ipfrs 0.2.0

Next-generation distributed file system with content-addressing, semantic search, and logic programming
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
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221

//! Benchmarks for semantic search operations

use criterion::{criterion_group, criterion_main, BenchmarkId, Criterion};
use ipfrs::{Node, NodeConfig, QueryFilter};
use ipfrs_semantic::{DistanceMetric, RouterConfig};
use std::hint::black_box;
use tokio::runtime::Runtime;

/// Generate a random embedding vector
fn random_embedding(dim: usize, seed: u64) -> Vec<f32> {
    // Simple deterministic pseudo-random generation for benchmarking
    (0..dim)
        .map(|i| {
            let x = ((seed
                .wrapping_mul(6364136223846793005)
                .wrapping_add(i as u64))
                >> 32) as f32;
            x / u32::MAX as f32
        })
        .collect()
}

/// Benchmark semantic indexing operations
fn bench_semantic_index(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();
    let mut group = c.benchmark_group("semantic_index");

    let dim = 768; // Standard embedding dimension
    for count in [10, 50, 100].iter() {
        group.bench_with_input(BenchmarkId::from_parameter(count), count, |b, &count| {
            b.to_async(&rt).iter(|| async {
                let config = NodeConfig::default().with_semantic(RouterConfig {
                    dimension: dim,
                    metric: DistanceMetric::Cosine,
                    max_connections: 16,
                    ef_construction: 200,
                    ef_search: 50,
                    cache_size: 1000,
                    ..RouterConfig::default()
                });

                let mut node = Node::new(config).unwrap();
                node.start().await.unwrap();

                // Index multiple blocks
                for i in 0..count {
                    let data = format!("block_{}", i).into_bytes();
                    let cid = node.add_bytes(data).await.unwrap();
                    let embedding = random_embedding(dim, i as u64);
                    node.index_content(&cid, &embedding).await.unwrap();
                }

                node.stop().await.unwrap();
                black_box(count)
            });
        });
    }
    group.finish();
}

/// Benchmark semantic search queries
fn bench_semantic_search(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();
    let mut group = c.benchmark_group("semantic_search");

    let dim = 768;
    let index_size = 100;

    // Setup: Create indexed node
    let setup_node = || async {
        let config = NodeConfig::default().with_semantic(RouterConfig {
            dimension: dim,
            metric: DistanceMetric::Cosine,
            max_connections: 16,
            ef_construction: 200,
            ef_search: 50,
            cache_size: 1000,
            ..RouterConfig::default()
        });

        let mut node = Node::new(config).unwrap();
        node.start().await.unwrap();

        // Index blocks
        for i in 0..index_size {
            let data = format!("block_{}", i).into_bytes();
            let cid = node.add_bytes(data).await.unwrap();
            let embedding = random_embedding(dim, i as u64);
            node.index_content(&cid, &embedding).await.unwrap();
        }

        node
    };

    // Benchmark different k values
    for k in [1, 5, 10, 20].iter() {
        group.bench_with_input(BenchmarkId::from_parameter(k), k, |b, &k| {
            let mut node = rt.block_on(setup_node());

            b.to_async(&rt).iter(|| async {
                let query = random_embedding(dim, 999);
                let results = black_box(node.search_similar(&query, k).await.unwrap());
                results
            });

            rt.block_on(async {
                node.stop().await.unwrap();
            });
        });
    }
    group.finish();
}

/// Benchmark filtered semantic search
fn bench_filtered_search(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();
    let mut group = c.benchmark_group("filtered_search");

    let dim = 768;
    let index_size = 100;

    // Setup: Create indexed node
    let config = NodeConfig::default().with_semantic(RouterConfig {
        dimension: dim,
        metric: DistanceMetric::Cosine,
        max_connections: 16,
        ef_construction: 200,
        ef_search: 50,
        cache_size: 1000,
        ..RouterConfig::default()
    });

    let mut node = rt.block_on(async {
        let mut node = Node::new(config).unwrap();
        node.start().await.unwrap();

        for i in 0..index_size {
            let data = format!("block_{}", i).into_bytes();
            let cid = node.add_bytes(data).await.unwrap();
            let embedding = random_embedding(dim, i as u64);
            node.index_content(&cid, &embedding).await.unwrap();
        }

        node
    });

    group.bench_function("filtered_search", |b| {
        b.to_async(&rt).iter(|| async {
            let query = random_embedding(dim, 999);
            let filter = QueryFilter {
                min_score: Some(0.5),
                max_score: None,
                max_results: Some(10),
                cid_prefix: None,
            };

            let results = black_box(node.search_hybrid(&query, 20, filter).await.unwrap());
            results
        });
    });

    rt.block_on(async {
        node.stop().await.unwrap();
    });

    group.finish();
}

/// Benchmark semantic statistics
fn bench_semantic_stats(c: &mut Criterion) {
    let rt = Runtime::new().unwrap();
    let mut group = c.benchmark_group("semantic_stats");

    let dim = 768;

    // Setup with different index sizes
    for size in [10, 50, 100].iter() {
        let config = NodeConfig::default().with_semantic(RouterConfig {
            dimension: dim,
            metric: DistanceMetric::Cosine,
            max_connections: 16,
            ef_construction: 200,
            ef_search: 50,
            cache_size: 1000,
            ..RouterConfig::default()
        });

        let mut node = rt.block_on(async {
            let mut node = Node::new(config).unwrap();
            node.start().await.unwrap();

            for i in 0..*size {
                let data = format!("block_{}", i).into_bytes();
                let cid = node.add_bytes(data).await.unwrap();
                let embedding = random_embedding(dim, i as u64);
                node.index_content(&cid, &embedding).await.unwrap();
            }

            node
        });

        group.bench_with_input(BenchmarkId::from_parameter(size), size, |b, _size| {
            b.to_async(&rt)
                .iter(|| async { black_box(node.semantic_stats().unwrap()) });
        });

        rt.block_on(async {
            node.stop().await.unwrap();
        });
    }
    group.finish();
}

criterion_group!(
    benches,
    bench_semantic_index,
    bench_semantic_search,
    bench_filtered_search,
    bench_semantic_stats
);
criterion_main!(benches);