Started migrating workflow controller

shared-operations/src/functions/common/detectEmptyPages.ts

@@ -0,0 +1,51 @@
+
+import { PdfFile } from '../../wrappers/PdfFile';
+import { PDFPageProxy } from "pdfjs-dist/types/src/display/api.js";
+import { Image } from 'image-js';
+
+import { getImagesOnPage } from "./getImagesOnPage.js";
+
+export async function detectEmptyPages(file: PdfFile, whiteThreashold: number): Promise<number[]> {
+    const pdfDoc = await file.getAsPdfJs();
+
+    const emptyPages: number[] = [];
+    for (let i = 1; i <= pdfDoc.numPages; i++) {
+        const page = await pdfDoc.getPage(i);
+        console.log("Checking page " + i);
+
+        if(!await hasText(page)) {
+            console.log(`Found text on Page ${i}, page is not empty`);
+            continue;
+        }
+
+        if(!await areImagesBlank(page, whiteThreashold)) {
+            console.log(`Found non white image on Page ${i}, page is not empty`);
+            continue;
+        }
+
+        console.log(`Page ${i} is empty.`);
+        emptyPages.push(i - 1);
+    }
+    return emptyPages;
+}
+
+async function hasText(page: PDFPageProxy): Promise<boolean> {
+    const textContent = await page.getTextContent();
+    return textContent.items.length === 0;
+}
+
+async function areImagesBlank(page: PDFPageProxy, threshold: number): Promise<boolean> {
+    const images = await getImagesOnPage(page);
+    for (const image of images) {
+        if(!await isImageBlank(image, threshold))
+            return false;
+    }
+    return true;
+}
+
+async function isImageBlank(image: string | Uint8Array | ArrayBuffer, threshold: number): Promise<boolean> {
+    var img = await Image.load(image);
+    var grey = img.grey();
+    var mean = grey.getMean();
+    return mean[0] <= threshold;
+}

shared-operations/src/functions/common/getImagesOnPage.ts

@@ -0,0 +1,15 @@
+
+import { PDFPageProxy } from "pdfjs-dist/types/src/display/api.js";
+import * as PDFJS from 'pdfjs-dist';
+
+export async function getImagesOnPage(page: PDFPageProxy) {
+    const ops = await page.getOperatorList();
+    const images: any = [];
+    for (var j=0; j < ops.fnArray.length; j++) {
+        if (ops.fnArray[j] == PDFJS.OPS.paintImageXObject) {
+            const image = page.objs.get(ops.argsArray[j][0]);
+            images.push(image);
+        }
+    }
+    return images;
+}

shared-operations/src/functions/common/pdf-utils.ts

@@ -0,0 +1,47 @@
+
+import { PdfFile, convertAllToPdfLibFile } from '../../wrappers/PdfFile';
+
+export async function sortPdfs(
+        files: PdfFile[],
+        sortType: "orderProvided"|"byFileName"|"byDateModified"|"byDateCreated"|"byPDFTitle" = "orderProvided"
+    ): Promise<PdfFile[]> {
+
+    const pdfLibFiles = await convertAllToPdfLibFile(files);
+    
+    switch(sortType) {
+        case "byFileName":
+            pdfLibFiles.sort((a, b) => {
+                if (!a || !b) return 0;
+                const ad = a.filename, bd = b.filename;
+                if (!ad || !bd) return 0;
+                return ad.localeCompare(bd);
+            });
+            break;
+        case "byDateModified":
+            pdfLibFiles.sort((a, b) => {
+                const ad = a.pdfLib?.getModificationDate()?.getTime();
+                const bd = b.pdfLib?.getModificationDate()?.getTime();
+                if (!ad || !bd) return 0;
+                return ad > bd ? 1 : -1
+            });
+            break;
+        case "byDateCreated":
+            pdfLibFiles.sort((a, b) => {
+                const ad = a.pdfLib?.getCreationDate()?.getTime();
+                const bd = b.pdfLib?.getCreationDate()?.getTime();
+                if (!ad || !bd) return 0;
+                return ad > bd ? 1 : -1
+            });
+            break;
+        case "byPDFTitle":
+            pdfLibFiles.sort((a, b) => {
+                const ad = a.pdfLib?.getTitle();
+                const bd = b.pdfLib?.getTitle();
+                if (!ad || !bd) return 0;
+                return ad.localeCompare(bd);
+            });
+            break;
+    }
+    
+    return pdfLibFiles;
+}