app/ui/schemas.py

from pydantic import BaseModel

from app.server.chat.schemas import Chunk


class Source(BaseModel):
    file: str
    page: str
    text: str

    class Config:
        frozen = True

    @staticmethod
    def curate_sources(sources: list[Chunk]) -> set["Source"]:
        curated_sources = set()

        for chunk in sources:
            doc_metadata = chunk.document.doc_metadata

            file_name = doc_metadata.get("file_name", "-") if doc_metadata else "-"
            page_label = doc_metadata.get("page_label", "-") if doc_metadata else "-"

            source = Source(file=file_name, page=page_label, text=chunk.text)
            curated_sources.add(source)

        return curated_sources