Secure your code as it's written. Use Snyk Code to scan source code in minutes - no build needed - and fix issues immediately.
def test_multiple(self):
if not hasattr(zstd, "BufferWithSegments"):
self.skipTest("BufferWithSegments not available")
b = zstd.BufferWithSegments(
b"foofooxfooxy", b"".join([ss.pack(0, 3), ss.pack(3, 4), ss.pack(7, 5)])
)
self.assertEqual(len(b), 3)
self.assertEqual(b.size, 12)
self.assertEqual(b.tobytes(), b"foofooxfooxy")
self.assertEqual(b[0].tobytes(), b"foo")
self.assertEqual(b[1].tobytes(), b"foox")
self.assertEqual(b[2].tobytes(), b"fooxy")
frames = cctx.multi_compress_to_buffer(original)
# Check round trip.
dctx = zstd.ZstdDecompressor()
decompressed = dctx.multi_decompress_to_buffer(frames, threads=3)
self.assertEqual(len(decompressed), len(original))
for i, data in enumerate(original):
self.assertEqual(data, decompressed[i].tobytes())
# And a manual mode.
b = b"".join([frames[0].tobytes(), frames[1].tobytes()])
b1 = zstd.BufferWithSegments(
b, struct.pack("=QQQQ", 0, len(frames[0]), len(frames[0]), len(frames[1]))
)
b = b"".join([frames[2].tobytes(), frames[3].tobytes(), frames[4].tobytes()])
b2 = zstd.BufferWithSegments(
b,
struct.pack(
"=QQQQQQ",
0,
len(frames[2]),
len(frames[2]),
len(frames[3]),
len(frames[2]) + len(frames[3]),
len(frames[4]),
),
)
def test_invalid_offset(self):
if not hasattr(zstd, "BufferWithSegments"):
self.skipTest("BufferWithSegments not available")
with self.assertRaisesRegex(
ValueError, "offset within segments array references memory"
):
zstd.BufferWithSegments(b"foo", ss.pack(0, 4))
b"foo3" * 3,
b"foo4" * 4,
b"foo5" * 5,
]
frames = [cctx.compress(c) for c in original]
b = b"".join([original[0], original[1]])
b1 = zstd.BufferWithSegments(
b,
struct.pack(
"=QQQQ", 0, len(original[0]), len(original[0]), len(original[1])
),
)
b = b"".join([original[2], original[3], original[4]])
b2 = zstd.BufferWithSegments(
b,
struct.pack(
"=QQQQQQ",
0,
len(original[2]),
len(original[2]),
len(original[3]),
len(original[2]) + len(original[3]),
len(original[4]),
),
)
c = zstd.BufferWithSegmentsCollection(b1, b2)
result = cctx.multi_compress_to_buffer(c)
def test_arguments(self):
if not hasattr(zstd, "BufferWithSegments"):
self.skipTest("BufferWithSegments not available")
with self.assertRaises(TypeError):
zstd.BufferWithSegments()
with self.assertRaises(TypeError):
zstd.BufferWithSegments(b"foo")
# Segments data should be a multiple of 16.
with self.assertRaisesRegex(
ValueError, "segments array size is not a multiple of 16"
):
zstd.BufferWithSegments(b"foo", b"\x00\x00")
def test_buffer_with_segments_input(self):
cctx = zstd.ZstdCompressor(write_checksum=True)
if not hasattr(cctx, "multi_compress_to_buffer"):
self.skipTest("multi_compress_to_buffer not available")
original = [b"foo" * 4, b"bar" * 6]
frames = [cctx.compress(c) for c in original]
offsets = struct.pack(
"=QQQQ", 0, len(original[0]), len(original[0]), len(original[1])
)
segments = zstd.BufferWithSegments(b"".join(original), offsets)
result = cctx.multi_compress_to_buffer(segments)
self.assertEqual(len(result), 2)
self.assertEqual(result.size(), 47)
self.assertEqual(result[0].tobytes(), frames[0])
self.assertEqual(result[1].tobytes(), frames[1])
def test_length(self):
if not hasattr(zstd, "BufferWithSegmentsCollection"):
self.skipTest("BufferWithSegmentsCollection not available")
b1 = zstd.BufferWithSegments(b"foo", ss.pack(0, 3))
b2 = zstd.BufferWithSegments(
b"barbaz", b"".join([ss.pack(0, 3), ss.pack(3, 3)])
)
c = zstd.BufferWithSegmentsCollection(b1)
self.assertEqual(len(c), 1)
self.assertEqual(c.size(), 3)
c = zstd.BufferWithSegmentsCollection(b2)
self.assertEqual(len(c), 2)
self.assertEqual(c.size(), 6)
c = zstd.BufferWithSegmentsCollection(b1, b2)
self.assertEqual(len(c), 3)
self.assertEqual(c.size(), 9)
kwargs = {}
if fn.threads_arg:
kwargs["threads"] = batch_threads
# Pass compressed frames in a BufferWithSegments rather than a list
# of bytes.
if fn.chunks_as_buffer:
s = struct.Struct("=QQ")
offsets = io.BytesIO()
current_offset = 0
for chunk in compressed_chunks:
offsets.write(s.pack(current_offset, len(chunk)))
current_offset += len(chunk)
chunks_arg = zstd.BufferWithSegments(
b"".join(compressed_chunks), offsets.getvalue()
)
if fn.decompressed_sizes_arg:
# Ideally we'd use array.array here. But Python 2 doesn't support the
# Q format.
s = struct.Struct("=Q")
kwargs["decompressed_sizes"] = b"".join(s.pack(len(c)) for c in orig_chunks)
results = timer(lambda: fn(chunks_arg, dopts, **kwargs))
format_results(results, fn.title, prefix, total_size)
for fn in get_benches("discrete", "compress"):
chunks_arg = chunks
kwargs = {}
if fn.threads_arg:
kwargs["threads"] = batch_threads
if fn.chunks_as_buffer:
s = struct.Struct("=QQ")
offsets = io.BytesIO()
current_offset = 0
for chunk in chunks:
offsets.write(s.pack(current_offset, len(chunk)))
current_offset += len(chunk)
chunks_arg = zstd.BufferWithSegments(b"".join(chunks), offsets.getvalue())
results = timer(lambda: fn(chunks_arg, zparams, **kwargs))
format_results(results, fn.title, prefix, total_size)