import time
import os
import re
import gradio as gr
from langchain.document_loaders import PyPDFLoader
#from langchain.document_loaders import PyMuPDFLoader

ABS_PATH = os.path.dirname(os.path.abspath(__file__))
DB_DIR = os.path.join(ABS_PATH, "db")
 
on_load="""
async()=>{
    console.log("HELLO");
}
"""

def get_documents():
    return PyPDFLoader("doc1.pdf").load()
    """
    loader = PyMuPDFLoader(
    "example.pdf", extract_images=True,
    )
    docs = loader.load()
    """

def extract_pdfs(x, request: gr.Request, progress=gr.Progress()):
    progress(0, desc="Test", unit = "Files")
    print("request", request)

    # Delete existing index directory and recreate the directory
    if os.path.exists(DB_DIR):
        import shutil
        shutil.rmtree(DB_DIR, ignore_errors=True)
        os.mkdir(DB_DIR)
    
    documents = []
    all_text = ""
    for num, doc in enumerate(progress.tqdm(get_documents())):
        print(" {num} DocPg : ", doc.page_content)
        doc.page_content = replace_newlines_and_spaces(doc.page_content)
        documents.append(doc)
        all_text += doc.page_content
        time.sleep(0.1)

    return documents, all_text

def replace_newlines_and_spaces(text):
    # Replace all newline characters with spaces
    text = text.replace("\n", " ")
    # Replace multiple spaces with a single space
    text = re.sub(r'\s+', ' ', text)
    return text


def test(x, request: gr.Request, progress=gr.Progress()):
    progress(0, desc="Test", unit = "Files")
    print("request", request)
    a = "abcdefghijklmnopqrstuv"
    for letter in progress.tqdm(a, desc = "TEST", unit = "Files"):
        time.sleep(0.1)
    return a  

with gr.Blocks() as demo:

    selected = gr.Dataframe(
        interactive=False,
        col_count=(1, "fixed"),
        headers=["Selected Files"],
    )
    prog = gr.HTML(
        value="<h3 style='text-align: center'> Processing...<h1>"
    )
    #gr.Interface(test, inputs=[selected])
    b = gr.Button()
    
    b.click(test, selected, prog)
    
    demo.load(extract_pdfs, inputs=None, outputs=[prog, selected]) #, _js=on_load)    
    
demo.launch()