v0.11.0

SVG pictures

• .svg files open as the picture alone: no page around it, scaled down to fit the view and shown at its own size otherwise (width / height in any unit, else the viewBox; 300 × 150 when the file says nothing). From any src and from File → Open; the bar shows an SVG badge. The file is kept as it is (given the SVG namespace when it leaves it out, since an <img> shows nothing without it); its <title> is the picture's alt text. A file with no <svg> element says so.
• The Code button shows the markup beside the picture; editing it redraws the picture as you type, and Save writes the .svg. Download as gives a Word document holding the picture, a PDF, or the picture itself.
• .svgz, and gzip-compressed plain files in general (.csv.gz, .md.gz…), are unpacked in the worker (native DecompressionStream) and opened as the file inside, also from File → Open and Import CSV into new sheet. A gzip file of another kind (.tar.gz) gets a message saying so. A .svgz the server already unpacked (Content-Encoding: gzip) opens too.

Markdown and HTML saved as themselves

• A .md or .html file is saved as .md / .html (the name onSave receives too), no longer as a Word document. Unedited, the file is written back as it is (in UTF-8). Edited as a document, the Markdown / HTML is built again from the document: headings, nested lists, bold / italic / underline / strike, links, inline code, colors, tables, pictures (embedded as data URLs; a diagram block becomes its drawing); the text reads with its tracked changes accepted. Download as Word still gives the .docx.
• The source pane and the document never disagree: edit the source and the document is rebuilt (as before); edit the document and the pane shows the source built from it when it is opened again. Diagram files are unchanged: the document around the drawing stays a Word file.
• Opened from File → Open, a Markdown / HTML / SVG file no longer starts as unsaved.
• The Markdown export (Download as Markdown too) keeps inline code and pictures.

CSV

• Measured: 500 million rows (55 GB) open in about 4.5 minutes in Chrome, the tab staying under 1 GB of memory (rows past 12 million cells stay in the file and are read by pages). The CSV / TSV export gathers the rows into Blobs every 8 MB rather than one array of strings, so a sheet of that size exports without filling the heap.

For integrators: doc.info.imported can be 'svg'; onSave(blob, name) receives a .md / .html / .svg file for those formats (.docx / .xlsx otherwise); src and File → Open accept .svg, .svgz and gzip-compressed plain files. No new dependency.

What's Changed

• feat(docx): table formatting, page setup, headers/footers with page numbers, tracked changes by @luongtrieuvy202 in #1

New Contributors

• @luongtrieuvy202 made their first contribution in #1

Full Changelog: v0.1.0...v0.11.0