Registry

Filter

Narrows the list by title or short title as you type.

Status

The state of the most recent import attempt.

• Pending — created, import not yet started.
• Processing — the import job is extracting and indexing pages.
• Ready — import finished successfully; pages are searchable.
• Failed — the import did not complete. The book's show page displays the diagnostic error. A book is never left looking Ready with only some of its pages imported.

Searchable

A global, admin-only kill switch, separate from any campaign's own per-book Enabled toggle. Turning this off removes the book from every campaign's search immediately and stops it appearing as an option to enable on any Library page — but does not delete the book or its imported pages.

Creating a Rulebook

PDF

Required when creating a rulebook. Saving the form queues the import as a background job immediately, using the uploaded file. The upload itself is never retained — it's deleted from the server once the import job finishes, whether it succeeds or fails. Only extracted page text and metadata are stored afterward, so keep the source PDF in your own archive.

PDF Page Offset

A PDF's internal page index almost never matches the number printed on the page — covers, blank pages, and roman-numeral front matter all shift things. The offset is PDF page − printed page, so:

printed page = PDF page − offset

To find the right value: open the PDF, find any page where you can read the printed page number, and note its PDF page index. For example, if PDF page 8 is printed as page 3, the offset is 8 − 3 = 5. Get this right before importing — it determines the default printed page number for every page that doesn't have an explicit override.

Header/Footer Patterns

One regular expression per line, matched against each page's extracted text and stripped before it's indexed for search. Extraction is often noisy — repeated headers/footers, page numbers bleeding into the text, purchaser watermarks — and these patterns clean that up.

Search Rank Modifier

Added directly to every search hit's relevance score for this book, before the minimum-relevance cutoff is applied. Defaults to 0, which changes nothing. Useful for books like FAQs or errata, where a correction's matching text is short enough that it would otherwise never clear the cutoff — a positive modifier lets its matches surface anyway. Negative values push a book's matches down instead.

Managing an Existing Rulebook

Start Import

Uploads a fresh PDF and reprocesses the book, replacing its previously imported pages. This always reprocesses — even a byte-identical re-upload of the current file — so it's how you force a reimport after an extractor change or edited header/footer patterns. The book's ID never changes across reimports.

Rebuild Search Vectors

Re-runs text normalisation and re-indexes every already-extracted page without re-reading the PDF. Use this after editing Header/Footer Patterns — the raw extracted text is always kept separately from the cleaned, searchable text specifically so normalisation can be redone without a full reimport.

Edit

Changes metadata, the page offset, header/footer patterns, and the Searchable flag. Does not itself trigger a reimport or a search-vector rebuild — use the buttons above for that.

Delete

Permanently removes the rulebook and all its imported pages. This cannot be undone.

Page Numbering

Not every page in a book fits the offset cleanly — covers, inserted maps, and renumbered sections need manual handling. Use the page mapping table, linked from a rulebook's show page, to correct individual pages.