from openai import OpenAI
import numpy as np
from numpy.linalg import norm
client = OpenAI()

def get_embedding(text, model="text-embedding-3-small"):
   text = text.replace("\n", " ")
   return client.embeddings.create(input = [text], model=model).data[0].embedding

item_1_with_field_prefix = get_embedding("LINE_ITEM_DESCRIPTION: DA Ext Door - Install Satin Nickel Single Cylinder Deadbolt Featuring Smartkey Security") #default airbyte embedding
item_1_with_symbol_prefix  = get_embedding("_: DA Ext Door - Install Satin Nickel Single Cylinder Deadbolt Featuring Smartkey Security") #airbyte embedding when field name is mapped to '_'
item_1_no_prefix = get_embedding("DA Ext Door - Install Satin Nickel Single Cylinder Deadbolt Featuring Smartkey Security") #preferred behavior when field name is omitted

item_2_with_field_prefix  = get_embedding("LINE_ITEM_DESCRIPTION: DA Trash Out (Adjust for Severity)") #default airbyte embedding
item_2_with_symbol_prefix = get_embedding("_: DA Trash Out (Adjust for Severity)") #airbyte embedding when field name is mapped to '_'
item_2_no_prefix = get_embedding("DA Trash Out (Adjust for Severity)") #preferred behavior when field name is omitted

def cosine_similarity(a, b):
   return np.dot(a, b) / (norm(a) * norm(b))

print(cosine_similarity(item_1_with_field_prefix, item_2_with_field_prefix))#0.3356051697289621
print(cosine_similarity(item_1_with_symbol_prefix, item_2_with_symbol_prefix))#0.2732279689867596
print(cosine_similarity(item_1_no_prefix, item_2_no_prefix))#0.15904993145660737