@jingyi0605/codingns 0.8.5 → 0.9.0

package/dist/server/modules/affairs-indexer/core/src/sqlite/migration-runner.d.ts

@@ -0,0 +1,22 @@
+import type { RuntimeConfig } from "../../../contracts/src/index.js";
+export interface CatalogMigration {
+    version: number;
+    name: string;
+    sql: string;
+    columns?: Array<{
+        table: string;
+        name: string;
+        definition: string;
+    }>;
+}
+export interface MigrationRunResult {
+    dbPath: string;
+    schemaVersion: number;
+    appliedMigrations: string[];
+    executedAt: string;
+}
+/**
+ * 最小迁移执行器。
+ * 先把版本与历史机制收整齐，避免后续 schema 演进继续失控。
+ */
+export declare function runCatalogMigrations(config: RuntimeConfig): MigrationRunResult;

package/dist/server/modules/affairs-indexer/core/src/sqlite/migration-runner.js

@@ -0,0 +1,430 @@
+import crypto from "node:crypto";
+import fs from "node:fs";
+import { openDatabase } from "./open-database.js";
+const CATALOG_MIGRATIONS = [
+    {
+        version: 1,
+        name: "bootstrap_node_schema_v1",
+        sql: `
+CREATE TABLE IF NOT EXISTS schema_meta (
+  key TEXT PRIMARY KEY,
+  value TEXT NOT NULL,
+  updated_at TEXT NOT NULL
+);
+
+CREATE TABLE IF NOT EXISTS migration_history (
+  id TEXT PRIMARY KEY,
+  version INTEGER NOT NULL,
+  name TEXT NOT NULL,
+  applied_at TEXT NOT NULL,
+  checksum TEXT,
+  status TEXT NOT NULL
+);
+
+CREATE TABLE IF NOT EXISTS files (
+  id TEXT PRIMARY KEY,
+  path TEXT UNIQUE NOT NULL,
+  dir_path TEXT NOT NULL,
+  name TEXT NOT NULL,
+  extension TEXT NOT NULL,
+  size INTEGER NOT NULL,
+  mtime TEXT NOT NULL,
+  ctime TEXT,
+  inode_key TEXT,
+  content_hash TEXT,
+  status TEXT NOT NULL,
+  last_seen_at TEXT NOT NULL
+);
+
+CREATE TABLE IF NOT EXISTS documents (
+  id TEXT PRIMARY KEY,
+  file_id TEXT UNIQUE NOT NULL,
+  title TEXT,
+  summary TEXT,
+  language TEXT,
+  parse_status TEXT NOT NULL,
+  parse_error TEXT,
+  index_status TEXT NOT NULL,
+  chunk_count INTEGER NOT NULL,
+  last_indexed_at TEXT,
+  FOREIGN KEY(file_id) REFERENCES files(id)
+);
+
+CREATE TABLE IF NOT EXISTS chunks (
+  id TEXT PRIMARY KEY,
+  document_id TEXT NOT NULL,
+  chunk_index INTEGER NOT NULL,
+  content TEXT NOT NULL,
+  content_hash TEXT,
+  page_no INTEGER,
+  sheet_name TEXT,
+  heading_path TEXT,
+  token_count INTEGER,
+  vector_point_id TEXT,
+  FOREIGN KEY(document_id) REFERENCES documents(id)
+);
+
+CREATE TABLE IF NOT EXISTS tags (
+  id TEXT PRIMARY KEY,
+  root_type TEXT NOT NULL,
+  path TEXT UNIQUE NOT NULL,
+  name TEXT NOT NULL,
+  parent_id TEXT,
+  canonical_name TEXT NOT NULL,
+  description TEXT,
+  status TEXT NOT NULL,
+  created_by TEXT NOT NULL,
+  created_at TEXT NOT NULL
+);
+
+CREATE TABLE IF NOT EXISTS tag_aliases (
+  id TEXT PRIMARY KEY,
+  tag_id TEXT NOT NULL,
+  alias TEXT NOT NULL,
+  alias_type TEXT NOT NULL,
+  confidence REAL NOT NULL,
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+CREATE TABLE IF NOT EXISTS document_tags (
+  id TEXT PRIMARY KEY,
+  document_id TEXT NOT NULL,
+  tag_id TEXT NOT NULL,
+  confidence REAL NOT NULL,
+  source TEXT NOT NULL,
+  evidence TEXT,
+  manual_override INTEGER NOT NULL,
+  updated_at TEXT NOT NULL,
+  UNIQUE(document_id, tag_id),
+  FOREIGN KEY(document_id) REFERENCES documents(id),
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+CREATE TABLE IF NOT EXISTS derived_document_tags (
+  id TEXT PRIMARY KEY,
+  document_id TEXT NOT NULL,
+  tag_id TEXT NOT NULL,
+  rule_name TEXT NOT NULL,
+  computed_at TEXT NOT NULL,
+  expires_at TEXT,
+  FOREIGN KEY(document_id) REFERENCES documents(id),
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+CREATE TABLE IF NOT EXISTS reindex_jobs (
+  id TEXT PRIMARY KEY,
+  job_type TEXT NOT NULL,
+  target_path TEXT,
+  status TEXT NOT NULL,
+  priority INTEGER NOT NULL,
+  error_message TEXT,
+  created_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL
+);
+
+CREATE TABLE IF NOT EXISTS watch_events (
+  id TEXT PRIMARY KEY,
+  event_type TEXT NOT NULL,
+  path TEXT NOT NULL,
+  created_at TEXT NOT NULL
+);
+
+CREATE INDEX IF NOT EXISTS idx_files_hash ON files(content_hash);
+CREATE INDEX IF NOT EXISTS idx_files_inode_key ON files(inode_key);
+CREATE INDEX IF NOT EXISTS idx_chunks_document ON chunks(document_id, chunk_index);
+CREATE INDEX IF NOT EXISTS idx_tags_path ON tags(path);
+CREATE INDEX IF NOT EXISTS idx_tag_aliases_alias ON tag_aliases(alias);
+CREATE INDEX IF NOT EXISTS idx_reindex_jobs_status ON reindex_jobs(status, priority, created_at);
+    `,
+    },
+    {
+        version: 2,
+        name: "add_parser_skip_catalog_v2",
+        sql: `
+CREATE TABLE IF NOT EXISTS parser_skip_catalog (
+  skip_key TEXT PRIMARY KEY,
+  adapter TEXT NOT NULL,
+  reason_code TEXT NOT NULL,
+  extension TEXT NOT NULL,
+  sample_paths_json TEXT NOT NULL,
+  sample_count INTEGER NOT NULL,
+  total_count INTEGER NOT NULL,
+  last_message TEXT,
+  first_seen_at TEXT NOT NULL,
+  last_seen_at TEXT NOT NULL,
+  last_run_at TEXT NOT NULL
+);
+
+CREATE INDEX IF NOT EXISTS idx_parser_skip_catalog_reason ON parser_skip_catalog(reason_code, extension, last_run_at);
+    `,
+    },
+    {
+        version: 3,
+        name: "add_tag_write_indexes_v3",
+        sql: `
+CREATE INDEX IF NOT EXISTS idx_document_tags_document ON document_tags(document_id);
+CREATE INDEX IF NOT EXISTS idx_derived_document_tags_document ON derived_document_tags(document_id);
+CREATE UNIQUE INDEX IF NOT EXISTS idx_derived_document_tags_pair ON derived_document_tags(document_id, tag_id);
+    `,
+    },
+    {
+        version: 4,
+        name: "affairs_tag_management_v4",
+        columns: [
+            { table: "tags", name: "updated_at", definition: "TEXT" },
+            { table: "tags", name: "disabled_at", definition: "TEXT" },
+            { table: "tag_aliases", name: "updated_at", definition: "TEXT" },
+            { table: "document_tags", name: "source_ref", definition: "TEXT" },
+            { table: "derived_document_tags", name: "source", definition: "TEXT" },
+            { table: "derived_document_tags", name: "source_ref", definition: "TEXT" },
+            { table: "derived_document_tags", name: "evidence", definition: "TEXT" },
+            { table: "derived_document_tags", name: "updated_at", definition: "TEXT" },
+        ],
+        sql: `
+UPDATE tags
+SET updated_at = COALESCE(updated_at, created_at);
+
+UPDATE tag_aliases
+SET updated_at = COALESCE(updated_at, (
+  SELECT tags.created_at
+  FROM tags
+  WHERE tags.id = tag_aliases.tag_id
+), CURRENT_TIMESTAMP);
+
+UPDATE derived_document_tags
+SET source = COALESCE(source, 'system_derived'),
+    source_ref = COALESCE(source_ref, NULL),
+    evidence = COALESCE(evidence, NULL),
+    updated_at = COALESCE(updated_at, computed_at);
+
+CREATE TABLE IF NOT EXISTS tag_rules (
+  id TEXT PRIMARY KEY,
+  tag_id TEXT NOT NULL,
+  enabled INTEGER NOT NULL,
+  rule_type TEXT NOT NULL,
+  scope_json TEXT NOT NULL,
+  matcher_json TEXT NOT NULL,
+  min_score REAL,
+  priority INTEGER NOT NULL,
+  source TEXT NOT NULL,
+  created_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL,
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+CREATE TABLE IF NOT EXISTS manual_document_tag_bindings (
+  id TEXT PRIMARY KEY,
+  document_id TEXT NOT NULL,
+  tag_id TEXT NOT NULL,
+  source TEXT NOT NULL,
+  created_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL,
+  UNIQUE(document_id, tag_id),
+  FOREIGN KEY(document_id) REFERENCES documents(id),
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+INSERT OR IGNORE INTO manual_document_tag_bindings(
+  id,
+  document_id,
+  tag_id,
+  source,
+  created_at,
+  updated_at
+)
+SELECT
+  'manual_binding_' || document_id || '_' || tag_id,
+  document_id,
+  tag_id,
+  'manual_document',
+  updated_at,
+  updated_at
+FROM document_tags
+WHERE manual_override = 1;
+
+CREATE TABLE IF NOT EXISTS folder_tag_bindings (
+  id TEXT PRIMARY KEY,
+  folder_path TEXT NOT NULL,
+  tag_id TEXT NOT NULL,
+  apply_mode TEXT NOT NULL,
+  created_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL,
+  UNIQUE(folder_path, tag_id, apply_mode),
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+CREATE TABLE IF NOT EXISTS tag_recommendation_batches (
+  id TEXT PRIMARY KEY,
+  status TEXT NOT NULL,
+  summary TEXT,
+  evidence_snapshot_json TEXT,
+  generated_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL
+);
+
+CREATE TABLE IF NOT EXISTS tag_recommendation_items (
+  id TEXT PRIMARY KEY,
+  batch_id TEXT NOT NULL,
+  proposed_path TEXT NOT NULL,
+  proposed_name TEXT NOT NULL,
+  proposed_parent_path TEXT,
+  document_count INTEGER NOT NULL,
+  evidence_json TEXT NOT NULL,
+  selected_by_default INTEGER NOT NULL,
+  status TEXT NOT NULL,
+  created_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL,
+  FOREIGN KEY(batch_id) REFERENCES tag_recommendation_batches(id)
+);
+
+CREATE INDEX IF NOT EXISTS idx_document_tags_source ON document_tags(document_id, source);
+CREATE INDEX IF NOT EXISTS idx_tag_rules_tag ON tag_rules(tag_id, enabled, priority);
+CREATE INDEX IF NOT EXISTS idx_manual_document_tag_bindings_document ON manual_document_tag_bindings(document_id);
+CREATE INDEX IF NOT EXISTS idx_folder_tag_bindings_folder ON folder_tag_bindings(folder_path);
+CREATE INDEX IF NOT EXISTS idx_tag_recommendation_items_batch ON tag_recommendation_items(batch_id, status);
+    `,
+    },
+    {
+        version: 5,
+        name: "files_identity_tracking_v5",
+        columns: [
+            { table: "files", name: "inode_key", definition: "TEXT" },
+        ],
+        sql: `
+CREATE INDEX IF NOT EXISTS idx_files_inode_key ON files(inode_key);
+    `,
+    },
+    {
+        version: 6,
+        name: "manual_file_tag_bindings_v6",
+        sql: `
+CREATE TABLE IF NOT EXISTS manual_file_tag_bindings (
+  id TEXT PRIMARY KEY,
+  inode_key TEXT,
+  content_hash TEXT,
+  file_size INTEGER NOT NULL,
+  extension TEXT NOT NULL,
+  tag_id TEXT NOT NULL,
+  source TEXT NOT NULL,
+  created_at TEXT NOT NULL,
+  updated_at TEXT NOT NULL,
+  FOREIGN KEY(tag_id) REFERENCES tags(id)
+);
+
+INSERT OR IGNORE INTO manual_file_tag_bindings(
+  id,
+  inode_key,
+  content_hash,
+  file_size,
+  extension,
+  tag_id,
+  source,
+  created_at,
+  updated_at
+)
+SELECT
+  'manual_file_binding_' || mdtb.document_id || '_' || mdtb.tag_id,
+  f.inode_key,
+  f.content_hash,
+  f.size,
+  f.extension,
+  mdtb.tag_id,
+  mdtb.source,
+  mdtb.created_at,
+  mdtb.updated_at
+FROM manual_document_tag_bindings mdtb
+JOIN documents d ON d.id = mdtb.document_id
+JOIN files f ON f.id = d.file_id
+WHERE f.status = 'active'
+  AND d.index_status IN ('indexed', 'failed', 'skipped');
+
+CREATE INDEX IF NOT EXISTS idx_manual_file_tag_bindings_inode ON manual_file_tag_bindings(inode_key);
+CREATE INDEX IF NOT EXISTS idx_manual_file_tag_bindings_content ON manual_file_tag_bindings(content_hash, file_size, extension);
+CREATE INDEX IF NOT EXISTS idx_manual_file_tag_bindings_tag ON manual_file_tag_bindings(tag_id);
+    `,
+    },
+];
+function hasTable(db, tableName) {
+    const row = db.prepare(`SELECT name FROM sqlite_master WHERE type = 'table' AND name = ?`).get(tableName);
+    return row?.name === tableName;
+}
+function hasColumn(db, tableName, columnName) {
+    const rows = db.prepare(`PRAGMA table_info(${tableName})`).all();
+    return rows.some((row) => row.name === columnName);
+}
+function ensureMigrationColumns(db, migration) {
+    for (const column of migration.columns ?? []) {
+        if (hasColumn(db, column.table, column.name)) {
+            continue;
+        }
+        db.exec(`ALTER TABLE ${column.table} ADD COLUMN ${column.name} ${column.definition}`);
+    }
+}
+function readCurrentSchemaVersion(db) {
+    if (!hasTable(db, "schema_meta")) {
+        return 0;
+    }
+    const row = db.prepare(`SELECT value FROM schema_meta WHERE key = 'schema_version'`).get();
+    if (!row?.value) {
+        return 0;
+    }
+    const parsed = Number(row.value);
+    return Number.isFinite(parsed) ? parsed : 0;
+}
+/**
+ * 最小迁移执行器。
+ * 先把版本与历史机制收整齐，避免后续 schema 演进继续失控。
+ */
+export function runCatalogMigrations(config) {
+    fs.mkdirSync(config.indexDir, { recursive: true });
+    fs.mkdirSync(config.exportDir, { recursive: true });
+    const db = openDatabase(config.dbPath);
+    const now = new Date().toISOString();
+    const appliedMigrations = [];
+    try {
+        db.exec("BEGIN IMMEDIATE");
+        let currentVersion = readCurrentSchemaVersion(db);
+        for (const migration of CATALOG_MIGRATIONS) {
+            if (migration.version <= currentVersion) {
+                continue;
+            }
+            ensureMigrationColumns(db, migration);
+            db.exec(migration.sql);
+            const checksum = crypto.createHash("sha1").update(migration.sql).digest("hex");
+            db.prepare(`
+        INSERT OR REPLACE INTO schema_meta(key, value, updated_at)
+        VALUES(?, ?, ?)
+      `).run("schema_version", String(migration.version), now);
+            db.prepare(`
+        INSERT OR REPLACE INTO schema_meta(key, value, updated_at)
+        VALUES(?, ?, ?)
+      `).run("managed_by", "node", now);
+            db.prepare(`
+        INSERT OR REPLACE INTO migration_history(id, version, name, applied_at, checksum, status)
+        VALUES(?, ?, ?, ?, ?, ?)
+      `).run(`migration_${migration.version}`, migration.version, migration.name, now, checksum, "done");
+            appliedMigrations.push(migration.name);
+            currentVersion = migration.version;
+        }
+        db.exec("COMMIT");
+        return {
+            dbPath: config.dbPath,
+            schemaVersion: currentVersion,
+            appliedMigrations,
+            executedAt: now,
+        };
+    }
+    catch (error) {
+        try {
+            db.exec("ROLLBACK");
+        }
+        catch {
+            // ignore rollback failure
+        }
+        throw error;
+    }
+    finally {
+        db.close();
+    }
+}
+//# sourceMappingURL=migration-runner.js.map

package/dist/server/modules/affairs-indexer/core/src/sqlite/migration-runner.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"migration-runner.js","sourceRoot":"","sources":["../../../../../../../src/modules/affairs-indexer/core/src/sqlite/migration-runner.ts"],"names":[],"mappings":"AAAA,OAAO,MAAM,MAAM,aAAa,CAAC;AACjC,OAAO,EAAE,MAAM,SAAS,CAAC;AAGzB,OAAO,EAAE,YAAY,EAAE,MAAM,oBAAoB,CAAC;AAoBlD,MAAM,kBAAkB,GAAuB;IAC7C;QACE,OAAO,EAAE,CAAC;QACV,IAAI,EAAE,0BAA0B;QAChC,GAAG,EAAE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAkIJ;KACF;IACD;QACE,OAAO,EAAE,CAAC;QACV,IAAI,EAAE,4BAA4B;QAClC,GAAG,EAAE;;;;;;;;;;;;;;;;KAgBJ;KACF;IACD;QACE,OAAO,EAAE,CAAC;QACV,IAAI,EAAE,0BAA0B;QAChC,GAAG,EAAE;;;;KAIJ;KACF;IACD;QACE,OAAO,EAAE,CAAC;QACV,IAAI,EAAE,2BAA2B;QACjC,OAAO,EAAE;YACP,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,MAAM,EAAE;YACzD,EAAE,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,aAAa,EAAE,UAAU,EAAE,MAAM,EAAE;YAC1D,EAAE,KAAK,EAAE,aAAa,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,MAAM,EAAE;YAChE,EAAE,KAAK,EAAE,eAAe,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,MAAM,EAAE;YAClE,EAAE,KAAK,EAAE,uBAAuB,EAAE,IAAI,EAAE,QAAQ,EAAE,UAAU,EAAE,MAAM,EAAE;YACtE,EAAE,KAAK,EAAE,uBAAuB,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,MAAM,EAAE;YAC1E,EAAE,KAAK,EAAE,uBAAuB,EAAE,IAAI,EAAE,UAAU,EAAE,UAAU,EAAE,MAAM,EAAE;YACxE,EAAE,KAAK,EAAE,uBAAuB,EAAE,IAAI,EAAE,YAAY,EAAE,UAAU,EAAE,MAAM,EAAE;SAC3E;QACD,GAAG,EAAE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KAsGJ;KACF;IACD;QACE,OAAO,EAAE,CAAC;QACV,IAAI,EAAE,4BAA4B;QAClC,OAAO,EAAE;YACP,EAAE,KAAK,EAAE,OAAO,EAAE,IAAI,EAAE,WAAW,EAAE,UAAU,EAAE,MAAM,EAAE;SAC1D;QACD,GAAG,EAAE;;KAEJ;KACF;IACD;QACE,OAAO,EAAE,CAAC;QACV,IAAI,EAAE,6BAA6B;QACnC,GAAG,EAAE;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;KA4CJ;KACF;CACF,CAAC;AAEF,SAAS,QAAQ,CAAC,EAAmC,EAAE,SAAiB;IACtE,MAAM,GAAG,GAAG,EAAE,CAAC,OAAO,CAAC,kEAAkE,CAAC,CAAC,GAAG,CAAC,SAAS,CAAkC,CAAC;IAC3I,OAAO,GAAG,EAAE,IAAI,KAAK,SAAS,CAAC;AACjC,CAAC;AAED,SAAS,SAAS,CAAC,EAAmC,EAAE,SAAiB,EAAE,UAAkB;IAC3F,MAAM,IAAI,GAAG,EAAE,CAAC,OAAO,CAAC,qBAAqB,SAAS,GAAG,CAAC,CAAC,GAAG,EAA8B,CAAC;IAC7F,OAAO,IAAI,CAAC,IAAI,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,GAAG,CAAC,IAAI,KAAK,UAAU,CAAC,CAAC;AACrD,CAAC;AAED,SAAS,sBAAsB,CAAC,EAAmC,EAAE,SAA2B;IAC9F,KAAK,MAAM,MAAM,IAAI,SAAS,CAAC,OAAO,IAAI,EAAE,EAAE,CAAC;QAC7C,IAAI,SAAS,CAAC,EAAE,EAAE,MAAM,CAAC,KAAK,EAAE,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC;YAC7C,SAAS;QACX,CAAC;QACD,EAAE,CAAC,IAAI,CAAC,eAAe,MAAM,CAAC,KAAK,eAAe,MAAM,CAAC,IAAI,IAAI,MAAM,CAAC,UAAU,EAAE,CAAC,CAAC;IACxF,CAAC;AACH,CAAC;AAED,SAAS,wBAAwB,CAAC,EAAmC;IACnE,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,aAAa,CAAC,EAAE,CAAC;QACjC,OAAO,CAAC,CAAC;IACX,CAAC;IACD,MAAM,GAAG,GAAG,EAAE,CAAC,OAAO,CAAC,4DAA4D,CAAC,CAAC,GAAG,EAAoC,CAAC;IAC7H,IAAI,CAAC,GAAG,EAAE,KAAK,EAAE,CAAC;QAChB,OAAO,CAAC,CAAC;IACX,CAAC;IACD,MAAM,MAAM,GAAG,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;IACjC,OAAO,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC;AAC9C,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,oBAAoB,CAAC,MAAqB;IACxD,EAAE,CAAC,SAAS,CAAC,MAAM,CAAC,QAAQ,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IACnD,EAAE,CAAC,SAAS,CAAC,MAAM,CAAC,SAAS,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAEpD,MAAM,EAAE,GAAG,YAAY,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;IACvC,MAAM,GAAG,GAAG,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;IACrC,MAAM,iBAAiB,GAAa,EAAE,CAAC;IAEvC,IAAI,CAAC;QACH,EAAE,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC;QAC3B,IAAI,cAAc,GAAG,wBAAwB,CAAC,EAAE,CAAC,CAAC;QAElD,KAAK,MAAM,SAAS,IAAI,kBAAkB,EAAE,CAAC;YAC3C,IAAI,SAAS,CAAC,OAAO,IAAI,cAAc,EAAE,CAAC;gBACxC,SAAS;YACX,CAAC;YAED,sBAAsB,CAAC,EAAE,EAAE,SAAS,CAAC,CAAC;YACtC,EAAE,CAAC,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC;YAEvB,MAAM,QAAQ,GAAG,MAAM,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,MAAM,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;YAC/E,EAAE,CAAC,OAAO,CAAC;;;OAGV,CAAC,CAAC,GAAG,CAAC,gBAAgB,EAAE,MAAM,CAAC,SAAS,CAAC,OAAO,CAAC,EAAE,GAAG,CAAC,CAAC;YACzD,EAAE,CAAC,OAAO,CAAC;;;OAGV,CAAC,CAAC,GAAG,CAAC,YAAY,EAAE,MAAM,EAAE,GAAG,CAAC,CAAC;YAElC,EAAE,CAAC,OAAO,CAAC;;;OAGV,CAAC,CAAC,GAAG,CACJ,aAAa,SAAS,CAAC,OAAO,EAAE,EAChC,SAAS,CAAC,OAAO,EACjB,SAAS,CAAC,IAAI,EACd,GAAG,EACH,QAAQ,EACR,MAAM,CACP,CAAC;YAEF,iBAAiB,CAAC,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;YACvC,cAAc,GAAG,SAAS,CAAC,OAAO,CAAC;QACrC,CAAC;QAED,EAAE,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;QAClB,OAAO;YACL,MAAM,EAAE,MAAM,CAAC,MAAM;YACrB,aAAa,EAAE,cAAc;YAC7B,iBAAiB;YACjB,UAAU,EAAE,GAAG;SAChB,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QACf,IAAI,CAAC;YACH,EAAE,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC;QACtB,CAAC;QAAC,MAAM,CAAC;YACP,0BAA0B;QAC5B,CAAC;QACD,MAAM,KAAK,CAAC;IACd,CAAC;YAAS,CAAC;QACT,EAAE,CAAC,KAAK,EAAE,CAAC;IACb,CAAC;AACH,CAAC"}

package/dist/server/modules/affairs-indexer/core/src/sqlite/open-database.d.ts

@@ -0,0 +1,9 @@
+import { DatabaseSync } from "node:sqlite";
+export interface OpenDatabaseOptions {
+    tempStore?: "FILE" | "MEMORY";
+}
+/**
+ * 打开 SQLite 数据库。
+ * 当前阶段统一使用 Node 内置 sqlite，同步接口足够支撑 CLI 初始化与检测。
+ */
+export declare function openDatabase(dbPath: string, options?: OpenDatabaseOptions): DatabaseSync;

package/dist/server/modules/affairs-indexer/core/src/sqlite/open-database.js

@@ -0,0 +1,19 @@
+import fs from "node:fs";
+import path from "node:path";
+import { DatabaseSync } from "node:sqlite";
+/**
+ * 打开 SQLite 数据库。
+ * 当前阶段统一使用 Node 内置 sqlite，同步接口足够支撑 CLI 初始化与检测。
+ */
+export function openDatabase(dbPath, options = {}) {
+    fs.mkdirSync(path.dirname(dbPath), { recursive: true });
+    const db = new DatabaseSync(dbPath);
+    const tempStore = options.tempStore ?? "FILE";
+    db.exec("PRAGMA journal_mode=WAL;");
+    db.exec("PRAGMA synchronous=NORMAL;");
+    db.exec(`PRAGMA temp_store=${tempStore};`);
+    db.exec("PRAGMA foreign_keys=ON;");
+    db.exec("PRAGMA busy_timeout=5000;");
+    return db;
+}
+//# sourceMappingURL=open-database.js.map

package/dist/server/modules/affairs-indexer/core/src/sqlite/open-database.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"open-database.js","sourceRoot":"","sources":["../../../../../../../src/modules/affairs-indexer/core/src/sqlite/open-database.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,SAAS,CAAC;AACzB,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,YAAY,EAAE,MAAM,aAAa,CAAC;AAM3C;;;GAGG;AACH,MAAM,UAAU,YAAY,CAC1B,MAAc,EACd,UAA+B,EAAE;IAEjC,EAAE,CAAC,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC,MAAM,CAAC,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IACxD,MAAM,EAAE,GAAG,IAAI,YAAY,CAAC,MAAM,CAAC,CAAC;IACpC,MAAM,SAAS,GAAG,OAAO,CAAC,SAAS,IAAI,MAAM,CAAC;IAC9C,EAAE,CAAC,IAAI,CAAC,0BAA0B,CAAC,CAAC;IACpC,EAAE,CAAC,IAAI,CAAC,4BAA4B,CAAC,CAAC;IACtC,EAAE,CAAC,IAAI,CAAC,qBAAqB,SAAS,GAAG,CAAC,CAAC;IAC3C,EAAE,CAAC,IAAI,CAAC,yBAAyB,CAAC,CAAC;IACnC,EAAE,CAAC,IAAI,CAAC,2BAA2B,CAAC,CAAC;IACrC,OAAO,EAAE,CAAC;AACZ,CAAC"}

package/dist/server/modules/affairs-indexer/core/src/tagging/simple-tag-inference.d.ts

@@ -0,0 +1,21 @@
+import type { FileScanResult } from "../scanner/file-scanner.js";
+import type { ParsedDocument } from "../parser/plain-text-parser.js";
+export interface TagAssignment {
+    tagPath: string;
+    confidence: number;
+    source: string;
+    evidence: string;
+    manualOverride?: boolean;
+}
+export interface TagInferenceResult {
+    tags: TagAssignment[];
+    derivedTags: TagAssignment[];
+}
+/**
+ * 最小系统标签推断器。
+ * 这里只保留稳定、可解释的系统派生标签，不再做业务规则识别。
+ */
+export declare class SimpleTagInferenceEngine {
+    constructor();
+    infer(file: FileScanResult, parsed: ParsedDocument): TagInferenceResult;
+}

package/dist/server/modules/affairs-indexer/core/src/tagging/simple-tag-inference.js

@@ -0,0 +1,94 @@
+const EXTENSION_TYPE_TAGS = new Map([
+    [".md", "类型/文本/Markdown"],
+    [".mdx", "类型/文本/Markdown"],
+    [".txt", "类型/文本/纯文本"],
+    [".rtf", "类型/文本/RTF"],
+    [".html", "类型/文本/HTML"],
+    [".htm", "类型/文本/HTML"],
+    [".pdf", "类型/办公/PDF"],
+    [".doc", "类型/办公/Word"],
+    [".docx", "类型/办公/Word"],
+    [".wps", "类型/办公/Word"],
+    [".ppt", "类型/办公/PPT"],
+    [".pptx", "类型/办公/PPT"],
+    [".xls", "类型/办公/Excel"],
+    [".xlsx", "类型/办公/Excel"],
+    [".csv", "类型/表格/CSV"],
+]);
+function setTag(target, assignment) {
+    const current = target.get(assignment.tagPath);
+    if (!current || assignment.confidence >= current.confidence) {
+        target.set(assignment.tagPath, assignment);
+    }
+}
+function startOfLocalDay(value) {
+    return new Date(value.getFullYear(), value.getMonth(), value.getDate(), 0, 0, 0, 0);
+}
+function resolveRecentDayDistance(now, modifiedAt) {
+    const todayStart = startOfLocalDay(now).getTime();
+    const modifiedStart = startOfLocalDay(modifiedAt).getTime();
+    if (!Number.isFinite(todayStart) || !Number.isFinite(modifiedStart)) {
+        return null;
+    }
+    return Math.max(0, Math.floor((todayStart - modifiedStart) / 86400000));
+}
+/**
+ * 最小系统标签推断器。
+ * 这里只保留稳定、可解释的系统派生标签，不再做业务规则识别。
+ */
+export class SimpleTagInferenceEngine {
+    constructor() { }
+    infer(file, parsed) {
+        void parsed;
+        const derived = new Map();
+        const typeTag = EXTENSION_TYPE_TAGS.get(file.extension);
+        if (typeTag) {
+            setTag(derived, {
+                tagPath: typeTag,
+                confidence: 0.92,
+                source: "derived_extension",
+                evidence: `扩展名推导: ${file.extension}`,
+            });
+        }
+        const modifiedAt = new Date(file.mtime);
+        if (!Number.isNaN(modifiedAt.getTime())) {
+            const now = new Date();
+            const recentDayDistance = resolveRecentDayDistance(now, modifiedAt);
+            setTag(derived, {
+                tagPath: `时间/${modifiedAt.getFullYear()}/${String(modifiedAt.getMonth() + 1).padStart(2, "0")}`,
+                confidence: 1,
+                source: "derived_time",
+                evidence: "由修改时间推导的绝对时间标签",
+            });
+            if (recentDayDistance !== null && recentDayDistance <= 29) {
+                setTag(derived, {
+                    tagPath: "时间/最近30天",
+                    confidence: 1,
+                    source: "derived_time_window",
+                    evidence: "最近30天有修改",
+                });
+            }
+            if (recentDayDistance !== null && recentDayDistance <= 6) {
+                setTag(derived, {
+                    tagPath: "时间/最近7天",
+                    confidence: 1,
+                    source: "derived_time_window",
+                    evidence: "最近7天有修改",
+                });
+            }
+            if (recentDayDistance !== null && recentDayDistance <= 2) {
+                setTag(derived, {
+                    tagPath: "时间/最近3天",
+                    confidence: 1,
+                    source: "derived_time_window",
+                    evidence: "最近3天有修改",
+                });
+            }
+        }
+        return {
+            tags: [],
+            derivedTags: [...derived.values()].sort((a, b) => a.tagPath.localeCompare(b.tagPath, "zh-Hans-CN")),
+        };
+    }
+}
+//# sourceMappingURL=simple-tag-inference.js.map

package/dist/server/modules/affairs-indexer/core/src/tagging/simple-tag-inference.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"simple-tag-inference.js","sourceRoot":"","sources":["../../../../../../../src/modules/affairs-indexer/core/src/tagging/simple-tag-inference.ts"],"names":[],"mappings":"AAgBA,MAAM,mBAAmB,GAAG,IAAI,GAAG,CAAiB;IAClD,CAAC,KAAK,EAAE,gBAAgB,CAAC;IACzB,CAAC,MAAM,EAAE,gBAAgB,CAAC;IAC1B,CAAC,MAAM,EAAE,WAAW,CAAC;IACrB,CAAC,MAAM,EAAE,WAAW,CAAC;IACrB,CAAC,OAAO,EAAE,YAAY,CAAC;IACvB,CAAC,MAAM,EAAE,YAAY,CAAC;IACtB,CAAC,MAAM,EAAE,WAAW,CAAC;IACrB,CAAC,MAAM,EAAE,YAAY,CAAC;IACtB,CAAC,OAAO,EAAE,YAAY,CAAC;IACvB,CAAC,MAAM,EAAE,YAAY,CAAC;IACtB,CAAC,MAAM,EAAE,WAAW,CAAC;IACrB,CAAC,OAAO,EAAE,WAAW,CAAC;IACtB,CAAC,MAAM,EAAE,aAAa,CAAC;IACvB,CAAC,OAAO,EAAE,aAAa,CAAC;IACxB,CAAC,MAAM,EAAE,WAAW,CAAC;CACtB,CAAC,CAAC;AAEH,SAAS,MAAM,CAAC,MAAkC,EAAE,UAAyB;IAC3E,MAAM,OAAO,GAAG,MAAM,CAAC,GAAG,CAAC,UAAU,CAAC,OAAO,CAAC,CAAC;IAC/C,IAAI,CAAC,OAAO,IAAI,UAAU,CAAC,UAAU,IAAI,OAAO,CAAC,UAAU,EAAE,CAAC;QAC5D,MAAM,CAAC,GAAG,CAAC,UAAU,CAAC,OAAO,EAAE,UAAU,CAAC,CAAC;IAC7C,CAAC;AACH,CAAC;AAED,SAAS,eAAe,CAAC,KAAW;IAClC,OAAO,IAAI,IAAI,CAAC,KAAK,CAAC,WAAW,EAAE,EAAE,KAAK,CAAC,QAAQ,EAAE,EAAE,KAAK,CAAC,OAAO,EAAE,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC;AACtF,CAAC;AAED,SAAS,wBAAwB,CAAC,GAAS,EAAE,UAAgB;IAC3D,MAAM,UAAU,GAAG,eAAe,CAAC,GAAG,CAAC,CAAC,OAAO,EAAE,CAAC;IAClD,MAAM,aAAa,GAAG,eAAe,CAAC,UAAU,CAAC,CAAC,OAAO,EAAE,CAAC;IAC5D,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,UAAU,CAAC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,aAAa,CAAC,EAAE,CAAC;QACpE,OAAO,IAAI,CAAC;IACd,CAAC;IACD,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,CAAC,UAAU,GAAG,aAAa,CAAC,GAAG,QAAQ,CAAC,CAAC,CAAC;AAC1E,CAAC;AAED;;;GAGG;AACH,MAAM,OAAO,wBAAwB;IACnC,gBAAe,CAAC;IAEhB,KAAK,CAAC,IAAoB,EAAE,MAAsB;QAChD,KAAK,MAAM,CAAC;QACZ,MAAM,OAAO,GAAG,IAAI,GAAG,EAAyB,CAAC;QAEjD,MAAM,OAAO,GAAG,mBAAmB,CAAC,GAAG,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;QACxD,IAAI,OAAO,EAAE,CAAC;YACZ,MAAM,CAAC,OAAO,EAAE;gBACd,OAAO,EAAE,OAAO;gBAChB,UAAU,EAAE,IAAI;gBAChB,MAAM,EAAE,mBAAmB;gBAC3B,QAAQ,EAAE,UAAU,IAAI,CAAC,SAAS,EAAE;aACrC,CAAC,CAAC;QACL,CAAC;QAED,MAAM,UAAU,GAAG,IAAI,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACxC,IAAI,CAAC,MAAM,CAAC,KAAK,CAAC,UAAU,CAAC,OAAO,EAAE,CAAC,EAAE,CAAC;YACxC,MAAM,GAAG,GAAG,IAAI,IAAI,EAAE,CAAC;YACvB,MAAM,iBAAiB,GAAG,wBAAwB,CAAC,GAAG,EAAE,UAAU,CAAC,CAAC;YAEpE,MAAM,CAAC,OAAO,EAAE;gBACd,OAAO,EAAE,MAAM,UAAU,CAAC,WAAW,EAAE,IAAI,MAAM,CAAC,UAAU,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC,EAAE;gBAC/F,UAAU,EAAE,CAAC;gBACb,MAAM,EAAE,cAAc;gBACtB,QAAQ,EAAE,gBAAgB;aAC3B,CAAC,CAAC;YAEH,IAAI,iBAAiB,KAAK,IAAI,IAAI,iBAAiB,IAAI,EAAE,EAAE,CAAC;gBAC1D,MAAM,CAAC,OAAO,EAAE;oBACd,OAAO,EAAE,UAAU;oBACnB,UAAU,EAAE,CAAC;oBACb,MAAM,EAAE,qBAAqB;oBAC7B,QAAQ,EAAE,UAAU;iBACrB,CAAC,CAAC;YACL,CAAC;YAED,IAAI,iBAAiB,KAAK,IAAI,IAAI,iBAAiB,IAAI,CAAC,EAAE,CAAC;gBACzD,MAAM,CAAC,OAAO,EAAE;oBACd,OAAO,EAAE,SAAS;oBAClB,UAAU,EAAE,CAAC;oBACb,MAAM,EAAE,qBAAqB;oBAC7B,QAAQ,EAAE,SAAS;iBACpB,CAAC,CAAC;YACL,CAAC;YAED,IAAI,iBAAiB,KAAK,IAAI,IAAI,iBAAiB,IAAI,CAAC,EAAE,CAAC;gBACzD,MAAM,CAAC,OAAO,EAAE;oBACd,OAAO,EAAE,SAAS;oBAClB,UAAU,EAAE,CAAC;oBACb,MAAM,EAAE,qBAAqB;oBAC7B,QAAQ,EAAE,SAAS;iBACpB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;QAED,OAAO;YACL,IAAI,EAAE,EAAE;YACR,WAAW,EAAE,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,aAAa,CAAC,CAAC,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;SACpG,CAAC;IACJ,CAAC;CACF"}

package/dist/server/modules/affairs-indexer/core/src/utils/abort.d.ts

@@ -0,0 +1,2 @@
+export declare function throwIfAborted(signal?: AbortSignal, fallbackMessage?: string): void;
+export declare function yieldToEventLoop(signal?: AbortSignal, fallbackMessage?: string): Promise<void>;

package/dist/server/modules/affairs-indexer/core/src/utils/abort.js

@@ -0,0 +1,13 @@
+export function throwIfAborted(signal, fallbackMessage = "事务文档库任务已取消") {
+    if (signal?.aborted) {
+        throw signal.reason ?? new Error(fallbackMessage);
+    }
+}
+export async function yieldToEventLoop(signal, fallbackMessage) {
+    throwIfAborted(signal, fallbackMessage);
+    await new Promise((resolve) => {
+        setImmediate(resolve);
+    });
+    throwIfAborted(signal, fallbackMessage);
+}
+//# sourceMappingURL=abort.js.map

package/dist/server/modules/affairs-indexer/core/src/utils/abort.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"abort.js","sourceRoot":"","sources":["../../../../../../../src/modules/affairs-indexer/core/src/utils/abort.ts"],"names":[],"mappings":"AAAA,MAAM,UAAU,cAAc,CAAC,MAAoB,EAAE,eAAe,GAAG,YAAY;IACjF,IAAI,MAAM,EAAE,OAAO,EAAE,CAAC;QACpB,MAAM,MAAM,CAAC,MAAM,IAAI,IAAI,KAAK,CAAC,eAAe,CAAC,CAAC;IACpD,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,gBAAgB,CAAC,MAAoB,EAAE,eAAwB;IACnF,cAAc,CAAC,MAAM,EAAE,eAAe,CAAC,CAAC;IACxC,MAAM,IAAI,OAAO,CAAO,CAAC,OAAO,EAAE,EAAE;QAClC,YAAY,CAAC,OAAO,CAAC,CAAC;IACxB,CAAC,CAAC,CAAC;IACH,cAAc,CAAC,MAAM,EAAE,eAAe,CAAC,CAAC;AAC1C,CAAC"}

package/dist/server/modules/affairs-indexer/core/src/utils/file-streaming.d.ts

@@ -0,0 +1,9 @@
+export declare function iterateNdjsonFileSync<T>(filePath: string, onRecord: (value: T) => void): void;
+export interface JsonArrayFileWriter {
+    append(value: unknown): void;
+    close(): void;
+}
+export declare function createJsonArrayFileWriter(filePath: string, options: {
+    prefix: string;
+    suffix: string;
+}): JsonArrayFileWriter;