[Python] Efficiently Storing Transcript Language Metadata in Firestore

Previously, due to the 1 MB document size limitation, I had to break down transcript text into smaller chunks and store them across multiple documents. The structure looked like this:

users (Collection)
|
|-- {user_id} (Document)
|
|-- notes (Subcollection)
|-- {note_id} (Document)
|-- transcripts (Subcollection)
|-- RZ290XHh3DD3vzavab1m (Document)
|-- text: string
|-- order: int
|-- 8fKb3NhL2a5DXQYmZPjC (Document)
|-- text: string
|-- order: int


Now, I need to introduce a new field, transcript_language, to store the language of the transcript. Is the following design a good approach?

users (Collection)
|
|-- {user_id} (Document)
|
|-- notes (Subcollection)
|-- {note_id} (Document)
|-- transcripts (Subcollection)
|-- metadata (Document) <-- Fixed document ID for storing metadata
|-- transcript_language: string
|-- RZ290XHh3DD3vzavab1m (Document)
|-- text: string
|-- order: int
|-- 8fKb3NhL2a5DXQYmZPjC (Document)
|-- text: string
|-- order: int

Continue reading...