String field (VARCHAR)

String fields are how you store text attributes (metadata) next to your vectors — things like:

Document title, URL, file name
Customer / tenant IDs
Tags (“invoice”, “support”, “legal”)
Source identifiers (“slack”, “email”, “notion”)

In DODIL VBase, string fields map to a VARCHAR field type.

Key rules

1) You must set a maximum length

When you define a string field, you must provide:

datatype=DataType.VARCHAR
max_length=<number>

max_length is the maximum number of bytes a value can take. The allowed range is 1 → 65,535.

Because it is measured in bytes, not “characters”, using UTF-8 content (Arabic, emoji, etc.) means some values may consume more bytes per character.

2) Optional: allow nulls and/or set a default

You can make a string field optional by setting:

nullable=True

And you can provide a fallback value with:

default_value="..."

This is useful when you want inserts to succeed even if a specific attribute is missing.

3) Optional: dynamic fields

If you enable dynamic fields (enable_dynamic_fields=True), you can insert extra scalar fields that were not pre-declared in your schema.

This can be convenient during early prototyping, but it also makes schemas harder to reason about at scale (more inconsistent fields, more complex filters).

Example: add string fields to a collection schema

This example creates a collection with two string fields:

title: up to 256 bytes, optional, default to "Untitled"
source: up to 64 bytes, optional

Assumption: you already have a vbase connection (see the Connect section in the docs).


# Python 3.10+
 
from dodil.vbase import VBaseConfig
from dodil.vbase import DataType
from dodil import Client
 
c = Client(
    service_account_id="...",
    service_account_secret="...",
)
 
vbase = c.vbase.connect(
    VBaseConfig(
        host="vbase-db-<id>.infra.dodil.cloud",
        port=443,
        scheme="https",
        db_name="db_<id>",
    )
)
 
# Build a schema using the underlying schema builder.
# (We expose `vbase.raw` for full compatibility with the backend engine.)
schema = vbase.raw.create_schema(
    auto_id=False,
    enable_dynamic_fields=True,
)
 
schema.add_field(
    field_name="title",
    datatype=DataType.VARCHAR,
    max_length=256,
    nullable=True,
    default_value="Untitled",
)
 
schema.add_field(
    field_name="source",
    datatype=DataType.VARCHAR,
    max_length=64,
    nullable=True,
)
 
# Primary key
schema.add_field(
    field_name="id",
    datatype=DataType.INT64,
    is_primary=True,
)
 
# Vector field
schema.add_field(
    field_name="embedding",
    datatype=DataType.FLOAT_VECTOR,
    dim=768,
)
 
# Create the collection
vbase.create_collection(
    collection_name="docs",
    schema=schema,
)