OCR PDFs as fallback during upload (#3204)

* OCR PDFs as fallback in spawn thread

* wip

* build our own worker fanout and wrapper

* norm pkgs

* bump dev

.github/workflows/dev-build.yaml

@@ -6,7 +6,7 @@ concurrency:
 
 on:
   push:
-    branches: ['agent-builder'] # put your current branch to create a build. Core team only.
+    branches: ['ocr-parse-pdfs'] # put your current branch to create a build. Core team only.
     paths-ignore:
       - '**.md'
       - 'cloud-deployments/*'

collector/package.json

@@ -19,6 +19,7 @@
     "@xenova/transformers": "^2.11.0",
     "bcrypt": "^5.1.0",
     "body-parser": "^1.20.2",
+    "canvas": "^2.11.2",
     "cors": "^2.8.5",
     "dotenv": "^16.0.3",
     "epub2": "^3.0.2",
@@ -40,6 +41,7 @@
     "puppeteer": "~21.5.2",
     "sharp": "^0.33.5",
     "slugify": "^1.6.6",
+    "tesseract.js": "^6.0.0",
     "url-pattern": "^1.0.3",
     "uuid": "^9.0.0",
     "wavefile": "^11.0.0",
@@ -50,4 +52,4 @@
     "nodemon": "^2.0.22",
     "prettier": "^2.4.1"
   }
-}
+}
\ No newline at end of file

collector/processSingleFile/convert/asPDF/index.js

@@ -7,6 +7,7 @@ const {
 const { tokenizeString } = require("../../../utils/tokenizer");
 const { default: slugify } = require("slugify");
 const PDFLoader = require("./PDFLoader");
+const OCRLoader = require("../../../utils/OCRLoader");
 
 async function asPdf({ fullFilePath = "", filename = "" }) {
   const pdfLoader = new PDFLoader(fullFilePath, {
@@ -15,7 +16,14 @@ async function asPdf({ fullFilePath = "", filename = "" }) {
 
   console.log(`-- Working ${filename} --`);
   const pageContent = [];
-  const docs = await pdfLoader.load();
+  let docs = await pdfLoader.load();
+
+  if (docs.length === 0) {
+    console.log(
+      `[asPDF] No text content found for ${filename}. Will attempt OCR parse.`
+    );
+    docs = await new OCRLoader().ocrPDF(fullFilePath);
+  }
 
   for (const doc of docs) {
     console.log(
@@ -28,7 +36,7 @@ async function asPdf({ fullFilePath = "", filename = "" }) {
   }
 
   if (!pageContent.length) {
-    console.error(`Resulting text content was empty for ${filename}.`);
+    console.error(`[asPDF] Resulting text content was empty for ${filename}.`);
     trashFile(fullFilePath);
     return {
       success: false,

collector/utils/OCRLoader/CanvasFactory.js

+/**
+ * This is a factory for creating a canvas and context in Node.js
+ * it is used to create a canvas and context for the PDFLoader for turning the PDF into an image
+ * so we can later use the image to extract text from the PDF.
+ */
+class NodeCanvasFactory {
+  constructor() {
+    this.CanvasModule = null;
+  }
+
+  async init() {
+    this.CanvasModule = await import("canvas");
+    this.Image = this.CanvasModule.Image;
+  }
+
+  /**
+   * Creates a canvas and context for the PDFLoader
+   * @param {number} width - The width of the canvas
+   * @param {number} height - The height of the canvas
+   * @param {boolean} transparent - Whether the canvas is transparent
+   * @returns {{canvas: HTMLCanvasElement, context: CanvasRenderingContext2D}} - The canvas and context
+   */
+  create(width, height, transparent = false) {
+    const canvas = this.CanvasModule.createCanvas(width, height);
+    const context = canvas.getContext("2d", { alpha: transparent });
+    if (transparent) context.clearRect(0, 0, width, height);
+    return {
+      canvas,
+      context,
+    };
+  }
+
+  /**
+   * Required for the PDFLoader pdfjs interation - do not remove or use directly.
+   */
+  reset(canvasAndContext, width, height) {
+    canvasAndContext.canvas.width = width;
+    canvasAndContext.canvas.height = height;
+  }
+
+  /**
+   * Required for the PDFLoader pdfjs interation - do not remove or use directly.
+   */
+  destroy(canvasAndContext) {
+    canvasAndContext.canvas.width = 0;
+    canvasAndContext.canvas.height = 0;
+    canvasAndContext.canvas = null;
+    canvasAndContext.context = null;
+  }
+}
+
+module.exports = NodeCanvasFactory;

collector/utils/OCRLoader/index.js

+const fs = require("fs");
+const os = require("os");
+const path = require("path");
+const NodeCanvasFactory = require("./CanvasFactory");
+
+class OCRLoader {
+  constructor() {
+    this.cacheDir = path.resolve(
+      process.env.STORAGE_DIR
+        ? path.resolve(process.env.STORAGE_DIR, `models`, `tesseract`)
+        : path.resolve(__dirname, `../../../server/storage/models/tesseract`)
+    );
+  }
+
+  log(text, ...args) {
+    console.log(`\x1b[36m[OCRLoader]\x1b[0m ${text}`, ...args);
+  }
+
+  /**
+   * Loads a PDF file and returns an array of documents.
+   * This function is reserved to parsing for SCANNED documents - digital documents are not supported in this function
+   * @returns {Promise<{pageContent: string, metadata: object}[]>} An array of documents with page content and metadata.
+   */
+  async ocrPDF(
+    filePath,
+    { maxExecutionTime = 300_000, batchSize = 10, maxWorkers = null } = {}
+  ) {
+    if (
+      !filePath ||
+      !fs.existsSync(filePath) ||
+      !fs.statSync(filePath).isFile()
+    ) {
+      this.log(`File ${filePath} does not exist. Skipping OCR.`);
+      return [];
+    }
+
+    const documentTitle = path.basename(filePath);
+    this.log(`Starting OCR of ${documentTitle}`);
+    const pdfjs = await import("pdf-parse/lib/pdf.js/v2.0.550/build/pdf.js");
+    let buffer = fs.readFileSync(filePath);
+    const canvasFactory = new NodeCanvasFactory();
+    await canvasFactory.init();
+    global.Image = canvasFactory.Image;
+
+    const pdfDocument = await pdfjs.getDocument({
+      data: new Uint8Array(buffer),
+      canvasFactory,
+    }).promise;
+    buffer = null;
+
+    const documents = [];
+    const meta = await pdfDocument.getMetadata().catch(() => null);
+    const metadata = {
+      source: filePath,
+      pdf: {
+        version: "v2.0.550",
+        info: meta?.info,
+        metadata: meta?.metadata,
+        totalPages: pdfDocument.numPages,
+      },
+    };
+
+    async function getPageAsBuffer(pageNumber, scale = 1) {
+      let canvas = null;
+      let context = null;
+      try {
+        const page = await pdfDocument.getPage(pageNumber);
+        const viewport = page.getViewport(scale);
+        ({ canvas, context } = canvasFactory.create(
+          viewport.width,
+          viewport.height
+        ));
+        await page.render({
+          canvasFactory,
+          canvasContext: context,
+          viewport,
+        }).promise;
+        return canvas.toBuffer();
+      } catch (e) {
+        this.log(`Error getting page as buffer: ${e.message}`);
+        return null;
+      } finally {
+        canvas = null;
+        context = null;
+      }
+    }
+
+    const { createWorker, OEM } = require("tesseract.js");
+    const BATCH_SIZE = batchSize;
+    const MAX_EXECUTION_TIME = maxExecutionTime;
+    const NUM_WORKERS = maxWorkers ?? Math.min(os.cpus().length, 4);
+    const totalPages = pdfDocument.numPages;
+    const workerPool = await Promise.all(
+      Array(NUM_WORKERS)
+        .fill(0)
+        .map(() =>
+          createWorker("eng", OEM.LSTM_ONLY, {
+            cachePath: this.cacheDir,
+          })
+        )
+    );
+
+    const startTime = Date.now();
+    try {
+      this.log("Bootstrapping OCR completed successfully!", {
+        MAX_EXECUTION_TIME_MS: MAX_EXECUTION_TIME,
+        BATCH_SIZE,
+        MAX_CONCURRENT_WORKERS: NUM_WORKERS,
+        TOTAL_PAGES: totalPages,
+      });
+      const timeoutPromise = new Promise((_, reject) => {
+        setTimeout(() => {
+          reject(
+            new Error(
+              `OCR job took too long to complete (${
+                MAX_EXECUTION_TIME / 1000
+              } seconds)`
+            )
+          );
+        }, MAX_EXECUTION_TIME);
+      });
+
+      const processPages = async () => {
+        for (
+          let startPage = 1;
+          startPage <= totalPages;
+          startPage += BATCH_SIZE
+        ) {
+          const endPage = Math.min(startPage + BATCH_SIZE - 1, totalPages);
+          const pageNumbers = Array.from(
+            { length: endPage - startPage + 1 },
+            (_, i) => startPage + i
+          );
+          this.log(`Working on pages ${startPage} - ${endPage}`);
+
+          const pageQueue = [...pageNumbers];
+          const results = [];
+          const workerPromises = workerPool.map(async (worker, workerIndex) => {
+            while (pageQueue.length > 0) {
+              const pageNum = pageQueue.shift();
+              this.log(
+                `\x1b[34m[Worker ${
+                  workerIndex + 1
+                }]\x1b[0m assigned pg${pageNum}`
+              );
+              const imageBuffer = await getPageAsBuffer(pageNum, 5);
+              const { data } = await worker.recognize(imageBuffer, {}, "text");
+              this.log(
+                `✅ \x1b[34m[Worker ${
+                  workerIndex + 1
+                }]\x1b[0m completed pg${pageNum}`
+              );
+              results.push({
+                pageContent: data.text,
+                metadata: {
+                  ...metadata,
+                  loc: { pageNumber: pageNum },
+                },
+              });
+            }
+          });
+
+          await Promise.all(workerPromises);
+          documents.push(
+            ...results.sort(
+              (a, b) => a.metadata.loc.pageNumber - b.metadata.loc.pageNumber
+            )
+          );
+        }
+        return documents;
+      };
+
+      await Promise.race([timeoutPromise, processPages()]);
+    } catch (e) {
+      this.log(`Error: ${e.message}`);
+    } finally {
+      global.Image = undefined;
+      await Promise.all(workerPool.map((worker) => worker.terminate()));
+    }
+
+    this.log(`Completed OCR of ${documentTitle}!`, {
+      documentsParsed: documents.length,
+      totalPages: totalPages,
+      executionTime: `${((Date.now() - startTime) / 1000).toFixed(2)}s`,
+    });
+    return documents;
+  }
+}
+
+module.exports = OCRLoader;

collector/yarn.lock

@@ -280,7 +280,7 @@
     "@langchain/core" "~0.1"
     js-tiktoken "^1.0.11"
 
-"@mapbox/node-pre-gyp@^1.0.11":
+"@mapbox/node-pre-gyp@^1.0.0", "@mapbox/node-pre-gyp@^1.0.11":
   version "1.0.11"
   resolved "https://registry.yarnpkg.com/@mapbox/node-pre-gyp/-/node-pre-gyp-1.0.11.tgz#417db42b7f5323d79e93b34a6d7a2a12c0df43fa"
   integrity sha512-Yhlar6v9WQgUp/He7BdgzOz8lqMQ8sU+jkCq7Wx8Myc5YFJLbEe7lgui/V7G1qB1DJykHSGwreceSaD60Y0PUQ==
@@ -693,6 +693,11 @@ bluebird@~3.4.0:
   resolved "https://registry.yarnpkg.com/bluebird/-/bluebird-3.4.7.tgz#f72d760be09b7f76d08ed8fae98b289a8d05fab3"
   integrity sha512-iD3898SR7sWVRHbiQv+sHUtHnMvC1o3nW5rAcqnq3uOn07DSAppZYUkIGslDz6gXC7HfunPe7YVBgoEJASPcHA==
 
+bmp-js@^0.1.0:
+  version "0.1.0"
+  resolved "https://registry.yarnpkg.com/bmp-js/-/bmp-js-0.1.0.tgz#e05a63f796a6c1ff25f4771ec7adadc148c07233"
+  integrity sha512-vHdS19CnY3hwiNdkaqk93DvjVLfbEcI8mys4UjuWrlX1haDmroo8o4xCzh4wD6DGV6HxRCyauwhHRqMTfERtjw==
+
 body-parser@1.20.2, body-parser@^1.20.2:
   version "1.20.2"
   resolved "https://registry.yarnpkg.com/body-parser/-/body-parser-1.20.2.tgz#6feb0e21c4724d06de7ff38da36dad4f57a747fd"
@@ -788,6 +793,15 @@ camelcase@6:
   resolved "https://registry.yarnpkg.com/camelcase/-/camelcase-6.3.0.tgz#5685b95eb209ac9c0c177467778c9c84df58ba9a"
   integrity sha512-Gmy6FhYlCY7uOElZUSbxo2UCDH8owEk996gkbrpsgGtrJLM3J7jGxl9Ic7Qwwj4ivOE5AWZWRMecDdF7hqGjFA==
 
+canvas@^2.11.2:
+  version "2.11.2"
+  resolved "https://registry.yarnpkg.com/canvas/-/canvas-2.11.2.tgz#553d87b1e0228c7ac0fc72887c3adbac4abbd860"
+  integrity sha512-ItanGBMrmRV7Py2Z+Xhs7cT+FNt5K0vPL4p9EZ/UX/Mu7hFbkxSjKF2KVtPwX7UYWp7dRKnrTvReflgrItJbdw==
+  dependencies:
+    "@mapbox/node-pre-gyp" "^1.0.0"
+    nan "^2.17.0"
+    simple-get "^3.0.3"
+
 chalk@^2.4.2:
   version "2.4.2"
   resolved "https://registry.yarnpkg.com/chalk/-/chalk-2.4.2.tgz#cd42541677a54333cf541a49108c1432b44c9424"
@@ -1043,6 +1057,13 @@ decamelize@1.2.0:
   resolved "https://registry.yarnpkg.com/decamelize/-/decamelize-1.2.0.tgz#f6534d15148269b20352e7bee26f501f9a191290"
   integrity sha512-z2S+W9X73hAUUki+N+9Za2lBlun89zigOyGrsax+KUQ6wKW4ZoWpEYBkGhQjwAjjDCkWxhY0VKEhk8wzY7F5cA==
 
+decompress-response@^4.2.0:
+  version "4.2.1"
+  resolved "https://registry.yarnpkg.com/decompress-response/-/decompress-response-4.2.1.tgz#414023cc7a302da25ce2ec82d0d5238ccafd8986"
+  integrity sha512-jOSne2qbyE+/r8G1VU+G/82LBs2Fs4LAsTiLSHOCOMZQl2OKZ6i8i4IyHemTe+/yIXOtTcRQMzPcgyhoFlqPkw==
+  dependencies:
+    mimic-response "^2.0.0"
+
 decompress-response@^6.0.0:
   version "6.0.0"
   resolved "https://registry.yarnpkg.com/decompress-response/-/decompress-response-6.0.0.tgz#ca387612ddb7e104bd16d85aab00d5ecf09c66fc"
@@ -1780,6 +1801,11 @@ iconv-lite@0.6.3, iconv-lite@^0.6.3:
   dependencies:
     safer-buffer ">= 2.1.2 < 3.0.0"
 
+idb-keyval@^6.2.0:
+  version "6.2.1"
+  resolved "https://registry.yarnpkg.com/idb-keyval/-/idb-keyval-6.2.1.tgz#94516d625346d16f56f3b33855da11bfded2db33"
+  integrity sha512-8Sb3veuYCyrZL+VBt9LJfZjLUPWVvqn8tG28VqYNFCo43KHcKuq+b4EiXGeuaLAQWL2YmyDgMp2aSpH9JHsEQg==
+
 ieee754@^1.1.13, ieee754@^1.2.1:
   version "1.2.1"
   resolved "https://registry.yarnpkg.com/ieee754/-/ieee754-1.2.1.tgz#8eb7a10a63fff25d15a57b001586d177d1b0d352"
@@ -1903,6 +1929,11 @@ is-stream@^2.0.0:
   resolved "https://registry.yarnpkg.com/is-stream/-/is-stream-2.0.1.tgz#fac1e3d53b97ad5a9d0ae9cef2389f5810a5c077"
   integrity sha512-hFoiJiTl63nn+kstHGBtewWSKnQLpyb155KHheA1l39uvtO9nWIop1p3udqPcUd/xbF1VLMO4n7OI6p7RbngDg==
 
+is-url@^1.2.4:
+  version "1.2.4"
+  resolved "https://registry.yarnpkg.com/is-url/-/is-url-1.2.4.tgz#04a4df46d28c4cff3d73d01ff06abeb318a1aa52"
+  integrity sha512-ITvGim8FhRiYe4IQ5uHSkj7pVaPDrCTkNd3yq3cV7iZAcJdHTUMPMEHcqSOy9xZ9qFenQCvi+2wjH9a1nXqHww==
+
 isarray@~1.0.0:
   version "1.0.0"
   resolved "https://registry.yarnpkg.com/isarray/-/isarray-1.0.0.tgz#bb935d48582cba168c06834957a54a3e07124f11"
@@ -2276,6 +2307,11 @@ mime@^3.0.0:
   resolved "https://registry.yarnpkg.com/mime/-/mime-3.0.0.tgz#b374550dca3a0c18443b0c950a6a58f1931cf7a7"
   integrity sha512-jSCU7/VB1loIWBZe14aEYHU/+1UMEHoaO7qxCOVJOw9GgH72VAWppxNcjU+x9a2k3GSIBXNKxXQFqRvvZ7vr3A==
 
+mimic-response@^2.0.0:
+  version "2.1.0"
+  resolved "https://registry.yarnpkg.com/mimic-response/-/mimic-response-2.1.0.tgz#d13763d35f613d09ec37ebb30bac0469c0ee8f43"
+  integrity sha512-wXqjST+SLt7R009ySCglWBCFpjUygmCIfD790/kVbiGmUgfYGuB14PiTd5DwVxSV4NcYHjzMkoj5LjQZwTQLEA==
+
 mimic-response@^3.1.0:
   version "3.1.0"
   resolved "https://registry.yarnpkg.com/mimic-response/-/mimic-response-3.1.0.tgz#2d1d59af9c1b129815accc2c46a022a5ce1fa3c9"
@@ -2389,6 +2425,11 @@ mustache@^4.2.0:
   resolved "https://registry.yarnpkg.com/mustache/-/mustache-4.2.0.tgz#e5892324d60a12ec9c2a73359edca52972bf6f64"
   integrity sha512-71ippSywq5Yb7/tVYyGbkBggbU8H3u5Rz56fH60jGFgr8uHwxs+aSKeqmluIVzM0m0kB7xQjKS6qPfd0b2ZoqQ==
 
+nan@^2.17.0:
+  version "2.22.0"
+  resolved "https://registry.yarnpkg.com/nan/-/nan-2.22.0.tgz#31bc433fc33213c97bad36404bb68063de604de3"
+  integrity sha512-nbajikzWTMwsW+eSsNm3QwlOs7het9gGJU5dDZzRTQGk03vyBOauxgI4VakDzE0PtsGTmXPsXTbbjVhRwR5mpw==
+
 napi-build-utils@^1.0.1:
   version "1.0.2"
   resolved "https://registry.yarnpkg.com/napi-build-utils/-/napi-build-utils-1.0.2.tgz#b1fddc0b2c46e380a0b7a76f984dd47c41a13806"
@@ -2431,7 +2472,7 @@ node-ensure@^0.0.0:
   resolved "https://registry.yarnpkg.com/node-ensure/-/node-ensure-0.0.0.tgz#ecae764150de99861ec5c810fd5d096b183932a7"
   integrity sha512-DRI60hzo2oKN1ma0ckc6nQWlHU69RH6xN0sjQTjMpChPfTYvKZdcQFfdYK2RWbJcKyUizSIy/l8OTGxMAM1QDw==
 
-node-fetch@^2.6.12, node-fetch@^2.6.7:
+node-fetch@^2.6.12, node-fetch@^2.6.7, node-fetch@^2.6.9:
   version "2.7.0"
   resolved "https://registry.yarnpkg.com/node-fetch/-/node-fetch-2.7.0.tgz#d0f0fa6e3e2dc1d27efcd8ad99d550bda94d187d"
   integrity sha512-c4FRfUm/dbcWZ7U+1Wq0AwCyFL+3nt2bEw05wfxSz+DWpWsitgmSgYmy2dQdWyKC1694ELPqMs/YzUSNozLt8A==
@@ -2634,6 +2675,11 @@ openapi-types@^12.1.3:
   resolved "https://registry.yarnpkg.com/openapi-types/-/openapi-types-12.1.3.tgz#471995eb26c4b97b7bd356aacf7b91b73e777dd3"
   integrity sha512-N4YtSYJqghVu4iek2ZUvcN/0aqH1kRDuNqzcycDxhOUpg7GdvLa2F3DgS6yBNhInhv2r/6I0Flkn7CqL8+nIcw==
 
+opencollective-postinstall@^2.0.3:
+  version "2.0.3"
+  resolved "https://registry.yarnpkg.com/opencollective-postinstall/-/opencollective-postinstall-2.0.3.tgz#7a0fff978f6dbfa4d006238fbac98ed4198c3259"
+  integrity sha512-8AV/sCtuzUeTo8gQK5qDZzARrulB3egtLzFgteqB2tcT4Mw7B8Kt7JcDHmltjz6FOAHsvTevk70gZEbhM4ZS9Q==
+
 option@~0.2.1:
   version "0.2.4"
   resolved "https://registry.yarnpkg.com/option/-/option-0.2.4.tgz#fd475cdf98dcabb3cb397a3ba5284feb45edbfe4"
@@ -2990,6 +3036,11 @@ readdirp@~3.6.0:
   dependencies:
     picomatch "^2.2.1"
 
+regenerator-runtime@^0.13.3:
+  version "0.13.11"
+  resolved "https://registry.yarnpkg.com/regenerator-runtime/-/regenerator-runtime-0.13.11.tgz#f6dca3e7ceec20590d07ada785636a90cdca17f9"
+  integrity sha512-kY1AZVr2Ra+t+piVaJ4gxaFaReZVH40AKNo7UCX6W+dEwBo/2oZJzqfuN1qLq1oL45o56cPaTXELwrTh8Fpggg==
+
 require-directory@^2.1.1:
   version "2.1.1"
   resolved "https://registry.yarnpkg.com/require-directory/-/require-directory-2.1.1.tgz#8c64ad5fd30dab1c976e2344ffe7f792a6a6df42"
@@ -3204,6 +3255,15 @@ simple-concat@^1.0.0:
   resolved "https://registry.yarnpkg.com/simple-concat/-/simple-concat-1.0.1.tgz#f46976082ba35c2263f1c8ab5edfe26c41c9552f"
   integrity sha512-cSFtAPtRhljv69IK0hTVZQ+OfE9nePi/rtJmw5UjHeVyVroEqJXP1sFztKUy1qU+xvz3u/sfYJLa947b7nAN2Q==
 
+simple-get@^3.0.3:
+  version "3.1.1"
+  resolved "https://registry.yarnpkg.com/simple-get/-/simple-get-3.1.1.tgz#cc7ba77cfbe761036fbfce3d021af25fc5584d55"
+  integrity sha512-CQ5LTKGfCpvE1K0n2us+kuMPbk/q0EKl82s4aheV9oXjFEz6W/Y7oQFVJuU6QG77hRT4Ghb5RURteF5vnWjupA==
+  dependencies:
+    decompress-response "^4.2.0"
+    once "^1.3.1"
+    simple-concat "^1.0.0"
+
 simple-get@^4.0.0, simple-get@^4.0.1:
   version "4.0.1"
   resolved "https://registry.yarnpkg.com/simple-get/-/simple-get-4.0.1.tgz#4a39db549287c979d352112fa03fd99fd6bc3543"
@@ -3421,6 +3481,26 @@ tar@^6.1.11:
     mkdirp "^1.0.3"
     yallist "^4.0.0"
 
+tesseract.js-core@^6.0.0:
+  version "6.0.0"
+  resolved "https://registry.yarnpkg.com/tesseract.js-core/-/tesseract.js-core-6.0.0.tgz#6f25da94f70f8e8f02aff47a43be61d49e6f67c3"
+  integrity sha512-1Qncm/9oKM7xgrQXZXNB+NRh19qiXGhxlrR8EwFbK5SaUbPZnS5OMtP/ghtqfd23hsr1ZvZbZjeuAGcMxd/ooA==
+
+tesseract.js@^6.0.0:
+  version "6.0.0"
+  resolved "https://registry.yarnpkg.com/tesseract.js/-/tesseract.js-6.0.0.tgz#62ff7fffc8833b5810430a4067785e49d5ca8e7f"
+  integrity sha512-tqYCod1HwJzkeZw1l6XWx+ly2hhisGcBtak9MArhYwDAxL0NgeVhLJcUjqPxZMQtpgtVUzWcpZPryi+hnaQGVw==
+  dependencies:
+    bmp-js "^0.1.0"
+    idb-keyval "^6.2.0"
+    is-url "^1.2.4"
+    node-fetch "^2.6.9"
+    opencollective-postinstall "^2.0.3"
+    regenerator-runtime "^0.13.3"
+    tesseract.js-core "^6.0.0"
+    wasm-feature-detect "^1.2.11"
+    zlibjs "^0.3.1"
+
 text-hex@1.0.x:
   version "1.0.0"
   resolved "https://registry.yarnpkg.com/text-hex/-/text-hex-1.0.0.tgz#69dc9c1b17446ee79a92bf5b884bb4b9127506f5"
@@ -3602,6 +3682,11 @@ vary@^1, vary@~1.1.2:
   resolved "https://registry.yarnpkg.com/vary/-/vary-1.1.2.tgz#2299f02c6ded30d4a5961b0b9f74524a18f634fc"
   integrity sha512-BNGbWLfd0eUPabhkXUVm0j8uuvREyTh5ovRa/dyow/BqAbZJyC+5fU+IzQOzmAKzYqYRAISoRhdQr3eIZ/PXqg==
 
+wasm-feature-detect@^1.2.11:
+  version "1.8.0"
+  resolved "https://registry.yarnpkg.com/wasm-feature-detect/-/wasm-feature-detect-1.8.0.tgz#4e9f55b0a64d801f372fbb0324ed11ad3abd0c78"
+  integrity sha512-zksaLKM2fVlnB5jQQDqKXXwYHLQUVH9es+5TOOHwGOVJOCeRBCiPjwSg+3tN2AdTCzjgli4jijCH290kXb/zWQ==
+
 wavefile@^11.0.0:
   version "11.0.0"
   resolved "https://registry.yarnpkg.com/wavefile/-/wavefile-11.0.0.tgz#9302165874327ff63a704d00b154c753eaa1b8e7"
@@ -3766,6 +3851,11 @@ youtubei.js@^9.1.0:
     tslib "^2.5.0"
     undici "^5.19.1"
 
+zlibjs@^0.3.1:
+  version "0.3.1"
+  resolved "https://registry.yarnpkg.com/zlibjs/-/zlibjs-0.3.1.tgz#50197edb28a1c42ca659cc8b4e6a9ddd6d444554"
+  integrity sha512-+J9RrgTKOmlxFSDHo0pI1xM6BLVUv+o0ZT9ANtCxGkjIVCCUdx9alUF8Gm+dGLKbkkkidWIHFDZHDMpfITt4+w==
+
 zod-to-json-schema@^3.22.3, zod-to-json-schema@^3.22.5:
   version "3.23.0"
   resolved "https://registry.yarnpkg.com/zod-to-json-schema/-/zod-to-json-schema-3.23.0.tgz#4fc60e88d3c709eedbfaae3f92f8a7bf786469f2"

server/storage/models/.gitignore

@@ -6,4 +6,5 @@ apipie
 novita
 mixedbread-ai*
 gemini
-togetherAi
\ No newline at end of file
+togetherAi
+tesseract
\ No newline at end of file