vectordb 0.4.4

LanceDB: A serverless, low-latency vector database for AI applications
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

use std::sync::Arc;

use arrow_schema::Schema;

use lance::dataset::{ReadParams, WriteParams};
use lance::io::{ObjectStoreParams, WrappingObjectStore};

use crate::error::{Error, Result};

pub trait PatchStoreParam {
    fn patch_with_store_wrapper(
        self,
        wrapper: Arc<dyn WrappingObjectStore>,
    ) -> Result<Option<ObjectStoreParams>>;
}

impl PatchStoreParam for Option<ObjectStoreParams> {
    fn patch_with_store_wrapper(
        self,
        wrapper: Arc<dyn WrappingObjectStore>,
    ) -> Result<Option<ObjectStoreParams>> {
        let mut params = self.unwrap_or_default();
        if params.object_store_wrapper.is_some() {
            return Err(Error::Lance {
                message: "can not patch param because object store is already set".into(),
            });
        }
        params.object_store_wrapper = Some(wrapper);

        Ok(Some(params))
    }
}

pub trait PatchWriteParam {
    fn patch_with_store_wrapper(
        self,
        wrapper: Arc<dyn WrappingObjectStore>,
    ) -> Result<Option<WriteParams>>;
}

impl PatchWriteParam for Option<WriteParams> {
    fn patch_with_store_wrapper(
        self,
        wrapper: Arc<dyn WrappingObjectStore>,
    ) -> Result<Option<WriteParams>> {
        let mut params = self.unwrap_or_default();
        params.store_params = params.store_params.patch_with_store_wrapper(wrapper)?;
        Ok(Some(params))
    }
}

// NOTE: we have some API inconsistency here.
// WriteParam is found in the form of Option<WriteParam> and ReadParam is found in the form of ReadParam

pub trait PatchReadParam {
    fn patch_with_store_wrapper(self, wrapper: Arc<dyn WrappingObjectStore>) -> Result<ReadParams>;
}

impl PatchReadParam for ReadParams {
    fn patch_with_store_wrapper(
        mut self,
        wrapper: Arc<dyn WrappingObjectStore>,
    ) -> Result<ReadParams> {
        self.store_options = self.store_options.patch_with_store_wrapper(wrapper)?;
        Ok(self)
    }
}

/// Find one default column to create index.
pub(crate) fn default_vector_column(schema: &Schema, dim: Option<i32>) -> Result<String> {
    // Try to find one fixed size list array column.
    let candidates = schema
        .fields()
        .iter()
        .filter_map(|field| match field.data_type() {
            arrow_schema::DataType::FixedSizeList(f, d)
                if f.data_type().is_floating()
                    && dim.map(|expect| *d == expect).unwrap_or(true) =>
            {
                Some(field.name())
            }
            _ => None,
        })
        .collect::<Vec<_>>();
    if candidates.is_empty() {
        Err(Error::Store {
            message: "No vector column found to create index".to_string(),
        })
    } else if candidates.len() != 1 {
        Err(Error::Store {
            message: format!(
                "More than one vector columns found, \
                    please specify which column to create index: {:?}",
                candidates
            ),
        })
    } else {
        Ok(candidates[0].to_string())
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    use arrow_schema::{DataType, Field};

    #[test]
    fn test_guess_default_column() {
        let schema_no_vector = Schema::new(vec![
            Field::new("id", DataType::Int16, true),
            Field::new("tag", DataType::Utf8, false),
        ]);
        assert!(default_vector_column(&schema_no_vector, None)
            .unwrap_err()
            .to_string()
            .contains("No vector column"));

        let schema_with_vec_col = Schema::new(vec![
            Field::new("id", DataType::Int16, true),
            Field::new(
                "vec",
                DataType::FixedSizeList(Arc::new(Field::new("item", DataType::Float64, false)), 10),
                false,
            ),
        ]);
        assert_eq!(
            default_vector_column(&schema_with_vec_col, None).unwrap(),
            "vec"
        );

        let multi_vec_col = Schema::new(vec![
            Field::new("id", DataType::Int16, true),
            Field::new(
                "vec",
                DataType::FixedSizeList(Arc::new(Field::new("item", DataType::Float64, false)), 10),
                false,
            ),
            Field::new(
                "vec2",
                DataType::FixedSizeList(Arc::new(Field::new("item", DataType::Float64, false)), 50),
                false,
            ),
        ]);
        assert!(default_vector_column(&multi_vec_col, None)
            .unwrap_err()
            .to_string()
            .contains("More than one"));
    }
}