import gradio as gr
from sentence_transformers import SentenceTransformer, util

model_name = 'nq-distilbert-base-v1'
bi_encoder = SentenceTransformer("./")
top_k = 5
sentences = [
    "a happy person is a person how can do what he want with his money",
    "That is a happy dog ho bark alot",
    "Today is a sunny day so that a happy person can walk on the street"
]
# vector embeddings created from dataset
corpus_embeddings = bi_encoder.encode(sentences, convert_to_tensor=True, show_progress_bar=True)

def search(query):
    # Encode the query using the bi-encoder and find potentially relevant passages
    question_embedding = bi_encoder.encode(query)
    hits = util.semantic_search(question_embedding, corpus_embeddings, top_k=top_k)
    hits = hits[0]  # Get the hits for the first query

    # Output of top-k hits
    print("Input question:", query)
    print("Results")
    for hit in hits:
        print("\t{:.3f}\t{}".format(hit['score'], sentences[hit['corpus_id']]))
    return hits

def greet(name):
    hittt = search(query=name)
    x=dict()
    for hit in hittt:
        score=hit['score']
        sentence=sentences[hit['corpus_id']]
        buffer={sentence:score}
        x.update(buffer)
    return x
import dill
def greet1(data):
    # pdf=data.get('pdf')
    print(data)
    x=eval(data)
    y=x.get('pdf')
    print(y)
    print(type(y))
    print(type(dill.loads(eval(y))))
    print(dill.loads(eval(y)).read())
    return y
iface = gr.Blocks()
with iface:
    name = gr.Textbox(label="Name")
    output = gr.Textbox(label="Output Box")
    greet_btn = gr.Button("Greet")
    greet_btn.click(fn=greet, inputs=name, outputs=output, api_name="greet")
    greet1_btn = gr.Button("Greet1")
    greet1_btn.click(fn=greet1, inputs=name, outputs=output, api_name="testing")

iface.launch()