Use structured output with lists in the new SDK

[osanseviero]

In the old SDK, we could do

import google.generativeai as genai
import typing_extensions as typing

class Recipe(typing.TypedDict):
    recipe_name: str
    ingredients: list[str]

model = genai.GenerativeModel("gemini-1.5-pro-latest")
result = model.generate_content(
    "List a few popular cookie recipes.",
    generation_config=genai.GenerationConfig(
        response_mime_type="application/json", response_schema=list[Recipe]
    ),
)
print(result)

With the new API, doing something similar

from google import genai
from pydantic import BaseModel

class Recipe(BaseModel):
    recipe_name: str
    ingredients: list[str]

result = client.models.generate_content(
    model="gemini-2.0-flash-exp",
    contents="List a few popular cookie recipes.",
    config=genai.types.GenerateContentConfig(
        response_mime_type="application/json", response_schema=list[Recipe]
    ),
)
print(result)

I would get

AttributeError: type object 'list' has no attribute 'model_json_schema'

If I instead try to wrap with another model

class RecipeList(BaseModel):
    recipes: list[Recipe]

...

result = client.models.generate_content(
    model=MODEL_ID,
    contents="List a few popular cookie recipes.",
    config=genai.types.GenerateContentConfig(
        response_mime_type="application/json", response_schema=RecipeList
    ),
)

I get an error as well, but this time in validation stage, which is unfortunately not very clear (but it does hint the model did return something)

ValidationError: 2 validation errors for Schema
properties.recipes.items.$ref
  Extra inputs are not permitted [type=extra_forbidden, input_value='#/$defs/Recipe', input_type=str]
    For further information visit https://errors.pydantic.dev/2.10/v/extra_forbidden
$defs
  Extra inputs are not permitted [type=extra_forbidden, input_value={'Recipe': {'properties':...ts'], 'type': 'OBJECT'}}, input_type=dict]
    For further information visit https://errors.pydantic.dev/2.10/v/extra_forbidden

[tomy2502]

from typing import List

class Output_from_AIStudio(BaseModel):
keywords: List[str]

[osanseviero]

Related: #60