Import & Export

Most authors don’t start with an empty Ishvana. They have years of notes somewhere else — an Obsidian vault full of worldbuilding, a Google Docs folder full of scene drafts, a Scrivener project from the last novel, a stack of Word files from when they were using Word, a few PDFs of research material they scraped from the web. Starting over in a new tool with none of that content would be a non-starter. Ishvana handles this through two separate import systems and a multi-format export system. The lore ingestion pipeline brings reference documents into your Legendry — it parses, structures, and splits content into lore entries automatically. The document import brings manuscripts directly into the writing module as editable documents. And the export system handles sending your work back out in whichever format you need. This page covers all three.

For print-ready PDF compilation with full typographic control, see Bookmaker — the export system here is for quick single-document conversions, not book-level layout work.

Lore Ingestion

Import reference documents to populate the Legendry with worldbuilding knowledge. Ingested content is parsed, structured, and made searchable as proper Legendry entries with sections, tags, and relationships.

Supported formats

• Markdown (.md) — extracts frontmatter, headings, and wikilinks.
• Plain text (.txt) — uses the first line as the title, the rest as body.
• Obsidian (.md with Obsidian features) — handles [[wikilinks]] and #inline-tags automatically.
• HTML (.html, .htm) — strips tags while preserving text structure.
• Word (.docx, .doc) — extracts paragraph text and headings.
• PDF (.pdf) — extracts text from all pages. Formatting is lost but the text survives.
• RTF (.rtf) — strips formatting codes to extract plain text.

Automatic format detection

Ishvana doesn’t force you to tell it what kind of file you’re importing. It detects the format from the content:

• Files with [[wikilinks]] get treated as Obsidian notes.
• Files starting with --- get treated as Markdown with frontmatter.
• Files with HTML tags get treated as HTML.
• Everything else falls back to Markdown parsing.

This matters because most authors have mixed collections — some files are Markdown, some are text, some were exported from one tool and modified in another. You don’t have to sort them yourself.

Entry type detection

Metadata in your files maps to lore entry types automatically. The system recognizes all twelve entry types — character, location, item, event, faction, concept, species, language, religion, legends, article, and custom. If no type is detected, the entry defaults to article, which is the most generic type.

You can override the detected type during import if the system guesses wrong. Most of the time it gets it right, because the structural cues in most worldbuilding notes are consistent — character notes have personality sections, location notes have geography sections, and so on.

Section extraction

For entry types with section templates (characters, locations, items, factions, etc.), the system splits content by headings into structured sections. The heading style is detected automatically — Markdown ## headings, HTML <h2> tags, Word heading styles, whatever the source uses.

Each heading becomes a section in the resulting lore entry. Section names are matched against the template’s expected sections where possible, so a heading called “Appearance” on a character entry maps to the canonical Appearance section even if you used a slightly different phrasing.

Obsidian vault import

Import an entire Obsidian vault at once. Ishvana recursively discovers all .md files in the vault (skipping .obsidian, .trash, and .git directories), parses each with the Obsidian parser, and creates Legendry entries. [[wikilinks]] get resolved to relationships between entries.

This is probably the single highest-leverage thing you can do on day one if you already have an Obsidian-based worldbuilding system. A two-year-old vault with 500 notes becomes a 500-entry Legendry in a few minutes.

Obsidian vault import

Wikilinks convert to actual Legendry relationships during import. A note that links to [[Kent Musa]] becomes an entry that has a relationship to the Kent Musa entry. Preserves your existing network structure without requiring you to manually recreate it.

Entity extraction during import

When importing files, Ishvana can optionally run entity extraction on the content. The extractor finds proper nouns (characters, locations, items, factions) mentioned in the text and creates connections in your Legendry. Useful for bringing in prose content where the entities aren’t explicitly listed but are mentioned throughout the text.

Document Import

Separate from lore ingestion, you can import manuscript files directly into the writing module. These become editable documents in the editor, not entries in the Legendry.

• Open files from disk via the file dialog. Supports DOCX, Markdown, plain text, HTML.
• Obsidian vault import for writing — brings in .md files as manuscript documents with wikilink conversion and folder structure preservation.
• External file references — link to research materials without copying their content into your project. Useful for large PDFs you want to reference but don’t want to store inside Ishvana’s data directory.

Document import is the faster of the two imports because it doesn’t try to parse the content into structured sections — it just creates a document from the file and opens it in the editor.

Document export

Once your work is in Ishvana, you’ll eventually want to get some of it out — for sharing with an editor, for handing to a beta reader, for uploading to KDP, for any of the workflows that live outside the writing studio.

Supported export formats

Format	Requirements
DOCX	Works out of the box — no dependencies
PDF	Requires Pandoc installed on your system
Markdown	Works out of the box
HTML	Works out of the box
Plain text	Works out of the box

Export options

• Format — pick your target format.
• Include TOC — generate a table of contents. PDF and DOCX only.
• Include front matter — include title page, dedication, copyright page. PDF only.

Pandoc dependency

Pandoc is required for PDF export

PDF export uses Pandoc under the hood. Install it separately from pandoc.org if you need PDF output. Ishvana will tell you clearly when you try to export to PDF without Pandoc available and list which formats are working.

DOCX, Markdown, HTML, and plain text export work without Pandoc. Most quick-share workflows only need DOCX, which is fine.

For print-ready PDFs with full typographic control, use Bookmaker instead — it’s the dedicated compile-a-book-to-PDF workflow and has different (and much more capable) PDF tooling than the quick-export path here.

What about ePub and MOBI?

Neither is supported directly in this export system. For ebook formats you’d use Bookmaker for the compile step and then run the output through a dedicated ebook converter like Calibre or Kindle Create. Ishvana’s export is for sharing and reference, not final ebook publication.

Import history

Every import operation gets tracked in the import history with:

• Timestamp of the import.
• Source file path or vault path.
• Number of documents or entries created.
• Status (success, partial, failed).
• Batch ID so you can undo an import if needed.

Query the history from the Legendry’s import panel or the writing module’s document import surface. The ability to undo a batch import is worth more than it sounds — if you import a vault of 500 notes and realize half of them shouldn’t have become Legendry entries, rolling back the entire batch is a single click.

What’s next

Documents Manage the manuscripts that document import brings into Ishvana.

Legendry The worldbuilding database that lore ingestion populates.

Bookmaker The dedicated compile-a-book-to-PDF system for print-ready output.