Commit Graph

10 Commits

Author SHA1 Message Date
wpdev 1d70d13f45 fix(2meet-data-optimizer): close the gaps in deferred table upgrades found in the 2.3 review
- a deferred group whose stored definition matches again (reverted, or
  upgraded by another process) goes through end_deferral() instead of staying
  native for good (review F2)
- the tmdo_ensure_group_indexes handler upgrades deferred tables itself: with
  ALTERNATE_WP_CRON the event runs after init in a visitor's request (F3)
- a new group table created while the entity writes to flat stays native
  until the drift fix has filled it from native meta; an opted-in HivePress
  attribute no longer routes searches to an empty table (F6)
- direct_read() returns native values for a deferred group (F10); the Hook
  Bus ignores deferral in aeav_only, as the router does
- dbDelta failures: the definition is recorded only when every column exists
  (schema.upgrade_failed otherwise), so a failed ALTER is retried and the
  group is not released onto missing columns
- test bootstrap: get_metadata() reads the real meta table
- Verify: SchemaUpgradeDeferralTest (6), Hook Bus direct read, integration
  587, unit 703, phpcs, stan
- PR: PR-114 (2.3 acceptance correctness review)
2026-09-27 17:30:36 +08:00
wpdev 29adb05567 fix(2meet-data-optimizer): alter an existing flat table in one process at a time; keep a deferred group native until its drift fix
- security review H1: anonymous /wp-cron.php requests pass ddl_allowed(),
  and dbDelta on an existing table had no lock, so concurrent requests each
  ran the ALTER on a large table. The upgrade now runs under the same MySQL
  GET_LOCK as the index DDL (skipped when another process holds it) and
  re-checks the stored hash once locked; this covers aeav_only too. The table
  build moves unchanged to write_table(); lock()/unlock() are shared
- security review M1: after a deferred upgrade the group stays native until
  tmdo_drift_resync has rebuilt the rows written meanwhile
  (TMDO_Schema_Manager::resynced() from run_scheduled_fix()); with no flat
  writes (disabled) it is released at once
- Verify: SchemaUpgradeDeferralTest (lock held elsewhere, resync release),
  integration 585, unit 702, phpcs, stan
- PR: PR-111 (2.3 acceptance security review)
2026-09-27 17:22:57 +08:00
wpdev 06a3594417 fix(2meet-data-optimizer): take the index lock on the primary under read/write splitting
- when $wpdb has send_reads_to_masters() (HyperDB, LudicrousDB), it is called
  before SELECT GET_LOCK, which a replica would take without locking out the
  ALTER on the primary
- Verify: GroupCompositeIndexTest (call order through a wpdb proxy),
  integration 583, unit 701, phpcs, stan
- PR: PR-108
2026-09-27 17:09:10 +08:00
wpdev e18030b206 fix(2meet-data-optimizer): upgrade existing flat tables from WP-CLI or WP-Cron only
- a changed group definition no longer runs dbDelta on an existing table
  inside a web request (seconds to minutes on a large table, and every
  concurrent request tried it): the group is recorded in tmdo_schema_deferred
  and tmdo_ensure_group_indexes is scheduled, whose request upgrades it
- meanwhile the group stays native: Hook Bus field lookups return no field and
  the router treats it as drifting; afterwards a drift fix
  (tmdo_drift_resync) rebuilds the writes that skipped the flat table
- new tables are still created at once; aeav_only upgrades at once (the flat
  table is the only copy); doctor lists deferred groups
- Verify: SchemaUpgradeDeferralTest, Hook Bus and router tests, integration
  582, unit 701, phpcs, stan; dev36 live cycle in the 2.3 acceptance
- PR: PR-107
2026-09-27 17:07:48 +08:00
wpdev ec8d69a6f4 fix(2meet-data-optimizer): add declared flat-table indexes from WP-CLI or WP-Cron only, one process at a time
- ensure_group_indexes() runs the ALTER only under WP-CLI or WP-Cron (filter
  tmdo_schema_ddl_allowed); web requests, including /wp-admin/ before
  auth_redirect() and admin-post.php, schedule tmdo_ensure_group_indexes instead
- MySQL GET_LOCK keeps it to one process (add_option() is INSERT ... ON
  DUPLICATE KEY UPDATE and cannot serve as a lock); released by MySQL if the
  process dies
- the applied signature is recorded only when every ALTER succeeds; failures
  log schema.index_failed and stay in pending_indexes(), which doctor reports
- new tables get their declared cidx_ indexes in CREATE TABLE
- test stub wpdb::query()/insert() return false on SQL errors like wpdb
- Verify: GroupCompositeIndexTest 7/7, unit 698, integration 576, phpcs, stan
- PR: PR-090a
2026-09-27 16:07:23 +08:00
wpdev 65c75dfc8d feat(2meet-data-optimizer): support composite indexes on entity groups
- register_group() accepts $args['indexes'] (name => ordered keys); indexes on
  unknown keys are dropped. Kept outside the field definition, so schema
  hashes are unchanged.
- Schema_Manager::ensure_group_indexes() adds missing cidx_* indexes to the
  existing flat table in place; the applied definition is remembered in
  tmdo_entity_indexes so later requests skip SHOW INDEX.
- Bench (1M listings, scenario 2): Entity p50 337 ms -> 278 ms with
  (hp_verified, hp_price), on par with Hot (269 ms).
- Verify: core unit 656 / integration 459 green, phpcs clean
- PR: PR-027
2026-09-26 01:05:23 +08:00
wpdev 1653011f4f feat(2meet-data-optimizer): guard orphaned flat-table groups
- On init:1, re-register aeav_only groups whose flat table exists and whose definition is in wpdo_registry_meta but no active plugin registers them, so their data stays readable (routing only, never DDL)
- Keys owned by a live group, or with rows in native meta (newer data written while orphaned), are left alone; the native-meta check is cached for a day
- Surface restored groups via Site Health (wpdo_orphaned_entity_groups), the admin Doctor tab, and wp wpdo doctor
- Verify: EntityRegistryRestoreTest (9), OrphanGroupRestoreTest (7); core unit 622 / integration 439, HP 141/49, WC 10/22 OK; dry-run on dev20–22 restores 16/17/4 groups, keeps hp_taxonomy hp_sort_order/hp_icon on native
- PR: PR-018
2026-09-25 23:38:12 +08:00
wpdev 76c01e44df refactor: 全部 128 個生產檔加入 declare(strict_types=1)(PR-H)
對齊 A v3.2.0。型別強制會把隱式轉換變成 TypeError,所以一次全檔加入
並跑完整測試(unit 451 / integration 398 全綠,無迴歸)。

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TbG1keQQ7XBa7qMQY16KCY
2026-07-31 06:13:33 +08:00
wpdev c46814ce83 refactor(migration): 回填 Migration Phase Strategy 體系(PR-D)
A v3.0.1 把 orchestrator 的 11 個 phase 抽成可注入的 Phase 物件,B 仍是
1107 行單體、以 'phase_' . $current 字串魔法分派、completed 甚至 inline
在 tick() 裡。本 commit 對齊:

新增 13 檔
- includes/migration/interface-migration-phase.php
- includes/migration/class-tmdo-migration-phase-base.php(log/get_managed_keys/
  execute_bulk_pivot/values_loose_equal 等共用 helper)
- includes/migration/phases/ 11 個 phase 類別

orchestrator 1107 → 640 行
- tick() 改 make_phase() 工廠 + $phase->execute($job)
- 移除 final、self::ENTITY_TYPE → static::(A v3.3.0 late static binding)
- 保留 B 原有的 '✓ %s (%.2fs)' 耗時 log(改為 tick 自行量測,A 版已簡化掉)
- 公開介面(preflight/start/tick/get_status/cancel/resume/needs_attention/
  cron_tick)經 diff 確認與 A 完全一致,呼叫端零影響

連帶
- Schema_Manager 補 table_exists 的 request-scoped cache 與
  flush_table_exists_cache()(A v3.1.6 + v3.4.6),Phase 測試需要它
- back-compat 補 12 個 Phase 類別的 WPDO_ alias
- 移植 MigrationPhaseTest + MigrationPhaseRemainingTest(527 行)

測試隔離差異(B 的 unit bootstrap 會載入 Member_Fields / Post_Fields,
A 的不會):MigrationPhaseRemainingTest 的 setUp 需額外清空 Entity_Registry
與 wpdo_register_entity_fields listener,否則 install_schema 會真的走進
dbDelta。MigrationPhaseTest 的 interface 斷言改用 TMDO_ 正式名稱
(PHP 無法 class_alias 介面,且該契約是核心內部擴充點)。

unit 451 / integration 398 GREEN

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TbG1keQQ7XBa7qMQY16KCY
2026-07-31 05:54:49 +08:00
wpdev d36bb954d1 chore: initial snapshot of 2meet-data-optimizer v0.1.0
Baseline before backporting wp-data-optimizer v3.0.1-v3.4.6.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01TbG1keQQ7XBa7qMQY16KCY
2026-07-31 05:06:36 +08:00