Supported formats

The viewer and the editor open PDFs, Word, PowerPoint and Excel files and the text formats around them, images, video and audio, and web pages. Each opens as a doc type, which decides the tools the UI offers. Once a document is open, docType says which doc type it opened as, and sourceFormat which format it came from.

Documents

The docMentis engine reads and renders these. The Editing column says whether the editor edits a format and saves the edits back to it:

Format	Extensions	sourceFormat	docType	Editing
PDF	.pdf	pdf	portable	Yes
Word	.docx, .docm, .dotx, .dotm	the extension	wordprocessing	Yes
Plain text	.txt	txt	wordprocessing	Yes
Markdown	.md, .markdown	md	wordprocessing	Yes
HTML	.html, .htm	html	wordprocessing	Yes
Email	.eml	eml	wordprocessing	Yes, but saved as another format, not .eml
PowerPoint	.pptx, .pptm, .potx, .potm, .ppsx, .ppsm	the extension	presentation	Yes
Excel	.xlsx, .xlsm, .xltx, .xltm	the extension	spreadsheet	Yes
Delimited text	.csv, .tsv, .tab, .psv	csv, tsv or psv	spreadsheet	Yes
Images	.jpg, .jpeg, .png, .gif, .bmp, .tif, .tiff, .webp	jpg, png, gif, bmp, tiff or webp	image	Yes
Vector images	.svg, .emf, .wmf	the extension	image	Yes, but an edited one can't be saved

• A template, a slide show or a macro-enabled Office file opens as the document it is a variant of, and sourceFormat names the variant: dotx, ppsx, xlsm.
• A text file opens in Courier New. A text file, a Markdown file (CommonMark, with GitHub's tables and strikethrough) and an HTML file (styled by its <style> rules and style attributes) show pageless, their text flowed to the width of the pane. The View tab shows them as pages instead.
• A .udoc package, which saveToPackage saves, opens with open as the document it holds, its sourceFormat the format it recorded.

Video, audio and web pages

These are extension documents: the browser plays or shows them in the document area, not the engine. Two are built in, and you can add your own.

Kind	Sources	sourceFormat	docType	Editing
Video and audio	.mp4, .m4v, .mov, .webm, .mp3, .m4a, .aac, .wav, .ogg, .oga, .ogv, .opus and .flac files; HLS streams; YouTube and Vimeo videos by their address	mp4, webm, mp3 and the rest; hls, youtube, vimeo	media	No
Web pages	Any URL, given format: "url"	url	webpage	No

How a format is recognized

open tells most formats from the file's content: a PDF, an Office file (and which variant it is), an image, and an HTML file that starts with <!DOCTYPE html> or <html>. A CSV, TSV, PSV, text, Markdown or email file has no signature, so it's recognized by its name: a File's name, the last segment of a URL's path, or the filename option. Video and audio are recognized by a YouTube or Vimeo address, then the file's extension, the media type the server or the Blob declares, and finally the file's first bytes.

Where none of that can work, as with bytes from an API, name the format with the format option. It skips detection entirely:

import { UDocViewer } from "@docmentis/udoc-viewer";

const viewer = await UDocViewer.create(document.getElementById("udoc")!);

// Bytes have no file name to tell a CSV by, so name the format, and the
// character its fields are split by.
const response = await fetch("https://example.com/export");
const bytes = new Uint8Array(await response.arrayBuffer());
await viewer.open(bytes, { format: "csv", delimiter: ";", filename: "export.csv" });

console.log(viewer.docType, viewer.sourceFormat); // "spreadsheet" "csv"

A CSV opened without a delimiter is split by the comma, semicolon, tab or vertical bar that splits its first records evenly. When none does, or more than one does, a dialog shows its first lines and asks the reader which character splits them.

What people can do with each

The viewer edits nothing but a PDF's annotations and form fields: it's for reading. The editor edits every document the engine opens, in its edit mode. Video, audio and web pages open view only in both.

Edit mode needs the editing feature, which free usage includes and a commercial license includes if it lists it. Without it, the editor offers view mode alone, and setMode("edit") does nothing.

In the viewer, and in the editor's view mode:

Doc type	View mode
PDF	Thumbnails, outline, layers, search and comments. Annotate: highlight, underline, strikeout, squiggly, note, text box, stamp, pen and shapes. Fill in forms. Present it a page at a time, with the presenter view.
Word document	Thumbnails and search. Show it as pages or pageless, and show its formatting marks.
Presentation	Thumbnails with their sections and hidden slides, speaker notes, the Notes Page view and search. Present it, with its transitions, animations and the presenter view.
Spreadsheet	Sheet tabs, the formula bar, the Normal and Page Layout views, and search.
Image	Thumbnails of its pages.
Video, audio, web page	The player or the page, and full screen.

In the editor's edit mode:

Doc type	Edit mode
PDF	Edit the text of the page in place. Select, move, restyle (fill and line), order, align, rotate and group the page's objects. Insert shapes. Insert, duplicate, rotate, delete and reorder pages. Annotate.
Word document	Edit the text: fonts and character formatting, alignment, spacing, lists and paragraph styles. Restyle and arrange the drawings in its text. It has no Insert tab yet.
Presentation	Edit the text and the shapes. Insert slides, pictures and any of PowerPoint's preset shapes. Duplicate, delete, move and hide slides, and edit sections and speaker notes. Slide masters show, but aren't edited yet.
Spreadsheet	Type values and formulas into cells. Insert, delete, rename, hide, color and move sheets. Cells can't be formatted yet.
Image	Rotate, flip, crop and resize it, and draw shapes with text over it. Rotate, delete and move the pages of a multi-page image, and duplicate a TIFF's.
Video, audio, web page	Nothing: they open view only.

Saving

The toolbar's Download, and download(), save the document in its sourceFormat. In the viewer, a PDF saves with the annotations and form entries made in it, and any other document downloads as it was opened. In the editor, every document saves with its edits.

saveToBytes(format, strategy) saves to any format of the document's doc type, and rejects any other:

Doc type	Formats it saves to
PDF	pdf
Word document	docx, docm, dotx, dotm, txt, md, html
Presentation	pptx, pptm, potx, potm, ppsx, ppsm
Spreadsheet	xlsx, xlsm, xltx, xltm, csv, tsv, psv
Image	The format it was opened from

The strategy says how much of the original file the saved one keeps:

• "incremental" appends the edits to the original as a new revision, which keeps a PDF's signatures valid. Only a PDF saves this way, and it's a PDF's default.
• "delta" writes the parts the edits changed over the original, and keeps every other part as it was. It's the default for an Office file saved in its own format, and keeps a macro-enabled file's macros.
• "rewrite" writes the whole file afresh. It's the default for the text formats, CSV, TSV, PSV and images, and the only way to save to a format other than the one the document was opened from. It drops a file's macros.

import { UDocEditor } from "@docmentis/udoc-editor";

const editor = await UDocEditor.create(document.getElementById("udoc")!);
await editor.open("https://example.com/notes.md");

// Saved back as Markdown, and as a Word document: a format other than the one
// it opened from is written afresh, by "rewrite".
await upload("notes.md", await editor.saveToBytes("md"));
await upload("notes.docx", await editor.saveToBytes("docx", "rewrite"));

// A rendering of its pages, which doesn't open as the same document again.
await upload("notes.pdf", await editor.exportToBytes("pdf"));
await upload("notes-1.png", await editor.exportPageToBytes(0, { png: { scale: 2 } }));

/** Your own code, which sends the bytes where they go. */
declare function upload(name: string, bytes: Uint8Array): Promise<void>;

Some saves aren't possible yet:

• An email can't be saved as .eml, so the editor's Download fails for one. Save it as a Word document, Markdown, HTML or text instead.
• A CSV, TSV or PSV holds one sheet, so a workbook saved as one keeps its first worksheet only.
• An image saves only as the format it was opened from, and an SVG, EMF or WMF that has been edited can't be saved.

Exporting

exportToBytes("pdf") renders any document the engine opened as a PDF of its pages, and exportPageToBytes renders one page as a PNG or an SVG. A workbook exports every worksheet, on the pages it prints on. An export is a picture of the document as it stands: unlike a save, it doesn't reopen as the same document.

Not supported

These don't open:

• The binary Office formats of Office 97–2003: .doc, .xls and .ppt.
• OpenDocument files (.odt, .ods, .odp) and RTF.
• Word, PowerPoint and Excel files protected by a password. A PDF protected by a password opens: the UI asks for it, or open takes it as its password option.
• HEIC, AVIF, ICO, TGA, PPM, HDR, EXR and QOI images.
• MKV, AVI and WMV video, and DASH streams.

Other limits

• Annotations, comments, form filling, the outline and layers are a PDF's.
• Two pages side by side (layoutMode) is for PDFs and Word documents, and turning the view (pageRotation) for PDFs.
• An image has no search panel.
• A workbook prints its current sheet, on the pages its Page Layout view shows.
• An extension document has no pages: while one is open, pageCount is 0, and the calls that need pages, search, annotations, saving, exporting, downloading, printing and collaboration reject.
• A font a document uses but doesn't embed is fetched from Google Fonts, taken from the reader's own machine, or drawn in a font that stands in for it. A document whose fonts can't be had any of those ways can look different from how it looks in Office.

On a server

@docmentis/udoc-node, the udoc command line and @docmentis/udoc-server open, save and export the same documents as the viewer and the editor. Video, audio and web pages are the browser's, so they don't open there.