sipp-rs 0.1.0

Unified Rust library for extensible Sipp inference
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
132
133
134
135
136
137
138

//! Tests the `runtime::inference_runtime::request::api` module in `sipp`.
//!
//! Covers deterministic inference-runtime helpers, state transitions, and error paths while avoiding native model execution unless a test is explicitly ignored.

use crate::runtime::config::NativeRuntimeConfig;
use crate::runtime::inference_runtime::runtime_tests::test_runtime;
use crate::runtime::request::GenerateRequest;

use super::*;

#[test]
fn normalize_context_key_uses_default_for_empty_values() {
    assert_eq!(normalize_context_key(""), DEFAULT_PROMPT_CONTEXT_KEY);
    assert_eq!(normalize_context_key("ctx"), "ctx");
}

#[test]
fn generate_request_maps_all_fields_and_normalizes_stops() {
    let request = generate_request(GenerateRequestFields {
        request_id: 3,
        context_key: "ctx".to_string(),
        prompt: "prompt".to_string(),
        prompt_tokens: vec![1, 2],
        n_tokens_predict: 5,
        grammar: "root ::= \"ok\"".to_string(),
        json_schema: "{}".to_string(),
        stop: vec!["zz".to_string(), "aa".to_string(), String::new()],
        sampling: None,
        emit_tokens: true,
    });

    assert_eq!(request.id, 3);
    assert_eq!(request.context_key, "ctx");
    assert_eq!(request.original_prompt, "prompt");
    assert_eq!(request.prompt_tokens, vec![1, 2]);
    assert_eq!(request.max_output_tokens, 5);
    assert_eq!(request.grammar, "root ::= \"ok\"");
    assert_eq!(request.json_schema, "{}");
    assert_eq!(request.stop, vec!["aa", "zz"]);
    assert!(request.emit_tokens);
}

#[test]
fn request_tokenization_flags_match_text_and_multimodal_modes() {
    assert_eq!(
        request_tokenization_flags_for_tests("text"),
        Some((true, true))
    );
    assert_eq!(
        request_tokenization_flags_for_tests("multimodal"),
        Some((false, false))
    );
    assert_eq!(request_tokenization_flags_for_tests("bad"), None);
}

#[test]
fn normalize_stop_sequences_drops_empty_values_and_deduplicates() {
    assert_eq!(
        normalize_stop_sequences(vec![
            "z".to_string(),
            String::new(),
            "a".to_string(),
            "z".to_string(),
        ]),
        vec!["a", "z"]
    );
}

#[test]
fn next_generate_request_id_rejects_overflow() {
    let mut runtime = test_runtime(NativeRuntimeConfig::default());
    runtime.next_request_id = GenerateRequestId::MAX;

    let error = runtime
        .next_generate_request_id()
        .expect_err("request id overflow");

    assert!(matches!(
        error,
        Error::InvalidRequest(message) if message.contains("request id overflow")
    ));
}

#[test]
fn enqueue_prepared_request_rejects_zero_request_id() {
    let mut runtime = test_runtime(NativeRuntimeConfig::default());
    let mut request = GenerateRequest::new(0, "ctx");
    request.prompt_tokens = vec![1, 2, 3];

    let error = runtime
        .enqueue_prepared_request(request)
        .expect_err("zero request id is invalid");

    assert!(matches!(
        error,
        Error::InvalidRequest(message) if message.contains("failed to enqueue request")
    ));
    assert_eq!(runtime.total_input_tokens, 3);
}

#[test]
fn enqueue_prepared_request_saturates_total_input_token_counter() {
    let mut runtime = test_runtime(NativeRuntimeConfig::default());
    runtime.total_input_tokens = usize::MAX - 1;
    let mut request = GenerateRequest::new(8, "ctx");
    request.prompt_tokens = vec![1, 2, 3];

    let request_id = runtime
        .enqueue_prepared_request(request)
        .expect("enqueue request");

    assert_eq!(request_id, 8);
    assert_eq!(runtime.total_input_tokens, usize::MAX);
}

#[test]
fn enqueue_generation_requests_reject_empty_runtime_before_tokenization() {
    let mut runtime = test_runtime(NativeRuntimeConfig::default());

    assert!(matches!(
        runtime.enqueue_request("ctx", "prompt", 1, "", "", Vec::new(), None, false),
        Err(Error::RuntimeNotReady)
    ));
    assert!(matches!(
        runtime.enqueue_multimodal_request(
            "ctx",
            "prompt",
            1,
            vec![vec![1, 2, 3]],
            "",
            "",
            Vec::new(),
            None,
            false
        ),
        Err(Error::RuntimeNotReady)
    ));
}