Full-Text Search Examples

Worked examples of fuzzy search, boosting, boolean queries, and substring search with LanceDB full-text search.

These worked examples build on the concepts from the Full-Text Search guide. They walk through creating sample tables, building FTS indices, and running fuzzy, phrase, boosted, boolean, and substring queries.

Fuzzy Search and Boosting Example

Generate Data

First, let’s create a table with sample text data for testing fuzzy search:

Python

import lancedb
import numpy as np
import pandas as pd
import random

# Connect to LanceDB
db = lancedb.connect(
    uri="db://your-project-slug",
    api_key="your-api-key",
    region="us-east-1"
)

# Generate 100 rows of random "<noun> <verb> <adverb> <adjective>" text
table_name = "fts-fuzzy-boosting-test"
vectors = [np.random.randn(128) for _ in range(100)]
verbs = ("runs", "hits", "jumps", "drives", "barfs")
adv = ("crazily.", "dutifully.", "foolishly.", "merrily.", "occasionally.")
adj = ("adorable", "clueless", "dirty", "odd", "stupid")

def sentence(nouns):
    return " ".join(random.choice(words) for words in (nouns, verbs, adv, adj))

text = [sentence(("puppy", "car")) for _ in range(100)]
text2 = [sentence(("rabbit", "girl", "monkey")) for _ in range(100)]
count = [random.randint(1, 10000) for _ in range(100)]

import * as lancedb from "@lancedb/lancedb"

const db = await lancedb.connect({
    uri: "db://your-project-slug",
    apiKey: "your-api-key",
    region: "us-east-1"
});

// Generate 100 rows of random "<noun> <verb> <adverb> <adjective>" text
const tableName = "fts-fuzzy-boosting-test-ts";
const n = 100;
const verbs = ["runs", "hits", "jumps", "drives", "barfs"];
const adverbs = ["crazily", "dutifully", "foolishly", "merrily", "occasionally"];
const adjectives = ["adorable", "clueless", "dirty", "odd", "stupid"];

const pick = (words: string[]) => words[Math.floor(Math.random() * words.length)];
const sentence = (nouns: string[]) =>
    [nouns, verbs, adverbs, adjectives].map(pick).join(" ");

const vectors = Array.from({ length: n }, () =>
    Array.from({ length: 128 }, () => Math.random() * 2 - 1)
);
const text = Array.from({ length: n }, () => sentence(["puppy", "car"]));
const text2 = Array.from({ length: n }, () => sentence(["rabbit", "girl", "monkey"]));
const count = Array.from({ length: n }, () => Math.floor(Math.random() * 10000) + 1);

Create Table

Python

# Create table with sample data
table = db.create_table(
    table_name,
    data=pd.DataFrame({
        "vector": vectors,
        "id": [i % 2 for i in range(100)],
        "text": text,
        "text2": text2,
        "count": count,
    }),
    mode="overwrite"
)

// Create table with sample data
const data = makeArrowTable(
    vectors.map((vector, i) => ({
        vector,
        id: i % 2,
        text: text[i],
        text2: text2[i],
        count: count[i],
    }))
);

const table = await db.createTable(tableName, data, { mode: "overwrite" });

Construct FTS Index

Create a full-text search index on the first text column:

Python

# Create FTS index on first text column
table.create_fts_index("text")
wait_for_index(table, "text_idx")

// Create FTS index on first text column
await table.createIndex("text", { config: Index.fts() });
await waitForIndex(table, "text_idx");

Then, create an index on the second text column:

Python

# Create FTS index on second text column
table.create_fts_index("text2")
wait_for_index(table, "text2_idx")

// Create FTS index on second text column
await table.createIndex("text2", { config: Index.fts() });
await waitForIndex(table, "text2_idx");

Basic and Fuzzy Search

Now we can perform basic, fuzzy, and prefix match searches:

Basic Exact Search

Python

from lancedb.query import MatchQuery

# Basic match (exact search)
basic_match_results = (
    table.search(MatchQuery("crazily", "text"))
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

import { MatchQuery } from "@lancedb/lancedb";

// Basic match (exact search)
const basicMatchResults = await table.query()
    .fullTextSearch(new MatchQuery("crazily", "text"))
    .select(["id", "text"])
    .limit(100)
    .toArray();

Fuzzy Search with Typos

Python

# Fuzzy match (allows typos)
fuzzy_results = (
    table.search(MatchQuery("craziou", "text", fuzziness=2))
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

// Fuzzy match (allows typos)
const fuzzyResults = await table.query()
    .fullTextSearch(new MatchQuery("craziou", "text", {
        fuzziness: 2,
    }))
    .select(["id", "text"])
    .limit(100)
    .toArray();

Prefix based Match

Prefix-based match allows you to search for documents containing words that start with a specific prefix.

Python

# Fuzzy match (allows typos)
fuzzy_results = (
    table.search(MatchQuery("cra", "text", prefix_length=3))
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

// Fuzzy match (allows typos)
const fuzzyResults = await table.query()
    .fullTextSearch(new MatchQuery("cra", "text", {
        prefixLength: 3,
    }))
    .select(["id", "text"])
    .limit(100)
    .toArray();

Phrase Match

Phrase matching enables you to search for exact sequences of words. Unlike regular text search which matches individual terms independently, phrase matching requires words to appear in the specified order with no intervening terms.

Phrase queries are supported but only for a single column; providing multiple columns with a quoted phrase raises an error.

Phrase matching is particularly useful for:

• Searching for specific multi-word expressions
• Matching exact titles or quotes
• Finding precise word combinations in a specific order

Python

# Exact phrase match
from lancedb.query import PhraseQuery

phrase_results = (
    table.search(PhraseQuery("puppy runs", "text"))
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

import { PhraseQuery } from "@lancedb/lancedb";

// Exact phrase match
const phraseResults = await table.query()
  .fullTextSearch(new PhraseQuery("puppy runs", "text"))
  .select(["id", "text"])
  .limit(100)
  .toArray();

Flexible Phrase Match

To provide more flexible phrase matching, LanceDB supports the slop parameter. This allows you to match phrases where the terms appear close to each other, even if they are not directly adjacent or in the exact order, as long as they are within the specified slop value.

For example, the phrase query “puppy merrily” would not return any results by default. However, if you set slop=1, it will match phrases like “puppy jumps merrily”, “puppy runs merrily”, and similar variations where one word appears between “puppy” and “merrily”.

Python

# Flexible phrase match with slop=1 for 'puppy merrily'
from lancedb.query import PhraseQuery

phrase_results = (
    table.search(PhraseQuery("puppy merrily", "text", slop=1))
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

import { PhraseQuery } from "@lancedb/lancedb";

// Flexible phrase match with slop=1 for 'puppy runs'
const phraseResults = await table.query()
  .fullTextSearch(new PhraseQuery("puppy runs", "text", { slop: 1 }))
  .select(["id", "text"])
  .limit(100)
  .toArray();

Search with Boosting

Boosting allows you to control the relative importance of different search terms or fields in your queries. This feature is particularly useful when you need to:

• Prioritize matches in certain columns
• Promote specific terms while demoting others
• Fine-tune relevance scoring for better search results

Parameter	Type	Default	Description
positive	Query	required	The primary query terms to match and promote in results
negative	Query	required	Terms to demote in the search results
negative_boost	float	0.5	Multiplier for negative matches (lower values = stronger demotion)

Python

from lancedb.query import MatchQuery, BoostQuery, MultiMatchQuery

# Boost data with 'runs' in text more than 'puppy' in text
boosting_results = (
  table.search(
      BoostQuery(
          MatchQuery("runs", "text"),
          MatchQuery("puppy", "text"),
          negative_boost=0.2,
      ),
  )
  .select(["id", "text"])
  .limit(100)
  .to_pandas()
)

# Search across both text and text2
multi_match_results = (
    table.search(MultiMatchQuery("crazily", ["text", "text2"]))
    .select(["id", "text", "text2"])
    .limit(100)
    .to_pandas()
)

# Search with field boosting
multi_match_boosting_results = (
    table.search(
        MultiMatchQuery("crazily", ["text", "text2"], boosts=[1.0, 2.0]),
    )
    .select(["id", "text", "text2"])
    .limit(100)
    .to_pandas()
)

import { MatchQuery, BoostQuery, MultiMatchQuery } from "@lancedb/lancedb";

// Boosting Example
const boostingResults = await table.query()
  .fullTextSearch(new BoostQuery(new MatchQuery("runs", "text"), new MatchQuery("puppy", "text"), {
    negativeBoost: 0.2,
  }))
  .select(["id", "text"])
  .limit(100)
  .toArray();

// Search across both text fields
const multiMatchResults = await table.query()
  .fullTextSearch(new MultiMatchQuery("crazily", ["text", "text2"]))
  .select(["id", "text", "text2"])
  .limit(100)
  .toArray();

// Search with field boosting
const multiMatchBoostingResults = await table.query()
  .fullTextSearch(new MultiMatchQuery("crazily", ["text", "text2"], {
    boosts: [1.0, 2.0],
  }))
  .select(["id", "text", "text2"])
  .limit(100)
  .toArray();

Best practices

• Use fuzzy search when handling user input that may contain typos or variations
• Apply field boosting to prioritize matches in more important columns
• Combine fuzzy search with boosting for robust and precise search results

Recommendations for optimal FTS performance:

• Create full-text search indices on text columns that will be frequently searched
• For hybrid search combining text and vectors, see our hybrid search guide
• For performance benchmarks, check our benchmark results
• For complex queries, use SQL to combine FTS with other filter conditions

Boolean Queries

LanceDB supports boolean logic in full-text search, allowing you to combine multiple queries using and and or operators. This is useful when you want to match documents that satisfy multiple conditions (intersection) or at least one of several conditions (union).

Combining Two Match Queries

In Python, you can combine two MatchQuery objects using either the and function or the & operator (e.g., MatchQuery("puppy", "text") and MatchQuery("merrily", "text")); both methods are supported and yield the same result. Similarly, you can use either the or function or the | operator to perform an or query.

In TypeScript, boolean queries are constructed using the BooleanQuery class with a list of [Occur, subquery] pairs. For example, to perform an AND query:

SQL

BooleanQuery([
[Occur.Must, new MatchQuery("puppy", "text")],
[Occur.Must, new MatchQuery("merrily", "text")],
])

This approach allows you to specify complex boolean logic by combining multiple subqueries with different Occur values (such as Must, Should, or MustNot).

Which queries are allowed?

A boolean query must include at least one SHOULD or MUST clause. Queries that contain only a MUST_NOT clause are not allowed.

Python

from lancedb.query import MatchQuery

# Example: Find documents containing both "puppy" and "merrily"
and_query = MatchQuery("puppy", "text") & MatchQuery("merrily", "text")
and_results = (
    table.search(and_query)
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

# Example: Find documents containing either "puppy" or "merrily"
or_query = MatchQuery("puppy", "text") | MatchQuery("merrily", "text")
or_results = (
    table.search(or_query)
    .select(["id", "text"])
    .limit(100)
    .to_pandas()
)

import { MatchQuery, BooleanQuery, Occur } from "@lancedb/lancedb";

// Flexible boolean queries with MatchQuery

// Find documents containing both "puppy" and "merrily"
const mustResults = await table
    .search(
      new BooleanQuery([
        [Occur.Must, new MatchQuery("puppy", "text")],
        [Occur.Must, new MatchQuery("merrily", "text")],
      ]),
    )
    .select(["id", "text"])
    .limit(100)
    .toArray();

// Find documents containing either "puppy" or "merrily"
const shouldResults = await table
    .search(
      new BooleanQuery([
        [Occur.Should, new MatchQuery("puppy", "text")],
        [Occur.Should, new MatchQuery("merrily", "text")],
      ]),
    )
    .select(["id", "text"])
    .limit(100)
    .toArray();

How to use booleans?

• Use and/&(Python), Occur.Must(Typescript) for intersection (documents must match all queries).
• Use or/|(Python), Occur.Should(Typescript) for union (documents must match at least one query).

Substring Search Example

LanceDB supports searching for substrings in text columns using n-gram tokenization. This is useful for finding partial matches within text content.

Setting Up the Table

First, create a table with sample text data and configure n-gram tokenization:

Python

import pyarrow as pa
import lancedb

db = lancedb.connect(":memory:")

data = pa.table({"text": ["hello world", "lance database", "lance is cool"]})
table = db.create_table("test", data=data)
table.create_fts_index("text", base_tokenizer="ngram")

Basic Substring Search

With the default n-gram settings (minimum length of 3), you can search for substrings of length 3 or more:

Python

results = table.search("lan", query_type="fts").limit(10).to_list()
assert len(results) == 2
assert set(r["text"] for r in results) == {"lance database", "lance is cool"}

results = (
    table.search("nce", query_type="fts").limit(10).to_list()
)  # spellchecker:disable-line
assert len(results) == 2
assert set(r["text"] for r in results) == {"lance database", "lance is cool"}

Handling Short Substrings

By default, the minimum n-gram length is 3, so shorter substrings like “la” won’t match:

Python

results = table.search("la", query_type="fts").limit(10).to_list()
assert len(results) == 0

Customizing N-gram Parameters

You can customize the n-gram behavior by adjusting the minimum length and using prefix-only matching:

Python

table.create_fts_index(
    "text",
    base_tokenizer="ngram",
    replace=True,
    ngram_min_length=2,
    prefix_only=True,
)

Testing Custom N-gram Settings

With the new settings, you can now search for shorter substrings and use prefix-only matching:

Python

results = table.search("lan", query_type="fts").limit(10).to_list()
assert len(results) == 2
assert set(r["text"] for r in results) == {"lance database", "lance is cool"}

results = (
    table.search("nce", query_type="fts").limit(10).to_list()
)  # spellchecker:disable-line
assert len(results) == 0

results = table.search("la", query_type="fts").limit(10).to_list()
assert len(results) == 2
assert set(r["text"] for r in results) == {"lance database", "lance is cool"}