diff --git a/compression_utils.py b/compression_utils.py
new file mode 100644
index 0000000000..2a26a1c9d2
--- /dev/null
+++ b/compression_utils.py
@@ -0,0 +1,50 @@
+import lzma
+from typing import BinaryIO, Union
+
+def lzma2_compress(data: bytes, format: str = "x86", preset: int = 9) -> bytes:
+    """
+    Compress data using LZMA2 algorithm.
+
+    :param data: The data to compress.
+    :param format: The format of the filter chain. Default is "x86".
+    :param preset: The compression preset. Default is 9.
+    :return: The compressed data.
+    """
+    compressed_data = lzma.compress(data, format=format, preset=preset)
+    return compressed_data
+
+def lzma2_decompress(data: bytes) -> bytes:
+    """
+    Decompress data compressed with LZMA2 algorithm.
+
+    :param data: The compressed data.
+    :return: The decompressed data.
+    """
+    decompressed_data = lzma.decompress(data)
+    return decompressed_data
+
+def lzma2_compress_file(input_file: Union[str, BinaryIO], output_file: Union[str, BinaryIO], format: str = "x86", preset: int = 9) -> None:
+    """
+    Compress a file using LZMA2 algorithm.
+
+    :param input_file: The input file to compress. Can be a file path or a binary file object.
+    :param output_file: The output file to write the compressed data. Can be a file path or a binary file object.
+    :param format: The format of the filter chain. Default is "x86".
+    :param preset: The compression preset. Default is 9.
+    """
+    with open(input_file, "rb") as f_in:
+        with open(output_file, "wb") as f_out:
+            compressed_data = lzma2_compress(f_in.read(), format=format, preset=preset)
+            f_out.write(compressed_data)
+
+def lzma2_decompress_file(compressed_file: Union[str, BinaryIO], decompressed_file: Union[str, BinaryIO]) -> None:
+    """
+    Decompress a file compressed with LZMA2 algorithm.
+
+    :param compressed_file: The compressed file to decompress. Can be a file path or a binary file object.
+    :param decompressed_file: The output file to write the decompressed data. Can be a file path or a binary file object.
+    """
+    with open(compressed_file, "rb") as f_in:
+        with open(decompressed_file, "wb") as f_out:
+            decompressed_data = lzma2_decompress(f_in.read())
+            f_out.write(decompressed_data)
\ No newline at end of file
diff --git a/tests/test_compression.py b/tests/test_compression.py
new file mode 100644
index 0000000000..760e0fd581
--- /dev/null
+++ b/tests/test_compression.py
@@ -0,0 +1,75 @@
+ import py7zr
+import shutil
+
+def test_lzma2_compression():
+    # Test various input sizes
+    small_data = b"This is a small test string"
+    large_data = b"This is a large test string, repeated many times over. This is a large test string, repeated many times over."
+
+    # Verify compression and decompression
+    compressed_data = compress_lzma2(small_data)
+    decompressed_data = decompress_lzma2(compressed_data)
+    assert decompressed_data == small_data
+
+    compressed_data = compress_lzma2(large_data)
+    decompressed_data = decompress_lzma2(compressed_data)
+    assert decompressed_data == large_data
+
+    # Check compression ratio
+    compression_ratio = calculate_compression_ratio(small_data, compressed_data)
+    assert compression_ratio > 0.5, "Compression ratio for small data is too low"
+
+    compression_ratio = calculate_compression_ratio(large_data, compressed_data)
+    assert compression_ratio > 0.8, "Compression ratio for large data is too low"
+
+    # Validate error handling
+    with pytest.raises(py7zr.exceptions.ChecksumError):
+        decompress_lzma2(compressed_data[:-1])
+
+    with pytest.raises(TypeError):
+        compress_lzma2("This is invalid input")
+
+    # Test memory efficiency and resource usage
+    # (This is a best-effort test and may not be accurate on all systems)
+    memory_usage_before = resource.getrusage(resource.RUSAGE_SELF).ru_maxrss
+    compress_lzma2(large_data * 100)
+    memory_usage_after = resource.getrusage(resource.RUSAGE_SELF).ru_maxrss
+    assert memory_usage_after - memory_usage_before < 1e6, "Memory usage during compression is too high"
+
+def compress_lzma2(data: bytes) -> bytes:
+    """
+    Compress data using the LZMA2 algorithm.
+
+    :param data: The data to compress.
+    :return: The compressed data.
+    """
+    with py7zr.SevenZipFile(mode="w", format="7z", compression_level=9) as z:
+        z.write("data.bin", data)
+    with open("data.7z", "rb") as f:
+        compressed_data = f.read()
+    shutil.os.remove("data.7z")
+    return compressed_data
+
+def decompress_lzma2(compressed_data: bytes) -> bytes:
+    """
+    Decompress data using the LZMA2 algorithm.
+
+    :param compressed_data: The compressed data.
+    :return: The decompressed data.
+    """
+    with open("data.7z", "wb") as f:
+        f.write(compressed_data)
+    with py7zr.SevenZipFile("data.7z", mode="r") as z:
+        decompressed_data = z.read("data.bin")
+    shutil.os.remove("data.7z")
+    return decompressed_data
+
+def calculate_compression_ratio(original_data: bytes, compressed_data: bytes) -> float:
+    """
+    Calculate the compression ratio of compressed_data compared to original_data.
+
+    :param original_data: The original data.
+    :param compressed_data: The compressed data.
+    :return: The compression ratio as a float.
+    """
+    return len(original_data) / len(compressed_data)
\ No newline at end of file