The SheetJS Community Edition offers battle-tested open-source solutions for extracting useful data from almost any complex spreadsheet and generating new spreadsheets that will work with legacy and modern software alike.
SheetJS Pro offers solutions beyond data processing: Edit complex templates with ease; let out your inner Picasso with styling; make custom sheets with images/graphs/PivotTables; evaluate formula expressions and port calculations to web apps; automate common spreadsheet tasks, and much more!
Browser Test and Support Matrix
Supported File Formats
Table of Contents
Expand to show Table of Contents
- Getting Started
- Installation
- Usage
- The Zen of SheetJS
- JS Ecosystem Demos
- Acquiring and Extracting Data
- Parsing Workbooks
- Processing JSON and JS Data
- Processing HTML Tables
- Processing Data
- Modifying Workbook Structure
- Modifying Cell Values
- Modifying Other Worksheet / Workbook / Cell Properties
- Packaging and Releasing Data
- Writing Workbooks
- Writing Examples
- Streaming Write
- Generating JSON and JS Data
- Generating HTML Tables
- Generating Single-Worksheet Snapshots
- Interface
- Parsing functions
- Writing functions
- Utilities
- Common Spreadsheet Format
- General Structures
- Cell Object
- Data Types
- Dates
- Sheet Objects
- Worksheet Object
- Chartsheet Object
- Macrosheet Object
- Dialogsheet Object
- Workbook Object
- Workbook File Properties
- Workbook-Level Attributes
- Defined Names
- Workbook Views
- Miscellaneous Workbook Properties
- Document Features
- Formulae
- Row and Column Properties
- Number Formats
- Hyperlinks
- Cell Comments
- Sheet Visibility
- VBA and Macros
- Parsing Options
- Input Type
- Guessing File Type
- Writing Options
- Supported Output Formats
- Output Type
- Utility Functions
- Array of Arrays Input
- Array of Objects Input
- HTML Table Input
- Formulae Output
- Delimiter-Separated Output
- UTF-16 Unicode Text
- HTML Output
- JSON
- File Formats
- Testing
- Node
- Browser
- Tested Environments
- Test Files
- Contributing
- OSX/Linux
- Windows
- Tests
- License
- References
Getting Started
Installation
Standalone Browser Scripts
The complete browser standalone build is saved to dist/xlsx.full.min.js and can be directly added to a page with a script tag:
CDN Availability (click to show) CDN URL unpkg https://unpkg.com/xlsx/ jsDelivr https://jsdelivr.com/package/npm/xlsx CDNjs https://cdnjs.com/libraries/xlsx
For example, unpkg makes the latest version available at:
Browser builds (click to show)
The complete single-file version is generated at dist/xlsx.full.min.js
dist/xlsx.core.min.js omits codepage library (no support for XLS encodings)
A slimmer build is generated at dist/xlsx.mini.min.js. Compared to full build:
- codepage library skipped (no support for XLS encodings)
- no support for XLSB / XLS / Lotus 1-2-3 / SpreadsheetML 2003 / Numbers
- node stream utils removed
With bower:
ECMAScript Modules
The ECMAScript Module build is saved to xlsx.mjs and can be directly added to a page with a script tag using type=module:
The npm package also exposes the module with the module parameter, supported in Angular and other projects:
Deno
xlsx.mjs can be imported in Deno. It is available from unpkg:
NodeJS
With npm:
By default, the module supports require:
The module also ships with xlsx.mjs for use with import:
Photoshop and InDesign
dist/xlsx.extendscript.js is an ExtendScript build for Photoshop and InDesign that is included in the npm package. It can be directly referenced with a #include directive:
#include “xlsx.extendscript.js” Internet Explorer and ECMAScript 3 Compatibility (click to show)
For broad compatibility with JavaScript engines, the library is written using ECMAScript 3 language dialect as well as some ES5 features like Array#forEach. Older browsers require shims to provide missing functions.
To use the shim, add the shim before the script tag that loads xlsx.js:
The script also includes IE_LoadFile and IE_SaveFile for loading and saving files in Internet Explorer versions 6-9. The xlsx.extendscript.js script bundles the shim in a format suitable for Photoshop and other Adobe products.
Usage
Most scenarios involving spreadsheets and data can be broken into 5 parts:
-
Acquire Data: Data may be stored anywhere: local or remote files, databases, HTML TABLE, or even generated programmatically in the web browser.
-
Extract Data: For spreadsheet files, this involves parsing raw bytes to read the cell data. For general JS data, this involves reshaping the data.
-
Process Data: From generating summary statistics to cleaning data records, this step is the heart of the problem.
-
Package Data: This can involve making a new spreadsheet or serializing with JSON.stringify or writing XML or simply flattening data for UI tools.
-
Release Data: Spreadsheet files can be uploaded to a server or written locally. Data can be presented to users in an HTML TABLE or data grid.
A common problem involves generating a valid spreadsheet export from data stored in an HTML table. In this example, an HTML TABLE on the page will be scraped, a row will be added to the bottom with the date of the report, and a new file will be generated and downloaded locally. XLSX.writeFile takes care of packaging the data and attempting a local download:
This library tries to simplify steps 2 and 4 with functions to extract useful data from spreadsheet files (read / readFile) and generate new spreadsheet files from data (write / writeFile). Additional utility functions like table_to_book work with other common data sources like HTML tables.
This documentation and various demo projects cover a number of common scenarios and approaches for steps 1 and 5.
Utility functions help with step 3.
“Acquiring and Extracting Data” describes solutions for common data import scenarios.
“Packaging and Releasing Data” describes solutions for common data export scenarios.
“Processing Data” describes solutions for common workbook processing and manipulation scenarios.
“Utility Functions” details utility functions for translating JSON Arrays and other common JS structures into worksheet objects.
The Zen of SheetJS
Data processing should fit in any workflow
The library does not impose a separate lifecycle. It fits nicely in websites and apps built using any framework. The plain JS data objects play nice with Web Workers and future APIs.
JavaScript is a powerful language for data processing
The “Common Spreadsheet Format” is a simple object representation of the core concepts of a workbook. The various functions in the library provide low-level tools for working with the object.
For friendly JS processing, there are utility functions for converting parts of a worksheet to/from an Array of Arrays. The following example combines powerful JS Array methods with a network request library to download data, select the information we want and create a workbook file:
Get Data from a JSON Endpoint and Generate a Workbook (click to show)
The goal is to generate a XLSB workbook of US President names and birthdays.
Acquire Data
Raw Data
https://theunitedstates.io/congress-legislators/executive.json has the desired data. For example, John Adams:
Filtering for Presidents
The dataset includes Aaron Burr, a Vice President who was never President!
Array#filter creates a new array with the desired rows. A President served at least one term with type set to “prez”. To test if a particular row has at least one “prez” term, Array#some is another native JS function. The complete filter would be:
Lining up the data
For this example, the name will be the first name combined with the last name (row.name.first + ” ” + row.name.last) and the birthday will be the subfield row.bio.birthday. Using Array#map, the dataset can be massaged in one call:
The result is an array of “simple” objects with no nesting:
Extract Data
With the cleaned dataset, XLSX.utils.json_to_sheet generates a worksheet:
XLSX.utils.book_new creates a new workbook and XLSX.utils.book_append_sheet appends a worksheet to the workbook. The new worksheet will be called “Dates”:
Process Data
Fixing headers
By default, json_to_sheet creates a worksheet with a header row. In this case, the headers come from the JS object keys: “name” and “birthday”.
The headers are in cells A1 and B1. XLSX.utils.sheet_add_aoa can write text values to the existing worksheet starting at cell A1:
Fixing Column Widths
Some of the names are longer than the default column width. Column widths are set by setting the “!cols” worksheet property.
The following line sets the width of column A to approximately 10 characters:
One Array#reduce call over rows can calculate the maximum width:
Note: If the starting point was a file or HTML table, XLSX.utils.sheet_to_json will generate an array of JS objects.
Package and Release Data
XLSX.writeFile creates a spreadsheet file and tries to write it to the system. In the browser, it will try to prompt the user to download the file. In NodeJS, it will write to the local directory.
Complete Example
For use in the web browser, assuming the snippet is saved to snippet.js, script tags should be used to include the axios and xlsx standalone builds:
File formats are implementation details
The parser covers a wide gamut of common spreadsheet file formats to ensure that “HTML-saved-as-XLS” files work as well as actual XLS or XLSX files.
The writer supports a number of common output formats for broad compatibility with the data ecosystem.
To the greatest extent possible, data processing code should not have to worry about the specific file formats involved.
JS Ecosystem Demos
The demos directory includes sample projects for:
Frameworks and APIs
- angularjs
- angular and ionic
- knockout
- meteor
- react and react-native
- vue 2.x and weex
- XMLHttpRequest and fetch
- nodejs server
- databases and key/value stores
- typed arrays and math
Bundlers and Tooling
- browserify
- fusebox
- parcel
- requirejs
- rollup
- systemjs
- typescript
- webpack 2.x
Platforms and Integrations
- deno
- electron application
- nw.js application
- Chrome / Chromium extensions
- Download a Google Sheet locally
- Adobe ExtendScript
- Headless Browsers
- canvas-datagrid
- x-spreadsheet
- react-data-grid
- vue3-table-light
- Swift JSC and other engines
- “serverless” functions
- internet explorer
Other examples are included in the showcase.
https://sheetjs.com/demos/modify.html shows a complete example of reading, modifying, and writing files.
https://github.com/SheetJS/sheetjs/blob/HEAD/bin/xlsx.njs is the command-line tool included with node installations, reading spreadsheet files and exporting the contents in various formats.
Acquiring and Extracting Data
Parsing Workbooks
API
Extract data from spreadsheet bytes
The read method can extract data from spreadsheet bytes stored in a JS string, “binary string”, NodeJS buffer or typed array (Uint8Array or ArrayBuffer).
Read spreadsheet bytes from a local file and extract data
The readFile method attempts to read a spreadsheet file at the supplied path. Browsers generally do not allow reading files in this way (it is deemed a security risk), and attempts to read files in this way will throw an error.
The second opts argument is optional. “Parsing Options” covers the supported properties and behaviors.
Examples
Here are a few common scenarios (click on each subtitle to see the code):
Local file in a NodeJS server (click to show)
readFile uses fs.readFileSync under the hood:
For Node ESM, the readFile helper is not enabled. Instead, fs.readFileSync should be used to read the file data as a Buffer for use with XLSX.read:
Local file in a Deno application (click to show)
readFile uses Deno.readFileSync under the hood:
Applications reading files must be invoked with the -allow-read flag. The deno demo has more examples
User-submitted file in a web page (“Drag-and-Drop”) (click to show)
For modern websites targeting Chrome 76+, File#arrayBuffer is recommended:
For maximal compatibility, the FileReader API should be used:
https://oss.sheetjs.com/sheetjs/ demonstrates the FileReader technique.
User-submitted file with an HTML INPUT element (click to show)
Starting with an HTML INPUT element with type=”file”:
For modern websites targeting Chrome 76+, Blob#arrayBuffer is recommended:
For broader support (including IE10+), the FileReader approach is recommended:
The oldie demo shows an IE-compatible fallback scenario.
Fetching a file in the web browser (“Ajax”) (click to show)
For modern websites targeting Chrome 42+, fetch is recommended:
For broader support, the XMLHttpRequest approach is recommended:
The xhr demo includes a longer discussion and more examples.
http://oss.sheetjs.com/sheetjs/ajax.html shows fallback approaches for IE6+.
Local file in a PhotoShop or InDesign plugin (click to show)
readFile wraps the File logic in Photoshop and other ExtendScript targets. The specified path should be an absolute path:
The extendscript demo includes a more complex example.
Local file in an Electron app (click to show)
readFile can be used in the renderer process:
Electron APIs have changed over time. The electron demo shows a complete example and details the required version-specific settings.
Local file in a mobile app with React Native (click to show)
The react demo includes a sample React Native app.
Since React Native does not provide a way to read files from the filesystem, a third-party library must be used. The following libraries have been tested:
- react-native-file-access
The base64 encoding returns strings compatible with the base64 type:
- react-native-fs
The ascii encoding returns binary strings compatible with the binary type:
NodeJS Server File Uploads (click to show)
read can accept a NodeJS buffer. readFile can read files generated by a HTTP POST request body parser like formidable:
The server demo has more advanced examples.
Download files in a NodeJS process (click to show)
Node 17.5 and 18.0 have native support for fetch:
For broader compatibility, third-party modules are recommended.
request requires a null encoding to yield Buffers:
axios works the same way in browser and in NodeJS:
Download files in an Electron app (click to show)
The net module in the main process can make HTTP/HTTPS requests to external resources. Responses should be manually concatenated using Buffer.concat:
Readable Streams in NodeJS (click to show)
When dealing with Readable Streams, the easiest approach is to buffer the stream and process the whole thing at the end:
ReadableStream in the browser (click to show)
When dealing with ReadableStream, the easiest approach is to buffer the stream and process the whole thing at the end:
More detailed examples are covered in the included demos
Processing JSON and JS Data
JSON and JS data tend to represent single worksheets. This section will use a few utility functions to generate workbooks.
Create a new Workbook
The book_new utility function creates an empty workbook with no worksheets.
Spreadsheet software generally require at least one worksheet and enforce the requirement in the user interface. This library enforces the requirement at write time, throwing errors if an empty workbook is passed to write functions.
API
Create a worksheet from an array of arrays of JS values
The aoa_to_sheet utility function walks an “array of arrays” in row-major order, generating a worksheet object. The following snippet generates a sheet with cell A1 set to the string A1, cell B1 set to B1, etc:
“Array of Arrays Input” describes the function and the optional opts argument in more detail.
Create a worksheet from an array of JS objects
The json_to_sheet utility function walks an array of JS objects in order, generating a worksheet object. By default, it will generate a header row and one row per object in the array. The optional opts argument has settings to control the column order and header output.
“Array of Objects Input” describes the function and the optional opts argument in more detail.
Examples
“Zen of SheetJS” contains a detailed example “Get Data from a JSON Endpoint and Generate a Workbook”
x-spreadsheet is an interactive data grid for previewing and modifying structured data in the web browser. The xspreadsheet demo includes a sample script with the xtos function for converting from x-spreadsheet data object to a workbook. https://oss.sheetjs.com/sheetjs/x-spreadsheet is a live demo.
Records from a database query (SQL or no-SQL) (click to show)
The database demo includes examples of working with databases and query results.
Numerical Computations with TensorFlow.js (click to show)
@tensorflow/tfjs and other libraries expect data in simple arrays, well-suited for worksheets where each column is a data vector. That is the transpose of how most people use spreadsheets, where each row is a vector.
When recovering data from tfjs, the returned data points are stored in a typed array. An array of arrays can be constructed with loops. Array#unshift can prepend a title row before the conversion:
The array demo shows a complete example.
Processing HTML Tables
API
Create a worksheet by scraping an HTML TABLE in the page
The table_to_sheet utility function takes a DOM TABLE element and iterates through the rows to generate a worksheet. The opts argument is optional. “HTML Table Input” describes the function in more detail.
Create a workbook by scraping an HTML TABLE in the page
The table_to_book utility function follows the same logic as table_to_sheet. After generating a worksheet, it creates a blank workbook and appends the spreadsheet.
The options argument supports the same options as table_to_sheet, with the addition of a sheet property to control the worksheet name. If the property is missing or no options are specified, the default name Sheet1 is used.
Examples
Here are a few common scenarios (click on each subtitle to see the code):
HTML TABLE element in a webpage (click to show)
Multiple tables on a web page can be converted to individual worksheets:
Alternatively, the HTML code can be extracted and parsed:
Chrome/Chromium Extension (click to show)
The chrome demo shows a complete example and details the required permissions and other settings.
In an extension, it is recommended to generate the workbook in a content script and pass the object back to the extension:
Server-Side HTML Tables with Headless Chrome (click to show)
The headless demo includes a complete demo to convert HTML files to XLSB workbooks. The core idea is to add the script to the page, parse the table in the page context, generate a base64 workbook and send it back for further processing:
Server-Side HTML Tables with Headless WebKit (click to show)
The headless demo includes a complete demo to convert HTML files to XLSB workbooks using PhantomJS. The core idea is to add the script to the page, parse the table in the page context, generate a binary workbook and send it back for further processing:
NodeJS HTML Tables without a browser (click to show)
NodeJS does not include a DOM implementation and Puppeteer requires a hefty Chromium build. jsdom is a lightweight alternative:
Processing Data
The “Common Spreadsheet Format” is a simple object representation of the core concepts of a workbook. The utility functions work with the object representation and are intended to handle common use cases.
Modifying Workbook Structure
API
Append a Worksheet to a Workbook
The book_append_sheet utility function appends a worksheet to the workbook. The third argument specifies the desired worksheet name. Multiple worksheets can be added to a workbook by calling the function multiple times. If the worksheet name is already used in the workbook, it will throw an error.
Append a Worksheet to a Workbook and find a unique name
If the fourth argument is true, the function will start with the specified worksheet name. If the sheet name exists in the workbook, a new worksheet name will be chosen by finding the name stem and incrementing the counter:
List the Worksheet names in tab order
The SheetNames property of the workbook object is a list of the worksheet names in “tab order”. API functions will look at this array.
Replace a Worksheet in place
The Sheets property of the workbook object is an object whose keys are names and whose values are worksheet objects. By reassigning to a property of the Sheets object, the worksheet object can be changed without disrupting the rest of the worksheet structure.
Examples
Add a new worksheet to a workbook (click to show)
This example uses XLSX.utils.aoa_to_sheet.
Modifying Cell Values
API
Modify a single cell value in a worksheet
Modify multiple cell values in a worksheet
The sheet_add_aoa utility function modifies cell values in a worksheet. The first argument is the worksheet object. The second argument is an array of arrays of values. The origin key of the third argument controls where cells will be written. The following snippet sets B3=1 and E5=”abc”:
“Array of Arrays Input” describes the function and the optional opts argument in more detail.
Examples
Appending rows to a worksheet (click to show)
The special origin value -1 instructs sheet_add_aoa to start in column A of the row after the last row in the range, appending the data:
Modifying Other Worksheet / Workbook / Cell Properties
The “Common Spreadsheet Format” section describes the object structures in greater detail.
Packaging and Releasing Data
Writing Workbooks
API
Generate spreadsheet bytes (file) from data
The write method attempts to package data from the workbook into a file in memory. By default, XLSX files are generated, but that can be controlled with the bookType property of the opts argument. Based on the type option, the data can be stored as a “binary string”, JS string, Uint8Array or Buffer.
The second opts argument is required. “Writing Options” covers the supported properties and behaviors.
Generate and attempt to save file
The writeFile method packages the data and attempts to save the new file. The export file format is determined by the extension of filename (SheetJS.xlsx signals XLSX export, SheetJS.xlsb signals XLSB export, etc).
The writeFile method uses platform-specific APIs to initiate the file save. In NodeJS, fs.readFileSync can create a file. In the web browser, a download is attempted using the HTML5 download attribute, with fallbacks for IE.
Generate and attempt to save an XLSX file
The writeFile method embeds a number of different export functions. This is great for developer experience but not amenable to tree shaking using the current developer tools. When only XLSX exports are needed, this method avoids referencing the other export functions.
The second opts argument is optional. “Writing Options” covers the supported properties and behaviors.
Examples
Local file in a NodeJS server (click to show)
writeFile uses fs.writeFileSync in server environments:
For Node ESM, the writeFile helper is not enabled. Instead, fs.writeFileSync should be used to write the file data to a Buffer for use with XLSX.write:
Local file in a Deno application (click to show)
writeFile uses Deno.writeFileSync under the hood:
Applications writing files must be invoked with the -allow-write flag. The deno demo has more examples
Local file in a PhotoShop or InDesign plugin (click to show)
writeFile wraps the File logic in Photoshop and other ExtendScript targets. The specified path should be an absolute path:
The extendscript demo includes a more complex example.
Download a file in the browser to the user machine (click to show)
XLSX.writeFile wraps a few techniques for triggering a file save:
- URL browser API creates an object URL for the file, which the library uses by creating a link and forcing a click. It is supported in modern browsers.
- msSaveBlob is an IE10+ API for triggering a file save.
- IE_FileSave uses VBScript and ActiveX to write a file in IE6+ for Windows XP and Windows 7. The shim must be included in the containing HTML page.
There is no standard way to determine if the actual file has been downloaded.
Download a file in legacy browsers (click to show)
XLSX.writeFile techniques work for most modern browsers as well as older IE. For much older browsers, there are workarounds implemented by wrapper libraries.
FileSaver.js implements saveAs. Note: XLSX.writeFile will automatically call saveAs if available.
Downloadify uses a Flash SWF button to generate local files, suitable for environments where ActiveX is unavailable:
The oldie demo shows an IE-compatible fallback scenario.
Browser upload file (ajax) (click to show)
A complete example using XHR is included in the XHR demo, along with examples for fetch and wrapper libraries. This example assumes the server can handle Base64-encoded files (see the demo for a basic nodejs server):
PhantomJS (Headless Webkit) File Generation (click to show)
The headless demo includes a complete demo to convert HTML files to XLSB workbooks using PhantomJS. PhantomJS fs.write supports writing files from the main process but has a different interface from the NodeJS fs module:
Note: The section “Processing HTML Tables” shows how to generate a workbook from HTML tables in a page in “Headless WebKit”.
The included demos cover mobile apps and other special deployments.
Writing Examples
- http://sheetjs.com/demos/table.html exporting an HTML table
- http://sheetjs.com/demos/writexlsx.html generates a simple file
Streaming Write
The streaming write functions are available in the XLSX.stream object. They take the same arguments as the normal write functions but return a NodeJS Readable Stream.
- XLSX.stream.to_csv is the streaming version of XLSX.utils.sheet_to_csv.
- XLSX.stream.to_html is the streaming version of XLSX.utils.sheet_to_html.
- XLSX.stream.to_json is the streaming version of XLSX.utils.sheet_to_json.
nodejs convert to CSV and write file (click to show) nodejs write JSON stream to screen (click to show) Exporting NUMBERS files (click to show)
The NUMBERS writer requires a fairly large base. The supplementary xlsx.zahl scripts provide support. xlsx.zahl.js is designed for standalone and NodeJS use, while xlsx.zahl.mjs is suitable for ESM.
Browser
Node
Deno
https://github.com/sheetjs/sheetaki pipes write streams to nodejs response.
Generating JSON and JS Data
JSON and JS data tend to represent single worksheets. The utility functions in this section work with single worksheets.
The “Common Spreadsheet Format” section describes the object structure in more detail. workbook.SheetNames is an ordered list of the worksheet names. workbook.Sheets is an object whose keys are sheet names and whose values are worksheet objects.
The “first worksheet” is stored at workbook.Sheets[workbook.SheetNames[0]].
API
Create an array of JS objects from a worksheet
Create an array of arrays of JS values from a worksheet
The sheet_to_json utility function walks a workbook in row-major order, generating an array of objects. The second opts argument controls a number of export decisions including the type of values (JS values or formatted text). The “JSON” section describes the argument in more detail.
By default, sheet_to_json scans the first row and uses the values as headers. With the header: 1 option, the function exports an array of arrays of values.
Examples
x-spreadsheet is an interactive data grid for previewing and modifying structured data in the web browser. The xspreadsheet demo includes a sample script with the stox function for converting from a workbook to x-spreadsheet data object. https://oss.sheetjs.com/sheetjs/x-spreadsheet is a live demo.
Previewing data in a React data grid (click to show)
react-data-grid is a data grid tailored for react. It expects two properties: rows of data objects and columns which describe the columns. For the purposes of massaging the data to fit the react data grid API it is easiest to start from an array of arrays.
This demo starts by fetching a remote file and using XLSX.read to extract:
Previewing data in a VueJS data grid (click to show)
vue3-table-lite is a simple VueJS 3 data table. It is featured in the VueJS demo.
Populating a database (SQL or no-SQL) (click to show)
The database demo includes examples of working with databases and query results.
Numerical Computations with TensorFlow.js (click to show)
@tensorflow/tfjs and other libraries expect data in simple arrays, well-suited for worksheets where each column is a data vector. That is the transpose of how most people use spreadsheets, where each row is a vector.
A single Array#map can pull individual named rows from sheet_to_json export:
All fields can be processed at once using a transpose of the 2D tensor generated with the sheet_to_json export with header: 1. The first row, if it contains header labels, should be removed with a slice:
The array demo shows a complete example.
Generating HTML Tables
API
Generate HTML Table from Worksheet
The sheet_to_html utility function generates HTML code based on the worksheet data. Each cell in the worksheet is mapped to a <TD> element. Merged cells in the worksheet are serialized by setting colspan and rowspan attributes.
Examples
The sheet_to_html utility function generates HTML code that can be added to any DOM element by setting the innerHTML:
Combining with fetch, constructing a site from a workbook is straightforward:
Vanilla JS + HTML fetch workbook and generate table previews (click to show) React fetch workbook and generate HTML table previews (click to show)
It is generally recommended to use a React-friendly workflow, but it is possible to generate HTML and use it in React with dangerouslySetInnerHTML:
The react demo includes more React examples.
VueJS fetch workbook and generate HTML table previews (click to show)
It is generally recommended to use a VueJS-friendly workflow, but it is possible to generate HTML and use it in VueJS with the v-html directive:
The vuejs demo includes more React examples.
Generating Single-Worksheet Snapshots
The sheet_to_* functions accept a worksheet object.
API
Generate a CSV from a single worksheet
This snapshot is designed to replicate the “CSV UTF8 (.csv)” output type. “Delimiter-Separated Output” describes the function and the optional opts argument in more detail.
Generate “Text” from a single worksheet
This snapshot is designed to replicate the “UTF16 Text (.txt)” output type. “Delimiter-Separated Output” describes the function and the optional opts argument in more detail.
Generate a list of formulae from a single worksheet
This snapshot generates an array of entries representing the embedded formulae. Array formulae are rendered in the form range=formula while plain cells are rendered in the form cell=formula or value. String literals are prefixed with an apostrophe ‘, consistent with Excel’s formula bar display.
“Formulae Output” describes the function in more detail.
Interface
XLSX is the exposed variable in the browser and the exported node variable
XLSX.version is the version of the library (added by the build script).
XLSX.SSF is an embedded version of the format library.
Parsing functions
XLSX.read(data, read_opts) attempts to parse data.
XLSX.readFile(filename, read_opts) attempts to read filename and parse.
Parse options are described in the Parsing Options section.
Writing functions
XLSX.write(wb, write_opts) attempts to write the workbook wb
XLSX.writeFile(wb, filename, write_opts) attempts to write wb to filename. In browser-based environments, it will attempt to force a client-side download.
XLSX.writeFileAsync(wb, filename, o, cb) attempts to write wb to filename. If o is omitted, the writer will use the third argument as the callback.
XLSX.stream contains a set of streaming write functions.
Write options are described in the Writing Options section.
Utilities
Utilities are available in the XLSX.utils object and are described in the Utility Functions section:
Constructing:
- book_new creates an empty workbook
- book_append_sheet adds a worksheet to a workbook
Importing:
- aoa_to_sheet converts an array of arrays of JS data to a worksheet.
- json_to_sheet converts an array of JS objects to a worksheet.
- table_to_sheet converts a DOM TABLE element to a worksheet.
- sheet_add_aoa adds an array of arrays of JS data to an existing worksheet.
- sheet_add_json adds an array of JS objects to an existing worksheet.
Exporting:
- sheet_to_json converts a worksheet object to an array of JSON objects.
- sheet_to_csv generates delimiter-separated-values output.
- sheet_to_txt generates UTF16 formatted text.
- sheet_to_html generates HTML output.
- sheet_to_formulae generates a list of the formulae (with value fallbacks).
Cell and cell address manipulation:
- format_cell generates the text value for a cell (using number formats).
- encode_row / decode_row converts between 0-indexed rows and 1-indexed rows.
- encode_col / decode_col converts between 0-indexed columns and column names.
- encode_cell / decode_cell converts cell addresses.
- encode_range / decode_range converts cell ranges.
Common Spreadsheet Format
SheetJS conforms to the Common Spreadsheet Format (CSF):
General Structures
Cell address objects are stored as {c:C, r:R} where C and R are 0-indexed column and row numbers, respectively. For example, the cell address B5 is represented by the object {c:1, r:4}.
Cell range objects are stored as {s:S, e:E} where S is the first cell and E is the last cell in the range. The ranges are inclusive. For example, the range A3:B7 is represented by the object {s:{c:0, r:2}, e:{c:1, r:6}}. Utility functions perform a row-major order walk traversal of a sheet range:
Cell Object
Cell objects are plain JS objects with keys and values following the convention:
Key Description v raw value (see Data Types section for more info) w formatted text (if applicable) t type: b Boolean, e Error, n Number, d Date, s Text, z Stub f cell formula encoded as an A1-style string (if applicable) F range of enclosing array if formula is array formula (if applicable) D if true, array formula is dynamic (if applicable) r rich text encoding (if applicable) h HTML rendering of the rich text (if applicable) c comments associated with the cell z number format string associated with the cell (if requested) l cell hyperlink object (.Target holds link, .Tooltip is tooltip) s the style/theme of the cell (if applicable)
Built-in export utilities (such as the CSV exporter) will use the w text if it is available. To change a value, be sure to delete cell.w (or set it to undefined) before attempting to export. The utilities will regenerate the w text from the number format (cell.z) and the raw value if possible.
The actual array formula is stored in the f field of the first cell in the array range. Other cells in the range will omit the f field.
Data Types
The raw value is stored in the v value property, interpreted based on the t type property. This separation allows for representation of numbers as well as numeric text. There are 6 valid cell types:
Type Description b Boolean: value interpreted as JS boolean e Error: value is a numeric code and w property stores common name ** n Number: value is a JS number ** d Date: value is a JS Date object or string to be parsed as Date ** s Text: value interpreted as JS string and written as text ** z Stub: blank stub cell that is ignored by data processing utilities ** Error values and interpretation (click to show) Value Error Meaning 0x00 #NULL! 0x07 #DIV/0! 0x0F #VALUE! 0x17 #REF! 0x1D #NAME? 0x24 #NUM! 0x2A #N/A 0x2B #GETTING_DATA
Type n is the Number type. This includes all forms of data that Excel stores as numbers, such as dates/times and Boolean fields. Excel exclusively uses data that can be fit in an IEEE754 floating point number, just like JS Number, so the v field holds the raw number. The w field holds formatted text. Dates are stored as numbers by default and converted with XLSX.SSF.parse_date_code.
Type d is the Date type, generated only when the option cellDates is passed. Since JSON does not have a natural Date type, parsers are generally expected to store ISO 8601 Date strings like you would get from date.toISOString(). On the other hand, writers and exporters should be able to handle date strings and JS Date objects. Note that Excel disregards timezone modifiers and treats all dates in the local timezone. The library does not correct for this error.
Type s is the String type. Values are explicitly stored as text. Excel will interpret these cells as “number stored as text”. Generated Excel files automatically suppress that class of error, but other formats may elicit errors.
Type z represents blank stub cells. They are generated in cases where cells have no assigned value but hold comments or other metadata. They are ignored by the core library data processing utility functions. By default these cells are not generated; the parser sheetStubs option must be set to true.
Dates
Excel Date Code details (click to show)
By default, Excel stores dates as numbers with a format code that specifies date processing. For example, the date 19-Feb-17 is stored as the number 42785 with a number format of d-mmm-yy. The SSF module understands number formats and performs the appropriate conversion.
XLSX also supports a special date type d where the data is an ISO 8601 date string. The formatter converts the date back to a number.
The default behavior for all parsers is to generate number cells. Setting cellDates to true will force the generators to store dates.
Time Zones and Dates (click to show)
Excel has no native concept of universal time. All times are specified in the local time zone. Excel limitations prevent specifying true absolute dates.
Following Excel, this library treats all dates as relative to local time zone.
Epochs: 1900 and 1904 (click to show)
Excel supports two epochs (January 1 1900 and January 1 1904). The workbook’s epoch can be determined by examining the workbook’s wb.Workbook.WBProps.date1904 property:
Sheet Objects
Each key that does not start with ! maps to a cell (using A-1 notation)
sheet[address] returns the cell object for the specified address.
Special sheet keys (accessible as sheet[key], each starting with !):
-
sheet[‘!ref’]: A-1 based range representing the sheet range. Functions that work with sheets should use this parameter to determine the range. Cells that are assigned outside of the range are not processed. In particular, when writing a sheet by hand, cells outside of the range are not included
Functions that handle sheets should test for the presence of !ref field. If the !ref is omitted or is not a valid range, functions are free to treat the sheet as empty or attempt to guess the range. The standard utilities that ship with this library treat sheets as empty (for example, the CSV output is empty string).
When reading a worksheet with the sheetRows property set, the ref parameter will use the restricted range. The original range is set at ws[‘!fullref’]
-
sheet[‘!margins’]: Object representing the page margins. The default values follow Excel’s “normal” preset. Excel also has a “wide” and a “narrow” preset but they are stored as raw measurements. The main properties are listed below:
Page margin details (click to show) key description “normal” “wide” “narrow” left left margin (inches) 0.7 1.0 0.25 right right margin (inches) 0.7 1.0 0.25 top top margin (inches) 0.75 1.0 0.75 bottom bottom margin (inches) 0.75 1.0 0.75 header header margin (inches) 0.3 0.5 0.3 footer footer margin (inches) 0.3 0.5 0.3
Worksheet Object
In addition to the base sheet keys, worksheets also add:
-
ws[‘!cols’]: array of column properties objects. Column widths are actually stored in files in a normalized manner, measured in terms of the “Maximum Digit Width” (the largest width of the rendered digits 0-9, in pixels). When parsed, the column objects store the pixel width in the wpx field, character width in the wch field, and the maximum digit width in the MDW field.
-
ws[‘!rows’]: array of row properties objects as explained later in the docs. Each row object encodes properties including row height and visibility.
-
ws[‘!merges’]: array of range objects corresponding to the merged cells in the worksheet. Plain text formats do not support merge cells. CSV export will write all cells in the merge range if they exist, so be sure that only the first cell (upper-left) in the range is set.
-
ws[‘!outline’]: configure how outlines should behave. Options default to the default settings in Excel 2019:
key Excel feature default above Uncheck “Summary rows below detail” false left Uncheck “Summary rows to the right of detail” false
- ws[‘!protect’]: object of write sheet protection properties. The password key specifies the password for formats that support password-protected sheets (XLSX/XLSB/XLS). The writer uses the XOR obfuscation method. The following keys control the sheet protection – set to false to enable a feature when sheet is locked or set to true to disable a feature:
Worksheet Protection Details (click to show) key feature (true=disabled / false=enabled) default selectLockedCells Select locked cells enabled selectUnlockedCells Select unlocked cells enabled formatCells Format cells disabled formatColumns Format columns disabled formatRows Format rows disabled insertColumns Insert columns disabled insertRows Insert rows disabled insertHyperlinks Insert hyperlinks disabled deleteColumns Delete columns disabled deleteRows Delete rows disabled sort Sort disabled autoFilter Filter disabled pivotTables Use PivotTable reports disabled objects Edit objects enabled scenarios Edit scenarios enabled
- ws[‘!autofilter’]: AutoFilter object following the schema:
Chartsheet Object
Chartsheets are represented as standard sheets. They are distinguished with the !type property set to “chart”.
The underlying data and !ref refer to the cached data in the chartsheet. The first row of the chartsheet is the underlying header.
Macrosheet Object
Macrosheets are represented as standard sheets. They are distinguished with the !type property set to “macro”.
Dialogsheet Object
Dialogsheets are represented as standard sheets. They are distinguished with the !type property set to “dialog”.
Workbook Object
workbook.SheetNames is an ordered list of the sheets in the workbook
wb.Sheets[sheetname] returns an object representing the worksheet.
wb.Props is an object storing the standard properties. wb.Custprops stores custom properties. Since the XLS standard properties deviate from the XLSX standard, XLS parsing stores core properties in both places.
wb.Workbook stores workbook-level attributes.
Workbook File Properties
The various file formats use different internal names for file properties. The workbook Props object normalizes the names:
File Properties (click to show) JS Name Excel Description Title Summary tab “Title” Subject Summary tab “Subject” Author Summary tab “Author” Manager Summary tab “Manager” Company Summary tab “Company” Category Summary tab “Category” Keywords Summary tab “Keywords” Comments Summary tab “Comments” LastAuthor Statistics tab “Last saved by” CreatedDate Statistics tab “Created”
For example, to set the workbook title property:
Custom properties are added in the workbook Custprops object:
Writers will process the Props key of the options object:
Workbook-Level Attributes
wb.Workbook stores workbook-level attributes.
Defined Names
wb.Workbook.Names is an array of defined name objects which have the keys:
Defined Name Properties (click to show) Key Description Sheet Name scope. Sheet Index (0 = first sheet) or null (Workbook) Name Case-sensitive name. Standard rules apply ** Ref A1-style Reference (“Sheet1!$A$1:$D$20”) Comment Comment (only applicable for XLS/XLSX/XLSB)
Excel allows two sheet-scoped defined names to share the same name. However, a sheet-scoped name cannot collide with a workbook-scope name. Workbook writers may not enforce this constraint.
Workbook Views
wb.Workbook.Views is an array of workbook view objects which have the keys:
Key Description RTL If true, display right-to-left
Miscellaneous Workbook Properties
wb.Workbook.WBProps holds other workbook properties:
Key Description CodeName VBA Project Workbook Code Name date1904 epoch: 0/false for 1900 system, 1/true for 1904 filterPrivacy Warn or strip personally identifying info on save
Document Features
Even for basic features like date storage, the official Excel formats store the same content in different ways. The parsers are expected to convert from the underlying file format representation to the Common Spreadsheet Format. Writers are expected to convert from CSF back to the underlying file format.
Formulae
The A1-style formula string is stored in the f field. Even though different file formats store the formulae in different ways, the formats are translated. Even though some formats store formulae with a leading equal sign, CSF formulae do not start with =.
Formulae File Format Support (click to show) Storage Representation Formats Read Write A1-style strings XLSX ✔ ✔ RC-style strings XLML and plain text ✔ ✔ BIFF Parsed formulae XLSB and all XLS formats ✔ OpenFormula formulae ODS/FODS/UOS ✔ ✔ Lotus Parsed formulae All Lotus WK_ formats ✔
Since Excel prohibits named cells from colliding with names of A1 or RC style cell references, a (not-so-simple) regex conversion is possible. BIFF Parsed formulae and Lotus Parsed formulae have to be explicitly unwound. OpenFormula formulae can be converted with regular expressions.
Shared formulae are decompressed and each cell has the formula corresponding to its cell. Writers generally do not attempt to generate shared formulae.
Single-Cell Formulae
For simple formulae, the f key of the desired cell can be set to the actual formula text. This worksheet represents A1=1, A2=2, and A3=A1+A2:
Utilities like aoa_to_sheet will accept cell objects in lieu of values:
Cells with formula entries but no value will be serialized in a way that Excel and other spreadsheet tools will recognize. This library will not automatically compute formula results! For example, the following worksheet will include the BESSELJ function but the result will not be available in JavaScript:
If the actual results are needed in JS, SheetJS Pro offers a formula calculator component for evaluating expressions, updating values and dependent cells, and refreshing entire workbooks.
Array Formulae
Assign an array formula
Array formulae are stored in the top-left cell of the array block. All cells of an array formula have a F field corresponding to the range. A single-cell formula can be distinguished from a plain formula by the presence of F field.
For example, setting the cell C1 to the array formula {=SUM(A1:A3*B1:B3)}:
For a multi-cell array formula, every cell has the same array range but only the first cell specifies the formula. Consider D1:D3=A1:A3*B1:B3:
Utilities and writers are expected to check for the presence of a F field and ignore any possible formula element f in cells other than the starting cell. They are not expected to perform validation of the formulae!
Dynamic Array Formulae
Assign a dynamic array formula
Released in 2020, Dynamic Array Formulae are supported in the XLSX/XLSM and XLSB file formats. They are represented like normal array formulae but have special cell metadata indicating that the formula should be allowed to adjust the range.
An array formula can be marked as dynamic by setting the cell’s D property to true. The F range is expected but can be the set to the current cell:
Localization with Function Names
SheetJS operates at the file level. Excel stores formula expressions using the English (United States) function names. For non-English users, Excel uses a localized set of function names.
For example, when the computer language and region is set to French (France), Excel interprets =SOMME(A1:C3) as if SOMME is the SUM function. However, in the actual file, Excel stores SUM(A1:C3).
Prefixed “Future Functions”
Functions introduced in newer versions of Excel are prefixed with _xlfn. when stored in files. When writing formula expressions using these functions, the prefix is required for maximal compatibility:
When reading a file, the xlfn option preserves the prefixes.
Functions requiring `_xlfn.` prefix (click to show)
This list is growing with each Excel release.
ACOT ACOTH AGGREGATE ARABIC BASE BETA.DIST BETA.INV BINOM.DIST BINOM.DIST.RANGE BINOM.INV BITAND BITLSHIFT BITOR BITRSHIFT BITXOR BYCOL BYROW CEILING.MATH CEILING.PRECISE CHISQ.DIST CHISQ.DIST.RT CHISQ.INV CHISQ.INV.RT CHISQ.TEST COMBINA CONFIDENCE.NORM CONFIDENCE.T COT COTH COVARIANCE.P COVARIANCE.S CSC CSCH DAYS DECIMAL ERF.PRECISE ERFC.PRECISE EXPON.DIST F.DIST F.DIST.RT F.INV F.INV.RT F.TEST FIELDVALUE FILTERXML FLOOR.MATH FLOOR.PRECISE FORMULATEXT GAMMA GAMMA.DIST GAMMA.INV GAMMALN.PRECISE GAUSS HYPGEOM.DIST IFNA IMCOSH IMCOT IMCSC IMCSCH IMSEC IMSECH IMSINH IMTAN ISFORMULA ISOMITTED ISOWEEKNUM LAMBDA LET LOGNORM.DIST LOGNORM.INV MAKEARRAY MAP MODE.MULT MODE.SNGL MUNIT NEGBINOM.DIST NORM.DIST NORM.INV NORM.S.DIST NORM.S.INV NUMBERVALUE PDURATION PERCENTILE.EXC PERCENTILE.INC PERCENTRANK.EXC PERCENTRANK.INC PERMUTATIONA PHI POISSON.DIST QUARTILE.EXC QUARTILE.INC QUERYSTRING RANDARRAY RANK.AVG RANK.EQ REDUCE RRI SCAN SEC SECH SEQUENCE SHEET SHEETS SKEW.P SORTBY STDEV.P STDEV.S T.DIST T.DIST.2T T.DIST.RT T.INV T.INV.2T T.TEST UNICHAR UNICODE UNIQUE VAR.P VAR.S WEBSERVICE WEIBULL.DIST XLOOKUP XOR Z.TEST
Row and Column Properties
Format Support (click to show)
Row Properties: XLSX/M, XLSB, BIFF8 XLS, XLML, SYLK, DOM, ODS
Column Properties: XLSX/M, XLSB, BIFF8 XLS, XLML, SYLK, DOM
Row and Column properties are not extracted by default when reading from a file and are not persisted by default when writing to a file. The option cellStyles: true must be passed to the relevant read or write function.
Column Properties
The !cols array in each worksheet, if present, is a collection of ColInfo objects which have the following properties:
Row Properties
The !rows array in each worksheet, if present, is a collection of RowInfo objects which have the following properties:
Outline / Group Levels Convention
The Excel UI displays the base outline level as 1 and the max level as 8. Following JS conventions, SheetJS uses 0-indexed outline levels wherein the base outline level is 0 and the max level is 7.
Why are there three width types? (click to show)
There are three different width types corresponding to the three different ways spreadsheets store column widths:
SYLK and other plain text formats use raw character count. Contemporaneous tools like Visicalc and Multiplan were character based. Since the characters had the same width, it sufficed to store a count. This tradition was continued into the BIFF formats.
SpreadsheetML (2003) tried to align with HTML by standardizing on screen pixel count throughout the file. Column widths, row heights, and other measures use pixels. When the pixel and character counts do not align, Excel rounds values.
XLSX internally stores column widths in a nebulous “Max Digit Width” form. The Max Digit Width is the width of the largest digit when rendered (generally the “0” character is the widest). The internal width must be an integer multiple of the the width divided by 256. ECMA-376 describes a formula for converting between pixels and the internal width. This represents a hybrid approach.
Read functions attempt to populate all three properties. Write functions will try to cycle specified values to the desired type. In order to avoid potential conflicts, manipulation should delete the other properties first. For example, when changing the pixel width, delete the wch and width properties.
Implementation details (click to show)
Row Heights
Excel internally stores row heights in points. The default resolution is 72 DPI or 96 PPI, so the pixel and point size should agree. For different resolutions they may not agree, so the library separates the concepts.
Even though all of the information is made available, writers are expected to follow the priority order:
- use hpx pixel height if available
- use hpt point height if available
Column Widths
Given the constraints, it is possible to determine the MDW without actually inspecting the font! The parsers guess the pixel width by converting from width to pixels and back, repeating for all possible MDW and selecting the MDW that minimizes the error. XLML actually stores the pixel width, so the guess works in the opposite direction.
Even though all of the information is made available, writers are expected to follow the priority order:
- use width field if available
- use wpx pixel width if available
- use wch character count if available
Number Formats
The cell.w formatted text for each cell is produced from cell.v and cell.z format. If the format is not specified, the Excel General format is used. The format can either be specified as a string or as an index into the format table. Parsers are expected to populate workbook.SSF with the number format table. Writers are expected to serialize the table.
Custom tools should ensure that the local table has each used format string somewhere in the table. Excel convention mandates that the custom formats start at index 164. The following example creates a custom format from scratch:
New worksheet with custom format (click to show)
The rules are slightly different from how Excel displays custom number formats. In particular, literal characters must be wrapped in double quotes or preceded by a backslash. For more info, see the Excel documentation article Create or delete a custom number format or ECMA-376 18.8.31 (Number Formats)
Default Number Formats (click to show)
The default formats are listed in ECMA-376 18.8.30:
ID Format 0 General 1 0 2 0.00 3 #,##0 4 #,##0.00 9 0% 10 0.00% 11 0.00E+00 12 # ?/? 13 # ??/?? 14 m/d/yy (see below) 15 d-mmm-yy 16 d-mmm 17 mmm-yy 18 h:mm AM/PM 19 h:mm:ss AM/PM 20 h:mm 21 h:mm:ss 22 m/d/yy h:mm 37 #,##0 ;(#,##0) 38 #,##0 ;[Red](#,##0) 39 #,##0.00;(#,##0.00) 40 #,##0.00;[Red](#,##0.00) 45 mm:ss 46 [h]:mm:ss 47 mmss.0 48 ##0.0E+0 49 @
Format 14 (m/d/yy) is localized by Excel: even though the file specifies that number format, it will be drawn differently based on system settings. It makes sense when the producer and consumer of files are in the same locale, but that is not always the case over the Internet. To get around this ambiguity, parse functions accept the dateNF option to override the interpretation of that specific format string.
Hyperlinks
Format Support (click to show)
Cell Hyperlinks: XLSX/M, XLSB, BIFF8 XLS, XLML, ODS
Tooltips: XLSX/M, XLSB, BIFF8 XLS, XLML
Hyperlinks are stored in the l key of cell objects. The Target field of the hyperlink object is the target of the link, including the URI fragment. Tooltips are stored in the Tooltip field and are displayed when you move your mouse over the text.
For example, the following snippet creates a link from cell A3 to https://sheetjs.com with the tip “Find us @ SheetJS.com!”:
Note that Excel does not automatically style hyperlinks – they will generally be displayed as normal text.
Remote Links
HTTP / HTTPS links can be used directly:
Excel also supports mailto email links with subject line:
Local Links
Links to absolute paths should use the file:// URI scheme:
Links to relative paths can be specified without a scheme:
Relative Paths have undefined behavior in the SpreadsheetML 2003 format. Excel 2019 will treat a .. parent mark as two levels up.
Internal Links
Links where the target is a cell or range or defined name in the same workbook (“Internal Links”) are marked with a leading hash character:
Cell Comments
Cell comments are objects stored in the c array of cell objects. The actual contents of the comment are split into blocks based on the comment author. The a field of each comment object is the author of the comment and the t field is the plain text representation.
For example, the following snippet appends a cell comment into cell A1:
Note: XLSB enforces a 54 character limit on the Author name. Names longer than 54 characters may cause issues with other formats.
To mark a comment as normally hidden, set the hidden property:
Threaded Comments
Introduced in Excel 365, threaded comments are plain text comment snippets with author metadata and parent references. They are supported in XLSX and XLSB.
To mark a comment as threaded, each comment part must have a true T property:
There is no Active Directory or Office 365 metadata associated with authors in a thread.
Sheet Visibility
Excel enables hiding sheets in the lower tab bar. The sheet data is stored in the file but the UI does not readily make it available. Standard hidden sheets are revealed in the “Unhide” menu. Excel also has “very hidden” sheets which cannot be revealed in the menu. It is only accessible in the VB Editor!
The visibility setting is stored in the Hidden property of sheet props array.
More details (click to show) Value Definition 0 Visible 1 Hidden 2 Very Hidden
With https://rawgit.com/SheetJS/test_files/HEAD/sheet_visibility.xlsx:
Non-Excel formats do not support the Very Hidden state. The best way to test if a sheet is visible is to check if the Hidden property is logical truth:
VBA and Macros
VBA Macros are stored in a special data blob that is exposed in the vbaraw property of the workbook object when the bookVBA option is true. They are supported in XLSM, XLSB, and BIFF8 XLS formats. The supported format writers automatically insert the data blobs if it is present in the workbook and associate with the worksheet names.
Custom Code Names (click to show)
The workbook code name is stored in wb.Workbook.WBProps.CodeName. By default, Excel will write ThisWorkbook or a translated phrase like DieseArbeitsmappe. Worksheet and Chartsheet code names are in the worksheet properties object at wb.Workbook.Sheets[i].CodeName. Macrosheets and Dialogsheets are ignored.
The readers and writers preserve the code names, but they have to be manually set when adding a VBA blob to a different workbook.
Macrosheets (click to show)
Older versions of Excel also supported a non-VBA “macrosheet” sheet type that stored automation commands. These are exposed in objects with the !type property set to “macro”.
Detecting macros in workbooks (click to show)
The vbaraw field will only be set if macros are present, so testing is simple:
Parsing Options
The exported read and readFile functions accept an options argument:
Option Name Default Description type Input data encoding (see Input Type below) raw false If true, plain text parsing will not parse values ** codepage If specified, use code page when appropriate ** cellFormula true Save formulae to the .f field cellHTML true Parse rich text and save HTML to the .h field cellNF false Save number format string to the .z field cellStyles false Save style/theme info to the .s field cellText true Generated formatted text to the .w field cellDates false Store dates as type d (default is n) dateNF If specified, use the string for date code 14 ** sheetStubs false Create cell objects of type z for stub cells sheetRows 0 If >0, read the first sheetRows rows ** bookDeps false If true, parse calculation chains bookFiles false If true, add raw files to book object ** bookProps false If true, only parse enough to get book metadata ** bookSheets false If true, only parse enough to get the sheet names bookVBA false If true, copy VBA blob to vbaraw field ** password “” If defined and file is encrypted, use password ** WTF false If true, throw errors on unexpected file features ** sheets If specified, only parse specified sheets ** PRN false If true, allow parsing of PRN files ** xlfn false If true, preserve _xlfn. prefixes in formulae ** FS DSV Field Separator override
- Even if cellNF is false, formatted text will be generated and saved to .w
- In some cases, sheets may be parsed even if bookSheets is false.
- Excel aggressively tries to interpret values from CSV and other plain text. This leads to surprising behavior! The raw option suppresses value parsing.
- bookSheets and bookProps combine to give both sets of information
- Deps will be an empty object if bookDeps is false
- bookFiles behavior depends on file type:
- keys array (paths in the ZIP) for ZIP-based formats
- files hash (mapping paths to objects representing the files) for ZIP
- cfb object for formats using CFB containers
- sheetRows-1 rows will be generated when looking at the JSON object output (since the header row is counted as a row when parsing the data)
- By default all worksheets are parsed. sheets restricts based on input type:
- number: zero-based index of worksheet to parse (0 is first worksheet)
- string: name of worksheet to parse (case insensitive)
- array of numbers and strings to select multiple worksheets.
- bookVBA merely exposes the raw VBA CFB object. It does not parse the data. XLSM and XLSB store the VBA CFB object in xl/vbaProject.bin. BIFF8 XLS mixes the VBA entries alongside the core Workbook entry, so the library generates a new XLSB-compatible blob from the XLS CFB container.
- codepage is applied to BIFF2 – BIFF5 files without CodePage records and to CSV files without BOM in type:”binary”. BIFF8 XLS always defaults to 1200.
- PRN affects parsing of text files without a common delimiter character.
- Currently only XOR encryption is supported. Unsupported error will be thrown for files employing other encryption methods.
- Newer Excel functions are serialized with the _xlfn. prefix, hidden from the user. SheetJS will strip _xlfn. normally. The xlfn option preserves them.
- WTF is mainly for development. By default, the parser will suppress read errors on single worksheets, allowing you to read from the worksheets that do parse properly. Setting WTF:true forces those errors to be thrown.
Input Type
Strings can be interpreted in multiple ways. The type parameter for read tells the library how to parse the data argument:
type expected input “base64” string: Base64 encoding of the file “binary” string: binary string (byte n is data.charCodeAt(n)) “string” string: JS string (characters interpreted as UTF8) “buffer” nodejs Buffer “array” array: array of 8-bit unsigned int (byte n is data[n]) “file” string: path of file that will be read (nodejs only)
Guessing File Type
Implementation Details (click to show)
Excel and other spreadsheet tools read the first few bytes and apply other heuristics to determine a file type. This enables file type punning: renaming files with the .xls extension will tell your computer to use Excel to open the file but Excel will know how to handle it. This library applies similar logic:
Byte 0 Raw File Type Spreadsheet Types 0xD0 CFB Container BIFF 5/8 or protected XLSX/XLSB or WQ3/QPW or XLR 0x09 BIFF Stream BIFF 2/3/4/5 0x3C XML/HTML SpreadsheetML / Flat ODS / UOS1 / HTML / plain text 0x50 ZIP Archive XLSB or XLSX/M or ODS or UOS2 or NUMBERS or text 0x49 Plain Text SYLK or plain text 0x54 Plain Text DIF or plain text 0xEF UTF8 Encoded SpreadsheetML / Flat ODS / UOS1 / HTML / plain text 0xFF UTF16 Encoded SpreadsheetML / Flat ODS / UOS1 / HTML / plain text 0x00 Record Stream Lotus WK* or Quattro Pro or plain text 0x7B Plain text RTF or plain text 0x0A Plain text SpreadsheetML / Flat ODS / UOS1 / HTML / plain text 0x0D Plain text SpreadsheetML / Flat ODS / UOS1 / HTML / plain text 0x20 Plain text SpreadsheetML / Flat ODS / UOS1 / HTML / plain text
DBF files are detected based on the first byte as well as the third and fourth bytes (corresponding to month and day of the file date)
Works for Windows files are detected based on the BOF record with type 0xFF
Plain text format guessing follows the priority order:
Format Test XML <?xml appears in the first 1024 characters HTML starts with < and HTML tags appear in the first 1024 characters * XML starts with < and the first tag is valid RTF starts with {rt DSV starts with /sep=.$/, separator is the specified character DSV more unquoted ` DSV more unquoted ; chars than t or , in the first 1024 TSV more unquoted t chars than , chars in the first 1024 CSV one of the first 1024 characters is a comma “,” ETH starts with socialcalc:version: PRN PRN option is set to true CSV (fallback)
- HTML tags include: html, table, head, meta, script, style, div
Why are random text files valid? (click to show)
Excel is extremely aggressive in reading files. Adding an XLS extension to any display text file (where the only characters are ANSI display chars) tricks Excel into thinking that the file is potentially a CSV or TSV file, even if it is only one column! This library attempts to replicate that behavior.
The best approach is to validate the desired worksheet and ensure it has the expected number of rows or columns. Extracting the range is extremely simple:
Writing Options
The exported write and writeFile functions accept an options argument:
Option Name Default Description type Output data encoding (see Output Type below) cellDates false Store dates as type d (default is n) bookSST false Generate Shared String Table ** bookType “xlsx” Type of Workbook (see below for supported formats) sheet “” Name of Worksheet for single-sheet formats ** compression false Use ZIP compression for ZIP-based formats ** Props Override workbook properties when writing ** themeXLSX Override theme XML when writing XLSX/XLSB/XLSM ** ignoreEC true Suppress “number as text” errors ** numbers Payload for NUMBERS export **
- bookSST is slower and more memory intensive, but has better compatibility with older versions of iOS Numbers
- The raw data is the only thing guaranteed to be saved. Features not described in this README may not be serialized.
- cellDates only applies to XLSX output and is not guaranteed to work with third-party readers. Excel itself does not usually write cells with type d so non-Excel tools may ignore the data or error in the presence of dates.
- Props is an object mirroring the workbook Props field. See the table from the Workbook File Properties section.
- if specified, the string from themeXLSX will be saved as the primary theme for XLSX/XLSB/XLSM files (to xl/theme/theme1.xml in the ZIP)
- Due to a bug in the program, some features like “Text to Columns” will crash Excel on worksheets where error conditions are ignored. The writer will mark files to ignore the error by default. Set ignoreEC to false to suppress.
- Due to the size of the data, the NUMBERS data is not included by default. The included xlsx.zahl.js and xlsx.zahl.mjs scripts include the data.
Supported Output Formats
For broad compatibility with third-party tools, this library supports many output formats. The specific file type is controlled with bookType option:
bookType file ext container sheets Description xlsx .xlsx ZIP multi Excel 2007+ XML Format xlsm .xlsm ZIP multi Excel 2007+ Macro XML Format xlsb .xlsb ZIP multi Excel 2007+ Binary Format biff8 .xls CFB multi Excel 97-2004 Workbook Format biff5 .xls CFB multi Excel 5.0/95 Workbook Format biff4 .xls none single Excel 4.0 Worksheet Format biff3 .xls none single Excel 3.0 Worksheet Format biff2 .xls none single Excel 2.0 Worksheet Format xlml .xls none multi Excel 2003-2004 (SpreadsheetML) numbers .numbers ZIP single Numbers 3.0+ Spreadsheet ods .ods ZIP multi OpenDocument Spreadsheet fods .fods none multi Flat OpenDocument Spreadsheet wk3 .wk3 none multi Lotus Workbook (WK3) csv .csv none single Comma Separated Values txt .txt none single UTF-16 Unicode Text (TXT) sylk .sylk none single Symbolic Link (SYLK) html .html none single HTML Document dif .dif none single Data Interchange Format (DIF) dbf .dbf none single dBASE II + VFP Extensions (DBF) wk1 .wk1 none single Lotus Worksheet (WK1) rtf .rtf none single Rich Text Format (RTF) prn .prn none single Lotus Formatted Text eth .eth none single Ethercalc Record Format (ETH)
- compression only applies to formats with ZIP containers.
- Formats that only support a single sheet require a sheet option specifying the worksheet. If the string is empty, the first worksheet is used.
- writeFile will automatically guess the output file format based on the file extension if bookType is not specified. It will choose the first format in the aforementioned table that matches the extension.
Output Type
The type argument for write mirrors the type argument for read:
type output “base64” string: Base64 encoding of the file “binary” string: binary string (byte n is data.charCodeAt(n)) “string” string: JS string (characters interpreted as UTF8) “buffer” nodejs Buffer “array” ArrayBuffer, fallback array of 8-bit unsigned int “file” string: path of file that will be created (nodejs only)
- For compatibility with Excel, csv output will always include the UTF-8 byte order mark.
Utility Functions
The sheet_to_* functions accept a worksheet and an optional options object.
The *_to_sheet functions accept a data object and an optional options object.
The examples are based on the following worksheet:
XXX| A | B | C | D | E | F | G | -+-+-+-+-+-+-+-+ 1 | S | h | e | e | t | J | S | 2 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 3 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
Array of Arrays Input
XLSX.utils.aoa_to_sheet takes an array of arrays of JS values and returns a worksheet resembling the input data. Numbers, Booleans and Strings are stored as the corresponding styles. Dates are stored as date or numbers. Array holes and explicit undefined values are skipped. null values may be stubbed. All other values are stored as strings. The function takes an options argument:
Option Name Default Description dateNF FMT 14 Use specified date format in string output cellDates false Store dates as type d (default is n) sheetStubs false Create cell objects of type z for null values nullError false If true, emit #NULL! error cells for null values Examples (click to show)
To generate the example sheet:
XLSX.utils.sheet_add_aoa takes an array of arrays of JS values and updates an existing worksheet object. It follows the same process as aoa_to_sheet and accepts an options argument:
Option Name Default Description dateNF FMT 14 Use specified date format in string output cellDates false Store dates as type d (default is n) sheetStubs false Create cell objects of type z for null values nullError false If true, emit #NULL! error cells for null values origin Use specified cell as starting point (see below)
origin is expected to be one of:
origin Description (cell object) Use specified cell (cell object) (string) Use specified cell (A1-style cell) (number >= 0) Start from the first column at specified row (0-indexed) -1 Append to bottom of worksheet starting on first column (default) Start from cell A1 Examples (click to show)
Consider the worksheet:
XXX| A | B | C | D | E | F | G | -+-+-+-+-+-+-+-+ 1 | S | h | e | e | t | J | S | 2 | 1 | 2 | | | 5 | 6 | 7 | 3 | 2 | 3 | | | 6 | 7 | 8 | 4 | 3 | 4 | | | 7 | 8 | 9 | 5 | 4 | 5 | 6 | 7 | 8 | 9 | 0 |
This worksheet can be built up in the order A1:G1, A2:B4, E2:G4, A5:G5:
Array of Objects Input
XLSX.utils.json_to_sheet takes an array of objects and returns a worksheet with automatically-generated “headers” based on the keys of the objects. The default column order is determined by the first appearance of the field using Object.keys. The function accepts an options argument:
Option Name Default Description header Use specified field order (default Object.keys) ** dateNF FMT 14 Use specified date format in string output cellDates false Store dates as type d (default is n) skipHeader false If true, do not include header row in output nullError false If true, emit #NULL! error cells for null values
- All fields from each row will be written. If header is an array and it does not contain a particular field, the key will be appended to the array.
- Cell types are deduced from the type of each value. For example, a Date object will generate a Date cell, while a string will generate a Text cell.
- Null values will be skipped by default. If nullError is true, an error cell corresponding to #NULL! will be written to the worksheet.
Examples (click to show)
The original sheet cannot be reproduced using plain objects since JS object keys must be unique. After replacing the second e and S with e_1 and S_1:
Alternatively, the header row can be skipped:
XLSX.utils.sheet_add_json takes an array of objects and updates an existing worksheet object. It follows the same process as json_to_sheet and accepts an options argument:
Option Name Default Description header Use specified column order (default Object.keys) dateNF FMT 14 Use specified date format in string output cellDates false Store dates as type d (default is n) skipHeader false If true, do not include header row in output nullError false If true, emit #NULL! error cells for null values origin Use specified cell as starting point (see below)
origin is expected to be one of:
origin Description (cell object) Use specified cell (cell object) (string) Use specified cell (A1-style cell) (number >= 0) Start from the first column at specified row (0-indexed) -1 Append to bottom of worksheet starting on first column (default) Start from cell A1 Examples (click to show)
Consider the worksheet:
XXX| A | B | C | D | E | F | G | -+-+-+-+-+-+-+-+ 1 | S | h | e | e | t | J | S | 2 | 1 | 2 | | | 5 | 6 | 7 | 3 | 2 | 3 | | | 6 | 7 | 8 | 4 | 3 | 4 | | | 7 | 8 | 9 | 5 | 4 | 5 | 6 | 7 | 8 | 9 | 0 |
This worksheet can be built up in the order A1:G1, A2:B4, E2:G4, A5:G5:
HTML Table Input
XLSX.utils.table_to_sheet takes a table DOM element and returns a worksheet resembling the input table. Numbers are parsed. All other data will be stored as strings.
XLSX.utils.table_to_book produces a minimal workbook based on the worksheet.
Both functions accept options arguments:
Option Name Default Description raw If true, every cell will hold raw strings dateNF FMT 14 Use specified date format in string output cellDates false Store dates as type d (default is n) sheetRows 0 If >0, read the first sheetRows rows of the table display false If true, hidden rows and cells will not be parsed Examples (click to show)
To generate the example sheet, start with the HTML table:
To process the table:
Note: XLSX.read can handle HTML represented as strings.
XLSX.utils.sheet_add_dom takes a table DOM element and updates an existing worksheet object. It follows the same process as table_to_sheet and accepts an options argument:
Option Name Default Description raw If true, every cell will hold raw strings dateNF FMT 14 Use specified date format in string output cellDates false Store dates as type d (default is n) sheetRows 0 If >0, read the first sheetRows rows of the table display false If true, hidden rows and cells will not be parsed
origin is expected to be one of:
origin Description (cell object) Use specified cell (cell object) (string) Use specified cell (A1-style cell) (number >= 0) Start from the first column at specified row (0-indexed) -1 Append to bottom of worksheet starting on first column (default) Start from cell A1 Examples (click to show)
A small helper function can create gap rows between tables:
Formulae Output
XLSX.utils.sheet_to_formulae generates an array of commands that represent how a person would enter data into an application. Each entry is of the form A1-cell-address=formula-or-value. String literals are prefixed with a ‘ in accordance with Excel.
Examples (click to show)
For the example sheet:
Delimiter-Separated Output
As an alternative to the writeFile CSV type, XLSX.utils.sheet_to_csv also produces CSV output. The function takes an options argument:
Option Name Default Description FS “,” “Field Separator” delimiter between fields RS “n” “Record Separator” delimiter between rows dateNF FMT 14 Use specified date format in string output strip false Remove trailing field separators in each record ** blankrows true Include blank lines in the CSV output skipHidden false Skips hidden rows/columns in the CSV output forceQuotes false Force quotes around fields
- strip will remove trailing commas from each line under default FS/RS
- blankrows must be set to false to skip blank lines.
- Fields containing the record or field separator will automatically be wrapped in double quotes; forceQuotes forces all cells to be wrapped in quotes.
- XLSX.write with csv type will always prepend the UTF-8 byte-order mark for Excel compatibility. sheet_to_csv returns a JS string and omits the mark. Using XLSX.write with type string will also skip the mark.
Examples (click to show)
For the example sheet:
UTF-16 Unicode Text
The txt output type uses the tab character as the field separator. If the codepage library is available (included in full distribution but not core), the output will be encoded in CP1200 and the BOM will be prepended.
XLSX.utils.sheet_to_txt takes the same arguments as sheet_to_csv.
HTML Output
As an alternative to the writeFile HTML type, XLSX.utils.sheet_to_html also produces HTML output. The function takes an options argument:
Option Name Default Description id Specify the id attribute for the TABLE element editable false If true, set contenteditable=”true” for every TD header Override header (default html body) footer Override footer (default /body /html) Examples (click to show)
For the example sheet:
JSON
XLSX.utils.sheet_to_json generates different types of JS objects. The function takes an options argument:
Option Name Default Description raw true Use raw values (true) or formatted strings (false) range from WS Override Range (see table below) header Control output format (see table below) dateNF FMT 14 Use specified date format in string output defval Use specified value in place of null or undefined blankrows ** Include blank lines in the output **
- raw only affects cells which have a format code (.z) field or a formatted text (.w) field.
- If header is specified, the first row is considered a data row; if header is not specified, the first row is the header row and not considered data.
- When header is not specified, the conversion will automatically disambiguate header entries by affixing _ and a count starting at 1. For example, if three columns have header foo the output fields are foo, foo_1, foo_2
- null values are returned when raw is true but are skipped when false.
- If defval is not specified, null and undefined values are skipped normally. If specified, all null and undefined points will be filled with defval
- When header is 1, the default is to generate blank rows. blankrows must be set to false to skip blank rows.
- When header is not 1, the default is to skip blank rows. blankrows must be true to generate blank rows
range is expected to be one of:
range Description (number) Use worksheet range but set starting row to the value (string) Use specified range (A1-style bounded range string) (default) Use worksheet range (ws[‘!ref’])
header is expected to be one of:
header Description 1 Generate an array of arrays (“2D Array”) “A” Row object keys are literal column labels array of strings Use specified strings as keys in row objects (default) Read and disambiguate first row as keys
- If header is not 1, the row object will contain the non-enumerable property __rowNum__ that represents the row of the sheet corresponding to the entry.
- If header is an array, the keys will not be disambiguated. This can lead to unexpected results if the array values are not unique!
Examples (click to show)
For the example sheet:
Example showing the effect of raw:
File Formats
Despite the library name xlsx, it supports numerous spreadsheet file formats:
Format Read Write Excel Worksheet/Workbook Formats :-: :-: Excel 2007+ XML Formats (XLSX/XLSM) ✔ ✔ Excel 2007+ Binary Format (XLSB BIFF12) ✔ ✔ Excel 2003-2004 XML Format (XML “SpreadsheetML”) ✔ ✔ Excel 97-2004 (XLS BIFF8) ✔ ✔ Excel 5.0/95 (XLS BIFF5) ✔ ✔ Excel 4.0 (XLS/XLW BIFF4) ✔ ✔ Excel 3.0 (XLS BIFF3) ✔ ✔ Excel 2.0/2.1 (XLS BIFF2) ✔ ✔ Excel Supported Text Formats :-: :-: Delimiter-Separated Values (CSV/TXT) ✔ ✔ Data Interchange Format (DIF) ✔ ✔ Symbolic Link (SYLK/SLK) ✔ ✔ Lotus Formatted Text (PRN) ✔ ✔ UTF-16 Unicode Text (TXT) ✔ ✔ Other Workbook/Worksheet Formats :-: :-: Numbers 3.0+ / iWork 2013+ Spreadsheet (NUMBERS) ✔ ✔ OpenDocument Spreadsheet (ODS) ✔ ✔ Flat XML ODF Spreadsheet (FODS) ✔ ✔ Uniform Office Format Spreadsheet (标文通 UOS1/UOS2) ✔ dBASE II/III/IV / Visual FoxPro (DBF) ✔ ✔ Lotus 1-2-3 (WK1/WK3) ✔ ✔ Lotus 1-2-3 (WKS/WK2/WK4/123) ✔ Quattro Pro Spreadsheet (WQ1/WQ2/WB1/WB2/WB3/QPW) ✔ Works 1.x-3.x DOS / 2.x-5.x Windows Spreadsheet (WKS) ✔ Works 6.x-9.x Spreadsheet (XLR) ✔ Other Common Spreadsheet Output Formats :-: :-: HTML Tables ✔ ✔ Rich Text Format tables (RTF) ✔ Ethercalc Record Format (ETH) ✔ ✔
Features not supported by a given file format will not be written. Formats with range limits will be silently truncated:
Format Last Cell Max Cols Max Rows Excel 2007+ XML Formats (XLSX/XLSM) XFD1048576 16384 1048576 Excel 2007+ Binary Format (XLSB BIFF12) XFD1048576 16384 1048576 Excel 97-2004 (XLS BIFF8) IV65536 256 65536 Excel 5.0/95 (XLS BIFF5) IV16384 256 16384 Excel 4.0 (XLS BIFF4) IV16384 256 16384 Excel 3.0 (XLS BIFF3) IV16384 256 16384 Excel 2.0/2.1 (XLS BIFF2) IV16384 256 16384 Lotus 1-2-3 R2 – R5 (WK1/WK3/WK4) IV8192 256 8192 Lotus 1-2-3 R1 (WKS) IV2048 256 2048
Excel 2003 SpreadsheetML range limits are governed by the version of Excel and are not enforced by the writer.
File Format Details (click to show)
Core Spreadsheet Formats
- Excel 2007+ XML (XLSX/XLSM)
XLSX and XLSM files are ZIP containers containing a series of XML files in accordance with the Open Packaging Conventions (OPC). The XLSM format, almost identical to XLSX, is used for files containing macros.
The format is standardized in ECMA-376 and later in ISO/IEC 29500. Excel does not follow the specification, and there are additional documents discussing how Excel deviates from the specification.
- Excel 2.0-95 (BIFF2/BIFF3/BIFF4/BIFF5)
BIFF 2/3 XLS are single-sheet streams of binary records. Excel 4 introduced the concept of a workbook (XLW files) but also had single-sheet XLS format. The structure is largely similar to the Lotus 1-2-3 file formats. BIFF5/8/12 extended the format in various ways but largely stuck to the same record format.
There is no official specification for any of these formats. Excel 95 can write files in these formats, so record lengths and fields were determined by writing in all of the supported formats and comparing files. Excel 2016 can generate BIFF5 files, enabling a full suite of file tests starting from XLSX or BIFF2.
- Excel 97-2004 Binary (BIFF8)
BIFF8 exclusively uses the Compound File Binary container format, splitting some content into streams within the file. At its core, it still uses an extended version of the binary record format from older versions of BIFF.
The MS-XLS specification covers the basics of the file format, and other specifications expand on serialization of features like properties.
- Excel 2003-2004 (SpreadsheetML)
Predating XLSX, SpreadsheetML files are simple XML files. There is no official and comprehensive specification, although MS has released documentation on the format. Since Excel 2016 can generate SpreadsheetML files, mapping features is pretty straightforward.
- Excel 2007+ Binary (XLSB, BIFF12)
Introduced in parallel with XLSX, the XLSB format combines the BIFF architecture with the content separation and ZIP container of XLSX. For the most part nodes in an XLSX sub-file can be mapped to XLSB records in a corresponding sub-file.
The MS-XLSB specification covers the basics of the file format, and other specifications expand on serialization of features like properties.
- Delimiter-Separated Values (CSV/TXT)
Excel CSV deviates from RFC4180 in a number of important ways. The generated CSV files should generally work in Excel although they may not work in RFC4180 compatible readers. The parser should generally understand Excel CSV. The writer proactively generates cells for formulae if values are unavailable.
Excel TXT uses tab as the delimiter and code page 1200.
Like in Excel, files starting with 0x49 0x44 (“ID”) are treated as Symbolic Link files. Unlike Excel, if the file does not have a valid SYLK header, it will be proactively reinterpreted as CSV. There are some files with semicolon delimiter that align with a valid SYLK file. For the broadest compatibility, all cells with the value of ID are automatically wrapped in double-quotes.
Miscellaneous Workbook Formats
Support for other formats is generally far behind XLS/XLSB/XLSX support, due in part to a lack of publicly available documentation. Test files were produced in the respective apps and compared to their XLS exports to determine structure. The main focus is data extraction.
- Lotus 1-2-3 (WKS/WK1/WK2/WK3/WK4/123)
The Lotus formats consist of binary records similar to the BIFF structure. Lotus did release a specification decades ago covering the original WK1 format. Other features were deduced by producing files and comparing to Excel support.
Generated WK1 worksheets are compatible with Lotus 1-2-3 R2 and Excel 5.0.
Generated WK3 workbooks are compatible with Lotus 1-2-3 R9 and Excel 5.0.
- Quattro Pro (WQ1/WQ2/WB1/WB2/WB3/QPW)
The Quattro Pro formats use binary records in the same way as BIFF and Lotus. Some of the newer formats (namely WB3 and QPW) use a CFB enclosure just like BIFF8 XLS.
- Works for DOS / Windows Spreadsheet (WKS/XLR)
All versions of Works were limited to a single worksheet.
Works for DOS 1.x – 3.x and Works for Windows 2.x extends the Lotus WKS format with additional record types.
Works for Windows 3.x – 5.x uses the same format and WKS extension. The BOF record has type FF
Works for Windows 6.x – 9.x use the XLR format. XLR is nearly identical to BIFF8 XLS: it uses the CFB container with a Workbook stream. Works 9 saves the exact Workbook stream for the XLR and the 97-2003 XLS export. Works 6 XLS includes two empty worksheets but the main worksheet has an identical encoding. XLR also includes a WksSSWorkBook stream similar to Lotus FM3/FMT files.
- Numbers 3.0+ / iWork 2013+ Spreadsheet (NUMBERS)
iWork 2013 (Numbers 3.0 / Pages 5.0 / Keynote 6.0) switched from a proprietary XML-based format to the current file format based on the iWork Archive (IWA). This format has been used up through the current release (Numbers 11.2).
The parser focuses on extracting raw data from tables. Numbers technically supports multiple tables in a logical worksheet, including custom titles. This parser will generate one worksheet per Numbers table.
The writer currently exports a small range from the first worksheet.
- OpenDocument Spreadsheet (ODS/FODS)
ODS is an XML-in-ZIP format akin to XLSX while FODS is an XML format akin to SpreadsheetML. Both are detailed in the OASIS standard, but tools like LO/OO add undocumented extensions. The parsers and writers do not implement the full standard, instead focusing on parts necessary to extract and store raw data.
- Uniform Office Spreadsheet (UOS1/2)
UOS is a very similar format, and it comes in 2 varieties corresponding to ODS and FODS respectively. For the most part, the difference between the formats is in the names of tags and attributes.
Miscellaneous Worksheet Formats
Many older formats supported only one worksheet:
- dBASE and Visual FoxPro (DBF)
DBF is really a typed table format: each column can only hold one data type and each record omits type information. The parser generates a header row and inserts records starting at the second row of the worksheet. The writer makes files compatible with Visual FoxPro extensions.
Multi-file extensions like external memos and tables are currently unsupported, limited by the general ability to read arbitrary files in the web browser. The reader understands DBF Level 7 extensions like DATETIME.
- Symbolic Link (SYLK)
There is no real documentation. All knowledge was gathered by saving files in various versions of Excel to deduce the meaning of fields. Notes:
-
Plain formulae are stored in the RC form.
-
Column widths are rounded to integral characters.
-
Lotus Formatted Text (PRN)
There is no real documentation, and in fact Excel treats PRN as an output-only file format. Nevertheless we can guess the column widths and reverse-engineer the original layout. Excel’s 240 character width limitation is not enforced.
- Data Interchange Format (DIF)
There is no unified definition. Visicalc DIF differs from Lotus DIF, and both differ from Excel DIF. Where ambiguous, the parser/writer follows the expected behavior from Excel. In particular, Excel extends DIF in incompatible ways:
-
Since Excel automatically converts numbers-as-strings to numbers, numeric string constants are converted to formulae: “0.3” -> “=””0.3″”
-
DIF technically expects numeric cells to hold the raw numeric data, but Excel permits formatted numbers (including dates)
-
DIF technically has no support for formulae, but Excel will automatically convert plain formulae. Array formulae are not preserved.
-
HTML
Excel HTML worksheets include special metadata encoded in styles. For example, mso-number-format is a localized string containing the number format. Despite the metadata the output is valid HTML, although it does accept bare & symbols.
The writer adds type metadata to the TD elements via the t tag. The parser looks for those tags and overrides the default interpretation. For example, text like <td>12345</td> will be parsed as numbers but <td t=”s”>12345</td> will be parsed as text.
- Rich Text Format (RTF)
Excel RTF worksheets are stored in clipboard when copying cells or ranges from a worksheet. The supported codes are a subset of the Word RTF support.
- Ethercalc Record Format (ETH)
Ethercalc is an open source web spreadsheet powered by a record format reminiscent of SYLK wrapped in a MIME multi-part message.
Testing
Node
(click to show)
make test will run the node-based tests. By default it runs tests on files in every supported format. To test a specific file type, set FMTS to the format you want to test. Feature-specific tests are available with make test_misc
To enable all errors, set the environment variable WTF=1:
flow and eslint checks are available:
Browser
(click to show)
The core in-browser tests are available at tests/index.html within this repo. Start a local server and navigate to that directory to run the tests. make ctestserv will start a server on port 8000.
make ctest will generate the browser fixtures. To add more files, edit the tests/fixtures.lst file and add the paths.
To run the full in-browser tests, clone the repo for oss.sheetjs.com and replace the xlsx.js file (then open a browser window and go to stress.html):
Tested Environments
(click to show)
- NodeJS 0.8, 0.10, 0.12, 4.x, 5.x, 6.x, 7.x, 8.x
- IE 6/7/8/9/10/11 (IE 6-9 require shims)
- Chrome 24+ (including Android 4.0+)
- Safari 6+ (iOS and Desktop)
- Edge 13+, FF 18+, and Opera 12+
Tests utilize the mocha testing framework.
- https://saucelabs.com/u/sheetjs for XLS* modules using Sauce Labs
The test suite also includes tests for various time zones. To change the timezone locally, set the TZ environment variable:
Test Files
Test files are housed in another repo.
Running make init will refresh the test_files submodule and get the files. Note that this requires svn, git, hg and other commands that may not be available. If make init fails, please download the latest version of the test files snapshot from the repo
Latest Snapshot (click to show)
Latest test files snapshot: http://github.com/SheetJS/test_files/releases/download/20170409/test_files.zip
(download and unzip to the test_files subdirectory)
Contributing
Due to the precarious nature of the Open Specifications Promise, it is very important to ensure code is cleanroom. Contribution Notes
File organization (click to show)
At a high level, the final script is a concatenation of the individual files in the bits folder. Running make should reproduce the final output on all platforms. The README is similarly split into bits in the docbits folder.
Folders:
folder contents bits raw source files that make up the final script docbits raw markdown files that make up README.md bin server-side bin scripts (xlsx.njs) dist dist files for web browsers and nonstandard JS environments demos demo projects for platforms like ExtendScript and Webpack tests browser tests (run make ctest to rebuild) types typescript definitions and tests misc miscellaneous supporting scripts test_files test files (pulled from the test files repository)
After cloning the repo, running make help will display a list of commands.
OSX/Linux
(click to show)
The xlsx.js file is constructed from the files in the bits subdirectory. The build script (run make) will concatenate the individual bits to produce the script. Before submitting a contribution, ensure that running make will produce the xlsx.js file exactly. The simplest way to test is to add the script:
To produce the dist files, run make dist. The dist files are updated in each version release and should not be committed between versions.
Windows
(click to show)
The included make.cmd script will build xlsx.js from the bits directory. Building is as simple as:
To prepare development environment:
The full list of commands available in Windows are displayed in make help:
make init – install deps and global modules make lint – run eslint linter make test – run mocha test suite make misc – run smaller test suite make book – rebuild README and summary make help – display this message
As explained in Test Files, on Windows the release ZIP file must be downloaded and extracted. If Bash on Windows is available, it is possible to run the OSX/Linux workflow. The following steps prepares the environment:
Tests
(click to show)
The test_misc target (make test_misc on Linux/OSX / make misc on Windows) runs the targeted feature tests. It should take 5-10 seconds to perform feature tests without testing against the entire test battery. New features should be accompanied with tests for the relevant file formats and features.
For tests involving the read side, an appropriate feature test would involve reading an existing file and checking the resulting workbook object. If a parameter is involved, files should be read with different values to verify that the feature is working as expected.
For tests involving a new write feature which can already be parsed, appropriate feature tests would involve writing a workbook with the feature and then opening and verifying that the feature is preserved.
For tests involving a new write feature without an existing read ability, please add a feature test to the kitchen sink tests/write.js.
License
Please consult the attached LICENSE file for details. All rights not explicitly granted by the Apache 2.0 License are reserved by the Original Author.
References
OSP-covered Specifications (click to show)
- MS-CFB: Compound File Binary File Format
- MS-CTXLS: Excel Custom Toolbar Binary File Format
- MS-EXSPXML3: Excel Calculation Version 2 Web Service XML Schema
- MS-ODATA: Open Data Protocol (OData)
- MS-ODRAW: Office Drawing Binary File Format
- MS-ODRAWXML: Office Drawing Extensions to Office Open XML Structure
- MS-OE376: Office Implementation Information for ECMA-376 Standards Support
- MS-OFFCRYPTO: Office Document Cryptography Structure
- MS-OI29500: Office Implementation Information for ISO/IEC 29500 Standards Support
- MS-OLEDS: Object Linking and Embedding (OLE) Data Structures
- MS-OLEPS: Object Linking and Embedding (OLE) Property Set Data Structures
- MS-OODF3: Office Implementation Information for ODF 1.2 Standards Support
- MS-OSHARED: Office Common Data Types and Objects Structures
- MS-OVBA: Office VBA File Format Structure
- MS-XLDM: Spreadsheet Data Model File Format
- MS-XLS: Excel Binary File Format (.xls) Structure Specification
- MS-XLSB: Excel (.xlsb) Binary File Format
- MS-XLSX: Excel (.xlsx) Extensions to the Office Open XML SpreadsheetML File Format
- XLS: Microsoft Office Excel 97-2007 Binary File Format Specification
- RTF: Rich Text Format
- ISO/IEC 29500:2012(E) “Information technology — Document description and processing languages — Office Open XML File Formats”
- Open Document Format for Office Applications Version 1.2 (29 September 2011)
- Worksheet File Format (From Lotus) December 1984