docs.sheetjs.com/docz/docs/03-demos/11-bigdata/02-worker.md
2023-06-05 21:49:28 -04:00

21 KiB

title pagination_prev pagination_next sidebar_custom_props
Web Workers demos/extensions/index demos/engines/index
summary
Run large data flows without freezing the browser

import current from '/version.js'; import CodeBlock from '@theme/CodeBlock';

Parsing and writing large spreadsheets takes time. During the process, if the SheetJS library is running in the web browser, the website may freeze.

Workers provide a way to off-load the hard work so that the website does not freeze during processing. The work is still performed locally. No data is sent to a remote server.

The following diagrams show the normal and Web Worker flows when exporting a dataset. The regions with a red background mark when the browser is frozen.

Normal ExportWeb Worker Export
sequenceDiagram
  autonumber
  actor User
  participant Page
  User->>Page: click button
  rect rgba(255,0,0,.2)
    activate Page
    Note over Page: collect dataset
    Note over Page: generate workbook
    Note over Page: create file
    Note over Page: setup download
  end
  Page->>User: download workbook
  deactivate Page
sequenceDiagram
  autonumber
  actor User
  participant Page
  participant Worker
  User->>Page: click button
  rect rgba(255,0,0,.2)
    activate Page
    Note over Page: collect dataset
    Page->>Worker: transfer dataset
  end
  deactivate Page
  activate Worker
  Note over Worker: generate workbook
  Note over Worker: create file
  Note over Worker: setup download
  Worker->>Page: URL to download
  deactivate Worker
  activate Page
  Page->>User: download workbook
  deactivate Page

:::note Browser Compatibility

IE10+ and modern browsers support basic Web Workers. Some APIs like fetch were added later. Feature testing is highly recommended.

:::

:::info Inline Workers

Due to limitations of the live code blocks, all of the workers in this section are in-line. The code is embedded in template literals. For production sites, typically workers are written in separate JS files.

Example (click to show)

For example, an in-line worker like

{\ const worker = new Worker(URL.createObjectURL(new Blob([\\ /* load standalone script from CDN / importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js"); \n
/
this callback will run once the main context sends a message / self.addEventListener('message', (e) => { / Pass the version string back */ postMessage({ version: XLSX.version }); }, false); `])));`}

would typically be stored in a separate JS file like "worker.js":

{\ /* load standalone script from CDN */ importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js"); \n\ /* this callback will run once the main context sends a message */ self.addEventListener('message', (e) => { /* Pass the version string back */ postMessage({ version: XLSX.version }); }, false);}

and the main script would pass a URL to the Worker constructor:

const worker = new Worker("./worker.js");

:::

Installation

In all cases, importScripts in a Worker can load the Standalone scripts

{\ importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js");}

For production use, it is highly encouraged to download and host the script.

ECMAScript Module Support (click to hide)

:::note Browser Compatibility

ESM is supported in Web Workers in the Chromium family of browsers (including Chrome and Edge) as well as in browsers powered by WebKit (including Safari).

For legacy browsers like Firefox and IE, importScripts should be used.

:::

Browser ESM imports require a complete URL including the .mjs extension:

{\ import * as XLSX from "https://cdn.sheetjs.com/xlsx-${current}/package/xlsx.mjs";}

When using Worker ESM, the Worker constructor must set the type option:

const worker = new Worker(
  url_to_worker_script,
  // highlight-next-line
  { type: "module" } // second argument to Worker constructor
);

Inline workers additionally require the Blob MIME type text/javascript:

{\ const worker_code = \\ /* load standalone script from CDN */ import * as XLSX from "https://cdn.sheetjs.com/xlsx-${current}/package/xlsx.mjs"; // ... do something with XLSX here ... `; const worker = new Worker( URL.createObjectURL( new Blob( [ worker_code ], // highlight-next-line { type: "text/javascript" } // second argument to the Blob constructor ) ), // highlight-next-line { type: "module" } // second argument to Worker constructor );`}

Live Demos

:::note

Each browser demo was tested in the following environments:

Browser Date Comments
Chrome 113 2023-05-22
Edge 113 2023-05-22
Safari 16.4 2023-05-22 File System Access API is not supported
Brave 1.51 2023-05-22 File System Access API is not supported
Firefox 113 2023-05-22 File System Access API is not supported

:::

Downloading a Remote File

:::note fetch in Web Workers

fetch was enabled in Web Workers in Chrome 42 and Safari 10.3

:::

Typically the Web Worker performs the fetch operation, processes the workbook, and sends a final result (HTML table or raw data) to the main browser context:

sequenceDiagram
  autonumber
  actor User
  participant Page
  participant Worker
  User->>Page: click button
  activate Page
  Page->>Worker: send URL
  deactivate Page
  activate Worker
  Note over Worker: fetch file
  Note over Worker: parse file
  Note over Worker: generate table
  Worker->>Page: HTML table
  deactivate Worker
  activate Page
  Note over Page: add to DOM
  Page->>User: table is visible
  deactivate Page
Live Demo (click to show)

In the following example, the script:

  • downloads https://sheetjs.com/pres.numbers in a Web Worker
  • loads the SheetJS library and parses the file in the Worker
  • generates an HTML string of the first table in the Worker
  • sends the string to the main browser context
  • adds the HTML to the page in the main browser context

{\ function SheetJSFetchDLWorker() { const [__html, setHTML] = React.useState(""); \n\ return ( <> <button onClick={() => { /* this mantra embeds the worker source in the function */ const worker = new Worker(URL.createObjectURL(new Blob([\\ /* load standalone script from CDN / importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js"); \n
/
this callback will run once the main context sends a message / self.addEventListener('message', async(e) => { try { / Fetch file / const res = await fetch("https://sheetjs.com/pres.numbers"); const ab = await res.arrayBuffer(); \n
/
Parse file / const wb = XLSX.read(ab, {dense: true}); const ws = wb.Sheets[wb.SheetNames[0]]; \n
/
Generate HTML / const html = XLSX.utils.sheet_to_html(ws); \n
/
Reply with result / postMessage({ html }); } catch(e) { / Pass the error message back / postMessage({html: String(e.message || e).bold() }); } }, false); `]))); \n
/
when the worker sends back the HTML, add it to the DOM / worker.onmessage = function(e) { setHTML(e.data.html); }; / post a message to the worker */ worker.postMessage({}); }}>Click to Start <div dangerouslySetInnerHTML={{ __html }}/> </> ); }`}

Creating a Local File

:::caution Writing files from Web Workers

XLSX.writeFile will not work in Web Workers! Raw file data can be passed from the Web Worker to the main browser context for downloading.

:::

Typically the Web Worker receives an array of JS objects, generates a workbook, and sends a URL to the main browser context for download:

sequenceDiagram
  autonumber
  actor User
  participant Page
  participant Worker
  User->>Page: click button
  activate Page
  Note over Page: collect dataset
  Page->>Worker: transfer dataset
  deactivate Page
  activate Worker
  Note over Worker: generate workbook
  Note over Worker: create file
  Note over Worker: setup download
  Worker->>Page: URL to download
  deactivate Worker
  activate Page
  Page->>User: download workbook
  deactivate Page
Live Demo (click to show)

In the following example, the script:

  • sends a dataset (array of JS objects) to the Web Worker
  • generates a workbook object in the Web Worker
  • generates a XLSB file using XLSX.write in the Web Worker
  • generates an object URL in the Web Worker
  • sends the object URL to the main browser context
  • performs a download action in the main browser context

{\ function SheetJSWriteFileWorker() { const [__html, setHTML] = React.useState(""); \n\ const data = [ { "SheetJS": "வணக்கம்", "in": "สวัสดี", "Web": "你好", "Workers": "가지마" }, { "SheetJS": 1, "in": 2, "Web": 3, "Workers": 4 }, ]; \n\ return ( <> <button onClick={() => { setHTML(""); /* this mantra embeds the worker source in the function */ const worker = new Worker(URL.createObjectURL(new Blob([\\ /* load standalone script from CDN / importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js"); \n
/
this callback will run once the main context sends a message / self.addEventListener('message', async(e) => { try { / Create a new workbook from the data / const ws = XLSX.utils.json_to_sheet(e.data.data); const wb = XLSX.utils.book_new(); XLSX.utils.book_append_sheet(wb, ws, "Data"); \n
/
Write XLSB data (Uint8Array) / const u8 = XLSX.write(wb, { bookType: "xlsb", type: "buffer" }); \n
/
Generate URL / const url = URL.createObjectURL(new Blob([u8])); \n
/
Reply with result / postMessage({ url }); } catch(e) { / Pass the error message back / postMessage({error: String(e.message || e).bold() }); } }, false); `]))); / when the worker sends back the data, create a download / worker.onmessage = function(e) { if(e.data.error) return setHTML(e.data.error); \n
/
this mantra is the standard HTML5 download attribute technique / const a = document.createElement("a"); a.download = "SheetJSWriteFileWorker.xlsb"; a.href = e.data.url; document.body.appendChild(a); a.click(); document.body.removeChild(a); }; / post a message to the worker */ worker.postMessage({ data }); }}>Click to Start <div dangerouslySetInnerHTML={{ __html }}/> </> ); }`}

User-Submitted File

:::note FileReaderSync

Typically FileReader is used in the main browser context. In Web Workers, the synchronous version FileReaderSync is more efficient.

:::

Typically the Web Worker receives a file pointer, reads and parses the file, and sends a final result (HTML table or raw data) to the main browser context:

sequenceDiagram
  autonumber
  actor User
  participant Page
  participant Worker
  User->>Page: submit file
  activate Page
  Page->>Worker: send pointer
  deactivate Page
  activate Worker
  Note over Worker: fetch file
  Note over Worker: parse file
  Note over Worker: generate table
  Worker->>Page: HTML table
  deactivate Worker
  activate Page
  Note over Page: add to DOM
  Page->>User: table is visible
  deactivate Page
Live Demo (click to show)

In the following example, when a file is dropped over the DIV or when the INPUT element is used to select a file, the script:

  • sends the File object to the Web Worker
  • loads the SheetJS library and parses the file in the Worker
  • generates an HTML string of the first table in the Worker
  • sends the string to the main browser context
  • adds the HTML to the page in the main browser context

{\ function SheetJSDragDropWorker() { const [__html, setHTML] = React.useState(""); /* suppress default behavior for drag and drop */ function suppress(e) { e.stopPropagation(); e.preventDefault(); } \n\ /* this worker is shared between drag-drop and file input element */ const worker = new Worker(URL.createObjectURL(new Blob([\\ /* load standalone script from CDN / importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js"); \n
/
this callback will run once the main context sends a message / self.addEventListener('message', (e) => { try { / Read file data / const ab = new FileReaderSync().readAsArrayBuffer(e.data.file); \n
/
Parse file / const wb = XLSX.read(ab, {dense: true}); const ws = wb.Sheets[wb.SheetNames[0]]; \n
/
Generate HTML / const html = XLSX.utils.sheet_to_html(ws); \n
/
Reply with result / postMessage({ html }); } catch(e) { / Pass the error message back / postMessage({html: String(e.message || e).bold() }); } }, false); `]))); / when the worker sends back the HTML, add it to the DOM / worker.onmessage = function(e) { setHTML(e.data.html); }; return ( <> <div onDragOver={suppress} onDragEnter={suppress} onDrop={(e) => { suppress(e); / post a message with the first File to the worker / worker.postMessage({ file: e.dataTransfer.files[0] }); }}>Drag a file to this DIV to process! (or use the file input) <input type="file" onChange={(e) => { suppress(e); / post a message with the first File to the worker */ worker.postMessage({ file: e.target.files[0] }); }}/> <div dangerouslySetInnerHTML={{ __html }}/> </> ); }`}

Streaming Write

A more general discussion, including row-oriented processing demos, is included in the "Large Datasets" demo.

XLSX.stream.to_csv incrementally generates CSV rows.

File System Access API

:::warning Browser Compatibility

At the time of writing, the File System Access API is only available in Chromium and Chromium-based browsers like Chrome and Edge.

:::

:::caution Performance

In local testing, committing each CSV row as it is generated is significantly slower than accumulating and writing once at the end.

When the target CSV is known to be less than 500MB, it is preferable to batch. Larger files may hit browser length limits.

:::

sequenceDiagram
  autonumber
  actor User
  participant Page
  participant Worker
  User->>Page: click button
  activate Page
  Page->>User: seek permission
  User->>Page: grant permission
  Note over Page: collect dataset
  Page->>Worker: transfer dataset
  deactivate Page
  activate Worker
  Note over Worker: setup stream
  Worker->>User: start download
  loop every 100 rows
    Note over Worker: generate rows
    Worker->>User: write rows to file (without freezing)
    Worker->>Page: progress message
    activate Page
    Page->>User: display progress
    deactivate Page
  end
  Worker->>User: finish download
  Worker->>Page: send completion message
  deactivate Worker
  activate Page
  Page->>User: download complete
  deactivate Page
Live Demo (click to show)

The following live demo fetches and parses a file in a Web Worker. The script:

  • prompts user to save file (window.showSaveFilePicker in the main thread)
  • passes the URL and the file object to the Web Worker
  • loads the SheetJS library in the Web Worker
  • fetches the requested URL and parses the workbook from the Worker
  • creates a Writable Stream from the file object.
  • uses XLSX.stream.to_csv to generate CSV rows of the first worksheet
    • every 100th row, a progress message is sent back to the main thread
    • at the end, a completion message is sent back to the main thread

The demo has a checkbox. If it is not checked (default), the Worker will collect each CSV row and write once at the end. If it is checked, the Worker will try to commit each row as it is generated.

The demo also has a URL input box. Feel free to change the URL. For example:

https://raw.githubusercontent.com/SheetJS/test_files/master/large_strings.xls is an XLS file over 50 MB. The generated CSV file is about 55 MB.

https://raw.githubusercontent.com/SheetJS/libreoffice_test-files/master/calc/xlsx-import/perf/8-by-300000-cells.xlsx is an XLSX file with 300000 rows (approximately 20 MB) yielding a CSV of 10 MB.

{\ function SheetJSFetchCSVStreamFile() { const [state, setState] = React.useState(""); const [__html, setHTML] = React.useState(""); const [cnt, setCnt] = React.useState(0); const [hz, setHz] = React.useState(0); const [url, setUrl] = React.useState("https://docs.sheetjs.com/test_files/large_strings.xlsx"); const ref = React.useRef(null); \n\ return ( <> <b>URL: </b><input type="text" value={url} onChange={(e) => setUrl(e.target.value)} size="80"/><br/> <b>Commit each row: </b><input type="checkbox" ref={ref}/><br/> <button onClick={async() => { /* this mantra embeds the worker source in the function */ const worker = new Worker(URL.createObjectURL(new Blob([\\ /* load standalone script from CDN / importScripts("https://cdn.sheetjs.com/xlsx-${current}/package/dist/xlsx.full.min.js"); \n
function sheet_to_csv_cb(ws, cb, opts, batch = 1000) { XLSX.stream.set_readable(() => ({ __done: false, // this function will be assigned by the SheetJS stream methods _read: function() { this.__done = true; }, // this function is called by the stream methods push: function(d) { if(!this.__done) cb(d); if(d == null) this.__done = true; }, resume: function pump() { for(var i = 0; i < batch && !this.__done; ++i) this._read(); if(!this.__done) setTimeout(pump.bind(this), 0); } })); return XLSX.stream.to_csv(ws, opts); } \n
/
this callback will run once the main context sends a message / self.addEventListener('message', async(e) => { try { / Fetch file / postMessage({state: "fetching"}); var t = Date.now(); const res = await fetch(e.data.url); const ab = await res.arrayBuffer(); postMessage({time: "fetch", ts: Date.now() - t}); \n
/
Parse file / postMessage({state: "parsing"}); t = Date.now(); const wb = XLSX.read(ab, {dense: true}); const ws = wb.Sheets[wb.SheetNames[0]]; postMessage({time: "parse", ts: Date.now() - t}); \n
/
Generate CSV rows / postMessage({state: "begin"}); t = Date.now(); const wstream = await e.data.wFile.createWritable(); let c = 0, buf = "", each = !!e.data.each; const strm = sheet_to_csv_cb(ws, async(csv) => { if(csv != null) { if(each) await wstream.write(csv); else buf += csv; if(!(++c % 100)) postMessage({ state: "writing", c, ts: Date.now() - t }); } else { if(buf) await wstream.write(buf); await wstream.close(); postMessage({state: "done", c, ts: Date.now() - t }); } }); strm.resume(); } catch(e) { / Pass the error message back / postMessage({error: String(e.message || e) }); } }, false); `]))); / when the worker sends back data, add it to the DOM / const log = (s, t) => setHTML(h => h + `${s}: ${(t/1000).toFixed(3).padStart(8)} sec\n`); worker.onmessage = function(e) { if(e.data.error) return setState(`Processing Error: ${e.data.error}`); else if(e.data.state) { setState(e.data.state); if(e.data.c) setCnt(e.data.c); if(e.data.ts) setHz((e.data.c || cnt) * 1000 / e.data.ts); if(e.data.state == "done") log("write", e.data.ts); } else if(e.data.time) log(e.data.time, e.data.ts); }; setCnt(0); setHz(0); setState(""); setHTML(""); try { / Show picker and get handle to file / const wFile = await window.showSaveFilePicker({ suggestedName: "SheetJSStream.csv", types: [ { description: 'csv', accept: { 'text/csv': ['.csv'] } } ] }); \n
/
post a message to the worker with the URL to fetch */ worker.postMessage({url, wFile, each: !!ref.current.checked}); } catch(e) { setState(`Selection Error: ${e && e.message || e}`); } }}>Click to Start

State: {state}
Count: {cnt} ({hz|0} Hz)
</> );
}`}