import pypdf
import docx
import pandas as pd
from pdf2image import convert_from_bytes
import pytesseract
import uuid

def process_file(uploaded_file):
    """
    Input: Streamlit UploadedFile
    Output: (full_text, filename, method)
    """
    text = ""
    filename = uploaded_file.name
    method = "Fast Text"
    
    try:
        # 1. PDF Handling
        if filename.endswith(".pdf"):
            pdf_bytes = uploaded_file.getvalue()
            reader = pypdf.PdfReader(uploaded_file)
            
            for i, page in enumerate(reader.pages):
                extracted = page.extract_text()
                if extracted: 
                    text += f"\n[PAGE {i+1}] {extracted}"
            
            # OCR Fallback
            if len(text.strip()) < 50: 
                method = "OCR (Slow)"
                images = convert_from_bytes(pdf_bytes)
                text = "" 
                for i, img in enumerate(images):
                    page_text = pytesseract.image_to_string(img)
                    text += f"\n[PAGE {i+1}] {page_text}"

        # 2. Word Handling
        elif filename.endswith(".docx"):
            doc = docx.Document(uploaded_file)
            text = "\n".join([para.text for para in doc.paragraphs])

        # 3. Excel/CSV Handling (NEW)
        elif filename.endswith(".csv"):
            df = pd.read_csv(uploaded_file)
            text = df.to_string(index=False)
            method = "Table Parse"
            
        elif filename.endswith(".xlsx") or filename.endswith(".xls"):
            df = pd.read_excel(uploaded_file)
            text = df.to_string(index=False)
            method = "Table Parse"

        # 4. Plain Text
        elif filename.endswith(".txt"):
            text = uploaded_file.read().decode("utf-8")
            
    except Exception as e: 
        return "", filename, f"Error: {str(e)}"
        
    return text, filename, method

def chunk_text(text, source, chunk_size=500, overlap=100):
    """
    Generates chunks AND assigns a unique doc_id to link them together.
    """
    words = text.split()
    chunks = []
    doc_id = str(uuid.uuid4()) # Generate ID once per document
    
    for i in range(0, len(words), chunk_size - overlap):
        chunk_text = " ".join(words[i:i + chunk_size])
        if len(chunk_text) > 20: # Minimal filter
            chunks.append({
                "text": chunk_text, 
                "source": source,
                "doc_id": doc_id,
                "chunk_id": str(uuid.uuid4())
            })
    
    return chunks, doc_id