​
extract_text

def extract_text(
    self,
    element_description: str,
    use_vision: bool = False
) -> str:
​
element_description
string
required

Natural language description of the element to extract text from.

​
use_vision
boolean
default:
"False"

Whether to use vision to find the text. Helpful if Simplex struggles to find the text when use_vision = False, otherwise not necessary.

​
return
str

Extracted text string(s) within the requested element. If there are multiple, separate text chunks within the element (i.e. the element description is all financial headlines on a page), this chunks will be separated by newline characters.

​
Sample Usage

In this example, we’re building a script to purchase 3D models from CGTrader (a 3D design marketplace) for a synthetic data project. Since we want to use this asset for creating synthetic data, we need to check if the license allows AI use.

Here’s a sample product page (license highlighted in green):

We can extract the license text to check if it allows AI use.

from simplex import Simplex

simplex = Simplex(api_key="your_api_key")
simplex.create_session()
simplex.goto("https://www.cgtrader.com/3d-models/aircraft/commercial-aircraft/boeing-787-8-5f56fa4f-4ca5-42b8-9ffa-154fb2215928")

succeeded, text = simplex.extract_text("type of license")
if succeeded and "no ai" not in text.lower():
    simplex.click("add to cart")

Was this page helpful?

ActionsManaging Website Credentials