codec-re: solve waveform body block framing; per-byte sample mapping still open

Decoded the structural framing of the Blastware waveform body — the bytes between the 21-byte STRT record and the 26-byte file footer. The body is a sequence of tagged variable-length blocks, NOT raw int16 LE. Five tag types (10/20/00/30/40 NN) and their lengths are now confirmed against the 4-event May 2026 fixture bundle. Body splits cleanly into ~16 segments (for a 1280-sample event) separated by 40 02 segment headers carrying a monotonically incrementing uint32 LE counter at bytes [8:12]. What's done: - minimateplus/waveform_codec.py — block walker, segment splitter, segment header parser. decode_waveform_v2 is a stub returning None until the byte-to-sample mapping is solved; client.py is unchanged. - tests/test_waveform_codec.py — 31 tests covering block detection, lengths, contiguous-walk, segment splitting, segment-header parsing, and counter monotonicity. All pass. - tests/fixtures/decode-re-5-8-26/ — bundled fixtures (4 events, BW binary + Blastware ASCII export each). - docs/instantel_protocol_reference.md §7.6.1 — replaced retraction box with the verified structural decoding plus an explicit list of what's still open. What's still open: the per-byte mapping inside 10 NN / 20 NN blocks. 96 channel-permutation × nibble-order × sign-convention combinations were brute-force tested; none match BW's ASCII export to within ±1 ADC count. The codec is more elaborate than uniform 4-bit deltas — likely a hybrid variable-bit-width scheme with segment-anchor resync points. Next recommended step: capture an event with a known calibration tone to pin down magnitude scaling. Walker also bails out partway through event-b (open issue documented in both the module and the protocol reference).
2026-05-08 20:44:37 +00:00
parent 7bd0f8badf
commit d3f77d1d96
29 changed files with 10102 additions and 105 deletions
@@ -0,0 +1,252 @@
+"""
+Tests for minimateplus.waveform_codec — Blastware waveform-file body block walker.
+
+These tests lock in the STRUCTURAL framing of the body codec.  The byte-to-sample
+mapping is open (see waveform_codec module docstring) — until that's nailed down,
+:func:`decode_waveform_v2` returns ``None`` and there is no per-sample assertion
+to make.
+"""
+from __future__ import annotations
+
+import os
+
+import pytest
+
+from minimateplus.waveform_codec import (
+    WaveformBlock,
+    find_data_start,
+    parse_segment_header,
+    split_segments,
+    walk_body,
+    decode_waveform_v2,
+)
+
+
+FIXTURES = os.path.join(
+    os.path.dirname(__file__), "fixtures", "decode-re-5-8-26"
+)
+
+
+def _bw_body(path):
+    """Strip the 22-byte header and 21-byte STRT and 26-byte footer to get the body."""
+    with open(path, "rb") as f:
+        binary = f.read()
+    return binary[43:-26]
+
+
+# Fixture metadata — bundled BW binaries from a real BE11529 unit, May 8 2026.
+# Each is paired with a Blastware TXT export (the ASCII ground truth).
+FIXTURES_INFO = {
+    "event-a": {
+        "filename": "M529LKVQ.6S0",
+        "n_samples": 3328,    # 3.0 s rectime + 0.25 s pretrig at 1024 sps
+        "rectime": 3.0,
+    },
+    "event-b": {
+        "filename": "M529LK5Q.RG0",
+        "n_samples": 2304,    # 2.0 s
+        "rectime": 2.0,
+    },
+    "event-c": {
+        "filename": "M529LK44.AB0",
+        "n_samples": 1280,    # 1.0 s
+        "rectime": 1.0,
+    },
+    "event-d": {
+        "filename": "M529LK2V.470",
+        "n_samples": 1280,
+        "rectime": 1.0,
+    },
+}
+
+
+def _fixture_path(event_name):
+    info = FIXTURES_INFO[event_name]
+    return os.path.join(FIXTURES, event_name, info["filename"])
+
+
+# ── Find data start ──────────────────────────────────────────────────────────
+
+
+@pytest.mark.parametrize("event_name", list(FIXTURES_INFO.keys()))
+def test_find_data_start_locates_first_block(event_name):
+    """The walker auto-detects the first ``10 NN`` tag within the first 20 bytes."""
+    path = _fixture_path(event_name)
+    if not os.path.exists(path):
+        pytest.skip(f"fixture missing: {path}")
+    body = _bw_body(path)
+    start = find_data_start(body)
+    assert 0 <= start < 20, f"expected start in [0, 20), got {start}"
+    assert body[start] == 0x10
+    assert body[start + 1] % 4 == 0
+    assert 0 < body[start + 1] <= 0xFC
+
+
+def test_find_data_start_preamble_lengths():
+    """All 4 events have either a 7-byte (single-shot) or 9-byte (continuous) preamble."""
+    starts = {}
+    for name in FIXTURES_INFO:
+        path = _fixture_path(name)
+        if not os.path.exists(path):
+            pytest.skip(f"fixture missing: {path}")
+        body = _bw_body(path)
+        starts[name] = find_data_start(body)
+    # Empirically: events a, b have 9-byte preamble; events c, d have 7-byte.
+    assert starts["event-a"] == 9
+    assert starts["event-b"] == 9
+    assert starts["event-c"] == 7
+    assert starts["event-d"] == 7
+
+
+# ── Block walker ─────────────────────────────────────────────────────────────
+
+
+def test_walk_body_empty_returns_empty():
+    assert walk_body(b"") == []
+
+
+def test_walk_body_invalid_start_returns_empty():
+    # Body that does not begin with a recognized tag.
+    assert walk_body(b"\xff\xff\xff\xff", start=0) == []
+
+
+@pytest.mark.parametrize("event_name", list(FIXTURES_INFO.keys()))
+def test_walk_body_produces_blocks(event_name):
+    """The walker should produce a non-empty stream of blocks for every fixture."""
+    path = _fixture_path(event_name)
+    if not os.path.exists(path):
+        pytest.skip(f"fixture missing: {path}")
+    body = _bw_body(path)
+    blocks = walk_body(body)
+    assert len(blocks) > 0
+    # All blocks have one of the 5 known tag types.
+    for b in blocks:
+        assert b.tag_hi in (0x10, 0x20, 0x00, 0x30, 0x40), (
+            f"unknown tag {b.tag_hi:#04x} at offset {b.offset}"
+        )
+
+
+@pytest.mark.parametrize("event_name", list(FIXTURES_INFO.keys()))
+def test_walk_body_block_lengths_consistent(event_name):
+    """Each block's recorded length matches its on-wire footprint."""
+    path = _fixture_path(event_name)
+    if not os.path.exists(path):
+        pytest.skip(f"fixture missing: {path}")
+    body = _bw_body(path)
+    blocks = walk_body(body)
+    for b in blocks:
+        # Tag (2 bytes) + payload should equal length.
+        assert 2 + len(b.data) == b.length, (
+            f"block at {b.offset} length mismatch: tag(2) + data({len(b.data)}) != length({b.length})"
+        )
+
+
+@pytest.mark.parametrize("event_name", list(FIXTURES_INFO.keys()))
+def test_walk_body_blocks_contiguous(event_name):
+    """Block n+1 starts exactly where block n ends (no gaps, no overlaps)."""
+    path = _fixture_path(event_name)
+    if not os.path.exists(path):
+        pytest.skip(f"fixture missing: {path}")
+    body = _bw_body(path)
+    blocks = walk_body(body)
+    for i in range(1, len(blocks)):
+        prev = blocks[i - 1]
+        cur = blocks[i]
+        assert cur.offset == prev.offset + prev.length, (
+            f"gap/overlap between block {i-1} (off={prev.offset} len={prev.length}) "
+            f"and block {i} (off={cur.offset})"
+        )
+
+
+# ── Segment splitting ────────────────────────────────────────────────────────
+
+
+@pytest.mark.parametrize("event_name", list(FIXTURES_INFO.keys()))
+def test_split_segments_yields_at_least_one(event_name):
+    path = _fixture_path(event_name)
+    if not os.path.exists(path):
+        pytest.skip(f"fixture missing: {path}")
+    body = _bw_body(path)
+    blocks = walk_body(body)
+    segments = split_segments(blocks)
+    assert len(segments) > 0
+
+
+def test_split_segments_segment_count_at_least_one_per_event():
+    """The walker should produce at least one ``40 02`` segment header per event.
+
+    Note: the walker currently bails out partway through event-b (still an
+    open issue — the body codec uses block lengths the walker doesn't
+    handle correctly past offset ~427).  The other 3 events walk farther
+    and have many segment headers.
+    """
+    for name in FIXTURES_INFO:
+        path = _fixture_path(name)
+        if not os.path.exists(path):
+            continue
+        body = _bw_body(path)
+        blocks = walk_body(body)
+        n_40 = sum(1 for b in blocks if b.tag_hi == 0x40)
+        assert n_40 >= 1, f"{name}: no 40 02 segment header found"
+
+
+# ── Segment header parsing ───────────────────────────────────────────────────
+
+
+def test_parse_segment_header_returns_none_for_non_40():
+    block = WaveformBlock(offset=0, tag_hi=0x10, tag_lo=0x04, data=b"\x00\x00", length=4)
+    assert parse_segment_header(block) is None
+
+
+def test_parse_segment_header_decodes_fields():
+    """Decode a known 40 02 block to verify field offsets."""
+    # First segment header from event-c at body offset 235:
+    # 40 02 00 00 00 00 0a 4b 01 1e 47 00 00 00 02 00 00 01 00 01
+    payload = bytes.fromhex("00000000 0a4b011e 47000000 02000001 0001".replace(" ", ""))
+    block = WaveformBlock(
+        offset=235, tag_hi=0x40, tag_lo=0x02, data=payload, length=20
+    )
+    decoded = parse_segment_header(block)
+    assert decoded is not None
+    assert decoded["counter"] == 0x47       # uint32 LE
+    assert decoded["fixed_pattern"] == b"\x02\x00\x00\x01"
+    assert decoded["anchor_bytes"] == b"\x00\x00\x00\x00"
+
+
+def test_segment_counter_increments():
+    """The 4-byte counter at bytes [8:12] of each 40 02 payload increments by 1."""
+    path = _fixture_path("event-c")
+    if not os.path.exists(path):
+        pytest.skip("fixture missing")
+    body = _bw_body(path)
+    blocks = walk_body(body)
+    headers = [b for b in blocks if b.tag_hi == 0x40 and b.tag_lo == 0x02]
+    counters = [parse_segment_header(b)["counter"] for b in headers]
+    assert len(counters) >= 5, "expect at least 5 segments to verify increments"
+    # First few counters should be strictly monotonic (the BW counter is global,
+    # incrementing across the whole flash buffer; some events may share counter
+    # values with the previous event's tail block, so allow non-strict).
+    for i in range(1, min(8, len(counters))):
+        assert counters[i] >= counters[i - 1], (
+            f"counter went backwards: {counters[i-1]} → {counters[i]}"
+        )
+
+
+# ── decode_waveform_v2: currently a stub ─────────────────────────────────────
+
+
+@pytest.mark.parametrize("event_name", list(FIXTURES_INFO.keys()))
+def test_decode_waveform_v2_returns_none_until_verified(event_name):
+    """
+    The verified per-byte sample decoder is not yet wired up.
+
+    This test ensures decode_waveform_v2 returns ``None`` so callers know
+    to keep using the legacy decoder.  When a verified decoder lands,
+    flip this assertion and add ground-truth tests against the bundled
+    TXT exports.
+    """
+    path = _fixture_path(event_name)
+    if not os.path.exists(path):
+        pytest.skip(f"fixture missing: {path}")
+    body = _bw_body(path)
+    assert decode_waveform_v2(body) is None