# app_logic.py
from transformers import pipeline, AutoTokenizer, AutoModelForSeq2SeqLM
import fitz, docx, pptx, openpyxl, re, nltk, tempfile, os, easyocr, hashlib, datetime
from nltk.tokenize import sent_tokenize
from fpdf import FPDF
from gtts import gTTS

nltk.download('punkt', quiet=True)

# Load once
MODEL_NAME = "facebook/bart-large-cnn"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForSeq2SeqLM.from_pretrained(MODEL_NAME)
summarizer = pipeline("summarization", model=model, tokenizer=tokenizer, device=-1, batch_size=4)
reader = easyocr.Reader(['en'], gpu=False)

summary_cache = {}

def clean_text(text):
    text = re.sub(r'\s+', ' ', text)
    text = re.sub(r'\u2022\s*|\d\.\s+', '', text)
    text = re.sub(r'\[.*?\]|\(.*?\)', '', text)
    text = re.sub(r'\bPage\s*\d+\b', '', text, flags=re.IGNORECASE)
    return text.strip()

def extract_text(file_path, file_extension):
    try:
        if file_extension in ["pdf"]:
            with fitz.open(file_path) as doc:
                text = "\n".join(page.get_text("text") for page in doc)
                if len(text.strip()) < 50:
                    images = [page.get_pixmap() for page in doc]
                    temp_img = tempfile.NamedTemporaryFile(suffix=".png", delete=False)
                    images[0].save(temp_img.name)
                    ocr_result = reader.readtext(temp_img.name, detail=0)
                    os.unlink(temp_img.name)
                    text = "\n".join(ocr_result) if ocr_result else text
        elif file_extension in ["docx"]:
            doc = docx.Document(file_path)
            text = "\n".join(p.text for p in doc.paragraphs)
        elif file_extension in ["pptx"]:
            prs = pptx.Presentation(file_path)
            text = "\n".join(shape.text for slide in prs.slides for shape in slide.shapes if hasattr(shape, "text"))
        elif file_extension in ["xlsx"]:
            wb = openpyxl.load_workbook(file_path, read_only=True)
            text = "\n".join([" ".join(str(cell) for cell in row if cell) for sheet in wb.sheetnames for row in wb[sheet].iter_rows(values_only=True)])
        else:
            return "", "Unsupported file type"
        
        return clean_text(text), ""
    except Exception as e:
        return "", f"Extraction error: {e}"

def chunk_text(text, max_tokens=950):
    sentences = sent_tokenize(text)
    chunks, current_chunk = [], ""
    for sentence in sentences:
        if len(tokenizer.encode(current_chunk + " " + sentence)) <= max_tokens:
            current_chunk += " " + sentence
        else:
            chunks.append(current_chunk.strip())
            current_chunk = sentence
    if current_chunk:
        chunks.append(current_chunk.strip())
    return chunks

def generate_summary(text, length="medium"):
    cache_key = hashlib.md5((text + length).encode()).hexdigest()
    if cache_key in summary_cache:
        return summary_cache[cache_key]

    params = {"short": (30, 80), "medium": (80, 200), "long": (210, 300)}[length]
    min_len, max_len = params

    chunks = chunk_text(text)
    summaries = summarizer(chunks, max_length=max_len, min_length=min_len, do_sample=False)
    final_summary = " ".join(s['summary_text'] for s in summaries)
    summary_cache[cache_key] = final_summary
    return final_summary

def text_to_speech(text):
    try:
        tts = gTTS(text)
        temp_audio = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
        tts.save(temp_audio.name)
        return temp_audio.name
    except:
        return ""

def create_pdf(summary, original_filename):
    try:
        pdf = FPDF()
        pdf.add_page()
        pdf.set_font("Arial", 'B', 16)
        pdf.cell(200, 10, "Summary", ln=True, align='C')
        pdf.set_font("Arial", size=12)
        pdf.multi_cell(0, 10, summary)
        temp_pdf = tempfile.NamedTemporaryFile(delete=False, suffix=".pdf")
        pdf.output(temp_pdf.name)
        return temp_pdf.name
    except:
        return ""