feat: add Value::Json value type (#7083)

* feat: struct value

Signed-off-by: Ning Sun <sunning@greptime.com>

* feat: update for proto module

* feat: wip struct type

* feat: implement more vector operations

* feat: make datatype and api

* feat: reoslve some compilation issues

* feat: resolve all compilation issues

* chore: format update

* test: resolve tests

* test: test and refactor value-to-pb

* feat: add more tests and fix for value types

* chore: remove dbg

* feat: test and fix iterator

* fix: resolve struct_type issue

* feat: pgwire 0.33 update

* refactor: use vec for struct items

* feat: conversion from json to value

* feat: add decode function

* fix: lint issue

* feat: update how we encode raw data

* feat: add convertion to fully strcutured StructValue

* refactor: take owned value in all encode/decode functions

* feat: add pg serialization of structvalue

* chore: toml format

* refactor: adopt new and try_new from struct value

* chore: cleanup residual issues

* docs: docs up

* fix lint issue

* Apply suggestion from @MichaelScofield

Co-authored-by: LFC <990479+MichaelScofield@users.noreply.github.com>

* Apply suggestion from @MichaelScofield

Co-authored-by: LFC <990479+MichaelScofield@users.noreply.github.com>

* Apply suggestion from @MichaelScofield

Co-authored-by: LFC <990479+MichaelScofield@users.noreply.github.com>

* Apply suggestion from @MichaelScofield

Co-authored-by: LFC <990479+MichaelScofield@users.noreply.github.com>

* chore: address review comment especially collection capacity

* refactor: remove unneeded processed keys collection

* feat: Value::Json type

* chore: add some work in progress changes

* feat: adopt new json type

* refactor: limit scope json conversion functions

* fix: self review update

* test: provide tests for value::json

* test: add tests for api/helper

* switch proto to main branch

* fix: implement is_null for ValueRef::Json

---------

Signed-off-by: Ning Sun <sunning@greptime.com>
Co-authored-by: LFC <990479+MichaelScofield@users.noreply.github.com>
This commit is contained in:
Ning Sun
2025-10-16 04:13:12 +08:00
committed by GitHub
parent 8073e552df
commit 145c1024d1
16 changed files with 590 additions and 150 deletions

View File

@@ -33,7 +33,7 @@ use datatypes::arrow::util::pretty;
use datatypes::prelude::{ConcreteDataType, VectorRef};
use datatypes::scalars::{ScalarVector, ScalarVectorBuilder};
use datatypes::schema::{ColumnSchema, Schema, SchemaRef};
use datatypes::types::json_type_value_to_string;
use datatypes::types::{JsonFormat, jsonb_to_string};
use datatypes::vectors::{BinaryVector, StringVectorBuilder};
use error::Result;
use futures::task::{Context, Poll};
@@ -90,32 +90,34 @@ pub fn map_json_type_to_string(
) -> Result<RecordBatch> {
let mut vectors = Vec::with_capacity(original_schema.column_schemas().len());
for (vector, schema) in batch.columns.iter().zip(original_schema.column_schemas()) {
if let ConcreteDataType::Json(j) = schema.data_type {
let mut string_vector_builder = StringVectorBuilder::with_capacity(vector.len());
let binary_vector = vector
.as_any()
.downcast_ref::<BinaryVector>()
.with_context(|| error::DowncastVectorSnafu {
from_type: schema.data_type.clone(),
to_type: ConcreteDataType::binary_datatype(),
})?;
for value in binary_vector.iter_data() {
let Some(value) = value else {
string_vector_builder.push(None);
continue;
};
let string_value =
json_type_value_to_string(value, &j.format).with_context(|_| {
error::CastVectorSnafu {
if let ConcreteDataType::Json(j) = &schema.data_type {
if matches!(&j.format, JsonFormat::Jsonb) {
let mut string_vector_builder = StringVectorBuilder::with_capacity(vector.len());
let binary_vector = vector
.as_any()
.downcast_ref::<BinaryVector>()
.with_context(|| error::DowncastVectorSnafu {
from_type: schema.data_type.clone(),
to_type: ConcreteDataType::binary_datatype(),
})?;
for value in binary_vector.iter_data() {
let Some(value) = value else {
string_vector_builder.push(None);
continue;
};
let string_value =
jsonb_to_string(value).with_context(|_| error::CastVectorSnafu {
from_type: schema.data_type.clone(),
to_type: ConcreteDataType::string_datatype(),
}
})?;
string_vector_builder.push(Some(string_value.as_str()));
}
})?;
string_vector_builder.push(Some(string_value.as_str()));
}
let string_vector = string_vector_builder.finish();
vectors.push(Arc::new(string_vector) as VectorRef);
let string_vector = string_vector_builder.finish();
vectors.push(Arc::new(string_vector) as VectorRef);
} else {
vectors.push(vector.clone());
}
} else {
vectors.push(vector.clone());
}