Coverage matrix for pkg/tcf, CCSDS 301.0-B-4
CCSDS 301.0-B-4 ships no PICS proforma. It is a format definition standard, not a protocol with a conformance annex, so there is no official table to fill in. What follows is a coverage matrix written against the normative text: each row names a feature, the clause that defines it, and what this package does with it.
Every row here was checked against the standard's text rather than against a published test vector, because CCSDS 301.0-B-4 publishes none. That is worth saying plainly: each row is one implementer's reading of a clause, and a round-trip test cannot tell you it is wrong. The 14 vectors listed at the bottom pin the octets, but their derivations come from the same reading.
A. General information
| Implementation | github.com/ravisuhag/astro/pkg/tcf |
| Standard | CCSDS 301.0-B-4, Time Code Formats |
| Scope | Encoding and decoding of the four time code formats, with the P-field and bare T-field forms |
| Not in scope | Time correlation, clock discipline, and onboard time management, none of which this standard defines |
B. P-field
| Feature | Clause | Status | Notes |
|---|---|---|---|
| First octet: extension flag, 3-bit ID, 4 detail bits | clause 3.2.2 and per-format | Y | |
PField | |||
| Second octet when the extension flag is set | clause 3.2.2 | Y | |
7 detail bits | |||
Time code ID 001, CUC Level 1 | clause 3.2 | Y | |
Time | |||
Time code ID 010, CUC Level 2 | clause 3.2 | Y | |
Time | |||
Time code ID 100, CDS | clause 3.3 | Y | |
TimeCodeCDS; level from detail bit 4 | |||
Time code ID 101, CCS | clause 3.4 | Y | |
TimeCodeCCS; Level 1 only | |||
| Unrecognised time code ID | - | Y | |
Err | |||
| Implicit P-field (bare T-field) | - | Y | |
EncodeTField, Decode, Decode, Decode. The standard permits an agreed format without a preamble; the octet counts, variant, and epoch are then caller-supplied. | |||
C. CUC, Unsegmented Time Code (clause 3.2)
| Feature | Clause | Status | Notes |
|---|---|---|---|
| Coarse time, 1-4 basic octets | clause 3.2.2 | Y | WithCUCCoarseBytes |
| Fine time, 0-3 basic octets | clause 3.2.2 | Y | WithCUCFineBytes |
| Extension octet: additional coarse (2 bits), additional fine (3 bits) | clause 3.2.2 | Y | Coarse to 7 octets, fine to 10 |
| Reserved bits 6-7 of the extension octet | clause 3.2.2 | Y | A non-zero value is rejected |
| Out-of-range octet counts | clause 3.2.2 | Y | ErrInvalidCoarseOctets, ErrInvalidFineOctets |
| Level 1, CCSDS epoch, 1958-01-01 TAI | clause 3.2 | Y | Coarse count is true TAI seconds: the offset in effect at the encoded instant is added on encode and removed on decode |
| Level 2, agency-defined epoch | clause 3.2 | Y | Purely arithmetic, no leap-second correction. WithCUCEpoch; ErrEpochRequired if omitted |
| Coarse value exceeding the configured width | clause 3.2.2 | Y | ErrOverflow |
| Fine-time wire fidelity across an encode/decode cycle | clause 3.2.2 | Y | Fine octets are preserved as transmitted, not re-derived from a rounded duration |
D. CDS, Day Segmented Time Code (clause 3.3)
| Feature | Clause | Status | Notes |
|---|---|---|---|
| Day segment, 16-bit | clause 3.3.2 | Y | |
With | |||
| Day segment, 24-bit | clause 3.3.2 | Y | |
With | |||
| Milliseconds of day, 0-86399999 | clause 3.3.2 | Y | |
Err outside range | |||
Sub-millisecond absent (code 00) | clause 3.3.2 | Y | |
With | |||
Sub-millisecond microseconds (code 01, 2 octets) | clause 3.3.2 | Y | |
With | |||
Sub-millisecond picoseconds (code 10, 4 octets) | clause 3.3.2 | Y | |
With | |||
Reserved sub-millisecond code 11 | clause 3.3.2 | Y | |
Rejected with Err | |||
| Sub-millisecond value outside the declared resolution | clause 3.3.2 | Y | |
Err | |||
| Extension flag on a CDS P-field | clause 3.3.2 | Y | |
The CDS P-field is a single octet; a set extension flag is rejected | |||
| Level 1 and Level 2 epochs | clause 3.3 | Y | |
WithCDSEpoch | |||
| Leap seconds | clause 3.3 | N/A | |
Day and millisecond counting is purely arithmetic. The fields act as a UTC day and time-of-day label; no leap-second table is applied. | |||
E. CCS, Calendar Segmented Time Code (clause 3.4)
| Feature | Clause | Status | Notes |
|---|---|---|---|
| Month/Day variant (detail bit 4 = 0) | clause 3.4.2 | Y | WithCCSMonthDay |
| Day-of-Year variant (detail bit 4 = 1) | clause 3.4.2 | Y | Default |
| BCD encoding of every segment | clause 3.4.1 | Y | |
| Upper 4 bits of the year's first octet are zero | clause 3.4.1.2 | Y | Enforced on encode and decode |
| Non-BCD nibble (value above 9) | clause 3.4.1 | Y | ErrInvalidBCD |
| Sub-second octets, 0-6 | clause 3.4.2 | Y | WithCCSSubSecBytes; each octet holds two decimal digits |
| Second value 60 (the leap second) | clause 3.4 | Y | Accepted and flagged; Time() normalizes it, because Go's time.Time cannot represent second 60 |
| Calendar field range checks | clause 3.4.1 | Y | Month 1-12, day of month 1-31, day of year 1-366, hour 0-23, minute 0-59; ErrInvalidCalendarTime |
| Cross-checks between calendar fields | clause 3.4.1 | Y | Derived: the standard gives per-field ranges; consistency between them is checked here as a decoder safeguard |
| Level 2 | clause 3.4 | N/A | CCS is Level 1 UTC only, by definition |
F. ASCII time codes (clause 3.5)
| Feature | Clause | Status | Notes |
|---|---|---|---|
Type A, YYYY-MM-DDThh:mm:ss.d...dZ | clause 3.5.1.1 | Y | ASCIITypeA |
Type B, YYYY-DDDThh:mm:ss.d...dZ | clause 3.5 | Y | ASCIITypeB |
| Fractional seconds, 1-9 digits | clause 3.5 | Y | WithASCIIPrecision; default 3 |
Optional Z terminator | clause 3.5 | Y | Accepted on decode, written on encode |
| Fixed field widths and mandatory separators | clause 3.5 | Y | Decode enforces the subset strictly and rejects anything ISO 8601 allows but clause 3.5 does not |
| Second 60 | clause 3.5 | Y | Accepted only at 23:59:60 |
| P-field | clause 3.5 | N/A | ASCII codes carry no P-field |
G. Time scales and leap seconds
| Feature | Status | Notes |
|---|---|---|
| Integer TAI-UTC offset table | Y | |
Embedded in leapseconds.; TAIUTCOffset returns the offset in effect at any instant. The values are historical facts and cannot change. | ||
| Current offset | Y | |
37 s from 2017-01-01. No leap second has been announced since. | ||
| Pre-1972 era | Partial | |
UTC used fractional "rubber second" adjustments with no integer representation. TAI and UTC are treated as identical (offset 0). TAI counts falling inside the 10-second step at 1972-01-01 do not round-trip exactly. Documented limitation. | ||
| An instant inside a leap second | Partial | |
A CUC Level 1 count landing on UTC 23:59:60 decodes to the following 00:00:00. Go's time.Time cannot represent second 60. Documented limitation. | ||
| Future leap seconds | N | |
Not predictable. A new IERS announcement needs a one-line addition to leapseconds.. | ||
H. Not implemented
| Item | Reason |
|---|---|
| Time correlation and clock discipline | Not defined by this standard |
| Agency-specific epoch registries | A mission supplies its own epoch through WithCUCEpoch or WithCDSEpoch |
| P-field octets beyond the second | None are defined by CCSDS 301.0-B-4 |
I. Verification
The conformance regressions live in pkg/tcf/audit_test.go, tagged TCF-n
per defect. They cover the TAI-UTC table against its historical values, CUC
Level 1 leap-second handling, CUC Level 2 arithmetic behaviour, fine-time wire
fidelity, the rejected P-field extension bit, the reserved CDS sub-millisecond
code, CCS BCD and calendar validation, strict ASCII decoding, bare T-field
round trips, and overflow guards on large coarse and day values.
pkg/tcf/fuzz_test.go fuzzes the decoders: arbitrary bytes must never panic
and never allocate from a length field an attacker controls.
Reference
- CCSDS 301.0-B-4, Time Code Formats (Blue Book)
- Protocol page | CLI
Wire test vectors
The octets backing this statement live in the vector corpus — 14 vectors. Each vector names the clause it comes from and carries the derivation that produced it.
| File | |
|---|---|
tcf/timecode.json | 14 vectors |
These are data files, so any implementation can check itself against the same octets. See CONTRACT.md for how, and how this is verified for what rests on a published vector versus a reading of the clause.