Importing notes

Import filesCopy Anchor Link

The import button beside New folder turns files into notes, one note per file. You can also drop files or whole folders onto the notes list. A dropped folder becomes a note folder with the same name, and any folders inside it become a prefix of each note's title, such as Week 2 / slides.

WorkRoad reads PDF, Word (DOCX), PowerPoint (PPTX), Excel (XLSX), OpenDocument text, spreadsheet and presentation files, RTF, EPUB, Markdown, HTML, CSV, LaTeX, Jupyter notebooks, plain text, and source code. JPEG, PNG, WebP and GIF images each become a note that shows the image. Each file can be up to 50 MiB.

Headings, lists, tables, code blocks and maths become the same blocks in the note, as far as the file records them. A PDF or slide deck with more than one page gets a Page 3 or Slide 3 heading before each page. Markdown is read as written; HTML inside it shows as text.

Import limitsCopy Anchor Link

An import brings in the first 200,000 characters of a file's text, or less if the text is mostly very short list items or styled spans, and Import notes says where it stopped. The rest stays readable in the original file, which stays attached to the first note. Text too long for one note is split into numbered notes, such as Report (1/2), in the same folder, at headings or pages where it can. A table, list or code block too long for one note continues in the next, with a table's header row repeated. A single table row or list item too long for any note keeps its text as plain paragraphs. Images stored inside a Markdown or HTML file's text, as data: URLs, are left out and reported.

Missing text or imagesCopy Anchor Link

When the importer knows it lost something, such as an image it could not read, a LaTeX command it does not know, or a notebook output it cannot show, the last note ends with an Import notes section that says what. Charts, drawings, embedded objects, and a web page's frames and videos are listed there too. The PDF and Office reader does not report everything else it drops, such as text boxes or a complex layout, so check the original for those.

HEIC photos, Keynote, Pages and Numbers files, OpenDocument drawings (ODG), and the old .doc, .ppt and .xls formats can't be read. Export them to JPEG, PDF or the newer Office formats first.

For scanned pages, unreadable words and partial scan failures, see Scans, photos and handwriting.

Scans, photos and handwritingCopy Anchor Link

With Plus, WorkRoad reads the words in images and scanned PDFs. An imported photo of notes, a whiteboard or a printed page becomes a note with the image at the top and its text, as headings, lists, tables and maths, underneath. In a PDF, the reader takes only the scanned pages, the ones with no text of their own, and puts each page's words under its Page N heading, or in the note itself for a one-page scan. It reads up to 50 scanned pages per file and lists the rest in Import notes. The reader writes [illegible] for a word it cannot make out. A page it cannot read at all, such as one whose scan is over 60 megapixels, gets a placeholder in the note and a line in Import notes. If a scan's pages cannot be read, the rest of the file still imports. Each read counts as one AI operation, and a file can be read at most 3 times.

Files already attached to a note can be read too. A scanned PDF shows Read the pages in its file list, and a PDF with some scanned pages shows Read N scanned pages. An image shows Read text, then Insert text to put its words at the cursor.

Without Plus, an imported image becomes a note that shows it, and a PDF with no text at all is refused with a message saying so. A PDF with some scanned pages imports the text it has and lists the scanned pages in Import notes.

Notes from other appsCopy Anchor Link

An export from another note app imports with the same button and becomes a folder of notes. In Notion, Obsidian and Markdown files, links between notes become mentions, so each note's backlinks show where it is linked from. The images the notes show stay in them, each as an image file of its note, so they count toward your file storage like any upload. Once storage is full, the rest of the images are left out of their notes and counted in Import notes. The import's own note lists the new notes and keeps the export attached. An export can be up to 200 MiB and bring in up to 1,000 of its notes; a long one splits into numbered notes as any long file does. A zip may hold up to 5,000 files and 1 GiB unpacked. A zip with several folders at its top keeps each as a note folder of its own.

• Notion. Open the page's ••• menu, choose Export, pick Markdown & CSV and include subpages, then import the zip as it is. Page names lose the ID Notion adds to them, a subpage's title starts with its parent's, such as Biology / Cells, and each database becomes a note with its table, up to 500 rows.
• Obsidian. Zip the vault folder and import the zip. [[Note]] and [[Note|alias]] links become mentions, ![[image.png]] embeds become images, and ![[Note]] embeds become links, since notes do not show other notes inside them. The importer omits front matter. A link to a note that is not in the vault stays as text and is counted in Import notes.
• Evernote. Export a notebook as an ENEX file and import it. Each note keeps its title, its creation date and its images, and a checkbox comes through as ☑ or ☐ text. Other attachments are named in the note and stay in the original. Links between Evernote notes stay as they were and do not become mentions.
• OneNote. In OneNote for Windows, choose File, then Export, and export the page or section as a Word document. Import the .docx. On a Mac or the web, print the page to PDF instead.
• Google Docs. Choose File, then Download, then Microsoft Word (.docx), and import that. Web page (.html, zipped) imports too: import the zip as it is, and its page becomes a note without its images.

Any other zip imports like a dropped folder: each file WorkRoad reads becomes a note, and an image no note links to becomes a note that shows it. A document inside a zip brings its text but not its pictures, and a scan inside one is not read; import those files on their own to keep them.