Major quality pass on top of Phase 2. Five substantive changes plus
cross-cutting touches across audio, hotkey, transcription, and Tauri
command layers.
Transcription quality
- Long-audio chunking in commands/transcription.rs: Parakeet and large
file transcription now chunk-and-recompose with overlap trimming, so
the live-path chunking advantage extends to file-based workflows.
- Stateful live speech gate in commands/live.rs on top of the earlier
duplicate-boundary filtering — distinguishes start-of-speech from
mid-speech and holds state across chunks.
Auto-learning corrections
- New crates/ai-formatting/src/correction_learning.rs: extracts user
text corrections from viewer edits and proposes additions to the
active profile's vocabulary.
- src-tauri/src/commands/profiles.rs bridge for frontend-driven
confirmation of learned terms.
- src/routes/viewer/+page.svelte hooks the learning path into the
segment-edit flow so corrections feed profile_terms without a
separate 'train this profile' UX.
Transcript profile provenance
- Migration v8 (crates/storage/src/migrations.rs) adds profile_id to
transcripts, defaulting to DEFAULT_PROFILE_ID so existing rows stay
valid.
- crates/storage/src/database.rs: TranscriptRow + CRUD carry profile_id.
- src-tauri/src/commands/transcripts.rs: add_transcript accepts and
persists profile_id.
- DictationPage.svelte + FilesPage.svelte send activeProfileId on
capture so learned corrections are attributed to the right profile.
Cleanup prompt contract
- crates/ai-formatting/src/llm_client.rs hardened: the CLEANUP_PROMPT
now specifies concrete do/do-not rules, ready for a real model-backed
cleanup pass. The llm_client is still a stub — kon-llm remains unwired
— but the prompt shape is final.
Cross-cutting polish
- Minor touches in audio (capture/decode/resample), hotkey (lib/linux/stub),
core, transcription (concurrency/model_manager/local_engine/whisper_rs),
and the rest of src-tauri/src/commands/*: error-path tightening, log
clarity, TS-migration follow-ups (@ts-nocheck additions for incremental
typing).
Verified locally: npm run check, cargo test -p kon-ai-formatting,
cargo test -p kon-storage, cargo test -p kon --lib commands::live::tests,
cargo check — all green.
Scope boundary: kon-llm crate is still a stub; task extraction remains
rule-based. Bundled local-LLM runtime is the next clean step and is not
in this commit.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
101 lines
3.1 KiB
Rust
101 lines
3.1 KiB
Rust
use rubato::{
|
|
Resampler, SincFixedIn, SincInterpolationParameters, SincInterpolationType, WindowFunction,
|
|
};
|
|
|
|
use kon_core::constants::WHISPER_SAMPLE_RATE;
|
|
use kon_core::error::{KonError, Result};
|
|
use kon_core::types::AudioSamples;
|
|
|
|
/// Resample audio to 16kHz mono using sinc interpolation (rubato).
|
|
/// Returns a new AudioSamples at the target sample rate.
|
|
pub fn resample_to_16khz(audio: &AudioSamples) -> Result<AudioSamples> {
|
|
let from_rate = audio.sample_rate();
|
|
let target_rate = WHISPER_SAMPLE_RATE;
|
|
|
|
if from_rate == target_rate {
|
|
return Ok(AudioSamples::mono_16khz(audio.samples().to_vec()));
|
|
}
|
|
|
|
if from_rate == 0 {
|
|
return Err(KonError::AudioDecodeFailed(
|
|
"Cannot resample: source rate is 0".into(),
|
|
));
|
|
}
|
|
|
|
let ratio = target_rate as f64 / from_rate as f64;
|
|
let chunk_size = 1024;
|
|
|
|
let params = SincInterpolationParameters {
|
|
sinc_len: 256,
|
|
f_cutoff: 0.95,
|
|
oversampling_factor: 128,
|
|
interpolation: SincInterpolationType::Cubic,
|
|
window: WindowFunction::Blackman,
|
|
};
|
|
|
|
let mut resampler = SincFixedIn::<f32>::new(
|
|
ratio, 1.1, params, chunk_size, 1, // mono
|
|
)
|
|
.map_err(|e| KonError::AudioDecodeFailed(format!("Resampler init failed: {e}")))?;
|
|
|
|
let samples = audio.samples();
|
|
let mut output_samples: Vec<f32> = Vec::new();
|
|
|
|
let mut offset = 0;
|
|
while offset < samples.len() {
|
|
let end = (offset + chunk_size).min(samples.len());
|
|
let mut chunk = samples[offset..end].to_vec();
|
|
|
|
if chunk.len() < chunk_size {
|
|
chunk.resize(chunk_size, 0.0);
|
|
}
|
|
|
|
let input = vec![chunk];
|
|
let result = resampler
|
|
.process(&input, None)
|
|
.map_err(|e| KonError::AudioDecodeFailed(format!("Resample failed: {e}")))?;
|
|
|
|
if !result.is_empty() && !result[0].is_empty() {
|
|
output_samples.extend_from_slice(&result[0]);
|
|
}
|
|
|
|
offset += chunk_size;
|
|
}
|
|
|
|
// Trim to expected length (padding may have added extra samples)
|
|
let expected_len = (samples.len() as f64 * ratio) as usize;
|
|
output_samples.truncate(expected_len);
|
|
|
|
Ok(AudioSamples::mono_16khz(output_samples))
|
|
}
|
|
|
|
#[cfg(test)]
|
|
mod tests {
|
|
use super::*;
|
|
|
|
#[test]
|
|
fn resample_passthrough_at_16khz() {
|
|
let input = AudioSamples::mono_16khz(vec![0.1, 0.2, 0.3]);
|
|
let output = resample_to_16khz(&input).unwrap();
|
|
assert_eq!(output.sample_rate(), 16000);
|
|
assert_eq!(output.samples().len(), 3);
|
|
}
|
|
|
|
#[test]
|
|
fn resample_preserves_approximate_duration() {
|
|
let rate = 48000;
|
|
let duration_secs = 1.0;
|
|
let num_samples = (rate as f64 * duration_secs) as usize;
|
|
let samples: Vec<f32> = (0..num_samples).map(|i| (i as f32 * 0.001).sin()).collect();
|
|
|
|
let input = AudioSamples::new(samples, rate, 1);
|
|
let output = resample_to_16khz(&input).unwrap();
|
|
|
|
let output_duration = output.samples().len() as f64 / 16000.0;
|
|
assert!(
|
|
(output_duration - duration_secs).abs() < 0.1,
|
|
"Duration mismatch: expected ~{duration_secs}s, got {output_duration}s"
|
|
);
|
|
}
|
|
}
|