import os
#from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline
from transformers import pipeline
from huggingface_hub import login

# Safe writable cache dirs in Hugging Face Spaces move to space settings vars`
#os.environ["TRANSFORMERS_CACHE"] = "/tmp/hf_cache"
#os.environ["HF_HOME"] = "/tmp/hf_home"
#os.environ["HF_HUB_CACHE"] = "/tmp/hf_hub"



MODEL_ID = "TypicaAI/magbert-ner"

#HF_TOKEN = os.getenv("HF_TOKEN")
#tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, token=HF_TOKEN)
#model = AutoModelForTokenClassification.from_pretrained(MODEL_ID, token=HF_TOKEN)
#ner_pipeline = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="first")

# Initialize global pipeline
ner_pipeline = None

# Authenticate using the secret `HFTOKEN`
def authenticate_with_token():
    """Authenticate with the Hugging Face API using the HFTOKEN secret."""
    hf_token = os.getenv("HF_TOKEN")  # Retrieve the token from environment variables
    if not hf_token:
        raise ValueError("HF_TOKEN is not set. Please add it to the Secrets in your Space settings.")
    login(token=hf_token)
	
    
def load_healthcare_ner_pipeline():
    """Load the Hugging Face pipeline for Healthcare NER."""
    global ner_pipeline
    if ner_pipeline is None:        
        
        # Authenticate and initialize pipeline
        authenticate_with_token()
        
        ner_pipeline = pipeline(
            "token-classification",
            model=MODEL_ID,
            aggregation_strategy="first"  # Groups B- and I- tokens into entities
        )
    return ner_pipeline




# Get NER pipeline
ner_pipeline = load_healthcare_ner_pipeline()