Skip to content

Search

Search is designed to get you back to the right place, not just show matching text. Results can open notebooks, pages, Today captures, Collections, Cabinet items, and indexed document locations.

Before you start

Search works best after content has been saved and indexed. Document text search depends on index document contents being enabled in Settings and on the file type being extractable.

OCR for scanned PDFs is separate from normal PDF text extraction. It can help when a PDF has little readable embedded text, but OCR is best-effort and may not read every scanned or image-heavy file.

Step-by-step

  1. Open global search from the Shelf/editor search control or the shortcut shown in Settings.
  2. Type a word or phrase.
  3. Review grouped results such as notebooks, pages, Cabinet, Today, Collections, and document text.
  4. Open the result that matches the location you need.
  5. For document results, use the result locator or highlight to jump into the relevant preview when available.
  6. If document results look stale, open Settings, check document search index, and use Rebuild document search index or a per-document reindex action when available.

Document text search

Onnoir can index supported document text when document indexing is enabled:

  • PDF text embedded in the document
  • DOCX text
  • Markdown files
  • plain-text and text-like files such as TXT, CSV, JSON, XML, YAML, and logs
  • OCR-readable PDF pages when OCR is enabled and extraction succeeds

PDF extraction currently indexes up to the first 25 pages by default. If the embedded PDF text is too sparse, Onnoir can try OCR for those pages when OCR scanned PDFs is enabled. OCR language options are controlled in Settings.

Document index entries can be queued, extracting, indexed, stale, failed, or unsupported. A stale result usually means the source document changed after the last index. Unsupported usually means the file type is not a previewable/indexable document type.

What to expect

Search covers normal writing and indexed document text. PDF, DOCX, and text-like documents can produce searchable text when extraction succeeds. Scanned or image-heavy documents may need OCR support, and some files may not produce useful text.

Large documents may avoid automatic preview loading. Search can still point at a document result, but previewing the full document may require an explicit Load preview action.

Safety notes

Search indexes are derived from local app records. Do not use search as proof of backup completeness; use backup verification for that. A missing search result does not prove the source file is missing, and a search hit does not prove the full file is recoverable.

Troubleshooting

If a normal note result is missing, confirm the item is not in Trash or excluded by the current search options.

If a document result is missing:

  1. Confirm the item is a supported document type.
  2. Confirm index document contents is enabled.
  3. Check whether the document index says failed, stale, or unsupported.
  4. Reindex the document or rebuild the document search index.
  5. If it is a scanned PDF, enable OCR scanned PDFs and the relevant OCR language, then reindex.
  6. If the document is very large, remember that preview behavior and search indexing are separate; search may index text even when preview loading is manual.