Text inside attachments

The words inside your attachments are found too: a scan by the number printed on it, a photo by what it shows.

With an Opensolr account, the words inside your attachments are found too: a scanned invoice by the number printed on it, a photo by what it shows.

01 · What is read
  • Documents: PDF, Word, RTF, OpenDocument text, plain text and HTML, with the pictures inside them read by OCR.
  • Pictures: the text printed in them, and what they show.
  • Up to 20 MB per file, and up to 10 files per message.
  • Never read: archives, spreadsheets, presentations and anything else.
02 · Where it happens

On Opensolr’s servers, not on your phone. The app hands the server a one-time link to each attachment; the server fetches it straight from Fastmail, reads its text into your Opensolr Index and does not keep the file. Nothing is downloaded to the phone for it, and no Wi-Fi is needed. A file that several messages carry is read once.

How many messages still have attachments to read is in Settings → Opensolr Index, on a progress bar of its own, Attachments read.

03 · How it is searched
  • Words read from an attachment count like the words in the body of the mail; words in the subject weigh most.
  • The filter Message is → Text inside attachments keeps only the messages with text read from their attachments, and Attachment types narrows by file type. See Filters and groups.
  • The AI answer reads the first 1,000 words of each message’s attachments.

Reading pictures, and the pictures inside documents, uses AI requests of your plan: see Plan and limits.

Search pages

Opensolr Mail is open source and MIT licensed. Questions about your Opensolr account, index or plan go to opensolr.com/contact; questions about the app itself belong on GitHub.

Opensolr Mail Documentation