processor.py

import layoutparser as lp
import pytesseract
from pdf2image import convert_from_path
from PIL import Image

def load_images(uploaded_file):
    if uploaded_file.name.endswith(".pdf"):
        return convert_from_path(uploaded_file)
    else:
        return [Image.open(uploaded_file)]

def analyze_layout(image):
    model = lp.EfficientDetLayoutModel(
        "lp://efficientdet/PubLayNet",
        extra_config=["MODEL.ROI_HEADS.SCORE_THRESH_TEST", 0.6],
        label_map={0: "Text", 1: "Title", 2: "List", 3: "Table", 4: "Figure"}
    )
    return model.detect(image)

def extract_text_from_blocks(image, layout):
    blocks = []
    for block in layout:
        segment_image = block.crop_image(image)
        text = pytesseract.image_to_string(segment_image)
        blocks.append({
            "type": block.type,
            "text": text.strip(),
            "coordinates": block.coordinates
        })
    return blocks

def rule_based_kv_extraction(blocks):
    data = {}
    for block in blocks:
        text = block["text"].lower()
        if "invoice" in text:
            data["Invoice Number"] = block["text"]
        elif "total" in text:
            data["Total Amount"] = block["text"]
        elif "customer" in text:
            data["Customer Name"] = block["text"]
    return data