paper-manager 0.5.0 → 0.7.0

package/dist/db/schema.d.ts

@@ -1,2 +1,371 @@
+export declare const knowledgeBases: import("drizzle-orm/sqlite-core").SQLiteTableWithColumns<{
+    name: "knowledge_bases";
+    schema: undefined;
+    columns: {
+        id: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "id";
+            tableName: "knowledge_bases";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: true;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        name: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "name";
+            tableName: "knowledge_bases";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        description: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "description";
+            tableName: "knowledge_bases";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        embeddingModelId: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "embedding_model_id";
+            tableName: "knowledge_bases";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        createdAt: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "created_at";
+            tableName: "knowledge_bases";
+            dataType: "date";
+            columnType: "SQLiteTimestamp";
+            data: Date;
+            driverParam: number;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: undefined;
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {}>;
+        updatedAt: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "updated_at";
+            tableName: "knowledge_bases";
+            dataType: "date";
+            columnType: "SQLiteTimestamp";
+            data: Date;
+            driverParam: number;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: undefined;
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {}>;
+    };
+    dialect: "sqlite";
+}>;
+export declare const literatures: import("drizzle-orm/sqlite-core").SQLiteTableWithColumns<{
+    name: "literatures";
+    schema: undefined;
+    columns: {
+        id: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "id";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: true;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        title: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "title";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        titleTranslation: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "title_translation";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        author: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "author";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        abstract: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "abstract";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        summary: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "summary";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        keywords: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "keywords";
+            tableName: "literatures";
+            dataType: "json";
+            columnType: "SQLiteTextJson";
+            data: string[];
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: undefined;
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            $type: string[];
+        }>;
+        url: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "url";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        doi: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "doi";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        notes: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "notes";
+            tableName: "literatures";
+            dataType: "json";
+            columnType: "SQLiteTextJson";
+            data: Record<string, string>;
+            driverParam: string;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: undefined;
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            $type: Record<string, string>;
+        }>;
+        knowledgeBaseId: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "knowledge_base_id";
+            tableName: "literatures";
+            dataType: "string";
+            columnType: "SQLiteText";
+            data: string;
+            driverParam: string;
+            notNull: false;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: [string, ...string[]];
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {
+            length: number | undefined;
+        }>;
+        createdAt: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "created_at";
+            tableName: "literatures";
+            dataType: "date";
+            columnType: "SQLiteTimestamp";
+            data: Date;
+            driverParam: number;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: undefined;
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {}>;
+        updatedAt: import("drizzle-orm/sqlite-core").SQLiteColumn<{
+            name: "updated_at";
+            tableName: "literatures";
+            dataType: "date";
+            columnType: "SQLiteTimestamp";
+            data: Date;
+            driverParam: number;
+            notNull: true;
+            hasDefault: false;
+            isPrimaryKey: false;
+            isAutoincrement: false;
+            hasRuntimeDefault: false;
+            enumValues: undefined;
+            baseColumn: never;
+            identity: undefined;
+            generated: undefined;
+        }, {}, {}>;
+    };
+    dialect: "sqlite";
+}>;
+export type KnowledgeBaseMetadata = typeof knowledgeBases.$inferSelect;
+export type LiteratureMetadata = typeof literatures.$inferSelect;
 export declare const CREATE_KNOWLEDGE_BASES_TABLE = "\nCREATE TABLE IF NOT EXISTS knowledge_bases (\n    id TEXT PRIMARY KEY,\n    name TEXT NOT NULL UNIQUE,\n    description TEXT NOT NULL,\n    embedding_model_id TEXT NOT NULL,\n    created_at INTEGER NOT NULL,\n    updated_at INTEGER NOT NULL\n)";
-export declare const CREATE_LITERATURES_TABLE = "\nCREATE TABLE IF NOT EXISTS literatures (\n    id TEXT PRIMARY KEY,\n    title TEXT NOT NULL,\n    title_translation TEXT,\n    author TEXT,\n    abstract TEXT,\n    summary TEXT,\n    keywords TEXT NOT NULL DEFAULT '[]',\n    url TEXT,\n    notes TEXT NOT NULL DEFAULT '{}',\n    knowledge_base_id TEXT,\n    created_at INTEGER NOT NULL,\n    updated_at INTEGER NOT NULL,\n    FOREIGN KEY (knowledge_base_id) REFERENCES knowledge_bases(id) ON DELETE SET NULL\n)";
+export declare const CREATE_LITERATURES_TABLE = "\nCREATE TABLE IF NOT EXISTS literatures (\n    id TEXT PRIMARY KEY,\n    title TEXT NOT NULL,\n    title_translation TEXT,\n    author TEXT,\n    abstract TEXT,\n    summary TEXT,\n    keywords TEXT NOT NULL DEFAULT '[]',\n    url TEXT,\n    doi TEXT,\n    notes TEXT NOT NULL DEFAULT '{}',\n    knowledge_base_id TEXT,\n    created_at INTEGER NOT NULL,\n    updated_at INTEGER NOT NULL,\n    FOREIGN KEY (knowledge_base_id) REFERENCES knowledge_bases(id) ON DELETE SET NULL\n)";

package/dist/db/schema.js

@@ -1,3 +1,31 @@
+import { integer, sqliteTable, text } from "drizzle-orm/sqlite-core";
+// ─── Drizzle Table Definitions ──────────────────────────────
+export const knowledgeBases = sqliteTable("knowledge_bases", {
+    id: text("id").primaryKey(),
+    name: text("name").notNull().unique(),
+    description: text("description").notNull(),
+    embeddingModelId: text("embedding_model_id").notNull(),
+    createdAt: integer("created_at", { mode: "timestamp_ms" }).notNull(),
+    updatedAt: integer("updated_at", { mode: "timestamp_ms" }).notNull(),
+});
+export const literatures = sqliteTable("literatures", {
+    id: text("id").primaryKey(),
+    title: text("title").notNull(),
+    titleTranslation: text("title_translation"),
+    author: text("author"),
+    abstract: text("abstract"),
+    summary: text("summary"),
+    keywords: text("keywords", { mode: "json" }).$type().notNull(),
+    url: text("url"),
+    doi: text("doi"),
+    notes: text("notes", { mode: "json" }).$type().notNull(),
+    knowledgeBaseId: text("knowledge_base_id").references(() => knowledgeBases.id, {
+        onDelete: "set null",
+    }),
+    createdAt: integer("created_at", { mode: "timestamp_ms" }).notNull(),
+    updatedAt: integer("updated_at", { mode: "timestamp_ms" }).notNull(),
+});
+// ─── Bootstrap SQL ──────────────────────────────────────────
 export const CREATE_KNOWLEDGE_BASES_TABLE = `
 CREATE TABLE IF NOT EXISTS knowledge_bases (
     id TEXT PRIMARY KEY,
@@ -17,6 +45,7 @@ CREATE TABLE IF NOT EXISTS literatures (
     summary TEXT,
     keywords TEXT NOT NULL DEFAULT '[]',
     url TEXT,
+    doi TEXT,
     notes TEXT NOT NULL DEFAULT '{}',
     knowledge_base_id TEXT,
     created_at INTEGER NOT NULL,

package/dist/db/schema.js.map

@@ -1 +1 @@
-{"version":3,"file":"schema.js","sourceRoot":"","sources":["../../src/db/schema.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,MAAM,4BAA4B,GAAG;;;;;;;;EAQ1C,CAAC;AAEH,MAAM,CAAC,MAAM,wBAAwB,GAAG;;;;;;;;;;;;;;;EAetC,CAAC"}
+{"version":3,"file":"schema.js","sourceRoot":"","sources":["../../src/db/schema.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,WAAW,EAAE,IAAI,EAAE,MAAM,yBAAyB,CAAC;AAErE,+DAA+D;AAE/D,MAAM,CAAC,MAAM,cAAc,GAAG,WAAW,CAAC,iBAAiB,EAAE;IAC3D,EAAE,EAAE,IAAI,CAAC,IAAI,CAAC,CAAC,UAAU,EAAE;IAC3B,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,OAAO,EAAE,CAAC,MAAM,EAAE;IACrC,WAAW,EAAE,IAAI,CAAC,aAAa,CAAC,CAAC,OAAO,EAAE;IAC1C,gBAAgB,EAAE,IAAI,CAAC,oBAAoB,CAAC,CAAC,OAAO,EAAE;IACtD,SAAS,EAAE,OAAO,CAAC,YAAY,EAAE,EAAE,IAAI,EAAE,cAAc,EAAE,CAAC,CAAC,OAAO,EAAE;IACpE,SAAS,EAAE,OAAO,CAAC,YAAY,EAAE,EAAE,IAAI,EAAE,cAAc,EAAE,CAAC,CAAC,OAAO,EAAE;CACrE,CAAC,CAAC;AAEH,MAAM,CAAC,MAAM,WAAW,GAAG,WAAW,CAAC,aAAa,EAAE;IACpD,EAAE,EAAE,IAAI,CAAC,IAAI,CAAC,CAAC,UAAU,EAAE;IAC3B,KAAK,EAAE,IAAI,CAAC,OAAO,CAAC,CAAC,OAAO,EAAE;IAC9B,gBAAgB,EAAE,IAAI,CAAC,mBAAmB,CAAC;IAC3C,MAAM,EAAE,IAAI,CAAC,QAAQ,CAAC;IACtB,QAAQ,EAAE,IAAI,CAAC,UAAU,CAAC;IAC1B,OAAO,EAAE,IAAI,CAAC,SAAS,CAAC;IACxB,QAAQ,EAAE,IAAI,CAAC,UAAU,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC,KAAK,EAAY,CAAC,OAAO,EAAE;IACxE,GAAG,EAAE,IAAI,CAAC,KAAK,CAAC;IAChB,GAAG,EAAE,IAAI,CAAC,KAAK,CAAC;IAChB,KAAK,EAAE,IAAI,CAAC,OAAO,EAAE,EAAE,IAAI,EAAE,MAAM,EAAE,CAAC,CAAC,KAAK,EAA0B,CAAC,OAAO,EAAE;IAChF,eAAe,EAAE,IAAI,CAAC,mBAAmB,CAAC,CAAC,UAAU,CAAC,GAAG,EAAE,CAAC,cAAc,CAAC,EAAE,EAAE;QAC7E,QAAQ,EAAE,UAAU;KACrB,CAAC;IACF,SAAS,EAAE,OAAO,CAAC,YAAY,EAAE,EAAE,IAAI,EAAE,cAAc,EAAE,CAAC,CAAC,OAAO,EAAE;IACpE,SAAS,EAAE,OAAO,CAAC,YAAY,EAAE,EAAE,IAAI,EAAE,cAAc,EAAE,CAAC,CAAC,OAAO,EAAE;CACrE,CAAC,CAAC;AAOH,+DAA+D;AAE/D,MAAM,CAAC,MAAM,4BAA4B,GAAG;;;;;;;;EAQ1C,CAAC;AAEH,MAAM,CAAC,MAAM,wBAAwB,GAAG;;;;;;;;;;;;;;;;EAgBtC,CAAC"}

package/dist/db/test-utils.d.ts

@@ -1,6 +1,6 @@
-import type BetterSqlite3 from "better-sqlite3";
+import type { AppDatabase } from "./index.js";
 /**
- * Creates an in-memory SQLite database with the schema initialized.
+ * Creates an in-memory SQLite database wrapped with Drizzle.
  * Each call returns a fresh, isolated database instance.
  */
-export declare function createTestDb(): BetterSqlite3.Database;
+export declare function createTestDb(): AppDatabase;

package/dist/db/test-utils.js

@@ -1,14 +1,12 @@
-import Database from "better-sqlite3";
-import { initializeDatabase } from "./index.js";
+import { drizzle } from "drizzle-orm/better-sqlite3";
+import { initializeDatabase, openDatabase } from "./index.js";
 /**
- * Creates an in-memory SQLite database with the schema initialized.
+ * Creates an in-memory SQLite database wrapped with Drizzle.
  * Each call returns a fresh, isolated database instance.
  */
 export function createTestDb() {
-    const db = new Database(":memory:");
-    db.pragma("journal_mode = WAL");
-    db.pragma("foreign_keys = ON");
-    initializeDatabase(db);
-    return db;
+    const client = openDatabase(":memory:");
+    initializeDatabase(client);
+    return drizzle(client);
 }
 //# sourceMappingURL=test-utils.js.map

package/dist/db/test-utils.js.map

@@ -1 +1 @@
-{"version":3,"file":"test-utils.js","sourceRoot":"","sources":["../../src/db/test-utils.ts"],"names":[],"mappings":"AACA,OAAO,QAAQ,MAAM,gBAAgB,CAAC;AAEtC,OAAO,EAAE,kBAAkB,EAAE,MAAM,YAAY,CAAC;AAEhD;;;GAGG;AACH,MAAM,UAAU,YAAY;IAC1B,MAAM,EAAE,GAAG,IAAI,QAAQ,CAAC,UAAU,CAAC,CAAC;IACpC,EAAE,CAAC,MAAM,CAAC,oBAAoB,CAAC,CAAC;IAChC,EAAE,CAAC,MAAM,CAAC,mBAAmB,CAAC,CAAC;IAC/B,kBAAkB,CAAC,EAAE,CAAC,CAAC;IACvB,OAAO,EAAE,CAAC;AACZ,CAAC"}
+{"version":3,"file":"test-utils.js","sourceRoot":"","sources":["../../src/db/test-utils.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,MAAM,4BAA4B,CAAC;AAGrD,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,MAAM,YAAY,CAAC;AAE9D;;;GAGG;AACH,MAAM,UAAU,YAAY;IAC1B,MAAM,MAAM,GAAG,YAAY,CAAC,UAAU,CAAC,CAAC;IACxC,kBAAkB,CAAC,MAAM,CAAC,CAAC;IAC3B,OAAO,OAAO,CAAC,MAAM,CAAC,CAAC;AACzB,CAAC"}

package/dist/extractor/index.d.ts

@@ -1,5 +1,6 @@
 import type { Document } from "@langchain/core/documents";
-import { extractPdfContent } from "./pdf.js";
+import { extractPdfContent, extractPdfMetadata } from "./pdf.js";
 import { extractTextContent } from "./text.js";
-export { extractPdfContent, extractTextContent };
+export { extractPdfContent, extractPdfMetadata, extractTextContent };
+export type { PdfMetadata } from "./pdf.js";
 export declare function extractContent(filePath: string): Promise<Document[]>;

package/dist/extractor/index.js

@@ -1,7 +1,7 @@
 import mime from "mime-types";
-import { extractPdfContent } from "./pdf.js";
+import { extractPdfContent, extractPdfMetadata } from "./pdf.js";
 import { extractTextContent } from "./text.js";
-export { extractPdfContent, extractTextContent };
+export { extractPdfContent, extractPdfMetadata, extractTextContent };
 const TEXT_LIKE_MIME_TYPES = new Set([
     "application/x-tex",
     "application/x-latex",

package/dist/extractor/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/extractor/index.ts"],"names":[],"mappings":"AACA,OAAO,IAAI,MAAM,YAAY,CAAC;AAE9B,OAAO,EAAE,iBAAiB,EAAE,MAAM,UAAU,CAAC;AAC7C,OAAO,EAAE,kBAAkB,EAAE,MAAM,WAAW,CAAC;AAE/C,OAAO,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,CAAC;AAEjD,MAAM,oBAAoB,GAAwB,IAAI,GAAG,CAAC;IACxD,mBAAmB;IACnB,qBAAqB;CACtB,CAAC,CAAC;AAEH,SAAS,UAAU,CAAC,QAAgB;IAClC,OAAO,QAAQ,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,oBAAoB,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC;AAC5E,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,cAAc,CAAC,QAAgB;IACnD,MAAM,QAAQ,GAAG,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC;IAEvC,IAAI,QAAQ,KAAK,iBAAiB,EAAE,CAAC;QACnC,OAAO,iBAAiB,CAAC,QAAQ,CAAC,CAAC;IACrC,CAAC;IAED,IAAI,OAAO,QAAQ,KAAK,QAAQ,IAAI,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;QACzD,OAAO,kBAAkB,CAAC,QAAQ,CAAC,CAAC;IACtC,CAAC;IAED,MAAM,GAAG,GAAG,QAAQ,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,IAAI,SAAS,CAAC;IACnD,MAAM,IAAI,KAAK,CAAC,2BAA2B,GAAG,KAAK,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;AAC1E,CAAC"}
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/extractor/index.ts"],"names":[],"mappings":"AACA,OAAO,IAAI,MAAM,YAAY,CAAC;AAE9B,OAAO,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,MAAM,UAAU,CAAC;AACjE,OAAO,EAAE,kBAAkB,EAAE,MAAM,WAAW,CAAC;AAE/C,OAAO,EAAE,iBAAiB,EAAE,kBAAkB,EAAE,kBAAkB,EAAE,CAAC;AAGrE,MAAM,oBAAoB,GAAwB,IAAI,GAAG,CAAC;IACxD,mBAAmB;IACnB,qBAAqB;CACtB,CAAC,CAAC;AAEH,SAAS,UAAU,CAAC,QAAgB;IAClC,OAAO,QAAQ,CAAC,UAAU,CAAC,OAAO,CAAC,IAAI,oBAAoB,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC;AAC5E,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,cAAc,CAAC,QAAgB;IACnD,MAAM,QAAQ,GAAG,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC;IAEvC,IAAI,QAAQ,KAAK,iBAAiB,EAAE,CAAC;QACnC,OAAO,iBAAiB,CAAC,QAAQ,CAAC,CAAC;IACrC,CAAC;IAED,IAAI,OAAO,QAAQ,KAAK,QAAQ,IAAI,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;QACzD,OAAO,kBAAkB,CAAC,QAAQ,CAAC,CAAC;IACtC,CAAC;IAED,MAAM,GAAG,GAAG,QAAQ,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,IAAI,SAAS,CAAC;IACnD,MAAM,IAAI,KAAK,CAAC,2BAA2B,GAAG,KAAK,MAAM,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC;AAC1E,CAAC"}

package/dist/extractor/pdf.d.ts

@@ -1,2 +1,13 @@
 import { Document } from "@langchain/core/documents";
+export interface PdfMetadata {
+    title: string | null;
+    author: string | null;
+    subject: string | null;
+    keywords: string[];
+    doi: string | null;
+    creator: string | null;
+    creationDate: Date | null;
+    modDate: Date | null;
+}
 export declare function extractPdfContent(pdfPath: string): Promise<Document[]>;
+export declare function extractPdfMetadata(pdfPath: string): Promise<PdfMetadata>;

package/dist/extractor/pdf.js

@@ -15,4 +15,69 @@ export async function extractPdfContent(pdfPath) {
         },
     }));
 }
+export async function extractPdfMetadata(pdfPath) {
+    const data = await readFile(pdfPath);
+    const parser = new PDFParse({ data });
+    const result = await parser.getInfo();
+    await parser.destroy();
+    const info = result.info;
+    const custom = info?.["Custom"];
+    const title = nonEmptyStringOrNull(info?.["Title"]);
+    const author = nonEmptyStringOrNull(info?.["Author"]);
+    const subject = nonEmptyStringOrNull(info?.["Subject"]);
+    const creator = nonEmptyStringOrNull(info?.["Creator"]);
+    const creationDate = parsePdfDate(nonEmptyStringOrNull(info?.["CreationDate"]));
+    const modDate = parsePdfDate(nonEmptyStringOrNull(info?.["ModDate"]));
+    const rawKeywords = nonEmptyStringOrNull(info?.["Keywords"]);
+    const keywords = rawKeywords
+        ? rawKeywords
+            .split(/[,;]/)
+            .map((k) => k.trim())
+            .filter(Boolean)
+        : [];
+    // DOI can appear in Custom fields (case-insensitive lookup)
+    const doi = findCustomField(custom, "doi");
+    return { title, author, subject, keywords, doi, creator, creationDate, modDate };
+}
+function nonEmptyStringOrNull(value) {
+    if (typeof value === "string" && value.trim().length > 0) {
+        return value.trim();
+    }
+    return null;
+}
+/**
+ * Parse PDF date format: D:YYYYMMDDHHmmSSOHH'mm
+ * Examples: "D:20231215120000Z", "D:20231215", "D:20231215120000+08'00"
+ */
+function parsePdfDate(value) {
+    if (!value)
+        return null;
+    const cleaned = value.replace(/^D:/, "");
+    // Extract components: YYYY[MM[DD[HH[mm[SS]]]]]
+    const match = /^(\d{4})(\d{2})?(\d{2})?(\d{2})?(\d{2})?(\d{2})?/.exec(cleaned);
+    if (!match)
+        return null;
+    const year = match[1];
+    const month = match[2] ?? "01";
+    const day = match[3] ?? "01";
+    const hour = match[4] ?? "00";
+    const min = match[5] ?? "00";
+    const sec = match[6] ?? "00";
+    const date = new Date(`${year}-${month}-${day}T${hour}:${min}:${sec}Z`);
+    return Number.isNaN(date.getTime()) ? null : date;
+}
+/**
+ * Case-insensitive lookup in the Custom fields dictionary.
+ */
+function findCustomField(custom, key) {
+    if (!custom)
+        return null;
+    const lowerKey = key.toLowerCase();
+    for (const [k, v] of Object.entries(custom)) {
+        if (k.toLowerCase() === lowerKey) {
+            return nonEmptyStringOrNull(v);
+        }
+    }
+    return null;
+}
 //# sourceMappingURL=pdf.js.map

package/dist/extractor/pdf.js.map

@@ -1 +1 @@
-{"version":3,"file":"pdf.js","sourceRoot":"","sources":["../../src/extractor/pdf.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAE5C,OAAO,EAAE,QAAQ,EAAE,MAAM,2BAA2B,CAAC;AACrD,OAAO,EAAE,QAAQ,EAAE,MAAM,WAAW,CAAC;AAErC,MAAM,CAAC,KAAK,UAAU,iBAAiB,CAAC,OAAe;IACrD,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,MAAM,GAAG,IAAI,QAAQ,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;IACtC,MAAM,MAAM,GAAG,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;IACtC,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;IAEvB,OAAO,MAAM,CAAC,KAAK,CAAC,GAAG,CACrB,CAAC,IAAI,EAAE,EAAE,CACP,IAAI,QAAQ,CAAC;QACX,WAAW,EAAE,IAAI,CAAC,IAAI;QACtB,QAAQ,EAAE;YACR,MAAM,EAAE,OAAO;YACf,GAAG,EAAE,EAAE,UAAU,EAAE,MAAM,CAAC,KAAK,EAAE;YACjC,GAAG,EAAE,EAAE,UAAU,EAAE,IAAI,CAAC,GAAG,EAAE;SAC9B;KACF,CAAC,CACL,CAAC;AACJ,CAAC"}
+{"version":3,"file":"pdf.js","sourceRoot":"","sources":["../../src/extractor/pdf.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAE5C,OAAO,EAAE,QAAQ,EAAE,MAAM,2BAA2B,CAAC;AACrD,OAAO,EAAE,QAAQ,EAAE,MAAM,WAAW,CAAC;AAarC,MAAM,CAAC,KAAK,UAAU,iBAAiB,CAAC,OAAe;IACrD,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,MAAM,GAAG,IAAI,QAAQ,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;IACtC,MAAM,MAAM,GAAG,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;IACtC,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;IAEvB,OAAO,MAAM,CAAC,KAAK,CAAC,GAAG,CACrB,CAAC,IAAI,EAAE,EAAE,CACP,IAAI,QAAQ,CAAC;QACX,WAAW,EAAE,IAAI,CAAC,IAAI;QACtB,QAAQ,EAAE;YACR,MAAM,EAAE,OAAO;YACf,GAAG,EAAE,EAAE,UAAU,EAAE,MAAM,CAAC,KAAK,EAAE;YACjC,GAAG,EAAE,EAAE,UAAU,EAAE,IAAI,CAAC,GAAG,EAAE;SAC9B;KACF,CAAC,CACL,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,OAAe;IACtD,MAAM,IAAI,GAAG,MAAM,QAAQ,CAAC,OAAO,CAAC,CAAC;IACrC,MAAM,MAAM,GAAG,IAAI,QAAQ,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;IACtC,MAAM,MAAM,GAAG,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;IACtC,MAAM,MAAM,CAAC,OAAO,EAAE,CAAC;IAEvB,MAAM,IAAI,GAAG,MAAM,CAAC,IAA2C,CAAC;IAChE,MAAM,MAAM,GAAG,IAAI,EAAE,CAAC,QAAQ,CAAwC,CAAC;IAEvE,MAAM,KAAK,GAAG,oBAAoB,CAAC,IAAI,EAAE,CAAC,OAAO,CAAC,CAAC,CAAC;IACpD,MAAM,MAAM,GAAG,oBAAoB,CAAC,IAAI,EAAE,CAAC,QAAQ,CAAC,CAAC,CAAC;IACtD,MAAM,OAAO,GAAG,oBAAoB,CAAC,IAAI,EAAE,CAAC,SAAS,CAAC,CAAC,CAAC;IACxD,MAAM,OAAO,GAAG,oBAAoB,CAAC,IAAI,EAAE,CAAC,SAAS,CAAC,CAAC,CAAC;IACxD,MAAM,YAAY,GAAG,YAAY,CAAC,oBAAoB,CAAC,IAAI,EAAE,CAAC,cAAc,CAAC,CAAC,CAAC,CAAC;IAChF,MAAM,OAAO,GAAG,YAAY,CAAC,oBAAoB,CAAC,IAAI,EAAE,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC;IAEtE,MAAM,WAAW,GAAG,oBAAoB,CAAC,IAAI,EAAE,CAAC,UAAU,CAAC,CAAC,CAAC;IAC7D,MAAM,QAAQ,GAAG,WAAW;QAC1B,CAAC,CAAC,WAAW;aACR,KAAK,CAAC,MAAM,CAAC;aACb,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;aACpB,MAAM,CAAC,OAAO,CAAC;QACpB,CAAC,CAAC,EAAE,CAAC;IAEP,4DAA4D;IAC5D,MAAM,GAAG,GAAG,eAAe,CAAC,MAAM,EAAE,KAAK,CAAC,CAAC;IAE3C,OAAO,EAAE,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,QAAQ,EAAE,GAAG,EAAE,OAAO,EAAE,YAAY,EAAE,OAAO,EAAE,CAAC;AACnF,CAAC;AAED,SAAS,oBAAoB,CAAC,KAAc;IAC1C,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACzD,OAAO,KAAK,CAAC,IAAI,EAAE,CAAC;IACtB,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED;;;GAGG;AACH,SAAS,YAAY,CAAC,KAAoB;IACxC,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAC;IACxB,MAAM,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC;IAEzC,+CAA+C;IAC/C,MAAM,KAAK,GAAG,kDAAkD,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IAC/E,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAC;IAExB,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAE,CAAC;IACvB,MAAM,KAAK,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;IAC/B,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;IAC7B,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;IAC9B,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;IAC7B,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,IAAI,CAAC;IAE7B,MAAM,IAAI,GAAG,IAAI,IAAI,CAAC,GAAG,IAAI,IAAI,KAAK,IAAI,GAAG,IAAI,IAAI,IAAI,GAAG,IAAI,GAAG,GAAG,CAAC,CAAC;IACxE,OAAO,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC;AACpD,CAAC;AAED;;GAEG;AACH,SAAS,eAAe,CAAC,MAA2C,EAAE,GAAW;IAC/E,IAAI,CAAC,MAAM;QAAE,OAAO,IAAI,CAAC;IACzB,MAAM,QAAQ,GAAG,GAAG,CAAC,WAAW,EAAE,CAAC;IACnC,KAAK,MAAM,CAAC,CAAC,EAAE,CAAC,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,EAAE,CAAC;QAC5C,IAAI,CAAC,CAAC,WAAW,EAAE,KAAK,QAAQ,EAAE,CAAC;YACjC,OAAO,oBAAoB,CAAC,CAAC,CAAC,CAAC;QACjC,CAAC;IACH,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC"}

package/dist/extractor/pdf.test.d.ts

@@ -0,0 +1 @@
+export {};