validator.py

import os
import sys
import csv
import pandas as pd

# --- Step 1: Input Validation ---
def validate_input_file(file_path):
    """Checks if the file exists and is accessible."""
    if not os.path.exists(file_path):
        print(f"Error: {file_path} does not exist.")
        sys.exit(1)

def validate_directory(directory):
    """Checks if the directory exists."""
    if not os.path.isdir(directory):
        print(f"Error: {directory} is not a valid directory.")
        sys.exit(1)

# --- Step 2: CSV File Processing ---
def check_files_in_csv_exist(csv_filename):
    """Checks if all files listed in the first column of the CSV exist."""
    validate_input_file(csv_filename)
    cwd = os.getcwd()
    
    # Open and read the CSV file
    with open(csv_filename, mode='r', newline='') as csvfile:
        csvreader = csv.reader(csvfile)
        next(csvreader)  # Skip the header
        missing_files = []
        for row in csvreader:
            if row:
                file_path = row[0]
                full_path = os.path.join(cwd, file_path)
                if not os.path.exists(full_path):
                    missing_files.append(full_path)
        
    if missing_files:
        print("The following files do not exist:")
        for missing_file in missing_files:
            print(missing_file)
    else:
        print("All files listed in the CSV exist.")

# --- Step 3: Checking for Duplicate Keys in ASEG Files ---
def check_repeated_keys_in_file(file_path):
    """Check if the first column of a tab-separated file has repeated keys."""
    try:
        df = pd.read_csv(file_path, sep='\t', index_col=0)
        duplicated = df.index.duplicated(keep=False)
        return duplicated.any()  # Returns True if duplicates exist
    except Exception as e:
        print(f"Error reading {file_path}: {e}")
        return False

def find_files_with_repeated_keys(directory):
    """Finds all .txt files in the directory that contain repeated keys."""
    validate_directory(directory)
    files_with_repeats = []
    for root, _, files in os.walk(directory):
        for file in files:
            if file.endswith(".txt"):
                file_path = os.path.join(root, file)
                if check_repeated_keys_in_file(file_path):
                    print(f"Repeated rows found in: {file_path}")
                    files_with_repeats.append(file_path)
    
    if not files_with_repeats:
        print("No files with repeated keys found.")
    return files_with_repeats

# --- Step 4: Check if ASEG Files Contain All ROIs ---
def read_first_column(file_path):
    """Reads and returns the first column from a tab-delimited file."""
    try:
        df = pd.read_csv(file_path, delimiter='\t', header=None)
        return df.iloc[:, 0]
    except Exception as e:
        print(f"Error reading {file_path}: {e}")
        return pd.Series()

def check_rois(rois_file, aseg_file):
    """Checks if the ASEG file contains all ROIs listed in the ROIs file."""
    try:
        rois = set(read_first_column(rois_file))
        aseg = set(read_first_column(aseg_file))
        missing_rois = rois - aseg  # ROIs in rois_file but not in aseg_file

        if missing_rois:
            print(f"{aseg_file} is missing the following ROIs: {', '.join(missing_rois)}")
        else:
            print(f"{aseg_file} contains all the ROIs.")
    except Exception as e:
        print(f"An error occurred while checking ROIs: {e}")

def check_aseg_files_for_rois(rois_file, directory):
    """Checks all .txt ASEG files in a directory for missing ROIs."""
    validate_input_file(rois_file)
    validate_directory(directory)
    
    txt_files = [f for f in os.listdir(directory) if f.endswith('.txt') and f != rois_file]
    
    for file in txt_files:
        aseg_file_path = os.path.join(directory, file)
        check_rois(rois_file, aseg_file_path)

# --- Main Execution Flow ---
if __name__ == "__main__":
    # Example usage based on command-line arguments
    if len(sys.argv) < 2:
        print("Usage: python validator.py <csv_filename> <rois_file> <directory>")
        sys.exit(1)
    
    # Extract input arguments
    csv_filename = sys.argv[1]
    rois_file = sys.argv[2]
    directory = sys.argv[3]

    # Validate CSV file for existence of paths
    check_files_in_csv_exist(csv_filename)
    
    # Check for repeated keys in .txt files in the directory
    find_files_with_repeated_keys(directory)
    
    # Check ASEG files for missing ROIs
    check_aseg_files_for_rois(rois_file, directory)