Skip to content

Unify nested variant handling #5

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.

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

Merged
Changes from 1 commit
Commits
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
Next Next commit
Unify nested variant handling
  • Loading branch information
alamb committed Jul 14, 2025
commit 78dc65565bc2a2a087d261eb8a1a95b3d5a7d1d4
129 changes: 40 additions & 89 deletions parquet-variant/src/builder.rs
Original file line number Diff line number Diff line change
Expand Up @@ -16,8 +16,7 @@
// under the License.
use crate::decoder::{VariantBasicType, VariantPrimitiveType};
use crate::{
ShortString, Variant, VariantDecimal16, VariantDecimal4, VariantDecimal8, VariantList,
VariantObject,
ShortString, Variant, VariantDecimal16, VariantDecimal4, VariantDecimal8,
};
use arrow_schema::ArrowError;
use indexmap::{IndexMap, IndexSet};
Expand Down Expand Up @@ -195,7 +194,28 @@ impl ValueBuffer {
self.0.len()
}

fn append_variant<'m, 'd>(&mut self, variant: Variant<'m, 'd>) {
fn new_object<'a>(&'a mut self, metadata_builder: &'a mut MetadataBuilder) -> ObjectBuilder<'a> {
Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is the major change - move the low level appending to the ValueBuffer itself

Then all the recursion seems to fall into place

let parent_state = ParentState::Variant {
buffer: self,
metadata_builder,
};
let validate_unique_fields = false;
ObjectBuilder::new(parent_state, validate_unique_fields)
}

fn new_list<'a>(
&'a mut self,
metadata_builder: &'a mut MetadataBuilder,
) -> ListBuilder<'a> {
let parent_state = ParentState::Variant {
buffer: self,
metadata_builder,
};
let validate_unique_fields = false;
ListBuilder::new(parent_state, validate_unique_fields)
}

fn append_variant<'m, 'd>(&mut self, variant: Variant<'m, 'd>, metadata_builder: &mut MetadataBuilder) {
match variant {
Variant::Null => self.append_null(),
Variant::BooleanTrue => self.append_bool(true),
Expand All @@ -215,14 +235,23 @@ impl ValueBuffer {
Variant::Binary(v) => self.append_binary(v),
Variant::String(s) => self.append_string(s),
Variant::ShortString(s) => self.append_short_string(s),
_ => unreachable!("Objects and lists must be appended using VariantBuilder::append_object and VariantBuilder::append_list"),
Variant::Object(obj) => {
let mut object_builder = self.new_object(metadata_builder);
for (field_name, value) in obj.iter() {
object_builder.insert(field_name, value);
}
object_builder.finish().unwrap();
}
Variant::List(list) => {
let mut list_builder = self.new_list(metadata_builder);
for value in list.iter() {
list_builder.append_value(value);
}
list_builder.finish();
}
}
}

fn append_non_nested_value<'m, 'd, T: Into<Variant<'m, 'd>>>(&mut self, value: T) {
self.append_variant(value.into());
}

/// Writes out the header byte for a variant object or list
fn append_header(&mut self, header_byte: u8, is_large: bool, num_items: usize) {
let buf = self.inner_mut();
Expand Down Expand Up @@ -699,79 +728,6 @@ impl VariantBuilder {
ObjectBuilder::new(parent_state, validate_unique_fields)
}

/// Appends a [`VariantObject`] to the builder.
///
/// # Panics
/// Will panic if the appended object has duplicate field names or any nested validation fails.
/// Use `try_append_object` if you need full validation for untrusted data.
pub fn append_object<'m, 'v>(&mut self, object: VariantObject<'m, 'v>) {
let (parent_state, validate_unique_fields) = self.parent_state();

let mut obj_builder = ObjectBuilder::new(parent_state, validate_unique_fields);

for (field_name, variant) in object.iter() {
obj_builder.insert(field_name, variant);
}

obj_builder.finish().unwrap();
}

/// Appends a [`VariantObject`] to the builder with full validation during iteration.
///
/// Recursively validates all nested variants in the object during iteration.
pub fn try_append_object<'m, 'v>(
&mut self,
object: VariantObject<'m, 'v>,
) -> Result<(), ArrowError> {
let (parent_state, validate_unique_fields) = self.parent_state();

let mut obj_builder = ObjectBuilder::new(parent_state, validate_unique_fields);

for res in object.iter_try() {
let (field_name, variant) = res?;

obj_builder.insert(field_name, variant);
}

obj_builder.finish()?;

Ok(())
}

/// Appends a [`VariantList`] to the builder.
///
/// # Panics
/// Will panic if any nested validation fails during list iteration.
/// Use `try_append_list` if you need full validation for untrusted data.
pub fn append_list<'m, 'v>(&mut self, list: VariantList<'m, 'v>) {
let (parent_state, validate_unique_fields) = self.parent_state();

let mut list_builder = ListBuilder::new(parent_state, validate_unique_fields);

for variant in list.iter() {
list_builder.append_value(variant);
}

list_builder.finish();
}

/// Appends a [`VariantList`] to the builder with full validation during iteration.
///
/// Recursively validates all nested variants in the list during iteration.
pub fn try_append_list<'m, 'v>(&mut self, list: VariantList<'m, 'v>) -> Result<(), ArrowError> {
let (parent_state, validate_unique_fields) = self.parent_state();

let mut list_builder = ListBuilder::new(parent_state, validate_unique_fields);

for variant in list.iter_try() {
list_builder.append_value(variant?);
}

list_builder.finish();

Ok(())
}

/// Append a non-nested value to the builder.
///
/// # Example
Expand All @@ -783,12 +739,7 @@ impl VariantBuilder {
/// ```
pub fn append_value<'m, 'd, T: Into<Variant<'m, 'd>>>(&mut self, value: T) {
let variant = value.into();

match variant {
Variant::Object(obj) => self.append_object(obj),
Variant::List(list) => self.append_list(list),
primitive => self.buffer.append_variant(primitive),
}
self.buffer.append_variant(variant, &mut self.metadata_builder);
}

/// Finish the builder and return the metadata and value buffers.
Expand Down Expand Up @@ -855,7 +806,7 @@ impl<'a> ListBuilder<'a> {
/// Appends a new primitive value to this list
pub fn append_value<'m, 'd, T: Into<Variant<'m, 'd>>>(&mut self, value: T) {
self.offsets.push(self.buffer.offset());
self.buffer.append_non_nested_value(value);
self.buffer.append_variant(value.into(), self.parent_state.metadata_builder())
}

/// Finalizes this list and appends it to its parent, which otherwise remains unmodified.
Expand Down Expand Up @@ -927,7 +878,7 @@ impl<'a> ObjectBuilder<'a> {
self.duplicate_fields.insert(field_id);
}

self.buffer.append_non_nested_value(value);
self.buffer.append_variant(value.into(), metadata_builder)
}

/// Enables validation for unique field keys when inserting into this object.
Expand Down
Loading
pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy