Skip to content

Add pgvector support to PostgreSQL dialect #2083

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
benfdking wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add pgvector support to PostgreSQL dialect #2083

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
169 changes: 169 additions & 0 deletions crates/lib-dialects/src/postgres.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6710,3 +6710,172 @@ pub fn statement_segment() -> Matchable {
false,
)
}

/// Apply pgvector extension support to the postgres dialect.
/// This adds VECTOR, HALFVEC, and SPARSEVEC types with optional dimension parameters.
pub fn apply_pgvector_extension(postgres: &mut Dialect) {
use sqruff_lib_core::parser::types::DialectElementType;

// Define the pgvector type keywords
const PGVECTOR_KEYWORDS: &[&str] = &["VECTOR", "HALFVEC", "SPARSEVEC"];

// Add keywords to the unreserved_keywords set
for &kw in PGVECTOR_KEYWORDS {
postgres.add_keyword_to_set("unreserved_keywords", kw);
}

// Also add the keywords directly to the library as StringParsers
// (since expand() has already been called and won't process newly added set members)
for &kw in PGVECTOR_KEYWORDS {
let parser = StringParser::new(kw, SyntaxKind::Keyword);
postgres.add([(kw.into(), DialectElementType::Matchable(parser.to_matchable()))]);
}

// Replace DatatypeSegment grammar to include pgvector types
postgres.replace_grammar(
"DatatypeSegment",
Sequence::new(vec_of_erased![
Sequence::new(vec_of_erased![
Ref::new("SingleIdentifierGrammar"),
Ref::new("DotSegment")
])
.config(|this| {
this.allow_gaps = false;
this.optional();
}),
one_of(vec_of_erased![
Ref::new("WellKnownTextGeometrySegment"),
Ref::new("DateTimeTypeIdentifier"),
Sequence::new(vec_of_erased![one_of(vec_of_erased![
Ref::keyword("SMALLINT"),
Ref::keyword("INTEGER"),
Ref::keyword("INT"),
Ref::keyword("INT2"),
Ref::keyword("INT4"),
Ref::keyword("INT8"),
Ref::keyword("BIGINT"),
Ref::keyword("FLOAT4"),
Ref::keyword("FLOAT8"),
Ref::keyword("REAL"),
Sequence::new(vec_of_erased![
Ref::keyword("DOUBLE"),
Ref::keyword("PRECISION")
]),
Ref::keyword("SMALLSERIAL"),
Ref::keyword("SERIAL"),
Ref::keyword("SERIAL2"),
Ref::keyword("SERIAL4"),
Ref::keyword("SERIAL8"),
Ref::keyword("BIGSERIAL"),
// Numeric types [(precision)]
Sequence::new(vec_of_erased![
one_of(vec_of_erased![Ref::keyword("FLOAT")]),
Ref::new("BracketedArguments").optional()
]),
// Numeric types [precision ["," scale])]
Sequence::new(vec_of_erased![
one_of(vec_of_erased![
Ref::keyword("DECIMAL"),
Ref::keyword("NUMERIC")
]),
Ref::new("BracketedArguments").optional()
]),
// Monetary type
Ref::keyword("MONEY"),
// Character types
one_of(vec_of_erased![
Sequence::new(vec_of_erased![
one_of(vec_of_erased![
Ref::keyword("BPCHAR"),
Ref::keyword("CHAR"),
Sequence::new(vec_of_erased![
Ref::keyword("CHAR"),
Ref::keyword("VARYING")
]),
Ref::keyword("CHARACTER"),
Sequence::new(vec_of_erased![
Ref::keyword("CHARACTER"),
Ref::keyword("VARYING")
]),
Ref::keyword("VARCHAR")
]),
Ref::new("BracketedArguments").optional()
]),
Ref::keyword("TEXT")
]),
// Binary type
Ref::keyword("BYTEA"),
// Boolean types
one_of(vec_of_erased![
Ref::keyword("BOOLEAN"),
Ref::keyword("BOOL")
]),
// Geometric types
one_of(vec_of_erased![
Ref::keyword("POINT"),
Ref::keyword("LINE"),
Ref::keyword("LSEG"),
Ref::keyword("BOX"),
Ref::keyword("PATH"),
Ref::keyword("POLYGON"),
Ref::keyword("CIRCLE")
]),
// Network address types
one_of(vec_of_erased![
Ref::keyword("CIDR"),
Ref::keyword("INET"),
Ref::keyword("MACADDR"),
Ref::keyword("MACADDR8")
]),
// Text search types
one_of(vec_of_erased![
Ref::keyword("TSVECTOR"),
Ref::keyword("TSQUERY")
]),
// pgvector types (dimension is optional)
Sequence::new(vec_of_erased![
one_of(vec_of_erased![
Ref::keyword("VECTOR"),
Ref::keyword("HALFVEC"),
Ref::keyword("SPARSEVEC")
]),
Ref::new("BracketedArguments").optional()
]),
// Bit string types
Sequence::new(vec_of_erased![
Ref::keyword("BIT"),
one_of(vec_of_erased![Ref::keyword("VARYING")])
.config(|this| this.optional()),
Ref::new("BracketedArguments").optional()
]),
// UUID type
Ref::keyword("UUID"),
// XML type
Ref::keyword("XML"),
// JSON types
one_of(vec_of_erased![Ref::keyword("JSON"), Ref::keyword("JSONB")]),
// Range types
Ref::keyword("INT4RANGE"),
Ref::keyword("INT8RANGE"),
Ref::keyword("NUMRANGE"),
Ref::keyword("TSRANGE"),
Ref::keyword("TSTZRANGE"),
Ref::keyword("DATERANGE"),
// pg_lsn type
Ref::keyword("PG_LSN")
])]),
Ref::new("DatatypeIdentifierSegment")
]),
one_of(vec_of_erased![
AnyNumberOf::new(vec_of_erased![
Bracketed::new(vec_of_erased![Ref::new("ExpressionSegment").optional()])
.config(|this| this.bracket_type("square"))
]),
Ref::new("ArrayTypeSegment"),
Ref::new("SizedArrayTypeSegment"),
])
.config(|this| this.optional()),
])
.to_matchable(),
);
}
2 changes: 2 additions & 0 deletions crates/lib/Cargo.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,8 +38,10 @@ name = "depth_map"
harness = false

[features]
default = ["postgres"]
parser = ["sqruff-lib-core/stringify"]
python = ["pyo3", "sqruff-lib-core/serde"]
postgres = ["sqruff-lib-dialects/postgres"]

[dependencies]
sqruff-lib-core.workspace = true
Expand Down
27 changes: 26 additions & 1 deletion crates/lib/src/core/config.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,6 +10,8 @@ use sqruff_lib_core::dialects::init::{DialectKind, dialect_readout};
use sqruff_lib_core::errors::SQLFluffUserError;
use sqruff_lib_core::parser::Parser;
use sqruff_lib_dialects::kind_to_dialect;
#[cfg(feature = "postgres")]
use sqruff_lib_dialects::postgres::apply_pgvector_extension;

use crate::utils::reflow::config::ReflowConfig;

Expand Down Expand Up @@ -121,7 +123,30 @@ impl FluffConfig {
_value => DialectKind::default(),
};

let dialect = kind_to_dialect(&dialect);
let mut dialect = kind_to_dialect(&dialect);

// Apply dialect extensions based on configuration
#[cfg(feature = "postgres")]
if let Some(ref mut dialect) = dialect {
if dialect.name() == DialectKind::Postgres {
// Check for postgres extensions configuration
if let Some(postgres_config) = configs.get("postgres") {
if let Some(extensions) = postgres_config
.as_map()
.and_then(|m| m.get("extensions"))
.and_then(|v| v.as_string())
{
// Parse comma-separated extension list
for ext in extensions.split(',').map(|s| s.trim()) {
if ext.eq_ignore_ascii_case("pgvector") {
apply_pgvector_extension(dialect);
}
}
}
}
}
}

for (in_key, out_key) in [
// Deal with potential ignore & warning parameters
("ignore", "ignore"),
Expand Down
Loading