pycadu 0.1.0

Satellite-agnostic CCSDS/CADU/VCDU frame handling

Satellite-agnostic CCSDS/CADU/VCDU frame handling in Python.

Reads raw downlink files containing CADU (Channel Access Data Unit) blocks, reassembles VCDU payloads, and yields fixed-length instrument frames delimited by a caller-supplied sync pattern. Works with any CCSDS-compliant downlink: VIIRS, MERSI, MODIS, etc.

Installation

pip install pycadu

Quick start

from pycadu import iter_vcdu_frames

MERSI_SYNC = bytes([0xAA, 0x55] * 6)
MERSI_FRAME_SIZE = 0x38E4F4
MERSI_VCID = 3

for frame in iter_vcdu_frames(
    "FY3D_20240101.dat",
    pattern=MERSI_SYNC,
    frame_size=MERSI_FRAME_SIZE,
    vcid=MERSI_VCID,
    stride=1024,          # or 896 / 1072 depending on ground station
):
    process(frame)

API

pycadu

• iter_vcdu_frames(path, pattern, frame_size, vcid, stride) — top-level convenience.

pycadu.cadu

• detect_cadu_stride(data) — infer stride (896 / 1024 / 1072) from a sample.

• extract_cadu_data(block, stride) — strip RS parity / annotation bytes.

• extract_vcdu_payload(cadu) — strip CCSDS primary header.

• extract_vcid(block) — read the 6-bit virtual channel identifier.

• CADU_STRIDE_1072 — 1072-byte stride (1024-byte CADU + 48 annotation bytes).

pycadu.sync

• find_sync_pattern(data, pattern, start) — search at all 8 bit offsets.

• _build_shifted(data, shift) — build a bit-shifted copy of data.

pycadu.reader

• cadu_blocks(path, stride) — stream physical blocks from a file.

• vcdu_payloads(blocks, stride, target_vcid) — filter by VCID, strip headers.

• vcdu_frames(payloads, pattern, frame_size) — reassemble variable-length chunks into fixed-length frames, handling 0-7 bit shifts automatically.

• read_frames_batch(path, stride, vcid, pattern, frame_size) — memory-mapped bulk reader, ~20x faster than the streaming path on large files.

pycadu.constants

CCSDS_SYNC_MARKER, CADU_STRIDE_896, CADU_STRIDE_1024, RS_PARITY_LENGTH, CADU_DATA_LENGTH, CCSDS_HEADER_LENGTH, VCDU_PAYLOAD_LENGTH.