import fitz # PyMuPDF
import pandas as pd

# Load the PDF
pdf_path = "/home/ok/Desktop/Diabetes/1-s2.0-S2405457724001633-mmc1.pdf"
doc = fitz.open(pdf_path)

# Extract all text from the PDF
all_text = ""
for page in doc:
    all_text += page.get_text()
    
with open("all_text", "w", encoding="utf-8") as txt_file:
    txt_file.write(all_text)

# Attempt to extract food names and insulin index (II) values
import re

# Match patterns where food/beverage name is followed by insulin index
# Simplified pattern matching lines like "Number Food / Beverage Name ... II ..."
# We'll look for lines that seem like: Number. Food name ... II value
matches = re.findall(r"\d+\.\s+(.+?)\s+\d+SINGLE|MIXED MEAL\s+([\d.]+)", all_text)

# Flatten and clean data
data = []
for match in matches:
    name = match[0].strip()
    print(name)
    try:
        ii = float(match[1])
        data.append((name, ii))
    except ValueError:
        continue

# Create and sort DataFrame
df = pd.DataFrame(data, columns=["Food Name", "Insulin Index"])
df_sorted = df.sort_values(by="Insulin Index", ascending=False)

# Save to CSV
csv_path = "/home/ok/Desktop/Diabetes/insulin_index_sorted.csv"
df_sorted.to_csv(csv_path, index=False)

csv_path