DocumentationGetting started
Importing
What is read from a Word or Markdown file, how chapters are detected, and what is reported as lost.
What is read from your file, how chapters are found, and what is reported as lost.
Word documents and Markdown. The file is read in your browser and never uploaded — you will see the panel say so while it works.
What can be imported
.docx from Word, and Markdown (.md, .markdown, .txt).
How chapters are found
This decides whether the import takes a minute or an afternoon.
If you used Word’s heading styles, they are read directly: Heading 1 is a chapter, Heading 2 is a section inside it, Title is the book’s title.
If you did not, the words are read instead. A line on its own reading Chapter Four,
Chapitre 4, Part Two, Prologue or XIV is recognised as a chapter opening. The recognition
covers English, French, German, Spanish and Italian, and it accepts arabic numerals, roman numerals
and numbers spelled out.
A third path catches the rest: if your manuscript opens with a contents listing, the titles in that listing are matched against the lines of the book, and where they line up, those lines become chapters. The listing itself is then removed, so you do not import a page of empty chapter stubs.
Check the chapter count first
The panel reports how many chapters were found before the book is created. Compare it to what your book actually has:
| It says | What it means | What to do |
|---|---|---|
| The right number | Heading styles or wording were recognised | Nothing |
| 1 | No heading styles and no recognisable chapter wording | Import anyway, then split in the editor |
| Far too many | Something looks like a chapter title and is not — usually scene breaks typed as short centred lines | Import anyway and merge, or fix the styles in Word first |
What survives
Bold, italic, underline, strikethrough. Footnotes, with their text, matched to where they are called from. Lists, including nested ones and their numbering. Tables. Images stored in the document. Charts.
What is left behind
The page. Your Word file states a page size, margins, paragraph indents and line spacing. None of it is imported, because a Word page and a book page are different objects: a manuscript is A4 or Letter with a wide measure and a 12-point face, and a novel is none of those.
The typeface, type size, line spacing and paragraph indent are read, and become a starting point you change in Formatting.
What gets reported
Anything that could not be brought across is listed before the book is created, and again afterwards.
| Report | Cause | Fix |
|---|---|---|
| A linked image | Added in Word with Link to File rather than Insert, so the picture is not inside the file | Re-insert it in Word, or place it in Nueel |
| An image over 30 MB | Too large to import | Save a smaller copy and place it in Nueel |
| A font that cannot be used | The typeface cannot be embedded in a book | Choose another in Formatting |
| Footnotes with no call | The reference was deleted but the note text remains in the file | Retype the call where it belongs |
| Unusually long sections | Over ~6,000 words with no heading inside — nearly always unstyled headings | Split in the editor |
A note with no call is dropped rather than imported, because a note nothing points at is unreachable in a book. The count is always reported: these are your words, so they are never dropped quietly.
Markdown
# headings become the structure. YAML front matter at the top (title, subtitle, author) is
read into the book’s details.
Two behaviours specific to Markdown: a Project Gutenberg header and licence block is detected and
removed, and image references are kept as captions rather than pictures, because a .md file carries
no image data.
After the import
Read the first chapter and the last one, checking that italics survived, scene breaks are there, and chapters begin where they should.
Then go to Themes.