Implement flatbuffer mode in log_to_mcap
This makes log_to_mcap output a flatbuffer-based MCAP format (if requested)
instead of a JSON format. This significantly speeds up conversion times
and reduces disk space used, and should at least put us in the ball-park
for working with images in foxglove (although we still need to do some
extra work to either use one of their pre-defined image formats or
support our existing image formats natively in studio).
Change-Id: Id1778c0b5b9ced90517ab34269511d78cd0c7e58
Signed-off-by: James Kuszmaul <james.kuszmaul@bluerivertech.com>
diff --git a/aos/util/log_to_mcap_test.py b/aos/util/log_to_mcap_test.py
index 86c3d1d..72c9078 100644
--- a/aos/util/log_to_mcap_test.py
+++ b/aos/util/log_to_mcap_test.py
@@ -22,7 +22,7 @@
subprocess.run([args.generate_log, "--output_folder", log_name]).check_returncode()
# Run with a really small chunk size, to force a multi-chunk file.
subprocess.run(
- [args.log_to_mcap, "--output_path", mcap_name, "--mcap_chunk_size", "1000",
+ [args.log_to_mcap, "--output_path", mcap_name, "--mcap_chunk_size", "1000", "--mode", "json",
log_name]).check_returncode()
# MCAP attempts to find $HOME/.mcap.yaml, and dies on $HOME not existing. So
# give it an arbitrary config location (it seems to be fine with a non-existent config).