app.py

import gradio as gr
import pandas as pd
import psycopg2
from algo import Algo
from psycopg2.extras import DictCursor
from db.db_utils import get_connection
from multi_food_item_detector import has_delimiters, extract_items

db_conn = get_connection()
db_cursor = db_conn.cursor(cursor_factory=psycopg2.extras.DictCursor)
algo = Algo(db_conn)

def process_input(input_text, csv_file):
    # if csv_file is not None:
    #     # Read the uploaded CSV file
    #     df = pd.read_csv(csv_file.name)
    #     # Check if 'description' column exists
    #     if 'description' in df.columns:
    #         descriptions = df['description'].tolist()
    #         results = algo.match_words(descriptions)
    #     else:
    #         return pd.DataFrame({"Error": ["CSV file must have a 'description' column"]})
    # else:
        # Process the single input text

    input_items = [[input_text, None]]
    if has_delimiters(input_text):
        items = extract_items(input_text)
        items = [[item, None] for item in items]
        input_items.extend(items)
    
    results = algo.match_words(input_items)


    print(f" - result -> {results}")
    df = pd.DataFrame(results, columns=["input_word", "cleaned_word", 'dictionary_word', "specificity", 'sr_legacy_food_category', 'wweia_category', 'dry_matter_content', 
    'water_content', 'similarity_score', 'confidence_score', 'similar_words', 'is_food', 'food_nonfood_score'])
    # Filter to only required columns
    df_filtered = df[["input_word", "dictionary_word", "specificity", "is_food", "sr_legacy_food_category", 'wweia_category', 'dry_matter_content', "water_content",  "similarity_score", "food_nonfood_score"]]
    return df_filtered

# Gradio interface
with gr.Blocks() as demo:
    with gr.Column():
        with gr.Row():
            input_text = gr.Textbox(label="Enter a food item", placeholder="e.g. apple")
            # csv_input = gr.File(label="Upload a CSV file (optional)")
            submit_button = gr.Button("Submit")
        output_table = gr.DataFrame(label="Processed Results")
    
    input_text.submit(fn=process_input, inputs=[input_text,], outputs=output_table)
    submit_button.click(fn=process_input, inputs=[input_text,], outputs=output_table)
    
demo.launch()
db_cursor.close()
db_conn.close()