Generate and serve files from Flask

Flask is one of the most used python frameworks for web development. Its simplicity and extensibility makes it convenient for both small and large applications alike.

In this blog we are going to create a simple flask web application that will generate and serve files without storing them in the server.

Note: For flask and python installation visit flask documentation

Let’s create a flask application with a basic route.

from flask import Flask

app = Flask(__name__)


@app.route("/")
def index():
    return "Hello Flask!"

and voila! We have our server up and running with only 5 lines of code.

Now we need to create a route which will accept a file name as parameter.

@app.route("/file/<file_name>")
def get_file(file_name):
    return file_name

For our use case we need to generate a csv file using fake data.We need to install faker to generate fake data such as name, address, birthdate etc. Also we are using pandas to generate dataframes that can be used to create both csv and spreadsheets.

python3 -m pip install faker pandas

Let’s add functions that will generate csv files using the fake data we get from Faker.

def generate_fake_data():
    fake_data = [fake.simple_profile() for item in range(5)]
    return pd.DataFrame(fake_data)


def generate_csv_file(file_df):
    # Create an o/p buffer
    file_buffer = StringIO()

    # Write the dataframe to the buffer
    file_df.to_csv(file_buffer, encoding="utf-8", index=False, sep=",")

    # Seek to the beginning of the stream
    file_buffer.seek(0)
    return file_buffer

Now we need to call these functions from our routing method and send the file as response.

@app.route("/file/<file_name>")
def get_file(file_name):
    fake_df = generate_fake_data()
    generated_file = generate_csv_file(fake_df)
    response = Response(generated_file, mimetype="text/csv")
    # add a filename
    response.headers.set(
        "Content-Disposition", "attachment", filename="{0}.csv".format(file_name)
    )
    return response

Once we hit the above route with a file name the browser will ask for permission to download the csv file.

Here is the full source code with a working example.

Feel free to edit and play around. Adios!