WritersNotebook/WritersNotebook
Watch
1
0
Fork
You've already forked WritersNotebook
 0
Code Issues Pull requests Projects 1 Releases 35 Packages Wiki Activity Actions
Page: Importing Manuscripts
Pages
Backup and Restore Characters Comments Editor Extension Development Extensions Find and Replace Getting Started Glossary Home Importing Manuscripts Install Outline Analysis Outlines Plots Projects Publishing Readability Analysis Reference Panel Review Mode Settings Updates World Building
3 Importing Manuscripts
Gwen Gullett edited this page 2026-08-08 13:33:42 -04:00
Table of contents

Importing Manuscripts

Already have a draft in another tool? Instead of starting from a blank scaffold, Writers Notebook can read an existing manuscript file and build a fully structured new project from it: a project record, an outline, a chapter/scene tree, and the prose for each node.

Supported files

Format Notes
EPUB (.epub) Chapters are recovered from the book's actual content, not just its table of contents.
PDF (.pdf) Only text-based PDFs. Scanned/image-only PDFs are rejected (no OCR is performed).
Word (.docx) Chapter titles are detected from Word's heading styles/outline level, or from "Chapter N"-style text. Tables and images aren't preserved and are flagged with a warning; text and basic formatting (bold/italic) are.
Plain text (.txt) Chapters detected heuristically (see below).
Markdown (.md) Headings map directly to chapters and scenes.

Legacy binary .doc files are not supported; you'll be told to re-save them as .docx in Word (or a compatible editor) and import that instead. .fdx (Final Draft screenplay) files aren't supported.

Where to start an import

Import is available in two places, and both use the same flow:

  • During onboarding: choose "Import a manuscript" at the project-creation step of the first-run wizard.
  • From the Workshop's project switcher: use the Import from File button to create an additional project later.

The import flow

  1. Pick your file: a native file dialog (desktop) or file picker (web), filtered to the supported formats.
  2. Choose a mode:
    • Automatic (default): chapter and scene detection runs hands-off and takes you straight to project creation.
    • Advanced: the same detection runs, but you review and adjust the detected structure before anything is written.
  3. Analyzing: the file is parsed. This can take a few seconds for a large EPUB or PDF.
  4. Preview & (in Advanced mode) adjust: see below.
  5. Commit: the new project is created and you land in the Workshop, ready to keep writing.

Import from file

Previewing detected content

In both modes, every detected chapter or part has a preview button. It opens the exact text the importer extracted for that piece, shown as faithful plain text so you can sanity-check the extraction (and spot any formatting quirks) before creating the project.

Advanced mode: reviewing boundaries

Advanced mode shows the detected chapters and scenes as an editable list where you can:

  • Merge a chapter into the previous one
  • Split a chapter at a chosen point
  • Retitle any entry
  • Reorder entries

Each row shows a recursive word count and a type-aware child count (e.g. "3 chapters" under a part, "2 scenes" under a chapter). Nothing is written to disk until you confirm.

Import Advanced View

How structure is detected

  • EPUB and Word: chapters are found from the document's actual content, including chapter titles that are styled paragraphs rather than headings, so books whose table of contents only lists coarse "Part 1 / Part 2" groupings (or nothing at all) still import as their real chapters.
  • Markdown: heading levels map to chapters (levels 1-2) and scenes (level 3+); a lone --- is a scene break.
  • Plain text and text-based PDF: detected by pattern, strongest first: Chapter N style headings, then short standalone Title-Case/ALL-CAPS lines, then blank-line or glyph runs as scene breaks.
  • The resulting tree can be up to three levels: Part → Chapter → Scene. A Part is only ever created when the manuscript's own content contains a container heading above the chapter level (e.g. an in-body "Part One" or "Book II" heading); Parts are never invented from a table of contents or a file's spine, since those splits are frequently arbitrary. Most manuscripts have no such heading and import as a flat list of chapters (with scenes nested underneath where detected).
  • Front matter (anything before the first detected chapter: title page, copyright, dedication) is kept as its own "Front Matter" node rather than being merged into chapter one, and is never nested inside a Part.

If no chapter breaks are found in Automatic mode, the import doesn't fail: the whole file becomes a single node with a warning, and you can split it later in the Outline.

What import does and doesn't do

  • Creates: a project, one outline, the chapter/scene node tree, and the prose for each node. EPUB/PDF metadata (title, author) is used to pre-fill the new project's title and description.
  • Does not: attempt to auto-populate Characters, Plots, World, or the Glossary. You build those yourself afterward, just like any new project. It also doesn't import into an existing project; import always creates a new one.

If something goes wrong

Import fails clearly (before any project is created) for: scanned/image-only PDFs, encrypted or DRM-protected EPUBs, password-protected PDFs, and unreadable or empty files. You'll see a specific message rather than a crash.