photos_words

The words you changed, up to fifty photos per call, with no picture.

Carries up the words you changed on photos already in your index: tags, people, wording, a place you set, the faces. No picture is attached and no photo is read again.

01 · The call
POST https://api.opensolr.com/solr_manager/api/photos_words

A JSON body with email, api_key, index_name and photos, up to 50 per call.

Each photo carries its id and only what changed: tags, persons, meaning, location, faces, plus its file_hash and where the file is on this phone. It runs in the sync that starts right after you save; saving itself is finished on the phone.

02 · What Opensolr does
  • Reads each document whole from your index and puts the new values in; everything else the document holds stays.
  • A new place is turned into place names again.
  • On a plan with vector search, makes the search vector again from the new words, one call for the batch. A photo whose faces alone changed keeps its vector.
  • Writes the documents back with one request.
03 · The answer
{"status": true, "results": [{"id": "…", "status": true, "custom_tags": […], "persons_ss": […], …}, …]}
  • Per photo, the document as written, without the vector and the similar-photo keys. The phone stores it in its copy.
  • NOT_INDEXED for a photo the index does not hold: the phone writes it back from its copy through photos_restore, or, if it has nothing for it, sends the photo the ordinary way.
  • A vector made again costs a tenth of an AI request; one Opensolr still holds costs nothing.

Every call it makes

Opensolr Photos 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 Photos Documentation