rust/lancedb
1
0
Fork 0
mirror of https://github.com/lancedb/lancedb.git synced 2026-09-01 02:58:36 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: expose list-element FTS document granularity (#4050)

Browse Source 
## Summary

LanceDB could not request Lance's list-element FTS document granularity
through Python or Remote APIs, and generic nested-field resolution
exposed Arrow's internal `item` segment instead of the public field
path.

This exposes typed `row | list_element` configuration for Python FTS
index creation and match/phrase queries, preserves `_doc_index`, and
keeps nested FTS paths public (for example, `docs.content`). Remote
list-element requests require server API version 0.6.0 so older servers
cannot silently execute them with row semantics; explicit row requests
remain compatible.

## Compatibility

Omitted index and query parameters retain row granularity. Remote row
index creation omits the new wire field.

## Tracking


[ENT-2342](https://linear.app/lancedb/issue/ENT-2342/expose-list-element-fts-document-granularity-end-to-end)
This commit is contained in:
Xuanwo
2026-08-26 23:27:39 +08:00
committed by GitHub
parent 06872463cf
commit b78f2a5044
15 changed files with 723 additions and 55 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/src/index.rs
+8 -2
View File
@@ -8,7 +8,7 @@ use lancedb::index::vector::{
};
use lancedb::index::{
Index as LanceDbIndex,
scalar::{BTreeIndexBuilder, FmIndexBuilder, FtsIndexBuilder},
scalar::{BTreeIndexBuilder, DocumentGranularity, FmIndexBuilder, FtsIndexBuilder},
};
use pyo3::IntoPyObject;
use pyo3::types::PyStringMethods;
@@ -60,7 +60,11 @@ pub fn extract_index_params(source: &Option<Bound<'_, PyAny>>) -> PyResult<Lance
.ngram_min_length(params.ngram_min_length)
.ngram_max_length(params.ngram_max_length)
.ngram_prefix_only(params.prefix_only)
.custom_stop_words(params.custom_stop_words);
.custom_stop_words(params.custom_stop_words)
.document_granularity(
DocumentGranularity::try_from(params.document_granularity.as_str())
.map_err(|err| PyValueError::new_err(err.to_string()))?,
);
if let Some(memory_limit) = params.memory_limit {
inner_opts = inner_opts.memory_limit_mb(memory_limit);
}
@@ -221,6 +225,7 @@ struct FtsParams {
block_size: usize,
memory_limit: Option<u64>,
num_workers: Option<usize>,
document_granularity: String,
}
#[derive(FromPyObject)]
@@ -481,6 +486,7 @@ mod tests {
block_size = 128
memory_limit = 2048
num_workers = 7
document_granularity = 'row'
config = FTS()",
None,
python/src/query.rs
+45 -12
View File
@@ -16,8 +16,8 @@ use arrow::pyarrow::FromPyArrow;
use arrow::pyarrow::IntoPyArrow;
use arrow::pyarrow::ToPyArrow;
use lancedb::index::scalar::{
BooleanQuery, BoostQuery, FtsQuery, FullTextSearchQuery, MatchQuery, MultiMatchQuery, Occur,
Operator, PhraseQuery,
BooleanQuery, BoostQuery, DocumentGranularity, FtsQuery, FullTextSearchQuery, MatchQuery,
MultiMatchQuery, Occur, Operator, PhraseQuery,
};
use lancedb::query::AnalyzePlanDistributedMetrics;
use lancedb::query::QueryBase;
@@ -76,8 +76,16 @@ impl<'a, 'py> FromPyObject<'a, 'py> for PyLanceDB<FtsQuery> {
let max_expansions = ob.getattr("max_expansions")?.extract()?;
let operator = ob.getattr("operator")?.extract::<String>()?;
let prefix_length = ob.getattr("prefix_length")?.extract()?;
let document_granularity = ob
.getattr("document_granularity")?
.extract::<Option<String>>()?
.map(|value| {
DocumentGranularity::try_from(value.as_str())
.map_err(|err| PyValueError::new_err(err.to_string()))
})
.transpose()?;
Ok(Self(
let mut query =
MatchQuery::new(query)
.with_column(Some(column))
.with_boost(boost)
@@ -86,21 +94,32 @@ impl<'a, 'py> FromPyObject<'a, 'py> for PyLanceDB<FtsQuery> {
.with_operator(Operator::try_from(operator.as_str()).map_err(|e| {
PyValueError::new_err(format!("Invalid operator: {}", e))
})?)
.with_prefix_length(prefix_length)
.into(),
))
.with_prefix_length(prefix_length);
if let Some(document_granularity) = document_granularity {
query = query.with_document_granularity(document_granularity);
}
Ok(Self(query.into()))
}
"PhraseQuery" => {
let query = ob.getattr("query")?.extract()?;
let column = ob.getattr("column")?.extract()?;
let slop = ob.getattr("slop")?.extract()?;
let document_granularity = ob
.getattr("document_granularity")?
.extract::<Option<String>>()?
.map(|value| {
DocumentGranularity::try_from(value.as_str())
.map_err(|err| PyValueError::new_err(err.to_string()))
})
.transpose()?;
Ok(Self(
PhraseQuery::new(query)
.with_column(Some(column))
.with_slop(slop)
.into(),
))
let mut query = PhraseQuery::new(query)
.with_column(Some(column))
.with_slop(slop);
if let Some(document_granularity) = document_granularity {
query = query.with_document_granularity(document_granularity);
}
Ok(Self(query.into()))
}
"BoostQuery" => {
let positive: Self = ob.getattr("positive")?.extract()?;
@@ -167,6 +186,13 @@ impl<'py> IntoPyObject<'py> for PyLanceDB<FtsQuery> {
kwargs.set_item("max_expansions", query.max_expansions)?;
kwargs.set_item::<_, &str>("operator", query.operator.into())?;
kwargs.set_item("prefix_length", query.prefix_length)?;
if let Some(document_granularity) = query.document_granularity {
let value = match document_granularity {
DocumentGranularity::Row => "row",
DocumentGranularity::ListElement => "list_element",
};
kwargs.set_item("document_granularity", value)?;
}
namespace
.getattr(intern!(py, "MatchQuery"))?
.call((query.terms, query.column.unwrap()), Some(&kwargs))
@@ -174,6 +200,13 @@ impl<'py> IntoPyObject<'py> for PyLanceDB<FtsQuery> {
FtsQuery::Phrase(query) => {
let kwargs = PyDict::new(py);
kwargs.set_item("slop", query.slop)?;
if let Some(document_granularity) = query.document_granularity {
let value = match document_granularity {
DocumentGranularity::Row => "row",
DocumentGranularity::ListElement => "list_element",
};
kwargs.set_item("document_granularity", value)?;
}
namespace
.getattr(intern!(py, "PhraseQuery"))?
.call((query.terms, query.column.unwrap()), Some(&kwargs))