What are documents?

A document is Moonjar’s word for anything you’ve captured: a webpage, a photo, a PDF, an email, a typed note. Whatever the source, once Moonjar has read it, it’s a document.

How it works

When you capture something, Moonjar:

1. Reads it. For URLs, it fetches the page and pulls out the readable text (skipping nav, ads, and chrome). For PDFs and images, it extracts the text. For emails, it consolidates the body and any PDF attachments into one block.
2. Classifies it. It picks one of 15 document types — receipt, travel document, utility bill, and so on.
3. Pulls out the fields that matter. A receipt gets vendor, total, and date. A travel document gets carrier, booking reference, departure, arrival. The fields differ per type.
4. Indexes it for search. Both full-text and semantic — see How search works.
5. Files it. The document, its fields, and the original source text all live together. You can reopen it any time.

Why it matters

The classification and field extraction are what make a Moonjar document different from a saved bookmark or a photo in your camera roll. You can ask things like “how much did I spend at the Apple Store last March?” — there’s nothing to set up, the fields were extracted at capture time and are available to search and to the chat assistant.

Limits

• Documents are private to you by default. To share with others, see Sharing & groups.
• If two captures look alike (same title, or near-identical content), Moonjar treats the second as a duplicate and asks before saving.
• Field extraction depends on what’s actually in the source. A faded receipt with no readable total leaves total blank — you can fill in missing fields by hand on the document detail screen.

See also

• Document types
• Capture
• How search works