quick_leftovers.py

import os
import shutil
import pandas as pd

# Configuration - modify these paths
csv_path = r"S:\qfl_australia_tls\Curtain Fig\trees_within_plot.csv"
source_folder = r"S:\qfl_australia_tls\Curtain Fig\output"
leftover_folder = r"S:\qfl_australia_tls\Curtain Fig\leftover_trees"

# Create leftover folder if it doesn't exist
os.makedirs(leftover_folder, exist_ok=True)

# Get CSV filenames (files to skip)
df = pd.read_csv(csv_path)
csv_filenames = set(df['filename'].tolist())

# Get all files in source
all_files = [f for f in os.listdir(source_folder) if os.path.isfile(os.path.join(source_folder, f))]

# Get already processed leftovers
already_leftover = set(os.listdir(leftover_folder)) if os.path.exists(leftover_folder) else set()

print(f"Total files in source: {len(all_files)}")
print(f"Files in CSV (to skip): {len(csv_filenames)}")
print(f"Already in leftover folder: {len(already_leftover)}")

copied = 0
skipped = 0

# Copy only leftover files (not in CSV, not already processed)
for file_name in all_files:
    if file_name in csv_filenames:
        # Skip - this file is in the selection CSV
        continue
    
    if file_name in already_leftover:
        # Skip - already copied to leftover folder
        skipped += 1
        continue
    
    # Copy to leftover folder
    source_path = os.path.join(source_folder, file_name)
    dest_path = os.path.join(leftover_folder, file_name)
    
    try:
        shutil.copy2(source_path, dest_path)
        copied += 1
        if copied % 100 == 0:
            print(f"Copied {copied} files...")
    except Exception as e:
        print(f"Error copying {file_name}: {e}")

print(f"\nDone! Copied {copied} new leftover files, skipped {skipped} already processed")