@credal/actions 0.2.133 → 0.2.135

package/dist/utils/pdf.js

@@ -7,27 +7,34 @@ var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, ge
         step((generator = generator.apply(thisArg, _arguments || [])).next());
     });
 };
-import { getDocument } from "pdfjs-dist/legacy/build/pdf.mjs";
-export function extractTextFromPdf(input) {
+import PDFParser from "pdf2json";
+export function extractTextFromPdf(buffer) {
     return __awaiter(this, void 0, void 0, function* () {
-        // Convert Buffer or ArrayBuffer -> plain Uint8Array
-        const data = input instanceof Uint8Array && !(typeof Buffer !== "undefined" && Buffer.isBuffer(input))
-            ? input
-            : typeof Buffer !== "undefined" && Buffer.isBuffer(input)
-                ? new Uint8Array(input) // copies bytes out of the Buffer
-                : new Uint8Array(input); // ArrayBuffer case
-        // Load PDF
-        const loadingTask = getDocument({ data });
-        const pdf = yield loadingTask.promise;
-        const pages = [];
-        for (let i = 1; i <= pdf.numPages; i++) {
-            const page = yield pdf.getPage(i);
-            const content = yield page.getTextContent();
-            // content.items is typed as TextItem | TextMarkedContent
-            const strings = content.items.map(item => ("str" in item ? item.str : "")).join(" ");
-            pages.push(strings.trim());
+        try {
+            const extractedText = yield new Promise((resolve, reject) => {
+                const pdfParser = new PDFParser();
+                pdfParser.on("pdfParser_dataError", (errData) => {
+                    reject(errData.parserError || new Error("PDF parsing failed"));
+                });
+                pdfParser.on("pdfParser_dataReady", (pdfData) => {
+                    try {
+                        const text = pdfData.Pages.map((page) => page.Texts.map((textItem) => {
+                            // Handle cases where R array might be empty or have multiple runs
+                            return textItem.R.map((run) => decodeURIComponent(run.T)).join("");
+                        }).join("")).join("\n");
+                        resolve(text);
+                    }
+                    catch (error) {
+                        reject(error);
+                    }
+                });
+                pdfParser.parseBuffer(Buffer.from(buffer));
+            });
+            return extractedText;
+        }
+        catch (error) {
+            console.error("Error extracting PDF text:", error);
+            throw error;
         }
-        yield pdf.destroy();
-        return pages.join("\n\n");
     });
 }

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@credal/actions",
-  "version": "0.2.133",
+  "version": "0.2.135",
   "type": "module",
   "description": "AI Actions by Credal AI",
   "sideEffects": false,
@@ -68,7 +68,6 @@
     "node-forge": "^1.3.1",
     "p-limit": "^7.1.1",
     "pdf2json": "^3.1.6",
-    "pdfjs-dist": "^5.4.149",
     "resend": "^4.7.0",
     "snowflake-sdk": "^2.0.2",
     "ts-node": "^10.9.2",

package/dist/actions/providers/jamf/types.d.ts

@@ -1,8 +0,0 @@
-import { z } from "zod";
-export declare const TokenResponseSchema: z.ZodObject<{
-    token: z.ZodString;
-}, "strip", z.ZodTypeAny, {
-    token: string;
-}, {
-    token: string;
-}>;

package/dist/actions/providers/jamf/types.js

@@ -1,7 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.TokenResponseSchema = void 0;
-const zod_1 = require("zod");
-exports.TokenResponseSchema = zod_1.z.object({
-    token: zod_1.z.string(),
-});