paper-manager 0.2.3 → 0.2.5

package/config.schema.json

@@ -17,9 +17,7 @@
         "properties": {
           "provider": {
             "type": "string",
-            "enum": [
-              "openai"
-            ]
+            "enum": ["openai"]
           },
           "model": {
             "type": "string"
@@ -37,11 +35,7 @@
             "maximum": 9007199254740991
           }
         },
-        "required": [
-          "provider",
-          "model",
-          "apiKey"
-        ],
+        "required": ["provider", "model", "apiKey"],
         "additionalProperties": false
       }
     },
@@ -50,8 +44,6 @@
       "minLength": 1
     }
   },
-  "required": [
-    "embeddingModels"
-  ],
+  "required": ["embeddingModels"],
   "additionalProperties": false
 }

package/dist/pdf/extractor.d.ts

@@ -1,2 +1,2 @@
-import type { Document } from "@langchain/core/documents";
+import { Document } from "@langchain/core/documents";
 export declare function extractPdfContent(pdfPath: string): Promise<Document[]>;

package/dist/pdf/extractor.js

@@ -1,6 +1,18 @@
-import { PDFLoader } from "@langchain/community/document_loaders/fs/pdf";
+import { readFile } from "node:fs/promises";
+import { Document } from "@langchain/core/documents";
+import { PDFParse } from "pdf-parse";
 export async function extractPdfContent(pdfPath) {
-    const loader = new PDFLoader(pdfPath, { splitPages: true });
-    return loader.load();
+    const data = await readFile(pdfPath);
+    const parser = new PDFParse({ data });
+    const result = await parser.getText();
+    await parser.destroy();
+    return result.pages.map((page) => new Document({
+        pageContent: page.text,
+        metadata: {
+            source: pdfPath,
+            pdf: { totalPages: result.total },
+            loc: { pageNumber: page.num },
+        },
+    }));
 }
 //# sourceMappingURL=extractor.js.map

package/dist/pdf/extractor.js.map

@@ -1 +1 @@
-{"version":3,"file":"extractor.js","sourceRoot":"","sources":["../../src/pdf/extractor.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,SAAS,EAAE,MAAM,8CAA8C,CAAC;AAGzE,MAAM,CAAC,KAAK,UAAU,iBAAiB,CAAC,OAAe;IACrD,MAAM,MAAM,GAAG,IAAI,SAAS,CAAC,OAAO,EAAE,EAAE,UAAU,EAAE,IAAI,EAAE,CAAC,CAAC;IAC5D,OAAO,MAAM,CAAC,IAAI,EAAE,CAAC;AACvB,CAAC"}
+{"version":3,"file":"extractor.js","sourceRoot":"","sources":["../../src/pdf/extractor.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"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "paper-manager",
-  "version": "0.2.3",
+  "version": "0.2.5",
   "description": "A paper management system.",
   "keywords": [],
   "homepage": "https://github.com/EurFelux/paper-manager",
@@ -31,6 +31,7 @@
     "commander": "^14.0.3",
     "faiss-node": "^0.5.1",
     "langchain": "^1.2.28",
+    "pdf-parse": "^2.4.5",
     "zod": "^4.3.6"
   },
   "devDependencies": {
@@ -48,7 +49,7 @@
   },
   "scripts": {
     "build": "tsc",
-    "generate:schema": "tsx scripts/generate-schema.ts",
+    "generate:schema": "tsx scripts/generate-schema.ts && oxfmt config.schema.json",
     "release": "tsx scripts/release.ts",
     "test": "echo \"Error: no test specified\" && exit 1",
     "typecheck": "tsc --noEmit",