[CHORE]: Move jq out of core (#2828)

Cargo.toml

@@ -12,6 +12,7 @@ daft-core = {path = "src/daft-core", default-features = false}
 daft-csv = {path = "src/daft-csv", default-features = false}
 daft-dsl = {path = "src/daft-dsl", default-features = false}
 daft-functions = {path = "src/daft-functions", default-features = false}
+daft-functions-json = {path = "src/daft-functions-json", default-features = false}
 daft-image = {path = "src/daft-image", default-features = false}
 daft-io = {path = "src/daft-io", default-features = false}
 daft-json = {path = "src/daft-json", default-features = false}
@@ -53,6 +54,7 @@ python = [
   "daft-sql/python",
   "daft-table/python",
   "daft-functions/python",
+  "daft-functions-json/python",
   "common-daft-config/python",
   "common-system-info/python",
   "common-display/python",
@@ -129,6 +131,7 @@ members = [
   "src/daft-scheduler",
   "src/daft-sketch",
   "src/daft-functions",
+  "src/daft-functions-json",
   "src/daft-sql",
   "src/hyperloglog"
 ]

daft/daft/__init__.pyi

@@ -1165,7 +1165,6 @@ class PyExpr:
     def partitioning_years(self) -> PyExpr: ...
     def partitioning_iceberg_bucket(self, n: int) -> PyExpr: ...
     def partitioning_iceberg_truncate(self, w: int) -> PyExpr: ...
-    def json_query(self, query: str) -> PyExpr: ...
 
     ###
     # Helper methods required by optimizer:
@@ -1256,6 +1255,11 @@ def image_encode(expr: PyExpr, image_format: ImageFormat) -> PyExpr: ...
 def image_resize(expr: PyExpr, w: int, h: int) -> PyExpr: ...
 def image_to_mode(expr: PyExpr, mode: ImageMode) -> PyExpr: ...
 
+# ---
+# expr.json namespace
+# ---
+def json_query(expr: PyExpr, query: str) -> PyExpr: ...
+
 class PyCatalog:
     @staticmethod
     def new() -> PyCatalog: ...

daft/expressions/expressions.py

@@ -3391,7 +3391,7 @@ def query(self, jq_query: str) -> Expression:
             Expression: Expression representing the result of the JQ query as a column of JSON-compatible strings
         """
 
-        return Expression._from_pyexpr(self._expr.json_query(jq_query))
+        return Expression._from_pyexpr(native.json_query(self._expr, jq_query))
 
 
 class ExpressionEmbeddingNamespace(ExpressionNamespace):

src/daft-core/Cargo.toml

@@ -34,10 +34,6 @@ html-escape = {workspace = true}
 hyperloglog = {path = "../hyperloglog"}
 indexmap = {workspace = true, features = ["serde"]}
 itertools = {workspace = true}
-jaq-core = {workspace = true}
-jaq-interpret = {workspace = true}
-jaq-parse = {workspace = true}
-jaq-std = {workspace = true}
 lazy_static = {workspace = true}
 log = {workspace = true}
 mur3 = "0.1.0"
@@ -46,7 +42,6 @@ num-traits = {workspace = true}
 pyo3 = {workspace = true, optional = true}
 regex = {workspace = true}
 serde = {workspace = true}
-serde_json = {workspace = true}
 sketches-ddsketch = {workspace = true}
 unicode-normalization = "0.1.23"
 

src/daft-core/src/array/ops/mod.rs

@@ -30,7 +30,6 @@ mod hll_merge;
 mod hll_sketch;
 mod if_else;
 mod is_in;
-mod json;
 mod len;
 mod list;
 mod list_agg;

src/daft-core/src/datatypes/mod.rs

@@ -362,3 +362,11 @@ impl<T: DaftNumericType> DataArray<T> {
         self.as_arrow().values().as_slice()
     }
 }
+
+impl<P: AsRef<str>> FromIterator<Option<P>> for Utf8Array {
+    #[inline]
+    fn from_iter<I: IntoIterator<Item = Option<P>>>(iter: I) -> Self {
+        let arrow_arr = arrow2::array::Utf8Array::<i64>::from_iter(iter);
+        Self::from(("", Box::new(arrow_arr)))
+    }
+}

src/daft-core/src/series/ops/mod.rs

@@ -22,7 +22,6 @@ pub mod groups;
 pub mod hash;
 pub mod if_else;
 pub mod is_in;
-pub mod json;
 pub mod len;
 pub mod list;
 pub mod log;

src/daft-dsl/src/functions/mod.rs

@@ -1,5 +1,4 @@
 pub mod float;
-pub mod json;
 pub mod list;
 pub mod map;
 pub mod numeric;
@@ -17,7 +16,6 @@ use std::hash::Hash;
 use crate::{Expr, ExprRef, Operator};
 
 use self::float::FloatExpr;
-use self::json::JsonExpr;
 use self::list::ListExpr;
 use self::map::MapExpr;
 use self::numeric::NumericExpr;
@@ -46,7 +44,6 @@ pub enum FunctionExpr {
     Map(MapExpr),
     Sketch(SketchExpr),
     Struct(StructExpr),
-    Json(JsonExpr),
     Python(PythonUDF),
     Partitioning(PartitioningExpr),
 }
@@ -75,7 +72,6 @@ impl FunctionExpr {
             Map(expr) => expr.get_evaluator(),
             Sketch(expr) => expr.get_evaluator(),
             Struct(expr) => expr.get_evaluator(),
-            Json(expr) => expr.get_evaluator(),
             Python(expr) => expr.get_evaluator(),
             Partitioning(expr) => expr.get_evaluator(),
         }