Convert into arrow without going via into_canonical (#1736)

spiraldb · Jan 2, 2025 · 14f51e8 · 14f51e8
1 parent 1b992fe
commit 14f51e8
Show file tree

Hide file tree

Showing 6 changed files with 14 additions and 15 deletions.
diff --git a/docs/guide.rst b/docs/guide.rst
@@ -91,7 +91,7 @@ If you have a struct array, use :meth:`.Array.to_arrow_table` to construct an Ar
    >>> struct_vtx.to_arrow_table()
    pyarrow.Table
    age: int64
-   name: string_view
+   name: string
    ----
    age: [[25,31,33,57]]
    name: [["Joseph","Narendra","Angela","Mikhail"]]
@@ -116,7 +116,7 @@ Pandas
    >>> vortex.array(df).to_arrow_table()
    pyarrow.Table
    age: int64
-   name: string_view
+   name: string
    ----
    age: [[25,31,33,57]]
    name: [["Joseph","Narendra","Angela","Mikhail"]]

diff --git a/pyvortex/python/vortex/encoding.py b/pyvortex/python/vortex/encoding.py
@@ -71,7 +71,7 @@ def _Array_to_arrow_table(self: _encoding.Array) -> pyarrow.Table:
     >>> array.to_arrow_table()
     pyarrow.Table
     age: int64
-    name: string_view
+    name: string
     ----
     age: [[25,31,33,57]]
     name: [["Joseph","Narendra","Angela","Mikhail"]]
@@ -357,7 +357,7 @@ def array(obj: pyarrow.Array | list | Any) -> Array:
     <pyarrow.lib.ChunkedArray object at ...>
     [
       -- is_valid: all not null
-      -- child 0 type: string_view
+      -- child 0 type: string
         [
           "Braund",
           "Allen",

diff --git a/pyvortex/src/expr.rs b/pyvortex/src/expr.rs
@@ -38,7 +38,7 @@ use crate::dtype::PyDType;
 ///   [
 ///     57
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Mikhail"
 ///   ]
@@ -56,7 +56,7 @@ use crate::dtype::PyDType;
 ///     25,
 ///     31
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Joseph",
 ///     null
@@ -73,7 +73,7 @@ use crate::dtype::PyDType;
 ///   [
 ///     25
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Joseph"
 ///   ]
@@ -90,7 +90,7 @@ use crate::dtype::PyDType;
 ///     null,
 ///     57
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Angela",
 ///     "Mikhail"
@@ -110,7 +110,7 @@ use crate::dtype::PyDType;
 ///     25,
 ///     null
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Joseph",
 ///     "Angela"

diff --git a/pyvortex/src/io.rs b/pyvortex/src/io.rs
@@ -48,7 +48,7 @@ use crate::{PyArray, TOKIO_RUNTIME};
 ///     57,
 ///     null
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Joseph",
 ///     null,
@@ -78,7 +78,7 @@ use crate::{PyArray, TOKIO_RUNTIME};
 /// >>> d.to_arrow_array()
 /// <pyarrow.lib.StructArray object at ...>
 /// -- is_valid: all not null
-/// -- child 0 type: string_view
+/// -- child 0 type: string
 ///   [
 ///     "Joseph",
 ///     null,
@@ -98,7 +98,7 @@ use crate::{PyArray, TOKIO_RUNTIME};
 ///   [
 ///     57
 ///   ]
-/// -- child 1 type: string_view
+/// -- child 1 type: string
 ///   [
 ///     "Mikhail"
 ///   ]

diff --git a/vortex-array/src/arrow/record_batch.rs b/vortex-array/src/arrow/record_batch.rs
@@ -49,7 +49,7 @@ impl TryFrom<StructArray> for RecordBatch {
     type Error = VortexError;
 
     fn try_from(value: StructArray) -> VortexResult<Self> {
-        let array_ref = value.into_canonical()?.into_arrow()?;
+        let array_ref = value.into_array().into_arrow()?;
         let struct_array = as_struct_array(array_ref.as_ref());
         Ok(Self::from(struct_array))
     }

diff --git a/vortex-array/src/canonical.rs b/vortex-array/src/canonical.rs
@@ -201,9 +201,8 @@ fn struct_to_arrow(struct_array: StructArray) -> VortexResult<ArrayRef> {
         .iter()
         .zip(struct_array.children())
         .map(|(name, f)| {
-            f.into_canonical()
+            f.into_arrow()
                 .map_err(|err| err.with_context(format!("Failed to canonicalize field {}", name)))
-                .and_then(|c| c.into_arrow())
         })
         .collect::<VortexResult<Vec<_>>>()?;