feat: move report artifacts to OSS storage

2026-08-04 12:37:26 +08:00
parent 727643f1f2
commit a2b86cdf10
45 changed files with 1288 additions and 1417 deletions
--- a/database/016_monthly_report_oss_artifacts.down.sql
+++ b/database/016_monthly_report_oss_artifacts.down.sql
@@ -0,0 +1,94 @@
+-- Restore the 012 local-only publication guard.
+
+BEGIN;
+
+DO $$
+BEGIN
+    IF current_database() <> 'booking_test' THEN
+        RAISE EXCEPTION
+            'ARR monthly OSS artifact rollback is allowed only in booking_test';
+    END IF;
+    IF to_regclass('reporting.monthly_runs') IS NULL
+       OR to_regprocedure('reporting.validate_monthly_run_publication()') IS NULL THEN
+        RAISE EXCEPTION
+            'ARR migration 016 is not applied';
+    END IF;
+    IF EXISTS (
+        SELECT 1
+        FROM reporting.monthly_runs AS run
+        JOIN ingestion.artifacts AS artifact
+          ON artifact.id IN (run.workbook_artifact_id, run.result_artifact_id)
+        WHERE run.report_status IN ('active', 'superseded')
+          AND artifact.storage_provider <> 'local'
+    ) THEN
+        RAISE EXCEPTION
+            'ARR monthly OSS artifact rollback refused while a published OSS artifact exists';
+    END IF;
+END;
+$$;
+
+CREATE OR REPLACE FUNCTION reporting.validate_monthly_run_publication()
+RETURNS trigger
+LANGUAGE plpgsql
+AS $$
+DECLARE
+    workbook_kind text;
+    workbook_provider text;
+    result_kind text;
+    result_provider text;
+    manifest_count integer;
+    manifest_rows bigint;
+    manifest_min integer;
+    manifest_max integer;
+    lineage_count integer;
+BEGIN
+    IF NEW.report_status NOT IN ('active', 'superseded') THEN
+        RETURN NEW;
+    END IF;
+
+    SELECT artifact_kind, storage_provider
+    INTO workbook_kind, workbook_provider
+    FROM ingestion.artifacts
+    WHERE id = NEW.workbook_artifact_id;
+
+    SELECT artifact_kind, storage_provider
+    INTO result_kind, result_provider
+    FROM ingestion.artifacts
+    WHERE id = NEW.result_artifact_id;
+
+    IF workbook_kind IS DISTINCT FROM 'monthly_xlsx'
+       OR workbook_provider IS DISTINCT FROM 'local'
+       OR result_kind IS DISTINCT FROM 'result_json'
+       OR result_provider IS DISTINCT FROM 'local' THEN
+        RAISE EXCEPTION
+            'published monthly run must reference controlled local monthly_xlsx and result_json artifacts';
+    END IF;
+
+    SELECT count(*), COALESCE(sum(row_count), 0), min(worksheet_order), max(worksheet_order)
+    INTO manifest_count, manifest_rows, manifest_min, manifest_max
+    FROM reporting.monthly_channel_manifest
+    WHERE report_id = NEW.id;
+
+    IF manifest_count <> NEW.channel_count
+       OR manifest_rows <> NEW.row_count
+       OR manifest_min <> 1
+       OR manifest_max <> manifest_count THEN
+        RAISE EXCEPTION
+            'published monthly run requires a continuous reconciled channel manifest';
+    END IF;
+
+    SELECT count(*)
+    INTO lineage_count
+    FROM reporting.monthly_run_daily_versions
+    WHERE report_id = NEW.id;
+
+    IF lineage_count < 1 THEN
+        RAISE EXCEPTION
+            'published monthly run requires daily-version lineage';
+    END IF;
+
+    RETURN NEW;
+END;
+$$;
+
+COMMIT;
--- a/database/016_monthly_report_oss_artifacts.sql
+++ b/database/016_monthly_report_oss_artifacts.sql
@@ -0,0 +1,88 @@
+-- Allow durable monthly publication artifacts in the existing private OSS.
+-- Historical local artifacts remain valid and readable.
+
+BEGIN;
+
+DO $$
+BEGIN
+    IF current_database() <> 'booking_test' THEN
+        RAISE EXCEPTION
+            'ARR monthly OSS artifact migration is allowed only in booking_test';
+    END IF;
+    IF to_regclass('reporting.monthly_runs') IS NULL
+       OR to_regclass('ingestion.artifacts') IS NULL
+       OR to_regprocedure('reporting.validate_monthly_run_publication()') IS NULL THEN
+        RAISE EXCEPTION
+            'ARR migration 012 must be applied first';
+    END IF;
+END;
+$$;
+
+CREATE OR REPLACE FUNCTION reporting.validate_monthly_run_publication()
+RETURNS trigger
+LANGUAGE plpgsql
+AS $$
+DECLARE
+    workbook_kind text;
+    workbook_provider text;
+    result_kind text;
+    result_provider text;
+    manifest_count integer;
+    manifest_rows bigint;
+    manifest_min integer;
+    manifest_max integer;
+    lineage_count integer;
+BEGIN
+    IF NEW.report_status NOT IN ('active', 'superseded') THEN
+        RETURN NEW;
+    END IF;
+
+    SELECT artifact_kind, storage_provider
+    INTO workbook_kind, workbook_provider
+    FROM ingestion.artifacts
+    WHERE id = NEW.workbook_artifact_id;
+
+    SELECT artifact_kind, storage_provider
+    INTO result_kind, result_provider
+    FROM ingestion.artifacts
+    WHERE id = NEW.result_artifact_id;
+
+    IF workbook_kind IS DISTINCT FROM 'monthly_xlsx'
+       OR COALESCE(workbook_provider, '') NOT IN ('oss', 's3', 'local')
+       OR result_kind IS DISTINCT FROM 'result_json'
+       OR COALESCE(result_provider, '') NOT IN ('oss', 's3', 'local') THEN
+        RAISE EXCEPTION
+            'published monthly run must reference OSS or controlled local monthly artifacts';
+    END IF;
+
+    SELECT count(*), COALESCE(sum(row_count), 0), min(worksheet_order), max(worksheet_order)
+    INTO manifest_count, manifest_rows, manifest_min, manifest_max
+    FROM reporting.monthly_channel_manifest
+    WHERE report_id = NEW.id;
+
+    IF manifest_count <> NEW.channel_count
+       OR manifest_rows <> NEW.row_count
+       OR manifest_min <> 1
+       OR manifest_max <> manifest_count THEN
+        RAISE EXCEPTION
+            'published monthly run requires a continuous reconciled channel manifest';
+    END IF;
+
+    SELECT count(*)
+    INTO lineage_count
+    FROM reporting.monthly_run_daily_versions
+    WHERE report_id = NEW.id;
+
+    IF lineage_count < 1 THEN
+        RAISE EXCEPTION
+            'published monthly run requires daily-version lineage';
+    END IF;
+
+    RETURN NEW;
+END;
+$$;
+
+COMMENT ON FUNCTION reporting.validate_monthly_run_publication() IS
+  'Published monthly artifacts may be private OSS/S3 or legacy controlled-local objects.';
+
+COMMIT;
--- a/database/APPLIED_MIGRATIONS.md
+++ b/database/APPLIED_MIGRATIONS.md
@@ -1,7 +1,7 @@
 # ARR 测试数据库已执行记录

-更新时间：2026-07-31  
-状态：008–015 已提交；014/015 已通过迁移回滚探针、正式应用、结构/数据复核及真实 PostgreSQL 草稿激活外层回滚验收。首次真实业务工作簿激活仍待操作员授权。
+更新时间：2026-08-04
+状态：008–015 已提交；014/015 已通过迁移回滚探针、正式应用、结构/数据复核及真实 PostgreSQL 草稿激活外层回滚验收。016 已提交代码、待在测试机 booking_test 应用并完成 OSS 月报验收。首次真实业务工作簿激活仍待操作员授权。

 ## 目标与隔离

@@ -29,6 +29,7 @@
 | `013_daily_upload_filename.sql` | `f7ea18d6b844d9bd90fa757a4cf8428d1dd5833c088ae23444ac81fbe204fb0a` | 已提交并验收 |
 | `014_booking_current_source_batch.sql` | `23bf0fcc880225ca276d4d7d871057be4f7950e761ddbeed1df2a3c6e25463b5` | 2026-07-31 已通过同事务回滚探针后由受控连接正式应用；数据复核通过 |
 | `015_booking_excel_review_drafts.sql` | `a80689c4ecc3b6b3502e8f094a8c175af38df8c09f9d2c64412a9aec55bae12a` | 2026-07-31 已由受控连接正式应用；真实 PostgreSQL 草稿编辑/激活外层回滚通过 |
+| `016_monthly_report_oss_artifacts.sql` | `70ca052f71da9f88e83fe8bccf0a6682cdd3138c3525981c80eebc868e401626` | 代码已提交；尚未在远程 `booking_test` 应用 |

 ## 014/015 最新核查状态