import pandas as pd
import pyarrow as pa
import pyarrow.ipc as ipc
import numpy as np
import os

# 确保输出目录存在
output_dir = 'generated_arrow_files'
os.makedirs(output_dir, exist_ok=True)

# 生成 Arrow 文件的函数
def generate_arrow_file(file_index):
    # 创建 Pandas DataFrame（每次生成不同的数据）
    data = {
        'integers': np.random.randint(0, 100, size=10),
        'floats': np.random.rand(10),
        'strings': [f'str_{i}' for i in range(10)],
        'booleans': np.random.choice([True, False], size=10)
    }

    df = pd.DataFrame(data)

    # 将 Pandas DataFrame 转换为 PyArrow Table
    table = pa.Table.from_pandas(df)

    # 写入内存中的字节流
    sink = pa.BufferOutputStream()
    with ipc.new_file(sink, table.schema) as writer:
        writer.write_table(table)

    # 获取字节流并保存为 .arrow 文件
    buffer = sink.getvalue()
    file_path = os.path.join(output_dir, f'example_{file_index}.arrow')
    with open(file_path, 'wb') as f:
        f.write(buffer.to_pybytes())

    print(f"Arrow file {file_index} generated successfully at {file_path}")

# 生成多个文件
for i in range(5):  # 生成5个文件，可以根据需要调整数量
    generate_arrow_file(i)