conproxy 0.2.1

Cache proxy server for heterogeneous RAG/vector search backends with cascade, scope, and MCP
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
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176

use super::*;

#[test]
fn test_smart_embedder_creation() {
    let embedder = SmartEmbedder::with_defaults();
    assert!(!embedder.is_loaded());
    assert!(embedder.dimensions().is_none());
}

#[test]
fn test_smart_embedder_config() {
    let config = EmbedderConfig::new("test-model")
        .with_cache_max_entries(100)
        .with_warmup(false);
    let embedder = SmartEmbedder::new(config);

    let stats = embedder.stats();
    assert_eq!(stats.model_name, "test-model");
    assert!(!stats.model_loaded);
}

#[test]
fn test_text_hash_deterministic() {
    let hash1 = SmartEmbedder::text_hash("onnx", "m", "hello world");
    let hash2 = SmartEmbedder::text_hash("onnx", "m", "hello world");
    let hash3 = SmartEmbedder::text_hash("onnx", "m", "different text");

    assert_eq!(hash1, hash2);
    assert_ne!(hash1, hash3);
}

#[test]
fn test_stats_initial() {
    let embedder = SmartEmbedder::with_defaults();
    let stats = embedder.stats();

    assert_eq!(stats.cache_size, 0);
    assert_eq!(stats.cache_hits, 0);
    assert_eq!(stats.cache_misses, 0);
    assert_eq!(stats.coalesced_requests, 0);
    assert_eq!(stats.embeddings_computed, 0);
    assert!((stats.hit_rate() - 0.0).abs() < f64::EPSILON);
}

#[test]
fn test_clear_cache() {
    let embedder = SmartEmbedder::with_defaults();

    // Manually insert some cache entries
    embedder.cache.insert(
        1,
        CachedEmbedding {
            vector: vec![1.0, 2.0],
            created_at: Instant::now(),
        },
    );
    embedder.cache.insert(
        2,
        CachedEmbedding {
            vector: vec![3.0, 4.0],
            created_at: Instant::now(),
        },
    );

    assert_eq!(embedder.cache.len(), 2);
    embedder.clear_cache();
    assert_eq!(embedder.cache.len(), 0);
}

#[test]
fn test_hit_rate_calculation() {
    let stats = SmartEmbedderStats {
        cache_size: 10,
        cache_hits: 80,
        cache_misses: 20,
        coalesced_requests: 5,
        embeddings_computed: 20,
        model_loaded: true,
        model_name: "test".to_string(),
        dimensions: Some(384),
    };

    assert!((stats.hit_rate() - 0.8).abs() < 0.001);
}

#[test]
fn test_hit_rate_zero_requests() {
    let stats = SmartEmbedderStats {
        cache_size: 0,
        cache_hits: 0,
        cache_misses: 0,
        coalesced_requests: 0,
        embeddings_computed: 0,
        model_loaded: false,
        model_name: "test".to_string(),
        dimensions: None,
    };

    // Should return 0.0 when no requests (avoid division by zero)
    assert!((stats.hit_rate() - 0.0).abs() < 0.001);
}

#[test]
fn test_cache_key_uniqueness() {
    // Different queries should produce different hashes
    let hash1 = SmartEmbedder::text_hash("onnx", "m", "what is rust programming");
    let hash2 = SmartEmbedder::text_hash("onnx", "m", "what is rust");
    let hash3 = SmartEmbedder::text_hash("onnx", "m", "What Is Rust Programming"); // case different
    let hash4 = SmartEmbedder::text_hash("onnx", "m", "what is rust programming"); // same as hash1

    assert_ne!(hash1, hash2);
    assert_ne!(hash1, hash3); // case matters
    assert_eq!(hash1, hash4); // identical should match
}

#[test]
fn test_embedder_config_builder() {
    use std::time::Duration;

    let config = EmbedderConfig::new("all-MiniLM-L6-v2")
        .with_cache_max_entries(500)
        .with_cache_ttl(Duration::from_secs(7200))
        .with_warmup(true);

    assert_eq!(config.model_name, "all-MiniLM-L6-v2");
    assert_eq!(config.cache_max_entries, 500);
    assert_eq!(config.cache_ttl, Duration::from_secs(7200));
    assert!(config.warmup_on_start);
}

#[test]
fn test_embedder_config_defaults() {
    use std::time::Duration;

    let config = EmbedderConfig::default();

    // Check defaults are reasonable
    assert!(!config.model_name.is_empty());
    assert!(config.cache_max_entries > 0);
    assert!(config.cache_ttl > Duration::ZERO);
}

#[test]
fn test_stats_model_info() {
    let embedder = SmartEmbedder::new(EmbedderConfig::new("test-model"));
    let stats = embedder.stats();

    assert_eq!(stats.model_name, "test-model");
    assert!(!stats.model_loaded); // Not loaded until first embed
    assert!(stats.dimensions.is_none()); // Unknown until loaded
}

#[test]
fn test_cache_key_includes_provider_and_model() {
    let same_text = "hello world";
    let a = SmartEmbedder::text_hash("onnx", "model-a", same_text);
    let b = SmartEmbedder::text_hash("onnx", "model-b", same_text);
    let c = SmartEmbedder::text_hash("openai", "model-a", same_text);
    let d = SmartEmbedder::text_hash("onnx", "model-a", same_text);
    let e = SmartEmbedder::text_hash("onnx", "model-a", "other text");

    assert_ne!(a, b, "different models must not collide");
    assert_ne!(a, c, "different providers must not collide");
    assert_eq!(a, d, "identical identity must match");
    assert_ne!(a, e, "different text must not collide");
}

#[test]
fn test_instance_cache_key_matches_helper() {
    let mut cfg = EmbedderConfig::new("my-model");
    cfg.provider = "openai".into();
    let emb = SmartEmbedder::new(cfg);
    let via_inst = emb.cache_key("q");
    let via_helper = SmartEmbedder::text_hash("openai", "my-model", "q");
    assert_eq!(via_inst, via_helper);
}