rag 0.1.4

A Rust library and CLI for Retrieval-Augmented Generation
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
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316

use rag::{
    aggregation::{count_by, group_by, sum_by},
    diversify::diversify,
    hybrid::{merge_hybrid, rrf_fusion},
    index::{DistanceMetric, FlatIndex, Index},
    keyword::{Bm25Config, Bm25Index, FieldBm25Index},
    sparse::{SparseIndex, SparseVector},
    vector_store::{Document, InMemoryVectorStore, JsonPersistentVectorStore, Similarity, VectorStore},
};
use std::collections::HashMap;
use tempfile::TempDir;

// --- Exact kNN filtered search ---

#[test]
fn exact_knn_filtered_search() {
    let index = FlatIndex::new();
    let mut d1 = Document::new("rust systems".to_string()).with_embedding(vec![1.0, 0.0, 0.0]);
    d1.metadata.insert("lang".to_string(), "rust".to_string());
    let mut d2 = Document::new("python scripts".to_string()).with_embedding(vec![0.0, 1.0, 0.0]);
    d2.metadata.insert("lang".to_string(), "python".to_string());
    let mut d3 = Document::new("rust cli".to_string()).with_embedding(vec![0.95, 0.05, 0.0]);
    d3.metadata.insert("lang".to_string(), "rust".to_string());

    index.add(d1);
    index.add(d2);
    index.add(d3);

    let results = index.search_exact_filtered(&[1.0, 0.0, 0.0], 5, &|doc: &Document| {
        doc.metadata.get("lang") == Some(&"rust".to_string())
    });

    assert_eq!(results.len(), 2);
    // Both rust docs should be returned, ordered by similarity
    assert!(results[0].score >= results[1].score);
}

#[test]
fn exact_knn_empty_filter() {
    let index = FlatIndex::new();
    index.add(Document::new("a".to_string()).with_embedding(vec![1.0, 0.0]));

    let results = index.search_exact_filtered(&[1.0, 0.0], 5, &|_| false);
    assert!(results.is_empty());
}

// --- RRF fusion ---

#[test]
fn rrf_fusion_combines_rankings() {
    let d1 = Document::new("first doc".to_string());
    let d2 = Document::new("second doc".to_string());
    let d3 = Document::new("third doc".to_string());

    let list1 = vec![
        Similarity { document: d1.clone(), score: 0.9 },
        Similarity { document: d2.clone(), score: 0.8 },
    ];
    let list2 = vec![
        Similarity { document: d3.clone(), score: 0.95 },
        Similarity { document: d1.clone(), score: 0.7 },
    ];

    let fused = rrf_fusion(&[list1, list2], 60, 10);
    assert!(!fused.is_empty());
    // d1 appears in both lists so should have highest RRF score
    assert_eq!(fused[0].document.id, d1.id);
}

#[test]
fn rrf_fusion_empty_input() {
    let fused = rrf_fusion(&[], 60, 10);
    assert!(fused.is_empty());
}

#[test]
fn rrf_fusion_single_list() {
    let d = Document::new("only".to_string());
    let fused = rrf_fusion(&[vec![Similarity { document: d, score: 1.0 }]], 60, 10);
    assert_eq!(fused.len(), 1);
}

// --- Merge hybrid with configurable BM25 ---

#[test]
fn merge_hybrid_with_custom_bm25() {
    let docs = vec![
        Document::new("rust programming language safety".to_string()),
        Document::new("python easy scripting dynamic".to_string()),
    ];
    let mut map = HashMap::new();
    for d in &docs {
        map.insert(d.id.clone(), d.clone());
    }

    let idx = Bm25Index::from_documents_with_config(&docs, Bm25Config::new(1.5, 0.5)).unwrap();
    let kw = idx.search("rust safety", 10);

    let vec_hits = vec![
        Similarity {
            document: docs[0].clone(),
            score: 0.95,
        },
        Similarity {
            document: docs[1].clone(),
            score: 0.3,
        },
    ];

    let merged = merge_hybrid(&map, &vec_hits, &kw, 0.7, 2).unwrap();
    assert_eq!(merged.len(), 2);
}

// --- Phrase and prefix search ---

#[test]
fn bm25_phrase_search_exact() {
    let docs = vec![
        Document::new("machine learning is great".to_string()),
        Document::new("deep learning and machine vision".to_string()),
    ];
    let idx = Bm25Index::from_documents(&docs).unwrap();
    let hits = idx.search_phrase("machine learning", 2);
    assert_eq!(hits.len(), 1);
    assert_eq!(hits[0].0, docs[0].id);
}

#[test]
fn bm25_prefix_search() {
    let docs = vec![
        Document::new("rust programming".to_string()),
        Document::new("python scripting".to_string()),
    ];
    let idx = Bm25Index::from_documents(&docs).unwrap();
    let hits = idx.search_prefix("rust", 2);
    assert_eq!(hits.len(), 1);
    assert_eq!(hits[0].0, docs[0].id);
}

// --- Field-level BM25 ---

#[test]
fn field_bm25_ranks_by_boost() {
    let mut d1 = Document::new("content1".to_string());
    d1.metadata.insert("title".to_string(), "rust systems".to_string());
    d1.metadata.insert("body".to_string(), "some text".to_string());

    let mut d2 = Document::new("content2".to_string());
    d2.metadata.insert("title".to_string(), "other topic".to_string());
    d2.metadata.insert("body".to_string(), "rust systems deep dive".to_string());

    let docs = vec![d1, d2];
    let mut idx = FieldBm25Index::new(vec![
        ("title".to_string(), 3.0),
        ("body".to_string(), 1.0),
    ]);
    idx.build(&docs).unwrap();

    let hits = idx.search("rust systems", 2);
    // d1 has rust systems in title (boosted 3x), d2 only in body
    assert!(!hits.is_empty());
}

// --- Lazy-flush persistence ---

#[tokio::test]
async fn lazy_flush_no_file_write_until_explicit() {
    let dir = TempDir::new().unwrap();
    let path = dir.path().join("vectors.json");

    let store = JsonPersistentVectorStore::open_lazy_flush(&path).await.unwrap();
    let doc = Document::new("hello".to_string()).with_embedding(vec![1.0, 0.0]);
    store.add(doc).await.unwrap();

    // File should not exist yet
    assert!(!path.exists());

    // Now flush explicitly
    store.flush().await.unwrap();
    assert!(path.exists());
}

#[tokio::test]
async fn lazy_flush_reload() {
    let dir = TempDir::new().unwrap();
    let path = dir.path().join("vectors.json");

    let store = JsonPersistentVectorStore::open_lazy_flush(&path).await.unwrap();
    let doc = Document::new("persist me".to_string()).with_embedding(vec![0.5, 0.5]);
    store.add(doc.clone()).await.unwrap();
    store.flush().await.unwrap();

    let reloaded = JsonPersistentVectorStore::open(&path).await.unwrap();
    let count = reloaded.count().await.unwrap();
    assert_eq!(count, 1);
}

// --- Diversification ---

#[test]
fn diversify_limits_per_source() {
    let mut docs = Vec::new();
    for i in 0..5 {
        let mut doc = Document::new(format!("doc {}", i));
        doc.metadata.insert("source".to_string(), "A".to_string());
        docs.push(Similarity { document: doc, score: 1.0 - i as f32 * 0.1 });
    }
    for i in 0..3 {
        let mut doc = Document::new(format!("doc b{}", i));
        doc.metadata.insert("source".to_string(), "B".to_string());
        docs.push(Similarity { document: doc, score: 0.5 - i as f32 * 0.05 });
    }

    let out = diversify(docs, "source", 2, 10);
    assert_eq!(out.len(), 4); // 2 from A, 2 from B
}

#[test]
fn diversify_zero_max_returns_empty() {
    let docs = vec![Similarity { document: Document::new("a".to_string()), score: 1.0 }];
    let out = diversify(docs, "source", 0, 10);
    assert!(out.is_empty());
}

// --- Aggregation ---

#[test]
fn count_by_groups_metadata() {
    let docs = vec![
        Document::new("a".to_string()).with_metadata("tag".to_string(), "x".to_string()),
        Document::new("b".to_string()).with_metadata("tag".to_string(), "x".to_string()),
        Document::new("c".to_string()).with_metadata("tag".to_string(), "y".to_string()),
    ];
    let counts = count_by(&docs, "tag");
    assert_eq!(counts.get("x"), Some(&2));
    assert_eq!(counts.get("y"), Some(&1));
}

#[test]
fn group_by_collects_documents() {
    let docs = vec![
        Document::new("a".to_string()).with_metadata("cat".to_string(), "A".to_string()),
        Document::new("b".to_string()).with_metadata("cat".to_string(), "A".to_string()),
        Document::new("c".to_string()).with_metadata("cat".to_string(), "B".to_string()),
    ];
    let groups = group_by(&docs, "cat");
    assert_eq!(groups.get("A").unwrap().len(), 2);
    assert_eq!(groups.get("B").unwrap().len(), 1);
}

#[test]
fn sum_by_numeric_metadata() {
    let docs = vec![
        Document::new("a".to_string())
            .with_metadata("cat".to_string(), "A".to_string())
            .with_metadata("score".to_string(), "10".to_string()),
        Document::new("b".to_string())
            .with_metadata("cat".to_string(), "A".to_string())
            .with_metadata("score".to_string(), "20".to_string()),
        Document::new("c".to_string())
            .with_metadata("cat".to_string(), "B".to_string())
            .with_metadata("score".to_string(), "5".to_string()),
    ];
    let sums = sum_by(&docs, "cat", "score");
    assert_eq!(sums.get("A"), Some(&30.0));
    assert_eq!(sums.get("B"), Some(&5.0));
}

// --- Sparse vectors ---

#[test]
fn sparse_vector_search_integration() {
    let mut idx = SparseIndex::new();
    idx.add("doc1".to_string(), SparseVector::new().insert(0, 1.0).insert(2, 0.5));
    idx.add("doc2".to_string(), SparseVector::new().insert(1, 1.0).insert(2, 0.3));
    idx.add("doc3".to_string(), SparseVector::new().insert(0, 0.8).insert(2, 0.6));

    let results = idx.search(&SparseVector::new().insert(0, 1.0).insert(2, 0.5), 2);
    assert!(!results.is_empty());
    // doc1 and doc3 share dims 0 and 2 with query
    assert!(results.iter().any(|(id, _)| id == "doc1"));
}

#[test]
fn sparse_vector_dot_product() {
    let a = SparseVector::new().insert(0, 2.0).insert(5, 3.0);
    let b = SparseVector::new().insert(5, 4.0).insert(10, 1.0);
    assert_eq!(a.dot(&b), 12.0); // 3.0 * 4.0 = 12.0 at dim 5
}

#[test]
fn sparse_vector_cosine() {
    let a = SparseVector::new().insert(0, 1.0);
    let b = SparseVector::new().insert(0, 1.0);
    assert!((a.cosine(&b) - 1.0).abs() < 1e-6);
}

// --- Fuzzy matching ---

#[test]
fn fuzzy_filter_typo_tolerance() {
    let terms = vec!["hello", "hallo", "helo", "world", "help"];
    let matches = rag::fuzzy::fuzzy_filter("hello", &terms, 1);
    assert!(matches.contains(&"hello"));
    assert!(matches.contains(&"hallo"));
    assert!(matches.contains(&"helo"));
    assert!(!matches.contains(&"world"));
}

#[test]
fn levenshtein_distance_basic() {
    use rag::fuzzy::levenshtein;
    assert_eq!(levenshtein("kitten", "sitting"), 3);
    assert_eq!(levenshtein("hello", "hello"), 0);
    assert_eq!(levenshtein("hello", "helo"), 1);
}