Skip to main content

Bundlers

SheetJS predates ECMAScript modules and bundler tools like Webpack. As best practices have evolved, stress testing SheetJS libraries have revealed bugs in the respective bundlers. This demo collects various notes and provides basic examples.

note

Issues should be reported to the respective bundler projects. Typically it is considered a bundler bug if the tool cannot properly handle JS libraries.

Browserify

browserify is compatible with the library and should "just work" with the require form in a main page or in a web worker:

var XLSX = require("xlsx");
// ... use XLSX ...

After installing the module, bundling is easy:

browserify app.js > browserify.js
uglifyjs browserify.js > browserify.min.js

Web Worker scripts can be bundled using the same approach.

Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

2) Download the following files:

3) Bundle the scripts:

npx browserify app.js > browserify.js
npx browserify xlsxworker.js > worker.js

4) Spin up a local web server:

npx http-server

5) Access the site http://localhost:8080/ and use the file input element to select a spreadsheet.

Bun

bun bun is capable of optimizing imported libraries in node_modules. In local testing, a bundled script can save tens of milliseconds per run.

Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

2) Save the following script to bun.js:

bun.js
import * as XLSX from 'xlsx/xlsx.mjs';
import * as fs from 'fs';
XLSX.set_fs(fs);

/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter((row) => row.terms.some((term) => term.type === "prez"));

/* flatten objects */
const rows = prez.map((row) => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = XLSX.utils.json_to_sheet(rows);
const workbook = XLSX.utils.book_new();
XLSX.utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
XLSX.utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
XLSX.writeFile(workbook, "Presidents.xlsx");

3) Bundle the script with bun bun:

bun bun bun.js

This procedure will generate node_modules.bun.

4) Run the script

bun bun.js

Dojo

Integration details are included in the "AMD" installation

Complete Examples are included in the "Dojo" demo

esbuild

The xlsx.mjs source file are written in a subset of ES6 that esbuild understands and is able to transpile down for older browsers.

Both the node and browser platforms work out of the box.

Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

2) Save the following to esbrowser.js:

esbrowser.js
import { utils, version, writeFileXLSX } from 'xlsx/xlsx.mjs';

(async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
})();

3) Create a small HTML page that loads the script. Save to index.html:

index.html
<body><script src="esb.browser.js"></script></body>

4) Create bundle:

npx esbuild esbrowser.js --bundle --outfile=esb.browser.js

5) Start a local HTTP server, then go to http://localhost:8080/

npx http-server .

Parcel

Parcel should play nice with SheetJS out of the box.

Parcel Bug

Errors of the form Could not statically evaluate fs call stem from a parcel bug. Upgrade to Parcel version 1.5.0 or later.

Complete Example (click to show)

This demo follows the Presidents Example.

1) Save the following to index.html:

index.html
<body>
<h3>SheetJS <span id="vers"></span> export demo</h3>
<button id="xport">Click to Export!</button>
<!-- the script tag must be marked as `type="module"` -->
<script type="module">
// ESM-style import from "xlsx"
import { utils, version, writeFileXLSX } from 'xlsx';

document.getElementById("vers").innerText = version;
document.getElementById("xport").onclick = async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
};
</script>
<body>

2) Install the SheetJS node module:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

3) Run the Parcel CLI tool:

npx -y parcel index.html

4) Access the page listed in the output (typically http://localhost:1234) and click the "Click to Export!" button to generate a file.

RequireJS

Standalone scripts comply with AMD define semantics, enabling use in RequireJS out of the box.

To enable use of the alias xlsx, the RequireJS config should set an alias in the paths property:

require.config({
baseUrl: ".",
name: "app",
paths: {
xlsx: "xlsx.full.min"
}
});
require(["xlsx"], function(XLSX) {
/* use XLSX here */
console.log(XLSX.version);
});

The Live demo loads RequireJS from the CDN, uses it to load the standalone script from the SheetJS CDN, and uses the XLSX variable to create a button click handler that creates a workbook.

The r.js optimizer also supports the standalone scripts.

Rollup

Rollup requires @rollup/plugin-node-resolve to support NodeJS modules:

Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz rollup @rollup/plugin-node-resolve

2) Save the following to index.js:

index.js
import { utils, version, writeFileXLSX } from 'xlsx/xlsx.mjs';

document.getElementById("xport").addEventListener("click", async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
});

3) Bundle the script:

npx rollup index.js --plugin @rollup/plugin-node-resolve --file bundle.js --format iife

4) Create a small HTML page that loads the script. Save to index.html:

index.html
<!DOCTYPE html>
<html lang="en">
<head></head>
<body>
<h1>SheetJS Presidents Demo</h1>
<button id="xport">Click here to export</button>
<script type="module" src="./bundle.js"></script>
</body>
</html>

5) Start a local HTTP server, then go to http://localhost:8080/

npx http-server .

Click on "Click here to export" to generate a file.

Snowpack

Snowpack works with no caveats.

Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

2) Save the following to index.js:

index.js
import { utils, version, writeFileXLSX } from 'xlsx/xlsx.mjs';

document.getElementById("xport").addEventListener("click", async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
});

3) Create a small HTML page that loads the script. Save to index.html:

index.html
<!DOCTYPE html>
<html lang="en">
<head></head>
<body>
<h1>SheetJS Presidents Demo</h1>
<button id="xport">Click here to export</button>
<script type="module" src="./index.js"></script>
</body>
</html>
note

Unlike other bundlers, Snowpack requires a full page including HEAD element.

4) Build for production:

npx snowpack build

5) Start a local HTTP server, then go to http://localhost:8080/

npx http-server build/

Click on "Click here to export" to generate a file.

SWC

SWC provides spack for bundling scripts.

Complete Example (click to show)

1) Install the dependencies using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz regenerator-runtime @swc/cli @swc/core
note

The regenerator-runtime dependency is used for transpiling fetch and is not required if the interface code does not use fetch or Promises.

2) Save the following to index.js:

index.js
import { utils, version, writeFileXLSX } from 'xlsx';

document.getElementById("xport").addEventListener("click", async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
});

3) Create an spack.config.js config file:

spack.config.js
const { config } = require('@swc/core/spack');

module.exports = ({
entry: {
'web': __dirname + '/index.js',
},
output: {
path: __dirname + '/lib'
},
module: {},
});

4) Build for production:

npx spack

This command will create the script lib/web.js

5) Create a small HTML page that loads the generated script:

index.html
<!DOCTYPE html>
<html lang="en">
<head></head>
<body>
<h1>SheetJS Presidents Demo</h1>
<button id="xport">Click here to export</button>
<script src="lib/web.js"></script>
</body>
</html>

6) Start a local HTTP server, then go to http://localhost:8080/

npx http-server build/

Click on "Click here to export" to generate a file.

SystemJS

With configuration, SystemJS supports both browser and NodeJS deployments.

caution

This demo was written against SystemJS 0.19, the most popular SystemJS version. used with Angular applications. In the years since the release, Angular and other tools using SystemJS have switched to Webpack.

SystemJS fails by default because the library does not export anything in the web browser. The meta configuration option can be used to expose XLSX:

SystemJS.config({
meta: {
'xlsx': {
exports: 'XLSX' // <-- tell SystemJS to expose the XLSX variable
}
},
map: {
'xlsx': 'https://cdn.sheetjs.com/xlsx-latest/package/dist/xlsx.full.min.js',
'fs': '', // <--|
'crypto': '', // <--| suppress native node modules
'stream': '' // <--|
}
});
SystemJS.import('main.js'); // load `main.js`

The main.js script can freely require("xlsx").

Web Workers

Web Workers can load the SystemJS library with importScripts, but the imported code cannot assign the original worker's onmessage callback. The recommended approach is to expose a global from the required script, For example, supposing the shared name is _cb, the primary worker script would call the callback:

worker.js
/* main worker script */
importScripts('system.js');

SystemJS.config({ /* ... browser config ... */ });

onmessage = function(evt) {
SystemJS.import('workermain.js').then(function() { _cb(evt); });
};

The worker script would define and expose the function:

workermain.js
/* Loaded with SystemJS import */
var XLSX = require('xlsx');

_cb = function(evt) { /* ... do work here ... */ };
Complete Example (click to show)

The Live demo loads SystemJS from the CDN, uses it to load the standalone script from the SheetJS CDN and emulate a require implementation when loading main.js

"View Source" works on the main HTML page and the main.js script.

Vite

caution

ViteJS adopted nascent package.json patterns. Version 0.18.10 implements the patterns required for ViteJS 3.0.3. These patterns are evolving and a future version of ViteJS may require more packaging changes.

ViteJS 3.0.3 is known to work with SheetJS version 0.18.10.

Complete Example (click to show)

1) Create a new ViteJS project:

npm create [email protected]
cd sheetjs-vite # (project name)
npm i

When prompted for Project Name, type sheetjs-vite

When prompted for Framework, select vue then vue-ts

2) Add the SheetJS dependency:

npm i --save https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

3) Replace src\components\HelloWorld.vue with:

src\components\HelloWorld.vue
<script setup lang="ts">
import { version, utils, writeFileXLSX } from 'xlsx';

interface President {
terms: { "type": "prez" | "viceprez"; }[];
name: { first: string; last: string; }
bio: { birthday: string; }
}

async function xport() {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data: President[] = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
}

</script>

<template>
<button type="button" @click="xport">Export with SheetJS version {{ version }}</button>
</template>

4) Run npm run dev and test functionality by opening a web browser to http://localhost:5173/ and clicking the button

5) Run npx vite build and verify the generated pages work by running a local web server in the dist folder:

npx http-server dist/

Access http://localhost:8080 in your web browser.

Webpack

The ECMAScript Module build has no require or import statements and does not use process or any variable that Webpack could interpret as a NodeJS feature. Various package.json fields have been added to appease various Webpack versions starting from the 2.x series.

CommonJS and ESM

Webpack bundled the CommonJS build in older versions of the library. Version 0.18.1 changed the NodeJS module package so that Webpack uses the ESM build.

The ESM build does not include the codepage support library for XLS reading. As described in the installation instructions, the codepage dependency should be imported explicitly:

import * as XLSX from 'xlsx';
import * as cptable from 'xlsx/dist/cpexcel.full.mjs';
set_cptable(cptable);
Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

2) Save the following to index.js:

index.js
import { utils, version, writeFileXLSX } from 'xlsx/xlsx.mjs';

document.getElementById("xport").addEventListener("click", async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
});

3) Create a small webpack.config.js script that writes to index.min.js:

webpack.config.js
module.exports = {
/* entry point index.js */
entry: './index.js',

/* write to index.min.js */
output: { path:__dirname, filename: './index.min.js' }
}

4) Build for production. The command depends on the version of webpack:

Pinning specific versions of webpack

The webpack tooling is not designed for switching between versions. A specific version above 4.0 can be pinned by locally installing webpack and the CLI tool.

Webpack 4.x

npm install --save [email protected] webpack-cli
npx webpack --mode=production

Webpack 5.x

npm install --save [email protected] webpack-cli
npx webpack --mode=production

Webpack latest

npm install --save webpack webpack-cli
npx webpack --mode=production

5) Create a small HTML page that loads the script. Save to index.html:

index.html
<!DOCTYPE html>
<html lang="en">
<head></head>
<body>
<h1>SheetJS Presidents Demo</h1>
<button id="xport">Click here to export</button>
<script src="./index.min.js"></script>
</body>
</html>

6) Start a local HTTP server and go to http://localhost:8080/

npx http-server .

Click on "Click here to export" to generate a file.

WMR

WMR follows the same structure as Snowpack

Complete Example (click to show)

1) Install the tarball using a package manager:

yarn add https://cdn.sheetjs.com/xlsx-0.18.12/xlsx-0.18.12.tgz

2) Save the following to index.js:

index.js
import { utils, version, writeFileXLSX } from 'xlsx/xlsx.mjs';

document.getElementById("xport").addEventListener("click", async() => {
/* fetch JSON data and parse */
const url = "https://sheetjs.com/data/executive.json";
const raw_data = await (await fetch(url)).json();

/* filter for the Presidents */
const prez = raw_data.filter(row => row.terms.some(term => term.type === "prez"));

/* flatten objects */
const rows = prez.map(row => ({
name: row.name.first + " " + row.name.last,
birthday: row.bio.birthday
}));

/* generate worksheet and workbook */
const worksheet = utils.json_to_sheet(rows);
const workbook = utils.book_new();
utils.book_append_sheet(workbook, worksheet, "Dates");

/* fix headers */
utils.sheet_add_aoa(worksheet, [["Name", "Birthday"]], { origin: "A1" });

/* calculate column width */
const max_width = rows.reduce((w, r) => Math.max(w, r.name.length), 10);
worksheet["!cols"] = [ { wch: max_width } ];

/* create an XLSX file and try to save to Presidents.xlsx */
writeFileXLSX(workbook, "Presidents.xlsx");
});

3) Create a small HTML page that loads the script. Save to index.html:

index.html
<!DOCTYPE html>
<html lang="en">
<head></head>
<body>
<h1>SheetJS Presidents Demo</h1>
<button id="xport">Click here to export</button>
<script type="module" src="./index.js"></script>
</body>
</html>

4) Build for production:

npx wmr build

5) Start a local HTTP server in dist folder and go to http://localhost:8080/

npx http-server dist/

Click on "Click here to export" to generate a file.