arrs-cli 0.1.3

Command-line tool for inspecting Lance and other Arrow-based datasets.
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

//! Write a Lance dataset to ./sample.lance with multiple versions on `main`,
//! a tag, a `dev` branch, and a BTree index — enough to exercise every Lance
//! feature the CLI exposes.
//!
//! Run:
//!     cargo run --example write_lance_fixture
//!
//! Then try the CLI:
//!     # Metadata commands
//!     cargo run -- versions sample.lance                       # every main version
//!     cargo run -- versions --tagged-only sample.lance         # only tagged versions
//!     cargo run -- versions --branch dev sample.lance          # dev branch (every version)
//!     cargo run -- branches sample.lance
//!     cargo run -- tags sample.lance                            # cross-branch tag listing
//!     cargo run -- indices sample.lance
//!
//!     # Time-travel reads
//!     cargo run -- rowcount sample.lance                       # latest of main
//!     cargo run -- rowcount --version 1 sample.lance           # just v1
//!     cargo run -- rowcount --tag release-v2 sample.lance      # at the tag
//!     cargo run -- head -n 5 --branch dev sample.lance         # dev rows
//!
//!     # Negative cases
//!     cargo run -- head --branch main --tag release-on-dev sample.lance  # error: tag/branch mismatch

use std::path::PathBuf;
use std::sync::Arc;

use arrow_array::builder::{ListBuilder, StringBuilder};
use arrow_array::{
    BinaryArray, BooleanArray, Float64Array, Int32Array, RecordBatch, RecordBatchIterator,
    StringArray, TimestampMicrosecondArray,
};
use arrow_schema::{DataType, Field, Schema, SchemaRef, TimeUnit};
use lance::Dataset;
use lance_index::DatasetIndexExt as _;
use lance_index::IndexType;
use lance_index::scalar::ScalarIndexParams;

fn schema() -> SchemaRef {
    Arc::new(Schema::new(vec![
        Field::new("id", DataType::Int32, false),
        Field::new("name", DataType::Utf8, true),
        Field::new("score", DataType::Float64, true),
        Field::new("flag", DataType::Boolean, true),
        Field::new("ts", DataType::Timestamp(TimeUnit::Microsecond, None), true),
        Field::new("data", DataType::Binary, true),
        Field::new_list(
            "tags",
            Arc::new(Field::new("item", DataType::Utf8, true)),
            true,
        ),
    ]))
}

/// Build a record batch with the rows whose `id` is in `ids`. Other columns
/// are derived deterministically so each row is identifiable in the output.
fn build_batch(ids: &[i32], name_prefix: &str) -> RecordBatch {
    let names: Vec<String> = ids.iter().map(|i| format!("{name_prefix}{i}")).collect();
    let scores: Vec<Option<f64>> = ids
        .iter()
        .map(|i| match i % 5 {
            0 => Some(f64::NAN),
            1 => None,
            2 => Some(f64::INFINITY),
            n => Some(f64::from(n) * 1.5),
        })
        .collect();
    let flags: Vec<Option<bool>> = ids.iter().map(|i| Some(i % 2 == 0)).collect();
    let base_ts = 1_700_000_000_000_000i64;
    let ts: Vec<Option<i64>> = ids
        .iter()
        .map(|i| Some(base_ts + i64::from(*i) * 1_000_000))
        .collect();
    let datas: Vec<Option<Vec<u8>>> = ids
        .iter()
        .map(|i| Some(vec![0xAB, (*i & 0xff) as u8]))
        .collect();

    let mut tags_builder = ListBuilder::new(StringBuilder::new());
    for i in ids {
        tags_builder
            .values()
            .append_value(name_prefix.trim_end_matches('-'));
        tags_builder.values().append_value(format!("id-{i}"));
        tags_builder.append(true);
    }

    RecordBatch::try_new(
        schema(),
        vec![
            Arc::new(Int32Array::from(ids.to_vec())),
            Arc::new(StringArray::from(
                names.iter().map(String::as_str).collect::<Vec<_>>(),
            )),
            Arc::new(Float64Array::from(scores)),
            Arc::new(BooleanArray::from(flags)),
            Arc::new(TimestampMicrosecondArray::from(ts)),
            Arc::new(BinaryArray::from_opt_vec(
                datas.iter().map(|o| o.as_deref()).collect::<Vec<_>>(),
            )),
            Arc::new(tags_builder.finish()),
        ],
    )
    .unwrap()
}

async fn append(ds: &mut Dataset, batch: RecordBatch) -> anyhow::Result<()> {
    let iter = RecordBatchIterator::new(vec![Ok(batch)].into_iter(), schema());
    ds.append(iter, None).await?;
    Ok(())
}

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    let path = PathBuf::from("sample.lance");
    if path.exists() {
        std::fs::remove_dir_all(&path)?;
    }
    let uri = path.to_string_lossy().into_owned();

    // ----- main: v1 -----------------------------------------------------------
    let v1 = build_batch(&[1, 2, 3], "main-");
    let iter = RecordBatchIterator::new(vec![Ok(v1)].into_iter(), schema());
    let mut ds = Dataset::write(iter, uri.as_str(), None).await?;
    println!("main v1: 3 rows (ids 1..=3)");

    // ----- main: v2 (append) --------------------------------------------------
    append(&mut ds, build_batch(&[4, 5, 6], "main-")).await?;
    println!("main v2: appended 3 rows (ids 4..=6)");

    // tag the v2 manifest as `release-v2`
    ds.tags().create("release-v2", 2u64).await?;
    println!("tag 'release-v2' -> main v2");

    // ----- main: v3 (append) --------------------------------------------------
    append(&mut ds, build_batch(&[7, 8, 9], "main-")).await?;
    println!("main v3: appended 3 rows (ids 7..=9)");

    // ----- branch: dev branched from main v2 ---------------------------------
    let mut dev = ds.create_branch("dev", 2u64, None).await?;
    println!("branch 'dev' from main v2");

    // dev v2 (append): rows clearly distinguishable by id range and prefix
    append(&mut dev, build_batch(&[100, 101, 102], "dev-")).await?;
    println!("dev: appended 3 rows (ids 100..=102) on top of inherited 6");

    // tag release-on-dev points at dev's latest version
    dev.tags().create("release-on-dev", ("dev", 2u64)).await?;
    println!("tag 'release-on-dev' -> dev v2");

    // ----- index on main's latest --------------------------------------------
    ds.create_index(
        &["id"],
        IndexType::BTree,
        Some("idx_id".to_string()),
        &ScalarIndexParams::default(),
        false,
    )
    .await?;
    println!("BTree index 'idx_id' on column 'id' (main latest)");

    println!("\nfixture ready at {}", path.display());
    Ok(())
}