High-performance pure Python ROS2 message serialization and deserialization for MCAP files
Verified details
These details have been verified by PyPIProject links
GitHub Statistics
Maintainers
Project description
mcap-ros2-support-fast
High-performance pure Python ROS2 message serialization and deserialization for MCAP files.
No dependencies on ROS2 or a ROS2 environment.
Installation
uv add mcap-ros2-support-fast
Usage
Decoding Messages
from small_mcap import read_message_decoded
from mcap_ros2_support_fast.decoder import DecoderFactory
decoder_factory = DecoderFactory()
with open("recording.mcap", "rb") as f:
for msg in read_message_decoded(f, decoder_factories=[decoder_factory]):
print(f"{msg.channel.topic}: {msg.decoded_message}")
Encoding Messages
from small_mcap import McapWriter
from mcap_ros2_support_fast import ROS2EncoderFactory
encoder_factory = ROS2EncoderFactory()
with open("output.mcap", "wb") as f:
writer = McapWriter(f, encoder_factory=encoder_factory)
writer.start(profile="ros2")
schema_id = writer.add_schema(
name="geometry_msgs/msg/Point",
encoding="ros2msg",
data=b"float64 x\nfloat64 y\nfloat64 z"
)
channel_id = writer.add_channel("/point", "cdr", schema_id=schema_id)
point = {"x": 1.0, "y": 2.0, "z": 3.0}
writer.add_message_encode(channel_id, log_time=0, data=point)
writer.finish()
Benchmarks
Full benchmark suite comparing against reference mcap-ros2 implementation, rosbags library, and pybag.
Read Performance
-------------------------------------------------------------------------------------- benchmark 'msgs-10': 4 tests --------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_decoder[mcap_ros2_fast-10] 5.9309 (1.0) 6.6827 (1.0) 6.2876 (1.0) 0.2078 (1.0) 6.2848 (1.0) 0.3145 (1.36) 159.0425 (1.0) 24 1
test_benchmark_decoder[rosbags-10] 7.8039 (1.32) 9.4448 (1.41) 8.2345 (1.31) 0.3564 (1.72) 8.1815 (1.30) 0.2311 (1.0) 121.4397 (0.76) 17 1
test_benchmark_decoder[pybag-10] 28.2604 (4.76) 34.4526 (5.16) 29.2232 (4.65) 1.5174 (7.30) 28.7849 (4.58) 0.7441 (3.22) 34.2194 (0.22) 17 1
test_benchmark_decoder[mcap_ros2-10] 344.9453 (58.16) 393.5084 (58.88) 371.4299 (59.07) 19.4372 (93.54) 368.3629 (58.61) 30.0467 (130.00) 2.6923 (0.02) 5 1
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------- benchmark 'msgs-100': 4 tests ------------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_decoder[mcap_ros2_fast-100] 36.1641 (1.0) 41.7886 (1.0) 38.7671 (1.0) 2.4714 (1.0) 37.0456 (1.0) 4.7878 (1.0) 25.7951 (1.0) 17 1
test_benchmark_decoder[rosbags-100] 64.0870 (1.77) 73.8624 (1.77) 67.7211 (1.75) 3.3350 (1.35) 66.9847 (1.81) 5.7329 (1.20) 14.7665 (0.57) 13 1
test_benchmark_decoder[pybag-100] 116.3701 (3.22) 126.1597 (3.02) 120.2669 (3.10) 3.8116 (1.54) 118.5586 (3.20) 6.1874 (1.29) 8.3148 (0.32) 7 1
test_benchmark_decoder[mcap_ros2-100] 1,304.4500 (36.07) 1,341.5722 (32.10) 1,325.5776 (34.19) 16.4693 (6.66) 1,332.5642 (35.97) 28.6133 (5.98) 0.7544 (0.03) 5 1
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------- benchmark 'msgs-1000': 3 tests -----------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_decoder[mcap_ros2_fast-1000] 126.9185 (1.0) 136.1736 (1.0) 132.6490 (1.0) 3.0011 (1.22) 133.3180 (1.0) 3.0429 (1.07) 7.5387 (1.0) 7 1
test_benchmark_decoder[rosbags-1000] 230.6469 (1.82) 236.9495 (1.74) 234.7352 (1.77) 2.4632 (1.0) 235.5722 (1.77) 2.8371 (1.0) 4.2601 (0.57) 5 1
test_benchmark_decoder[mcap_ros2-1000] 4,034.0901 (31.78) 4,092.9589 (30.06) 4,075.4871 (30.72) 24.8972 (10.11) 4,087.6284 (30.66) 31.2742 (11.02) 0.2454 (0.03) 5 1
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Read Summary:
- 31-59x faster than reference mcap-ros2
- 1.31-1.77x faster than rosbags
- 3.10-4.65x faster than pybag (10-100 msgs; pybag skipped at 1000 due to decoder limitations)
Write Performance (Read + Write)
------------------------------------------------------------------------------------------ benchmark 'write-msgs-10': 4 tests -----------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_read_and_write[mcap_ros2_fast_writer-10] 17.6656 (1.0) 24.6546 (1.17) 19.1740 (1.0) 1.8360 (5.76) 18.8823 (1.0) 1.1571 (2.74) 52.1540 (1.0) 12 1
test_benchmark_read_and_write[rosbags_writer-10] 19.9034 (1.13) 21.1216 (1.0) 20.3388 (1.06) 0.3189 (1.0) 20.2124 (1.07) 0.4219 (1.0) 49.1670 (0.94) 13 1
test_benchmark_read_and_write[pybag_writer-10] 61.6168 (3.49) 84.1206 (3.98) 66.9191 (3.49) 6.0055 (18.83) 65.0424 (3.44) 6.6212 (15.69) 14.9434 (0.29) 13 1
test_benchmark_read_and_write[mcap_ros2_writer-10] 920.7613 (52.12) 992.4728 (46.99) 948.9506 (49.49) 35.9130 (112.61) 925.6756 (49.02) 64.1231 (151.99) 1.0538 (0.02) 5 1
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------- benchmark 'write-msgs-100': 4 tests ----------------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_read_and_write[mcap_ros2_fast_writer-100] 108.6150 (1.0) 2,200.8597 (15.12) 380.2710 (2.84) 735.6531 (95.09) 121.3869 (1.0) 8.9596 (1.0) 2.6297 (0.35) 8 1
test_benchmark_read_and_write[rosbags_writer-100] 125.4855 (1.16) 145.5714 (1.0) 134.0954 (1.0) 7.7364 (1.0) 136.0569 (1.12) 12.2291 (1.36) 7.4574 (1.0) 7 1
test_benchmark_read_and_write[pybag_writer-100] 264.1394 (2.43) 290.3523 (1.99) 274.4185 (2.05) 9.9997 (1.29) 273.6181 (2.25) 12.1666 (1.36) 3.6441 (0.49) 5 1
test_benchmark_read_and_write[mcap_ros2_writer-100] 3,102.7205 (28.57) 3,390.4792 (23.29) 3,253.1039 (24.26) 126.2048 (16.31) 3,229.5814 (26.61) 225.6545 (25.19) 0.3074 (0.04) 5 1
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------- benchmark 'write-msgs-1000': 3 tests ----------------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_read_and_write[mcap_ros2_fast_writer-1000] 389.8025 (1.0) 418.4601 (1.0) 403.8736 (1.0) 12.1332 (1.0) 401.6407 (1.0) 21.0602 (1.0) 2.4760 (1.0) 5 1
test_benchmark_read_and_write[rosbags_writer-1000] 479.7473 (1.23) 513.6734 (1.23) 493.5608 (1.22) 15.2506 (1.26) 488.2160 (1.22) 27.0344 (1.28) 2.0261 (0.82) 5 1
test_benchmark_read_and_write[mcap_ros2_writer-1000] 8,899.2066 (22.83) 10,416.2178 (24.89) 9,589.6736 (23.74) 542.5575 (44.72) 9,552.1125 (23.78) 491.9584 (23.36) 0.1043 (0.04) 5 1
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Read+Write Summary:
- 23-49x faster than reference mcap-ros2
- 1.06-1.22x faster than rosbags
- 2.05-3.49x faster than pybag (10-100 msgs; pybag skipped at 1000 due to decoder limitations)
Write-Only Performance
-------------------------------------------------------------------------------- benchmark 'write-only-msgs-10': 4 tests ---------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_write_only[rosbags-10] 10.9500 (1.0) 14.6925 (1.0) 12.0007 (1.0) 0.8779 (1.0) 11.6257 (1.0) 1.3553 (6.55) 83.3284 (1.0) 24 1
test_benchmark_write_only[mcap_ros2_fast-10] 11.8910 (1.09) 15.5508 (1.06) 12.5770 (1.05) 0.9914 (1.13) 12.1515 (1.05) 0.2069 (1.0) 79.5100 (0.95) 23 1
test_benchmark_write_only[pybag-10] 32.2464 (2.94) 40.7300 (2.77) 34.2348 (2.85) 1.8887 (2.15) 33.4549 (2.88) 2.4743 (11.96) 29.2101 (0.35) 27 1
test_benchmark_write_only[mcap_ros2-10] 562.1606 (51.34) 564.9182 (38.45) 563.6723 (46.97) 1.1536 (1.31) 563.7195 (48.49) 1.9709 (9.52) 1.7741 (0.02) 5 1
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------- benchmark 'write-only-msgs-100': 4 tests --------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_write_only[rosbags-100] 53.7435 (1.0) 87.0201 (1.30) 58.4993 (1.0) 7.7466 (4.42) 56.7353 (1.0) 2.6326 (1.0) 17.0942 (1.0) 16 1
test_benchmark_write_only[mcap_ros2_fast-100] 61.0728 (1.14) 66.7065 (1.0) 64.2220 (1.10) 1.7516 (1.0) 64.0696 (1.13) 2.7867 (1.06) 15.5710 (0.91) 14 1
test_benchmark_write_only[pybag-100] 107.3560 (2.00) 117.4027 (1.76) 112.6513 (1.93) 2.9156 (1.66) 111.6172 (1.97) 3.1993 (1.22) 8.8769 (0.52) 9 1
test_benchmark_write_only[mcap_ros2-100] 785.2788 (14.61) 792.8765 (11.89) 789.4166 (13.49) 3.3582 (1.92) 790.9028 (13.94) 5.7939 (2.20) 1.2668 (0.07) 5 1
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------- benchmark 'write-only-msgs-1000': 3 tests -------------------------------------------------------------------------------------
Name (time in ms) Min Max Mean StdDev Median IQR OPS Rounds Iterations
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_write_only[rosbags-1000] 181.6391 (1.0) 204.9827 (1.0) 188.3532 (1.0) 8.4449 (2.24) 186.1796 (1.0) 4.3215 (1.0) 5.3092 (1.0) 6 1
test_benchmark_write_only[mcap_ros2_fast-1000] 207.8413 (1.14) 217.3372 (1.06) 213.3029 (1.13) 3.7784 (1.0) 214.1898 (1.15) 5.7395 (1.33) 4.6882 (0.88) 5 1
test_benchmark_write_only[mcap_ros2-1000] 1,387.1190 (7.64) 1,442.0496 (7.03) 1,402.5220 (7.45) 22.3435 (5.91) 1,394.4735 (7.49) 15.6472 (3.62) 0.7130 (0.13) 5 1
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Write-Only Summary:
- 7-47x faster than reference mcap-ros2
- 0.91-1.05x similar to rosbags (rosbags slightly faster for write-only)
- 1.76-2.85x faster than pybag (10-100 msgs; pybag skipped at 1000 due to decoder limitations)
Project details
Verified details
These details have been verified by PyPIProject links
GitHub Statistics
Maintainers
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file mcap_ros2_support_fast-0.1.0.tar.gz.
File metadata
- Download URL: mcap_ros2_support_fast-0.1.0.tar.gz
- Upload date:
- Size: 17.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via: uv/0.10.6 {"installer":{"name":"uv","version":"0.10.6","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
a028dcf082a09457d8aafa63a75678660735e7503540a37eeaf2d99fd2d9bf50
|
|
| MD5 |
faa9433da6f25411e1ecb019da64fdde
|
|
| BLAKE2b-256 |
48f948a9a86aed673fa76512525682d8148246970750fa06bbe5e333f30aec84
|
File details
Details for the file mcap_ros2_support_fast-0.1.0-py3-none-any.whl.
File metadata
- Download URL: mcap_ros2_support_fast-0.1.0-py3-none-any.whl
- Upload date:
- Size: 22.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via: uv/0.10.6 {"installer":{"name":"uv","version":"0.10.6","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
b8507c1261a2aa49696c3f36ec70e1c01818574599ae499e5334fc8eaa0aae69
|
|
| MD5 |
5c233fe59dffed11aad2e99adf40a02b
|
|
| BLAKE2b-256 |
8145c509227c8d8776ab00ea9bdd75b3e0a7e90b9c2f99caf58657246fae214d
|
