes/index_scan.rb +466 -470
  250. data/lib/rubydb/migrations/migration.rb +253 -254
  251. data/lib/rubydb/migrations/migration_lock.rb +146 -146
  252. data/lib/rubydb/migrations/migration_manager.rb +187 -176
  253. data/lib/rubydb/migrations/migration_version.rb +71 -71
  254. data/lib/rubydb/migrations/schema_diff.rb +211 -211
  255. data/lib/rubydb/migrations/schema_version.rb +64 -64
  256. data/lib/rubydb/monitoring/events.rb +155 -160
  257. data/lib/rubydb/monitoring/health.rb +216 -222
  258. data/lib/rubydb/monitoring/logger.rb +188 -193
  259. data/lib/rubydb/monitoring/metrics.rb +363 -359
  260. data/lib/rubydb/monitoring/performance.rb +176 -176
  261. data/lib/rubydb/monitoring/statistics.rb +168 -170
  262. data/lib/rubydb/mvcc/garbage_collector.rb +199 -199
  263. data/lib/rubydb/mvcc/mvcc.rb +16 -16
  264. data/lib/rubydb/mvcc/snapshot.rb +146 -147
  265. data/lib/rubydb/mvcc/vacuum.rb +180 -180
  266. data/lib/rubydb/mvcc/version.rb +106 -106
  267. data/lib/rubydb/mvcc/version_store.rb +396 -398
  268. data/lib/rubydb/mvcc/visibility.rb +107 -109
  269. data/lib/rubydb/protocol/capabilities.rb +125 -125
  270. data/lib/rubydb/protocol/decoder.rb +142 -145
  271. data/lib/rubydb/protocol/encoder.rb +131 -136
  272. data/lib/rubydb/protocol/handshake.rb +306 -305
  273. data/lib/rubydb/protocol/message.rb +121 -121
  274. data/lib/rubydb/protocol/parameter_binder.rb +101 -0
  275. data/lib/rubydb/protocol/protocol.rb +276 -277
  276. data/lib/rubydb/protocol/version.rb +54 -54
  277. data/lib/rubydb/rails/adapter.rb +245 -239
  278. data/lib/rubydb/rails/connection.rb +312 -314
  279. data/lib/rubydb/rails/database_statements.rb +122 -122
  280. data/lib/rubydb/rails/migration.rb +131 -131
  281. data/lib/rubydb/rails/quoting.rb +109 -109
  282. data/lib/rubydb/rails/result.rb +117 -117
  283. data/lib/rubydb/rails/schema_statements.rb +339 -339
  284. data/lib/rubydb/rails/transaction.rb +105 -105
  285. data/lib/rubydb/rails/type.rb +126 -126
  286. data/lib/rubydb/recovery/checkpoint.rb +261 -257
  287. data/lib/rubydb/recovery/consistency.rb +457 -467
  288. data/lib/rubydb/recovery/corruption_detector.rb +5 -5
  289. data/lib/rubydb/recovery/crash_recovery.rb +381 -387
  290. data/lib/rubydb/recovery/recovery_manager.rb +204 -206
  291. data/lib/rubydb/recovery/redo.rb +235 -237
  292. data/lib/rubydb/recovery/undo.rb +204 -206
  293. data/lib/rubydb/replication/failover.rb +5 -5
  294. data/lib/rubydb/replication/fencing.rb +63 -63
  295. data/lib/rubydb/replication/primary.rb +461 -450
  296. data/lib/rubydb/replication/replica.rb +382 -384
  297. data/lib/rubydb/replication/replication_log.rb +194 -200
  298. data/lib/rubydb/replication/replication_manager.rb +307 -308
  299. data/lib/rubydb/replication/replication_slot.rb +293 -295
  300. data/lib/rubydb/replication/replication_stream.rb +198 -201
  301. data/lib/rubydb/rubydb.rb +564 -560
  302. data/lib/rubydb/security/access_control.rb +252 -254
  303. data/lib/rubydb/security/audit_log.rb +209 -213
  304. data/lib/rubydb/security/authentication.rb +302 -302
  305. data/lib/rubydb/security/authorization.rb +282 -282
  306. data/lib/rubydb/security/credentials.rb +192 -196
  307. data/lib/rubydb/security/password.rb +205 -215
  308. data/lib/rubydb/security/permissions.rb +74 -74
  309. data/lib/rubydb/security/role.rb +99 -101
  310. data/lib/rubydb/security/user.rb +86 -86
  311. data/lib/rubydb/server/connection.rb +383 -366
  312. data/lib/rubydb/server/connection_pool.rb +193 -193
  313. data/lib/rubydb/server/lifecycle.rb +227 -228
  314. data/lib/rubydb/server/listener.rb +139 -136
  315. data/lib/rubydb/server/request_handler.rb +277 -276
  316. data/lib/rubydb/server/server.rb +363 -364
  317. data/lib/rubydb/server/session.rb +371 -369
  318. data/lib/rubydb/server/worker.rb +206 -210
  319. data/lib/rubydb/server/worker_pool.rb +168 -168
  320. data/lib/rubydb/sql/ast/alter_table.rb +169 -169
  321. data/lib/rubydb/sql/ast/begin_transaction.rb +47 -47
  322. data/lib/rubydb/sql/ast/commit.rb +37 -37
  323. data/lib/rubydb/sql/ast/constraint.rb +92 -83
  324. data/lib/rubydb/sql/ast/create_database.rb +41 -41
  325. data/lib/rubydb/sql/ast/create_index.rb +61 -61
  326. data/lib/rubydb/sql/ast/create_schema.rb +52 -52
  327. data/lib/rubydb/sql/ast/create_table.rb +187 -187
  328. data/lib/rubydb/sql/ast/delete.rb +54 -54
  329. data/lib/rubydb/sql/ast/drop_database.rb +41 -41
  330. data/lib/rubydb/sql/ast/drop_index.rb +41 -41
  331. data/lib/rubydb/sql/ast/drop_schema.rb +49 -49
  332. data/lib/rubydb/sql/ast/drop_table.rb +49 -49
  333. data/lib/rubydb/sql/ast/explain.rb +64 -64
  334. data/lib/rubydb/sql/ast/expression.rb +617 -604
  335. data/lib/rubydb/sql/ast/insert.rb +66 -66
  336. data/lib/rubydb/sql/ast/node.rb +42 -42
  337. data/lib/rubydb/sql/ast/rollback.rb +63 -63
  338. data/lib/rubydb/sql/ast/savepoint.rb +59 -59
  339. data/lib/rubydb/sql/ast/select.rb +88 -88
  340. data/lib/rubydb/sql/ast/set_operation.rb +22 -20
  341. data/lib/rubydb/sql/ast/trigger.rb +35 -29
  342. data/lib/rubydb/sql/ast/update.rb +88 -88
  343. data/lib/rubydb/sql/ast/vacuum.rb +19 -19
  344. data/lib/rubydb/sql/ast/view.rb +38 -32
  345. data/lib/rubydb/sql/ast/with.rb +32 -32
  346. data/lib/rubydb/sql/grammar.rb +86 -86
  347. data/lib/rubydb/sql/keywords.rb +156 -156
  348. data/lib/rubydb/sql/lexer.rb +209 -214
  349. data/lib/rubydb/sql/operators.rb +100 -100
  350. data/lib/rubydb/sql/parser.rb +1167 -1170
  351. data/lib/rubydb/sql/planner/analyzer.rb +283 -302
  352. data/lib/rubydb/sql/planner/binder.rb +537 -550
  353. data/lib/rubydb/sql/planner/type_checker.rb +427 -431
  354. data/lib/rubydb/sql/token.rb +210 -210
  355. data/lib/rubydb/storage/buffer_frame.rb +44 -44
  356. data/lib/rubydb/storage/buffer_pool.rb +155 -155
  357. data/lib/rubydb/storage/database_lock.rb +74 -74
  358. data/lib/rubydb/storage/deserializer.rb +332 -342
  359. data/lib/rubydb/storage/engine.rb +2347 -2330
  360. data/lib/rubydb/storage/file_manager.rb +191 -187
  361. data/lib/rubydb/storage/free_space_map.rb +79 -81
  362. data/lib/rubydb/storage/page.rb +92 -94
  363. data/lib/rubydb/storage/page_allocator.rb +852 -855
  364. data/lib/rubydb/storage/page_header.rb +63 -67
  365. data/lib/rubydb/storage/page_manager.rb +127 -131
  366. data/lib/rubydb/storage/record.rb +58 -58
  367. data/lib/rubydb/storage/row.rb +78 -78
  368. data/lib/rubydb/storage/serializer.rb +51 -51
  369. data/lib/rubydb/storage/storage_layout.rb +151 -151
  370. data/lib/rubydb/storage/storage_manager.rb +114 -114
  371. data/lib/rubydb/storage/tuple.rb +458 -461
  372. data/lib/rubydb/storage/visibility_map.rb +964 -973
  373. data/lib/rubydb/transactions/commit_manager.rb +219 -220
  374. data/lib/rubydb/transactions/isolation.rb +98 -98
  375. data/lib/rubydb/transactions/lock.rb +76 -76
  376. data/lib/rubydb/transactions/lock_manager.rb +359 -362
  377. data/lib/rubydb/transactions/savepoint.rb +142 -143
  378. data/lib/rubydb/transactions/transaction.rb +214 -215
  379. data/lib/rubydb/transactions/transaction_id.rb +84 -84
  380. data/lib/rubydb/transactions/transaction_log.rb +256 -257
  381. data/lib/rubydb/transactions/transaction_manager.rb +434 -435
  382. data/lib/rubydb/types/bigint.rb +36 -36
  383. data/lib/rubydb/types/blob.rb +37 -37
  384. data/lib/rubydb/types/boolean.rb +34 -34
  385. data/lib/rubydb/types/date.rb +39 -39
  386. data/lib/rubydb/types/decimal.rb +48 -48
  387. data/lib/rubydb/types/float.rb +34 -34
  388. data/lib/rubydb/types/integer.rb +36 -36
  389. data/lib/rubydb/types/json.rb +41 -41
  390. data/lib/rubydb/types/null.rb +34 -34
  391. data/lib/rubydb/types/smallint.rb +36 -36
  392. data/lib/rubydb/types/text.rb +37 -37
  393. data/lib/rubydb/types/time.rb +46 -46
  394. data/lib/rubydb/types/timestamp.rb +39 -39
  395. data/lib/rubydb/types/type.rb +119 -119
  396. data/lib/rubydb/types/uuid.rb +47 -47
  397. data/lib/rubydb/types/varchar.rb +37 -37
  398. data/lib/rubydb/version.rb +32 -32
  399. data/lib/rubydb/wal/archive.rb +190 -193
  400. data/lib/rubydb/wal/checkpoint.rb +181 -183
  401. data/lib/rubydb/wal/lsn.rb +94 -94
  402. data/lib/rubydb/wal/reader.rb +259 -260
  403. data/lib/rubydb/wal/record.rb +105 -105
  404. data/lib/rubydb/wal/segment.rb +193 -193
  405. data/lib/rubydb/wal/wal.rb +480 -452
  406. data/lib/rubydb/wal/writer.rb +236 -236
  407. data/lib/rubydb.rb +7 -7
  408. data/packaging/docker/docker-compose.failover.yml +43 -43
  409. data/packaging/homebrew/rubydb.rb +19 -19
  410. data/rubydb.gemspec +60 -57
  411. data/scripts/benchmark +7 -7
  412. data/scripts/durability_drill +37 -37
  413. data/scripts/fuzz +63 -63
  414. data/scripts/release +45 -40
  415. data/scripts/release_check +43 -43
  416. data/scripts/replication_failover_drill +268 -250
  417. data/scripts/replication_network_failover_drill +287 -255
  418. data/scripts/restore_drill +45 -45
  419. data/scripts/security +45 -0
  420. metadata +54 -1
@@ -1,159 +1,163 @@
1
- # frozen_string_literal: true
2
-
3
- module RubyDB
4
- module Execution
5
- # Planner - Converts AST to execution plan
6
- class Planner
7
- attr_reader :engine, :catalog
8
-
9
- def initialize(engine)
10
- @engine = engine
11
- @catalog = engine.catalog if engine.respond_to?(:catalog)
12
- @stats = {
13
- plans_created: 0,
14
- optimizations_applied: 0
15
- }
16
- end
17
-
18
- def plan(statement)
19
- @stats[:plans_created] += 1
20
-
21
- plan = case statement
22
- when SQL::AST::With
23
- Plan::With.new(statement.ctes.map { |name, query| [name, plan(query)] }, plan(statement.query), recursive: statement.recursive)
24
- when SQL::AST::Select
25
- plan_select(statement)
26
- when SQL::AST::SetOperation
27
- Plan::SetOperation.new(plan(statement.left), plan(statement.right), statement.operator, all: statement.all)
28
- when SQL::AST::Insert
29
- plan_insert(statement)
30
- when SQL::AST::Update
31
- plan_update(statement)
32
- when SQL::AST::Delete
33
- plan_delete(statement)
34
- when SQL::AST::CreateTable
35
- plan_create_table(statement)
36
- when SQL::AST::DropTable
37
- plan_drop_table(statement)
38
- when SQL::AST::CreateDatabase
39
- Plan::CreateDatabase.new(statement.name, if_not_exists: statement.if_not_exists)
40
- when SQL::AST::DropDatabase
41
- Plan::DropDatabase.new(statement.name, if_exists: statement.if_exists)
42
- when SQL::AST::CreateSchema
43
- Plan::CreateSchema.new(statement.name, if_not_exists: statement.if_not_exists, authorization: statement.authorization)
44
- when SQL::AST::DropSchema
45
- Plan::DropSchema.new(statement.name, if_exists: statement.if_exists, cascade: statement.cascade)
46
- when SQL::AST::CreateView
47
- Plan::CreateView.new(statement.name, statement.query, if_not_exists: statement.if_not_exists)
48
- when SQL::AST::DropView
49
- Plan::DropView.new(statement.name, if_exists: statement.if_exists)
50
- when SQL::AST::CreateTrigger
51
- Plan::CreateTrigger.new(statement.name, statement.timing, statement.event, statement.table_name, statement.function_name)
52
- when SQL::AST::DropTrigger
53
- Plan::DropTrigger.new(statement.name, if_exists: statement.if_exists)
54
- when SQL::AST::Vacuum
55
- Plan::Vacuum.new
56
- when SQL::AST::AlterTableAddColumn
57
- options = statement.options.transform_values { |value| value.respond_to?(:value) ? value.value : value }
58
- Plan::AlterTableAddColumn.new(statement.table_name, statement.column_name, statement.column_type, options)
59
- when SQL::AST::AlterTableDropColumn
60
- Plan::AlterTableDropColumn.new(statement.table_name, statement.column_name)
61
- when SQL::AST::AlterTableAddConstraint
62
- Plan::AlterTableAddConstraint.new(statement.table_name, statement.constraint)
63
- when SQL::AST::AlterTableDropConstraint
64
- Plan::AlterTableDropConstraint.new(statement.table_name, statement.constraint_name)
65
- when SQL::AST::CreateIndex
66
- plan_create_index(statement)
67
- when SQL::AST::DropIndex
68
- plan_drop_index(statement)
69
- when SQL::AST::BeginTransaction
70
- plan_begin_transaction(statement)
71
- when SQL::AST::Commit
72
- plan_commit(statement)
73
- when SQL::AST::Rollback
74
- plan_rollback(statement)
75
- when SQL::AST::Savepoint
76
- Plan::Savepoint.new(statement.name)
77
- when SQL::AST::RollbackToSavepoint
78
- Plan::RollbackToSavepoint.new(statement.savepoint_name)
79
- when SQL::AST::ReleaseSavepoint
80
- Plan::ReleaseSavepoint.new(statement.name)
81
- when SQL::AST::Explain
82
- plan_explain(statement)
83
- else
84
- raise ExecutionError, "Unknown statement type: #{statement.class}"
85
- end
86
-
87
- # Apply optimizations
88
- plan = optimize(plan)
89
-
90
- plan
91
- end
92
-
93
- def plan_select(statement)
94
- table_name = statement.table_name
95
- columns = statement.columns.map { |col| col.expression.name rescue col.to_s }
96
-
97
- plan = Plan::Select.new(table_name, columns)
98
- plan.set_source_reference(statement.from)
99
-
100
- if statement.joins&.any?
101
- plan.set_joins(statement.joins.map do |join|
102
- { type: join.type, table: join.table, predicate: build_predicate(join.condition) }
103
- end)
104
- end
105
-
106
- # Handle projections
107
- if statement.columns.any?
108
- plan.set_projections(statement.columns)
109
- end
110
-
111
- # Handle WHERE clause
112
- if statement.where
113
- predicate = build_predicate(statement.where)
114
- plan.set_predicate(predicate)
115
- end
116
-
117
- # Handle ORDER BY
118
- if statement.order_by&.any?
119
- order_by = statement.order_by.map do |order|
120
- column = order.expression.respond_to?(:name) ? order.expression.name : order.expression.to_s
121
- table = order.expression.respond_to?(:table) ? order.expression.table : nil
122
- { column: table ? "#{table}.#{column}" : column, direction: order.direction }
123
- end
124
- plan.set_order_by(order_by)
125
- end
126
-
127
- # Handle GROUP BY
128
- aggregates = statement.columns.select do |column|
129
- expression = column.respond_to?(:expression) ? column.expression : column
130
- expression.is_a?(SQL::AST::FunctionCall) && !expression.window &&
131
- %w[COUNT SUM AVG MIN MAX].include?(expression.name.to_s.upcase)
132
- end
133
- if statement.group_by&.any? || aggregates.any?
134
- plan.set_group_by(statement.group_by || [], aggregates)
135
- end
136
- plan.set_having(build_predicate(statement.having)) if statement.having
137
-
138
- # Handle LIMIT and OFFSET
139
- if statement.limit
140
- plan.set_limit(statement.limit.value, statement.offset&.value)
141
- end
142
-
143
- # Handle DISTINCT
144
- plan.set_distinct(statement.distinct) if statement.distinct
145
-
146
- # Choose scan method
147
- choose_scan_method(plan)
148
-
149
- # Estimate cost
150
- estimate_cost(plan)
151
-
152
- plan
153
- end
154
-
155
- def plan_insert(statement)
156
- Plan::Insert.new(
1
+ # frozen_string_literal: true
2
+
3
+ module RubyDB
4
+ module Execution
5
+ # Planner - Converts AST to execution plan
6
+ class Planner
7
+ attr_reader :engine, :catalog
8
+
9
+ def initialize(engine)
10
+ @engine = engine
11
+ @catalog = engine.catalog if engine.respond_to?(:catalog)
12
+ @stats = {
13
+ plans_created: 0,
14
+ optimizations_applied: 0
15
+ }
16
+ end
17
+
18
+ def plan(statement)
19
+ @stats[:plans_created] += 1
20
+
21
+ plan = case statement
22
+ when SQL::AST::With
23
+ Plan::With.new(statement.ctes.map { |name, query| [name, plan(query)] }, plan(statement.query), recursive: statement.recursive)
24
+ when SQL::AST::Select
25
+ plan_select(statement)
26
+ when SQL::AST::SetOperation
27
+ Plan::SetOperation.new(plan(statement.left), plan(statement.right), statement.operator, all: statement.all)
28
+ when SQL::AST::Insert
29
+ plan_insert(statement)
30
+ when SQL::AST::Update
31
+ plan_update(statement)
32
+ when SQL::AST::Delete
33
+ plan_delete(statement)
34
+ when SQL::AST::CreateTable
35
+ plan_create_table(statement)
36
+ when SQL::AST::DropTable
37
+ plan_drop_table(statement)
38
+ when SQL::AST::CreateDatabase
39
+ Plan::CreateDatabase.new(statement.name, if_not_exists: statement.if_not_exists)
40
+ when SQL::AST::DropDatabase
41
+ Plan::DropDatabase.new(statement.name, if_exists: statement.if_exists)
42
+ when SQL::AST::CreateSchema
43
+ Plan::CreateSchema.new(statement.name, if_not_exists: statement.if_not_exists, authorization: statement.authorization)
44
+ when SQL::AST::DropSchema
45
+ Plan::DropSchema.new(statement.name, if_exists: statement.if_exists, cascade: statement.cascade)
46
+ when SQL::AST::CreateView
47
+ Plan::CreateView.new(statement.name, statement.query, if_not_exists: statement.if_not_exists)
48
+ when SQL::AST::DropView
49
+ Plan::DropView.new(statement.name, if_exists: statement.if_exists)
50
+ when SQL::AST::CreateTrigger
51
+ Plan::CreateTrigger.new(statement.name, statement.timing, statement.event, statement.table_name, statement.function_name)
52
+ when SQL::AST::DropTrigger
53
+ Plan::DropTrigger.new(statement.name, if_exists: statement.if_exists)
54
+ when SQL::AST::Vacuum
55
+ Plan::Vacuum.new
56
+ when SQL::AST::AlterTableAddColumn
57
+ options = statement.options.transform_values { |value| value.respond_to?(:value) ? value.value : value }
58
+ Plan::AlterTableAddColumn.new(statement.table_name, statement.column_name, statement.column_type, options)
59
+ when SQL::AST::AlterTableDropColumn
60
+ Plan::AlterTableDropColumn.new(statement.table_name, statement.column_name)
61
+ when SQL::AST::AlterTableAddConstraint
62
+ Plan::AlterTableAddConstraint.new(statement.table_name, statement.constraint)
63
+ when SQL::AST::AlterTableDropConstraint
64
+ Plan::AlterTableDropConstraint.new(statement.table_name, statement.constraint_name)
65
+ when SQL::AST::CreateIndex
66
+ plan_create_index(statement)
67
+ when SQL::AST::DropIndex
68
+ plan_drop_index(statement)
69
+ when SQL::AST::BeginTransaction
70
+ plan_begin_transaction(statement)
71
+ when SQL::AST::Commit
72
+ plan_commit(statement)
73
+ when SQL::AST::Rollback
74
+ plan_rollback(statement)
75
+ when SQL::AST::Savepoint
76
+ Plan::Savepoint.new(statement.name)
77
+ when SQL::AST::RollbackToSavepoint
78
+ Plan::RollbackToSavepoint.new(statement.savepoint_name)
79
+ when SQL::AST::ReleaseSavepoint
80
+ Plan::ReleaseSavepoint.new(statement.name)
81
+ when SQL::AST::Explain
82
+ plan_explain(statement)
83
+ else
84
+ raise ExecutionError, "Unknown statement type: #{statement.class}"
85
+ end
86
+
87
+ # Apply optimizations
88
+ optimize(plan)
89
+ end
90
+
91
+ def plan_select(statement)
92
+ table_name = statement.table_name
93
+ columns = statement.columns.map { |col|
94
+ begin
95
+ col.expression.name
96
+ rescue
97
+ col.to_s
98
+ end
99
+ }
100
+
101
+ plan = Plan::Select.new(table_name, columns)
102
+ plan.set_source_reference(statement.from)
103
+
104
+ if statement.joins&.any?
105
+ plan.set_joins(statement.joins.map do |join|
106
+ {type: join.type, table: join.table, predicate: build_predicate(join.condition)}
107
+ end)
108
+ end
109
+
110
+ # Handle projections
111
+ if statement.columns.any?
112
+ plan.set_projections(statement.columns)
113
+ end
114
+
115
+ # Handle WHERE clause
116
+ if statement.where
117
+ predicate = build_predicate(statement.where)
118
+ plan.set_predicate(predicate)
119
+ end
120
+
121
+ # Handle ORDER BY
122
+ if statement.order_by&.any?
123
+ order_by = statement.order_by.map do |order|
124
+ column = order.expression.respond_to?(:name) ? order.expression.name : order.expression.to_s
125
+ table = order.expression.respond_to?(:table) ? order.expression.table : nil
126
+ {column: table ? "#{table}.#{column}" : column, direction: order.direction}
127
+ end
128
+ plan.set_order_by(order_by)
129
+ end
130
+
131
+ # Handle GROUP BY
132
+ aggregates = statement.columns.select do |column|
133
+ expression = column.respond_to?(:expression) ? column.expression : column
134
+ expression.is_a?(SQL::AST::FunctionCall) && !expression.window &&
135
+ %w[COUNT SUM AVG MIN MAX].include?(expression.name.to_s.upcase)
136
+ end
137
+ if statement.group_by&.any? || aggregates.any?
138
+ plan.set_group_by(statement.group_by || [], aggregates)
139
+ end
140
+ plan.set_having(build_predicate(statement.having)) if statement.having
141
+
142
+ # Handle LIMIT and OFFSET
143
+ if statement.limit
144
+ plan.set_limit(statement.limit.value, statement.offset&.value)
145
+ end
146
+
147
+ # Handle DISTINCT
148
+ plan.set_distinct(statement.distinct) if statement.distinct
149
+
150
+ # Choose scan method
151
+ choose_scan_method(plan)
152
+
153
+ # Estimate cost
154
+ estimate_cost(plan)
155
+
156
+ plan
157
+ end
158
+
159
+ def plan_insert(statement)
160
+ Plan::Insert.new(
157
161
  statement.table,
158
162
  statement.columns,
159
163
  statement.values,
@@ -161,32 +165,32 @@ module RubyDB
161
165
  on_conflict: statement.on_conflict,
162
166
  default_values: statement.default_values?
163
167
  )
164
- end
165
-
166
- def plan_update(statement)
167
- assignments = statement.assignments.map do |ass|
168
- { column: ass.column, value: ass.value }
169
- end
170
-
171
- plan = Plan::Update.new(statement.table, assignments)
172
-
173
- if statement.where
174
- plan.set_predicate(build_predicate(statement.where))
175
- end
176
-
177
- plan
178
- end
179
-
180
- def plan_delete(statement)
181
- plan = Plan::Delete.new(statement.table)
182
-
183
- if statement.where
184
- plan.set_predicate(build_predicate(statement.where))
185
- end
186
-
187
- plan
188
- end
189
-
168
+ end
169
+
170
+ def plan_update(statement)
171
+ assignments = statement.assignments.map do |ass|
172
+ {column: ass.column, value: ass.value}
173
+ end
174
+
175
+ plan = Plan::Update.new(statement.table, assignments)
176
+
177
+ if statement.where
178
+ plan.set_predicate(build_predicate(statement.where))
179
+ end
180
+
181
+ plan
182
+ end
183
+
184
+ def plan_delete(statement)
185
+ plan = Plan::Delete.new(statement.table)
186
+
187
+ if statement.where
188
+ plan.set_predicate(build_predicate(statement.where))
189
+ end
190
+
191
+ plan
192
+ end
193
+
190
194
  def plan_create_table(statement)
191
195
  columns = statement.columns.map do |col|
192
196
  options = col.options.dup
@@ -194,357 +198,359 @@ module RubyDB
194
198
  options[:default] = default.value if default.respond_to?(:value) && !default.is_a?(String)
195
199
  Catalog::Column.new(col.name, col.type_class, **options)
196
200
  end
197
-
198
- Plan::CreateTable.new(
199
- statement.name,
200
- columns,
201
- if_not_exists: statement.if_not_exists,
202
- constraints: statement.constraints
203
- )
204
- end
205
-
206
- def plan_drop_table(statement)
207
- Plan::DropTable.new(statement.name, if_exists: statement.if_exists)
208
- end
209
-
210
- def plan_create_index(statement)
211
- Plan::CreateIndex.new(
212
- statement.name,
213
- statement.table_name,
214
- statement.columns,
215
- unique: statement.unique,
216
- if_not_exists: statement.if_not_exists
217
- )
218
- end
219
-
220
- def plan_drop_index(statement)
221
- Plan::DropIndex.new(statement.name, if_exists: statement.if_exists)
222
- end
223
-
224
- def plan_begin_transaction(statement)
225
- Plan::BeginTransaction.new(statement.isolation_level)
226
- end
227
-
228
- def plan_commit(_statement)
229
- Plan::Commit.new
230
- end
231
-
232
- def plan_rollback(_statement)
233
- Plan::Rollback.new
234
- end
235
-
236
- def plan_explain(statement)
237
- Plan::Explain.new(statement.statement, statement.analyze)
238
- end
239
-
240
- # Predicate building
241
- def build_predicate(ast)
242
- case ast
243
- when SQL::AST::BinaryOp
244
- left = build_predicate(ast.left)
245
- right = build_predicate(ast.right)
246
-
247
- case ast.operator
248
- when :AND, :and
249
- Predicate::And.new(left, right)
250
- when :OR, :or
251
- Predicate::Or.new(left, right)
252
- when :EQ, :NE, :LT, :LTE, :GT, :GTE
253
- Predicate::Comparison.new(
254
- build_expression(ast.left),
255
- build_expression(ast.right),
256
- { EQ: :eq, NE: :ne, LT: :lt, LTE: :lte, GT: :gt, GTE: :gte }.fetch(ast.operator)
257
- )
258
- when :LIKE, :ILIKE
259
- Predicate::Like.new(
260
- build_expression(ast.left),
261
- build_expression(ast.right),
262
- ast.operator == :LIKE
263
- )
264
- else
265
- nil
266
- end
267
- when SQL::AST::UnaryOp
268
- operand = build_predicate(ast.operand)
269
- Predicate::Not.new(operand)
270
- when SQL::AST::Between
271
- Predicate::Between.new(
272
- build_expression(ast.expression),
273
- build_expression(ast.low),
274
- build_expression(ast.high)
275
- )
276
- when SQL::AST::In
277
- values = ast.values.flat_map do |value|
278
- if value.is_a?(SQL::AST::Subquery)
279
- Executor.new(@engine).execute(plan(value.query))[:rows].map { |row| Expression::Literal.new(row.values.first) }
280
- else
281
- [build_expression(value)]
282
- end
283
- end
284
- Predicate::In.new(build_expression(ast.expression), values)
285
- when SQL::AST::Exists
286
- exists = Executor.new(@engine).execute(plan(ast.query))[:rows].any?
287
- Predicate::Comparison.new(Expression::Literal.new(exists), Expression::Literal.new(true), :eq)
288
- when SQL::AST::IsNull
289
- Predicate::IsNull.new(build_expression(ast.expression), ast.negated)
290
- else
291
- Predicate::Comparison.new(
292
- build_expression(ast),
293
- build_expression(SQL::AST::Literal.new(true, :TRUE)),
294
- :eq
295
- )
296
- end
297
- end
298
-
299
- def build_expression(ast)
300
- case ast
301
- when SQL::AST::Literal
302
- Expression::Literal.new(ast.value)
303
- when SQL::AST::Identifier
304
- Expression::Column.new(ast.name, ast.table)
305
- when SQL::AST::BinaryOp
306
- left = build_expression(ast.left)
307
- right = build_expression(ast.right)
308
- operator = {
309
- PLUS: :plus, MINUS: :minus, STAR: :multiply, SLASH: :divide,
310
- PERCENT: :modulo, EQ: :eq
311
- }[ast.operator] || ast.operator
312
- Expression::BinaryOp.new(left, right, operator)
313
- when SQL::AST::UnaryOp
314
- operand = build_expression(ast.operand)
315
- Expression::UnaryOp.new(operand, ast.operator)
316
- when SQL::AST::FunctionCall
317
- args = ast.arguments.map { |arg| build_expression(arg) }
318
- Expression::Function.new(ast.name, args)
319
- when SQL::AST::Parameter
320
- Expression::Parameter.new(ast.index)
321
- else
322
- Expression::Literal.new(nil)
323
- end
324
- end
325
-
326
- # Scan method selection
327
- def choose_scan_method(plan)
328
- return unless plan.type == :select
329
-
330
- table_name = plan.table_name
331
- return plan.set_scan_type(:sequential) unless table_name
332
-
333
- # Check if there's an index that can be used
334
- if @engine.respond_to?(:index_manager)
335
- indexes = @engine.index_manager.get_indexes_for_table(table_name)
336
-
337
- # Try to find a matching index for the predicate
338
- if plan.predicate && indexes.any?
339
- index = find_matching_index(plan.predicate, indexes)
340
- if index
341
- plan.set_scan_type(:index, index)
342
- return
343
- end
344
- end
345
-
346
- # A full index scan is not a substitute for a table scan: it can
347
- # omit rows while an index is being maintained or rebuilt. Only use
348
- # an index when the predicate specifically matches its columns.
349
- end
350
-
351
- # Default to sequential scan
352
- plan.set_scan_type(:sequential)
353
- end
354
-
355
- def find_matching_index(predicate, indexes)
356
- # Get columns used in predicate
357
- columns = extract_predicate_columns(predicate)
358
-
359
- indexes.each do |index|
360
- # Check if predicate columns match index columns
361
- if index.columns.any? { |col| columns.map(&:to_s).include?(col.to_s) }
362
- return index
363
- end
364
- end
365
-
366
- nil
367
- end
368
-
369
- def extract_predicate_columns(predicate)
370
- return [] if predicate.nil?
371
-
372
- case predicate
373
- when Predicate::Comparison
374
- left_name = predicate.left.respond_to?(:name) ? predicate.left.name : nil
375
- [left_name].compact
376
- when Predicate::And, Predicate::Or
377
- extract_predicate_columns(predicate.left) + extract_predicate_columns(predicate.right)
378
- when Predicate::Not
379
- extract_predicate_columns(predicate.operand)
380
- when Predicate::Between, Predicate::In, Predicate::IsNull, Predicate::Like
381
- expr_name = predicate.expression.respond_to?(:name) ? predicate.expression.name : nil
382
- [expr_name].compact
383
- else
384
- []
385
- end
386
- end
387
-
388
- # Cost estimation
389
- def estimate_cost(plan)
390
- table_name = plan.table_name
391
- row_count = @engine.table_row_count(table_name) rescue 0
392
-
393
- case plan.scan_type
394
- when :sequential
395
- plan.set_cost(row_count, row_count)
396
- when :index
397
- # Index scan is cheaper for large tables with selective predicates
398
- if plan.predicate
399
- selectivity = estimate_selectivity(plan.predicate)
400
- estimated_rows = (row_count * selectivity).ceil
401
- plan.set_cost(estimated_rows * 0.1, estimated_rows)
402
- else
403
- plan.set_cost(row_count * 0.5, row_count)
404
- end
405
- else
406
- plan.set_cost(row_count, row_count)
407
- end
408
- end
409
-
410
- def estimate_selectivity(predicate)
411
- case predicate
412
- when Predicate::Comparison
413
- case predicate.operator
414
- when :EQ then 0.01
415
- when :NE then 0.9
416
- when :LT, :LTE, :GT, :GTE then 0.5
417
- else 0.5
418
- end
419
- when Predicate::And
420
- estimate_selectivity(predicate.left) * estimate_selectivity(predicate.right)
421
- when Predicate::Or
422
- estimate_selectivity(predicate.left) + estimate_selectivity(predicate.right) -
423
- estimate_selectivity(predicate.left) * estimate_selectivity(predicate.right)
424
- when Predicate::Not
425
- 1.0 - estimate_selectivity(predicate.operand)
426
- else
427
- 0.5
428
- end
429
- end
430
-
431
- # Optimization
432
- def optimize(plan)
433
- @stats[:optimizations_applied] += 1
434
-
435
- # Push down predicates
436
- plan = push_down_predicates(plan)
437
-
438
- # Remove redundant conditions
439
- plan = remove_redundant_conditions(plan)
440
-
441
- # Reorder only a pure inner-join chain. Outer joins are order
442
- # sensitive, and cross joins are an explicit user request, so those
443
- # remain in source order. Predicate dependencies prevent a join from
444
- # being moved before a table referenced by its ON clause.
445
- reorder_inner_joins(plan) if plan.is_a?(Plan::Select)
446
-
447
- plan
448
- end
449
-
450
- def push_down_predicates(plan)
451
- # Push predicates down to scan level
452
- # This is already done in the planner
453
- plan
454
- end
455
-
456
- def reorder_inner_joins(plan)
457
- joins = plan.joins
458
- return plan if joins.empty? || joins.any? { |join| join[:type] != :inner }
459
-
460
- available = [plan.source_reference&.name || plan.table_name].compact.map(&:to_s)
461
- pending = joins.dup
462
- ordered = []
463
-
464
- until pending.empty?
465
- eligible = pending.select do |join|
466
- table = join[:table]
467
- right_names = [table.name, table.alias_name].compact.map(&:to_s)
468
- referenced = predicate_table_names(join[:predicate])
469
- (referenced - (available + right_names)).empty?
470
- end
471
- eligible = [pending.first] if eligible.empty?
472
-
473
- chosen = eligible.min_by do |join|
474
- @engine.table_row_count(join[:table].name) rescue 1_000_000
475
- end
476
- pending.delete(chosen)
477
- ordered << chosen
478
- available.concat([chosen[:table].name, chosen[:table].alias_name].compact.map(&:to_s))
479
- end
480
-
481
- plan.set_joins(ordered)
482
- plan
483
- end
484
-
485
- def predicate_table_names(predicate)
486
- case predicate
487
- when Predicate::Comparison
488
- expression_table_names(predicate.left) + expression_table_names(predicate.right)
489
- when Predicate::And, Predicate::Or
490
- predicate_table_names(predicate.left) + predicate_table_names(predicate.right)
491
- when Predicate::Not
492
- predicate_table_names(predicate.operand)
493
- when Predicate::Between
494
- expression_table_names(predicate.expression) + expression_table_names(predicate.low) +
495
- expression_table_names(predicate.high)
496
- when Predicate::In, Predicate::IsNull, Predicate::Like
497
- expression_table_names(predicate.expression)
498
- else
499
- []
500
- end.uniq
501
- end
502
-
503
- def expression_table_names(expression)
504
- case expression
505
- when Expression::Column
506
- expression.table ? [expression.table.to_s] : []
507
- when Expression::BinaryOp
508
- expression_table_names(expression.left) + expression_table_names(expression.right)
509
- when Expression::UnaryOp
510
- expression_table_names(expression.operand)
511
- when Expression::Function
512
- expression.arguments.flat_map { |argument| expression_table_names(argument) }
513
- else
514
- []
515
- end
516
- end
517
-
518
- def remove_redundant_conditions(plan)
519
- if plan.predicate.is_a?(Predicate::And)
520
- # Simplify AND conditions
521
- simplified = simplify_and(plan.predicate)
522
- plan.set_predicate(simplified)
523
- end
524
- plan
525
- end
526
-
527
- def simplify_and(predicate)
528
- left = predicate.left
529
- right = predicate.right
530
-
531
- # Remove redundant TRUE conditions
532
- if left.is_a?(Predicate::Comparison) && left.left.is_a?(Expression::Literal) &&
533
- left.left.value == true && left.operator == :EQ
534
- return right
535
- end
536
-
537
- if right.is_a?(Predicate::Comparison) && right.left.is_a?(Expression::Literal) &&
538
- right.left.value == true && right.operator == :EQ
539
- return left
540
- end
541
-
542
- predicate
543
- end
544
-
545
- def stats
546
- @stats
547
- end
548
- end
549
- end
550
- end
201
+
202
+ Plan::CreateTable.new(
203
+ statement.name,
204
+ columns,
205
+ if_not_exists: statement.if_not_exists,
206
+ constraints: statement.constraints
207
+ )
208
+ end
209
+
210
+ def plan_drop_table(statement)
211
+ Plan::DropTable.new(statement.name, if_exists: statement.if_exists)
212
+ end
213
+
214
+ def plan_create_index(statement)
215
+ Plan::CreateIndex.new(
216
+ statement.name,
217
+ statement.table_name,
218
+ statement.columns,
219
+ unique: statement.unique,
220
+ if_not_exists: statement.if_not_exists
221
+ )
222
+ end
223
+
224
+ def plan_drop_index(statement)
225
+ Plan::DropIndex.new(statement.name, if_exists: statement.if_exists)
226
+ end
227
+
228
+ def plan_begin_transaction(statement)
229
+ Plan::BeginTransaction.new(statement.isolation_level)
230
+ end
231
+
232
+ def plan_commit(_statement)
233
+ Plan::Commit.new
234
+ end
235
+
236
+ def plan_rollback(_statement)
237
+ Plan::Rollback.new
238
+ end
239
+
240
+ def plan_explain(statement)
241
+ Plan::Explain.new(statement.statement, statement.analyze)
242
+ end
243
+
244
+ # Predicate building
245
+ def build_predicate(ast)
246
+ case ast
247
+ when SQL::AST::BinaryOp
248
+ left = build_predicate(ast.left)
249
+ right = build_predicate(ast.right)
250
+
251
+ case ast.operator
252
+ when :AND, :and
253
+ Predicate::And.new(left, right)
254
+ when :OR, :or
255
+ Predicate::Or.new(left, right)
256
+ when :EQ, :NE, :LT, :LTE, :GT, :GTE
257
+ Predicate::Comparison.new(
258
+ build_expression(ast.left),
259
+ build_expression(ast.right),
260
+ {EQ: :eq, NE: :ne, LT: :lt, LTE: :lte, GT: :gt, GTE: :gte}.fetch(ast.operator)
261
+ )
262
+ when :LIKE, :ILIKE
263
+ Predicate::Like.new(
264
+ build_expression(ast.left),
265
+ build_expression(ast.right),
266
+ ast.operator == :LIKE
267
+ )
268
+ end
269
+ when SQL::AST::UnaryOp
270
+ operand = build_predicate(ast.operand)
271
+ Predicate::Not.new(operand)
272
+ when SQL::AST::Between
273
+ Predicate::Between.new(
274
+ build_expression(ast.expression),
275
+ build_expression(ast.low),
276
+ build_expression(ast.high)
277
+ )
278
+ when SQL::AST::In
279
+ values = ast.values.flat_map do |value|
280
+ if value.is_a?(SQL::AST::Subquery)
281
+ Executor.new(@engine).execute(plan(value.query))[:rows].map { |row| Expression::Literal.new(row.values.first) }
282
+ else
283
+ [build_expression(value)]
284
+ end
285
+ end
286
+ Predicate::In.new(build_expression(ast.expression), values)
287
+ when SQL::AST::Exists
288
+ exists = Executor.new(@engine).execute(plan(ast.query))[:rows].any?
289
+ Predicate::Comparison.new(Expression::Literal.new(exists), Expression::Literal.new(true), :eq)
290
+ when SQL::AST::IsNull
291
+ Predicate::IsNull.new(build_expression(ast.expression), ast.negated)
292
+ else
293
+ Predicate::Comparison.new(
294
+ build_expression(ast),
295
+ build_expression(SQL::AST::Literal.new(true, :TRUE)),
296
+ :eq
297
+ )
298
+ end
299
+ end
300
+
301
+ def build_expression(ast)
302
+ case ast
303
+ when SQL::AST::Literal
304
+ Expression::Literal.new(ast.value)
305
+ when SQL::AST::Identifier
306
+ Expression::Column.new(ast.name, ast.table)
307
+ when SQL::AST::BinaryOp
308
+ left = build_expression(ast.left)
309
+ right = build_expression(ast.right)
310
+ operator = {
311
+ PLUS: :plus, MINUS: :minus, STAR: :multiply, SLASH: :divide,
312
+ PERCENT: :modulo, EQ: :eq
313
+ }[ast.operator] || ast.operator
314
+ Expression::BinaryOp.new(left, right, operator)
315
+ when SQL::AST::UnaryOp
316
+ operand = build_expression(ast.operand)
317
+ Expression::UnaryOp.new(operand, ast.operator)
318
+ when SQL::AST::FunctionCall
319
+ args = ast.arguments.map { |arg| build_expression(arg) }
320
+ Expression::Function.new(ast.name, args)
321
+ when SQL::AST::Parameter
322
+ Expression::Parameter.new(ast.index)
323
+ else
324
+ Expression::Literal.new(nil)
325
+ end
326
+ end
327
+
328
+ # Scan method selection
329
+ def choose_scan_method(plan)
330
+ return unless plan.type == :select
331
+
332
+ table_name = plan.table_name
333
+ return plan.set_scan_type(:sequential) unless table_name
334
+
335
+ # Check if there's an index that can be used
336
+ if @engine.respond_to?(:index_manager)
337
+ indexes = @engine.index_manager.get_indexes_for_table(table_name)
338
+
339
+ # Try to find a matching index for the predicate
340
+ if plan.predicate && indexes.any?
341
+ index = find_matching_index(plan.predicate, indexes)
342
+ if index
343
+ plan.set_scan_type(:index, index)
344
+ return
345
+ end
346
+ end
347
+
348
+ # A full index scan is not a substitute for a table scan: it can
349
+ # omit rows while an index is being maintained or rebuilt. Only use
350
+ # an index when the predicate specifically matches its columns.
351
+ end
352
+
353
+ # Default to sequential scan
354
+ plan.set_scan_type(:sequential)
355
+ end
356
+
357
+ def find_matching_index(predicate, indexes)
358
+ # Get columns used in predicate
359
+ columns = extract_predicate_columns(predicate)
360
+
361
+ indexes.each do |index|
362
+ # Check if predicate columns match index columns
363
+ if index.columns.any? { |col| columns.map(&:to_s).include?(col.to_s) }
364
+ return index
365
+ end
366
+ end
367
+
368
+ nil
369
+ end
370
+
371
+ def extract_predicate_columns(predicate)
372
+ return [] if predicate.nil?
373
+
374
+ case predicate
375
+ when Predicate::Comparison
376
+ left_name = predicate.left.respond_to?(:name) ? predicate.left.name : nil
377
+ [left_name].compact
378
+ when Predicate::And, Predicate::Or
379
+ extract_predicate_columns(predicate.left) + extract_predicate_columns(predicate.right)
380
+ when Predicate::Not
381
+ extract_predicate_columns(predicate.operand)
382
+ when Predicate::Between, Predicate::In, Predicate::IsNull, Predicate::Like
383
+ expr_name = predicate.expression.respond_to?(:name) ? predicate.expression.name : nil
384
+ [expr_name].compact
385
+ else
386
+ []
387
+ end
388
+ end
389
+
390
+ # Cost estimation
391
+ def estimate_cost(plan)
392
+ table_name = plan.table_name
393
+ row_count = begin
394
+ @engine.table_row_count(table_name)
395
+ rescue
396
+ 0
397
+ end
398
+
399
+ case plan.scan_type
400
+ when :sequential
401
+ plan.set_cost(row_count, row_count)
402
+ when :index
403
+ # Index scan is cheaper for large tables with selective predicates
404
+ if plan.predicate
405
+ selectivity = estimate_selectivity(plan.predicate)
406
+ estimated_rows = (row_count * selectivity).ceil
407
+ plan.set_cost(estimated_rows * 0.1, estimated_rows)
408
+ else
409
+ plan.set_cost(row_count * 0.5, row_count)
410
+ end
411
+ else
412
+ plan.set_cost(row_count, row_count)
413
+ end
414
+ end
415
+
416
+ def estimate_selectivity(predicate)
417
+ case predicate
418
+ when Predicate::Comparison
419
+ case predicate.operator
420
+ when :EQ then 0.01
421
+ when :NE then 0.9
422
+ when :LT, :LTE, :GT, :GTE then 0.5
423
+ else 0.5
424
+ end
425
+ when Predicate::And
426
+ estimate_selectivity(predicate.left) * estimate_selectivity(predicate.right)
427
+ when Predicate::Or
428
+ estimate_selectivity(predicate.left) + estimate_selectivity(predicate.right) -
429
+ estimate_selectivity(predicate.left) * estimate_selectivity(predicate.right)
430
+ when Predicate::Not
431
+ 1.0 - estimate_selectivity(predicate.operand)
432
+ else
433
+ 0.5
434
+ end
435
+ end
436
+
437
+ # Optimization
438
+ def optimize(plan)
439
+ @stats[:optimizations_applied] += 1
440
+
441
+ # Push down predicates
442
+ plan = push_down_predicates(plan)
443
+
444
+ # Remove redundant conditions
445
+ plan = remove_redundant_conditions(plan)
446
+
447
+ # Reorder only a pure inner-join chain. Outer joins are order
448
+ # sensitive, and cross joins are an explicit user request, so those
449
+ # remain in source order. Predicate dependencies prevent a join from
450
+ # being moved before a table referenced by its ON clause.
451
+ reorder_inner_joins(plan) if plan.is_a?(Plan::Select)
452
+
453
+ plan
454
+ end
455
+
456
+ def push_down_predicates(plan)
457
+ # Push predicates down to scan level
458
+ # This is already done in the planner
459
+ plan
460
+ end
461
+
462
+ def reorder_inner_joins(plan)
463
+ joins = plan.joins
464
+ return plan if joins.empty? || joins.any? { |join| join[:type] != :inner }
465
+
466
+ available = [plan.source_reference&.name || plan.table_name].compact.map(&:to_s)
467
+ pending = joins.dup
468
+ ordered = []
469
+
470
+ until pending.empty?
471
+ eligible = pending.select do |join|
472
+ table = join[:table]
473
+ right_names = [table.name, table.alias_name].compact.map(&:to_s)
474
+ referenced = predicate_table_names(join[:predicate])
475
+ (referenced - (available + right_names)).empty?
476
+ end
477
+ eligible = [pending.first] if eligible.empty?
478
+
479
+ chosen = eligible.min_by do |join|
480
+ @engine.table_row_count(join[:table].name)
481
+ rescue
482
+ 1_000_000
483
+ end
484
+ pending.delete(chosen)
485
+ ordered << chosen
486
+ available.concat([chosen[:table].name, chosen[:table].alias_name].compact.map(&:to_s))
487
+ end
488
+
489
+ plan.set_joins(ordered)
490
+ plan
491
+ end
492
+
493
+ def predicate_table_names(predicate)
494
+ case predicate
495
+ when Predicate::Comparison
496
+ expression_table_names(predicate.left) + expression_table_names(predicate.right)
497
+ when Predicate::And, Predicate::Or
498
+ predicate_table_names(predicate.left) + predicate_table_names(predicate.right)
499
+ when Predicate::Not
500
+ predicate_table_names(predicate.operand)
501
+ when Predicate::Between
502
+ expression_table_names(predicate.expression) + expression_table_names(predicate.low) +
503
+ expression_table_names(predicate.high)
504
+ when Predicate::In, Predicate::IsNull, Predicate::Like
505
+ expression_table_names(predicate.expression)
506
+ else
507
+ []
508
+ end.uniq
509
+ end
510
+
511
+ def expression_table_names(expression)
512
+ case expression
513
+ when Expression::Column
514
+ expression.table ? [expression.table.to_s] : []
515
+ when Expression::BinaryOp
516
+ expression_table_names(expression.left) + expression_table_names(expression.right)
517
+ when Expression::UnaryOp
518
+ expression_table_names(expression.operand)
519
+ when Expression::Function
520
+ expression.arguments.flat_map { |argument| expression_table_names(argument) }
521
+ else
522
+ []
523
+ end
524
+ end
525
+
526
+ def remove_redundant_conditions(plan)
527
+ if plan.predicate.is_a?(Predicate::And)
528
+ # Simplify AND conditions
529
+ simplified = simplify_and(plan.predicate)
530
+ plan.set_predicate(simplified)
531
+ end
532
+ plan
533
+ end
534
+
535
+ def simplify_and(predicate)
536
+ left = predicate.left
537
+ right = predicate.right
538
+
539
+ # Remove redundant TRUE conditions
540
+ if left.is_a?(Predicate::Comparison) && left.left.is_a?(Expression::Literal) &&
541
+ left.left.value == true && left.operator == :EQ
542
+ return right
543
+ end
544
+
545
+ if right.is_a?(Predicate::Comparison) && right.left.is_a?(Expression::Literal) &&
546
+ right.left.value == true && right.operator == :EQ
547
+ return left
548
+ end
549
+
550
+ predicate
551
+ end
552
+
553
+ attr_reader :stats
554
+ end
555
+ end
556
+ end