Self-updating search vector field using postgres 12 "generated" column feature

auto_search_vector_field.py

from django.contrib.postgres.search import SearchVectorField
from django.db.models.fields import NOT_PROVIDED


class AutoSearchVectorField(SearchVectorField):
    def __init__(self, language_field: str, source_field_weights: dict, *args, **kwargs) -> None:
        self.language_field = language_field
        self.source_field_weights = source_field_weights

        kwargs["editable"] = False
        kwargs["serialize"] = False
        kwargs["default"] = NOT_PROVIDED
        kwargs["null"] = True

        super().__init__(*args, **kwargs)

    def deconstruct(self):
        name, path, args, kwargs = super().deconstruct()
        kwargs["language_field"] = self.language_field
        kwargs["source_field_weights"] = self.source_field_weights

        del kwargs["editable"]
        del kwargs["serialize"]
        # del kwargs["default"]
        del kwargs["null"]

        return name, path, args, kwargs

    def db_type(self, connection):
        # TODO: this naive syntax only works for new columns, so the migrations need to be manually adjusted to
        # RemoveField+AddField
        return f"tsvector GENERATED ALWAYS AS ({self._get_search_vector()}) STORED"

    def _get_search_vector(self) -> str:
        regconfig = self._get_lazy_lang_to_regconfig()

        return "||".join(
            [
                f"setweight(to_tsvector({regconfig}, coalesce({field},'')), '{weight}')"
                for field, weight in self.source_field_weights.items()
            ],
        )

    def _get_lazy_lang_to_regconfig(self) -> str:
        lang_to_regconfig = {
            "en": "english",
            "de": "german",
        }

        return "CASE %s END" % " ".join(
            [
                f"WHEN {self.language_field} LIKE '{lang}%%' THEN '{regconfig}'::regconfig"
                for lang, regconfig in lang_to_regconfig.items()
            ]
        )


# We need this hacky mixin because there's currently no "right" way to make a field read-only on the DB level.
# See: https://code.djangoproject.com/ticket/21454
class AutoSearchModelMixin:
    def save(self, **kwargs) -> None:
        search_fields = []
        for i, field in enumerate(self._meta.local_fields):
            if isinstance(field, AutoSearchVectorField):
                search_fields.append(self._meta.local_fields.pop(i))
                self._meta._expire_cache(forward=True, reverse=True)

        super().save(**kwargs)

        for field in search_fields:
            self._meta.add_field(field)
        self._meta._expire_cache(forward=True, reverse=True)