ReconcileContract
What the module refuses rather than approximating
12/12 passed in 1ms
| Test | What it does | Result |
|---|---|---|
test_above_capacity | A code rate above channel capacity is refused naming the Shannon limit, not returned as a beta above one. | ✅ pass |
test_auth_saturated | auth_failure refuses a round whose forgery probabilities union-bound past one. | ✅ pass |
test_below_threshold | A code below its measured threshold is refused as not decoding, and pick() refuses an SNR under every threshold on file. | ✅ pass |
test_biconf_refused | The Sugimoto-Yamazaki variant, BICONF after two passes, is refused rather than run as more parity passes. | ✅ pass |
test_catalogue | Every catalogue entry is an exported callable listing parameters, under a unique name. | ✅ pass |
test_double_charge | Charging reconciliation leakage inside the key length and again as a padded parity cost is refused. | ✅ pass |
test_entropy_shared | reconcile._entropy equals impairments._entropy bit for bit at nine points from 0 to 1. | ✅ pass |
test_hash_absent | hash_charge refuses bb84, whose six log terms are one composite, rrdps, whose terms are bit counts with no epsilon, and the unrecorded cow. | ✅ pass |
test_no_interpolation | Cascade efficiency is refused at an error rate Table 3 does not carry, f_EC not being monotone across its grid. | ✅ pass |
test_short_block | hash_length refuses a min-entropy under the leftover hash toll rather than return a negative length. | ✅ pass |
test_subunit_fec | An inefficiency below one, under the Slepian-Wolf bound, is refused, as is f_EC at a zero error rate. | ✅ pass |
test_zero_fer | verify_failure refuses a zero frame error rate, which would certify the keys never differ, naming what zero counted failures bound. | ✅ pass |