Support XLSX files (#2403)

* support xlsx files

* lint

* create seperate docs for each xlsx sheet

* lint

* use node-xlsx pkg for parsing xslx files

* lint

* update error handling

---------

Co-authored-by: timothycarambat <rambat1010@gmail.com>

collector/package.json

@@ -33,6 +33,7 @@
     "mime": "^3.0.0",
     "moment": "^2.29.4",
     "node-html-parser": "^6.1.13",
+    "node-xlsx": "^0.24.0",
     "officeparser": "^4.0.5",
     "openai": "4.38.5",
     "pdf-parse": "^1.1.1",
@@ -48,4 +49,4 @@
     "nodemon": "^2.0.22",
     "prettier": "^2.4.1"
   }
-}
+}

collector/processSingleFile/convert/asXlsx.js

@@ -0,0 +1,113 @@
+const { v4 } = require("uuid");
+const xlsx = require("node-xlsx").default;
+const path = require("path");
+const fs = require("fs");
+const {
+  createdDate,
+  trashFile,
+  writeToServerDocuments,
+} = require("../../utils/files");
+const { tokenizeString } = require("../../utils/tokenizer");
+const { default: slugify } = require("slugify");
+
+function convertToCSV(data) {
+  return data
+    .map((row) =>
+      row
+        .map((cell) => {
+          if (cell === null || cell === undefined) return "";
+          if (typeof cell === "string" && cell.includes(","))
+            return `"${cell}"`;
+          return cell;
+        })
+        .join(",")
+    )
+    .join("\n");
+}
+
+async function asXlsx({ fullFilePath = "", filename = "" }) {
+  const documents = [];
+  const folderName = slugify(`${path.basename(filename)}-${v4().slice(0, 4)}`, {
+    lower: true,
+    trim: true,
+  });
+
+  const outFolderPath =
+    process.env.NODE_ENV === "development"
+      ? path.resolve(
+          __dirname,
+          `../../../server/storage/documents/${folderName}`
+        )
+      : path.resolve(process.env.STORAGE_DIR, `documents/${folderName}`);
+
+  try {
+    const workSheetsFromFile = xlsx.parse(fullFilePath);
+    if (!fs.existsSync(outFolderPath))
+      fs.mkdirSync(outFolderPath, { recursive: true });
+
+    for (const sheet of workSheetsFromFile) {
+      try {
+        const { name, data } = sheet;
+        const content = convertToCSV(data);
+
+        if (!content?.length) {
+          console.warn(`Sheet "${name}" is empty. Skipping.`);
+          continue;
+        }
+
+        console.log(`-- Processing sheet: ${name} --`);
+        const sheetData = {
+          id: v4(),
+          url: `file://${path.join(outFolderPath, `${slugify(name)}.csv`)}`,
+          title: `${filename} - Sheet:${name}`,
+          docAuthor: "Unknown",
+          description: `Spreadsheet data from sheet: ${name}`,
+          docSource: "an xlsx file uploaded by the user.",
+          chunkSource: "",
+          published: createdDate(fullFilePath),
+          wordCount: content.split(/\s+/).length,
+          pageContent: content,
+          token_count_estimate: tokenizeString(content).length,
+        };
+
+        const document = writeToServerDocuments(
+          sheetData,
+          `sheet-${slugify(name)}`,
+          outFolderPath
+        );
+        documents.push(document);
+        console.log(
+          `[SUCCESS]: Sheet "${name}" converted & ready for embedding.`
+        );
+      } catch (err) {
+        console.error(`Error processing sheet "${name}":`, err);
+        continue;
+      }
+    }
+  } catch (err) {
+    console.error("Could not process xlsx file!", err);
+    return {
+      success: false,
+      reason: `Error processing ${filename}: ${err.message}`,
+      documents: [],
+    };
+  } finally {
+    trashFile(fullFilePath);
+  }
+
+  if (documents.length === 0) {
+    console.error(`No valid sheets found in ${filename}.`);
+    return {
+      success: false,
+      reason: `No valid sheets found in ${filename}.`,
+      documents: [],
+    };
+  }
+
+  console.log(
+    `[SUCCESS]: ${filename} fully processed. Created ${documents.length} document(s).\n`
+  );
+  return { success: true, reason: null, documents };
+}
+
+module.exports = asXlsx;

collector/utils/constants.js

@@ -11,6 +11,10 @@ const ACCEPTED_MIMES = {
     ".pptx",
   ],
 
+  "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet": [
+    ".xlsx",
+  ],
+
   "application/vnd.oasis.opendocument.text": [".odt"],
   "application/vnd.oasis.opendocument.presentation": [".odp"],
 
@@ -41,6 +45,8 @@ const SUPPORTED_FILETYPE_CONVERTERS = {
   ".odt": "./convert/asOfficeMime.js",
   ".odp": "./convert/asOfficeMime.js",
 
+  ".xlsx": "./convert/asXlsx.js",
+
   ".mbox": "./convert/asMbox.js",
 
   ".epub": "./convert/asEPub.js",

collector/yarn.lock

@@ -2326,6 +2326,13 @@ node-html-parser@^6.1.13:
     css-select "^5.1.0"
     he "1.2.0"
 
+node-xlsx@^0.24.0:
+  version "0.24.0"
+  resolved "https://registry.yarnpkg.com/node-xlsx/-/node-xlsx-0.24.0.tgz#a6a365acb18ad37c66c2b254b6ebe0c22dc9dc6f"
+  integrity sha512-1olwK48XK9nXZsyH/FCltvGrQYvXXZuxVitxXXv2GIuRm51aBi1+5KwR4rWM4KeO61sFU+00913WLZTD+AcXEg==
+  dependencies:
+    xlsx "https://cdn.sheetjs.com/xlsx-0.20.2/xlsx-0.20.2.tgz"
+
 nodemailer@6.9.13:
   version "6.9.13"
   resolved "https://registry.yarnpkg.com/nodemailer/-/nodemailer-6.9.13.tgz#5b292bf1e92645f4852ca872c56a6ba6c4a3d3d6"
@@ -3528,6 +3535,10 @@ ws@8.14.2:
   resolved "https://registry.yarnpkg.com/ws/-/ws-8.14.2.tgz#6c249a806eb2db7a20d26d51e7709eab7b2e6c7f"
   integrity sha512-wEBG1ftX4jcglPxgFCMJmZ2PLtSbJ2Peg6TmpJFTbe9GZYOQCDPdMYu/Tm0/bGZkw8paZnJY45J4K2PZrLYq8g==
 
+"xlsx@https://cdn.sheetjs.com/xlsx-0.20.2/xlsx-0.20.2.tgz":
+  version "0.20.2"
+  resolved "https://cdn.sheetjs.com/xlsx-0.20.2/xlsx-0.20.2.tgz#0f64eeed3f1a46e64724620c3553f2dbd3cd2d7d"
+
 xml2js@^0.6.2:
   version "0.6.2"
   resolved "https://registry.yarnpkg.com/xml2js/-/xml2js-0.6.2.tgz#dd0b630083aa09c161e25a4d0901e2b2a929b499"