transcribe-cpp 0.3.0

Rust bindings for transcribe.cpp, a C/C++ speech-to-text library built on ggml
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131

//! `serde` feature: derive coverage plus the crate's custom serde behavior
//! (NaN confidences, sparse-document defaults).

#![cfg(feature = "serde")]

use transcribe_cpp::{
    Backend, Capabilities, CommitPolicy, DeviceType, Diarize, ExtSlot, Feature, Itn, KvType,
    MoonshineStreamingOptions, ParakeetBufferedStreamOptions, ParakeetStreamOptions, Pnc,
    RunExtension, RunOptions, Segment, SessionLimits, SessionOptions, SortformerPreset,
    SortformerStreamOptions, SpeakerSegment, StreamExtension, StreamOptions, StreamState,
    StreamText, StreamUpdate, Task, TimestampKind, Timings, Token, Transcript,
    VoxtralRealtimeStreamOptions, WhisperRunOptions, Word,
};

fn assert_serde<T: serde::Serialize + serde::de::DeserializeOwned>() {}

/// Fails to compile if any plain-data type loses its derives.
#[test]
fn plain_data_types_are_serializable() {
    // Options.
    assert_serde::<RunOptions>();
    assert_serde::<StreamOptions>();
    assert_serde::<SessionOptions>();
    assert_serde::<RunExtension>();
    assert_serde::<StreamExtension>();
    assert_serde::<WhisperRunOptions>();
    assert_serde::<MoonshineStreamingOptions>();
    assert_serde::<ParakeetStreamOptions>();
    assert_serde::<ParakeetBufferedStreamOptions>();
    assert_serde::<VoxtralRealtimeStreamOptions>();
    assert_serde::<SortformerStreamOptions>();
    assert_serde::<SortformerPreset>();
    // Results.
    assert_serde::<Transcript>();
    assert_serde::<Segment>();
    assert_serde::<SpeakerSegment>();
    assert_serde::<Word>();
    assert_serde::<Token>();
    assert_serde::<Timings>();
    assert_serde::<StreamUpdate>();
    assert_serde::<StreamText>();
    assert_serde::<Capabilities>();
    assert_serde::<SessionLimits>();
    // Enums.
    assert_serde::<Task>();
    assert_serde::<TimestampKind>();
    assert_serde::<KvType>();
    assert_serde::<Pnc>();
    assert_serde::<Itn>();
    assert_serde::<Diarize>();
    assert_serde::<Backend>();
    assert_serde::<Feature>();
    assert_serde::<CommitPolicy>();
    assert_serde::<StreamState>();
    assert_serde::<DeviceType>();
    assert_serde::<ExtSlot>();
}

fn nan_transcript() -> Transcript {
    Transcript {
        text: "and so".into(),
        speaker_segments: vec![SpeakerSegment {
            t1_ms: 900,
            speaker_id: 1,
            p: f32::NAN,
            ..Default::default()
        }],
        tokens: vec![
            Token {
                id: 42,
                p: f32::NAN,
                text: " and".into(),
                ..Default::default()
            },
            Token {
                id: 43,
                p: 0.75,
                text: " so".into(),
                ..Default::default()
            },
        ],
        ..Default::default()
    }
}

/// NaN != NaN, so compare by postcard bytes.
fn same_transcript(a: &Transcript, b: &Transcript) -> bool {
    postcard::to_allocvec(a).unwrap() == postcard::to_allocvec(b).unwrap()
}

#[test]
fn nan_confidence_round_trips() {
    let original = nan_transcript();

    let json = serde_json::to_string(&original).unwrap();
    assert!(
        json.contains("\"p\":null"),
        "NaN should encode as null: {json}"
    );
    let from_json: Transcript = serde_json::from_str(&json).unwrap();
    assert!(from_json.tokens[0].p.is_nan());
    assert!(from_json.speaker_segments[0].p.is_nan());
    assert_eq!(from_json.tokens[1].p, 0.75);
    assert!(same_transcript(&from_json, &original), "json");

    let bytes = postcard::to_allocvec(&original).unwrap();
    let from_postcard: Transcript = postcard::from_bytes(&bytes).unwrap();
    assert!(from_postcard.tokens[0].p.is_nan());
    assert!(from_postcard.speaker_segments[0].p.is_nan());
    assert!(same_transcript(&from_postcard, &original), "postcard");
}

#[test]
fn missing_fields_take_defaults() {
    // `RunOptions::default()` is hand-written; its -1 sentinel must survive.
    let run: RunOptions = serde_json::from_str(r#"{"language":"fr"}"#).unwrap();
    assert_eq!(run.spec_k_drafts, -1);
    assert_eq!(
        run,
        RunOptions {
            language: Some("fr".into()),
            ..Default::default()
        }
    );

    let token: Token = serde_json::from_str(r#"{"id":5}"#).unwrap();
    assert_eq!(token.id, 5);
    assert!(token.p.is_nan(), "missing p decoded as {}", token.p);
    let speaker: SpeakerSegment = serde_json::from_str(r#"{"speaker_id":2}"#).unwrap();
    assert!(speaker.p.is_nan(), "missing p decoded as {}", speaker.p);
}