Commit Graph
12 Commits
Author SHA1 Message Date
mathiasandClaude Sonnet 5.5 81f3d7ffc6 test(transcribe): a StatusError exposes its code as a kind, not its snippet (red; adversarial review F3)
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-02 16:39:25 +02:00
mathiasandClaude Sonnet 5.5 aefc8103f7 feat(transcribe): the ensemble declares and reports its phases; a failed engine shows failed, a shutdown does not (ADR-035)
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-02 08:54:31 +02:00
mathiasandClaude Sonnet 5.5 f064dec185 test(transcribe): the ensemble declares and reports its phases; a failed engine is shown failed, a shutdown is not (red)
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-02 08:54:07 +02:00
mathiasandClaude Sonnet 5.5 3d755fd57a fix(transcribe): check (discard) the deferred Close errors golangci-lint flags
Found because the project's lint step cannot fail: the Taskfile runs
'golangci-lint run ./... || echo "not installed -- skipping"', so a failing lint
prints that message and exits 0. CI's log showed these findings and still passed.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 23:26:02 +02:00
mathiasandClaude Sonnet 5.5 34bc8dbdee feat(transcribe): Ensemble -- two local engines, a local reconciler, fail-soft at every step
Runs the fast engine then the other from one WAV and reconciles; one engine down
uses the other and says so; the model down gives the base text unreconciled; the
cloud fallback runs only when BOTH local engines fail and is recorded as such; a
cancelled job stops everything and never reaches the cloud. The endpoint lists every
contributor. Mutation-checked, including both cancellation points.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 23:21:17 +02:00
mathiasandClaude Sonnet 5.5 8778376aa6 test(transcribe): a shutdown in the middle of either engine stops the job
Mutation-testing showed each of the two cancellation checks was redundant with the
other in my first test. Now a cancel during the first engine must not start the second
or the cloud, and a cancel during the second must not be recorded as a single-engine
success ('kb-whisper failed').

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 23:21:17 +02:00
mathiasandClaude Sonnet 5.5 ad39b6154c test(transcribe): two engines, then a reconciler -- and every way it can fail softly (red)
Both run in order (the fast one first) and are reconciled; identical texts cost no
model call; one engine down uses the other and says so, with the cloud fallback
untouched; only when BOTH local engines fail does the fallback run, and the row says
berget served it; both failing with no fallback reports both errors; a cancelled
job never reaches the cloud; a model that is down yields the base text unreconciled,
not a failed job.

Red: transcribe.Ensemble is undefined.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 23:20:40 +02:00
mathiasandClaude Sonnet 5.5 d11afb5ff1 fix(transcribe): rebuild the transcript from segments so split words are whole
CI / Lint / Test / Vet (pull_request) Successful in 22s
CI / Build & Import (pull_request) Skipped
CI / Deploy via GitOps (pull_request) Skipped
whisper.cpp wraps at ~60 chars and can wrap inside a word; the response's text
field joins segments with newlines, so chat saw 'nyt' / 't godkännande'. Text
is now rebuilt from the raw segments: concatenated when any segment carries a
leading space (whisper.cpp), joined by a space when none do (berget's sentence
level), whitespace collapsed; the text field is only a fallback.

Checked against the first real one-hour transcript with pianissimo's vocabulary
as an independent yardstick: 36 spans changed, 31 now consist entirely of words
pianissimo also produced, words unknown to pianissimo fell 224 -> 177 and word
agreement rose 92.4% -> 93.1%. Five rejoined spans still contain an unknown
word (names/loanwords or an unlucky glue) -- a residual risk, small.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 19:35:14 +02:00
mathiasandClaude Sonnet 5.5 500e11962b test(transcribe): rebuild Text from segments so a split word is not broken (red)
Found by comparing the first real one-hour transcript against pianissimo:
whisper.cpp wraps at ~60 chars and can wrap INSIDE a word. The continuation
segment has no leading space, the text field joins segments with newlines, and
a segment can be a lone '.'. Measured 2026-10-01: 7% of segments (96/1332) in a
synthetic hour, 42/1302 lines in the real one. My earlier tests used short
sentences and could not see it. berget's segments are whole sentences with no
leading space, so the two backends need different joins.

Red: Text still joins by newline.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 19:34:54 +02:00
mathiasandClaude Sonnet 5.5 2227cfff7e test(transcribe): opt-in live check against the real gateway
The httptest fake proves only what we THINK the gateway accepts. Skipped
unless TAPIR_LIVE_WAV and TAPIR_LIVE_GATEWAY_URL are set. Run once on
2026-10-01 with 11 min of synthetic Swedish (21 MB, above the 8-10 MB ingress
cap, via the NodePort): both iguana/kb-whisper (52.8 s) and berget/kb-whisper
(35.4 s) accepted the streamed, computed-Content-Length body and returned
usable verbose_json. Synthetic audio only; this repo is public.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 16:46:14 +02:00
mathiasandClaude Sonnet 5.5 cae423750c feat(transcribe): gateway client with an iguana->berget fallback chain (ADR-033)
Streams the WAV from disk with a computed Content-Length (no buffering, no
chunking), posts verbose_json with field `file` and language=sv, clamps the
final segment to the audio duration, and treats an empty transcript as a
failure. The chain falls back on ANY primary error, records the serving
endpoint, and never falls back once the context is cancelled.

gateway.go and chain.go are one commit: the chain's tests use the gateway's
StatusError and ErrEmptyTranscript, so neither compiles alone.

Mutation-checked: field name, clamp, cancel-fallback, content length and the
recorded endpoint each turn a test red.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 16:44:34 +02:00
mathiasandClaude Sonnet 5.5 bb285d8757 test(transcribe): gateway client and iguana->berget chain (red)
Gateway: field is `file`, no Expect header, a computed Content-Length (not
chunked, not buffered -- a 90 min WAV is ~170 MB), verbose_json normalised
with the last segment clamped to the audio duration, errors independent of
the body, an empty transcript is a failure. Chain: ANY primary failure falls
back and the row records the fallback endpoint; a cancelled job never falls
back to the cloud.

Red: package transcribe does not exist.

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
2026-10-01 16:43:51 +02:00