import os
import shutil
from pathlib import Path
import sys
import subprocess

n_splits = int(sys.argv[1]) if len(sys.argv) > 1 else 12   # 默认12份，可传参修改
src_dir = Path('./ligand/pdbqt')
dst_root = Path('./ligand')

all_files = sorted(src_dir.glob('*.pdbqt'))
total = len(all_files)
if total == 0:
    raise ValueError("No pdbqt files found in ./ligand/pdbqt")

def split_list_evenly(lst, n):
    k, m = divmod(len(lst), n)
    return [lst[i * k + min(i, m):(i + 1) * k + min(i + 1, m)] for i in range(n)]

chunks = split_list_evenly(all_files, n_splits)

for i, files in enumerate(chunks, 1):
    dst_dir = dst_root / f"pdbqt{i}"
    dst_dir.mkdir(parents=True, exist_ok=True)
    for f in files:
        shutil.copy2(f, dst_dir / f.name)

sh_template = """#!/bin/bash
#DSUB -n vina_job{idx}
#DSUB -R 'cpu=32' 
#DSUB -aa
#DSUB --label arm
#DSUB -o vina_docking{idx}_cpu_job_%J.out 
date
echo "autodock vina docking task{idx}, version: 1.2.7" 
cd /share/home/lyzeng24/rdkit_script/vina
./scripts/batch_docking.sh ./receptor/TrpE_entry_1.pdbqt \\
                          ./config/TrpE_entry_1.box.txt \\
                          ./ligand/pdbqt{idx} \\
                          ./result/poses{idx} \\
                          ./result/batch_docking{idx}.log ./vina 
cd ./result/poses{idx} 
module load mamba
micromamba run -n vina mk_export.py ./*_out.pdbqt --suffix _converted
date
"""

submit_sh_list = []
for i in range(1, n_splits + 1):
    sh_content = sh_template.format(idx=i)
    sh_path = Path(f'./submit_vina{i}.sh')
    with open(sh_path, 'w') as f:
        f.write(sh_content)
    os.chmod(sh_path, 0o755)
    submit_sh_list.append(sh_path)

print(f"Done! {total} pdbqt files split to {n_splits} folders, {n_splits} shell scripts generated.")

import time

# 自动提交
for i, sh_path in enumerate(submit_sh_list, 1):
    print(f"提交: dsub -s {sh_path}")
    subprocess.run(['dsub', '-s', str(sh_path)])
    time.sleep(1)  # 每次提交后等待1秒