Business
Jobs
  • About Us
  • Solutions
    • Job Postings
      Post your job and receive qualified candidates in 48h.
    • Candidate Assessments
      500+ technical and psychological tests, plus anti-fraud.
    • Headhunting
      Tailor-made executive search from start to finish.
    • Payroll + EOR
      Payroll dispersal and EOR across 15+ LATAM countries.
  • Pricing
  • Jobs

0

451
Views
In FastAPI, how can I expand a model's output with a reverse URL lookup based on a field?

Let's assume I have a simple file storage in the database. The SQLAlchemy model looks like this:

class Blob(Base):
    id = Column(Integer, primary_key=True)
    blob = deferred(Column(LargeBinary().with_variant(LONGBLOB, "mysql")))

The respective Pydantic model looks like this:

class BlobBase(sqlalchemy_to_pydantic(Blob, exclude=["blob"])):
    class Config:
        orm_mode = True

There is also a FastAPI route intented for fetching single files by ID:

@router.get("/blob/{blob_id}")
async def get_blob(blob_id: str):
   [...]

And a route for fetching a list of all files:

@router.get("/blobs", response_model=List[BlobBase])
async def get_blobs():
    return [BlobBase.from_orm(x) for x in db.session.query(Blob).all()]

Now, I would like to include the resolved get_blob URL in each entry in get_blobs. Naively, I suppose this should look like this:

class BlobBase(sqlalchemy_to_pydantic(Blob, exclude=["blob"])):
    class Config:
        orm_mode = True

    url: Optional[str]

    @validator("url")
    def make_url(cls, v, values):
        return request.url_for("get_blob", blob_id=values["id"])

However, I do not have access to a request or an app object in the validator, so that I cannot properly resolve the URL. NB: I do have access to the router object, which is an APIRouter for the current sub-URL, but get_blob in the actual application is in a different APIRouter, so I can't use that without jamming everything in a single file.

What is the correct way to solve this, i.e. include a resolved URL in model's output?

over 4 years ago · Santiago Trujillo
1 answers
Answer question

0

Since you have defined from_orm in the config for your model, you shouldn't have to do the manual transformation using from_orm(x) in your get_blobs view - returning the result from the query by itself should be enough.

@router.get("/blobs", response_model=List[BlobBase])
async def get_blobs():
    return db.session.query(Blob).all()

It's also suggested to use a dependency to resolve db for each async endpoint (there's an example in the FastAPI docs) instead of having a global-ish db entry.

Since the reverse URL isn't really a property of the schema itself, I think I'd go with adding a composite schema and then populate it in your view:

class BlobWithUrl(BaseModel):
    blob: BaseBlob  # Consider using Blob as the name instead - base indicates that it should only be inherited
    url: str


@router.get("/blobs", response_model=List[BlobWithUrl])
async def get_blobs(request: Request):
    return [
        {'url': url_for(...), 'blob': blob}
        for blob in db.session.query(Blob).all()
    ]

Another option is to use the strategy with manually calling from_orm, and then having a schema that extends BlobBase:

class BlobWithUrl(BaseBlob):
    url: Optional[str]


@router.get("/blobs", response_model=List[BlobWithUrl])
async def get_blobs(request: Request):
    blobs = []

    for retrieved_blob in db.session.query(Blob).all():
        blob = BlobWithUrl.from_orm(retrieved_blob)
        blob.url = url_for(...)
        blobs.append(blob)
    
    return blobs
over 4 years ago · Santiago Trujillo Report
Answer question
Find remote jobs

Discover the new way to find a job!

Top jobs
Top job categories
Business
Post vacancy Pricing Sales
Legal
Terms and conditions Privacy policy
© 2026 PeakU Inc. All Rights Reserved.
Andres GPT
Show me some job opportunities
There's an error!