conda create -n ncbi_dl -y -c conda-forge ncbi-datasets-cli jq

for f in batches/batch_*; do
  base=$(basename "$f")
  out="zips/${base}.zip"

  echo "Downloading $f -> $out"
  datasets download genome accession --inputfile "$f" \
    --include genome,gff3,gbff \
    --filename "$out"

  # 解压到独立目录（也可以最后统一解压）
  mkdir -p "genomes/$base"
  unzip -q "$out" -d "genomes/$base"

  # 关键：主动降速，避免触发 NCBI 限制
  sleep 10
done


#!/usr/bin/env bash

mkdir -p zips genomes

for f in batches/batch_*; do
  base=$(basename "$f")
  out="zips/${base}.zip"
  dir="genomes/$base"

  # 如果已经有对应解压目录，就认为完成，直接跳过
  if [[ -d "$dir" ]]; then
    echo "SKIP (exists): $dir"
    continue
  fi

  echo "Downloading $f -> $out"
  datasets download genome accession --inputfile "$f" \
    --include genome,gff3,gbff \
    --filename "$out"

  mkdir -p "$dir"
  unzip -q "$out" -d "$dir"

  # 主动降速，避免触发 NCBI 限制
  sleep 10
done