Struct arrow::datatypes::Schema

source ·

pub struct Schema {
    pub fields: Fields,
    pub metadata: HashMap<String, String>,
}

Expand description

Describes the meta-data of an ordered sequence of relative types.

Note that this information is only part of the meta-data and not part of the physical memory layout.

Fields§

§fields: Fields

A sequence of fields that describe the schema.

§metadata: HashMap<String, String>

A map of key-value pairs containing additional meta data.

Implementations§

source §

impl Schema

source

pub fn empty() -> Schema

Creates an empty Schema

source

pub fn new(fields: impl Into<Fields>) -> Schema

Creates a new Schema from a sequence of Field values.

§Example

let field_a = Field::new("a", DataType::Int64, false);
let field_b = Field::new("b", DataType::Boolean, false);

let schema = Schema::new(vec![field_a, field_b]);

source

pub fn new_with_metadata( fields: impl Into<Fields>, metadata: HashMap<String, String>, ) -> Schema

Creates a new Schema from a sequence of Field values and adds additional metadata in form of key value pairs.

§Example


let field_a = Field::new("a", DataType::Int64, false);
let field_b = Field::new("b", DataType::Boolean, false);

let mut metadata: HashMap<String, String> = HashMap::new();
metadata.insert("row_count".to_string(), "100".to_string());

let schema = Schema::new_with_metadata(vec![field_a, field_b], metadata);

source

pub fn with_metadata(self, metadata: HashMap<String, String>) -> Schema

Sets the metadata of this Schema to be metadata and returns self

source

pub fn project(&self, indices: &[usize]) -> Result<Schema, ArrowError>

Returns a new schema with only the specified columns in the new schema This carries metadata from the parent schema over as well

source

pub fn try_merge( schemas: impl IntoIterator<Item = Schema>, ) -> Result<Schema, ArrowError>

Merge schema into self if it is compatible. Struct fields will be merged recursively.

Example:


let merged = Schema::try_merge(vec![
    Schema::new(vec![
        Field::new("c1", DataType::Int64, false),
        Field::new("c2", DataType::Utf8, false),
    ]),
    Schema::new(vec![
        Field::new("c1", DataType::Int64, true),
        Field::new("c2", DataType::Utf8, false),
        Field::new("c3", DataType::Utf8, false),
    ]),
]).unwrap();

assert_eq!(
    merged,
    Schema::new(vec![
        Field::new("c1", DataType::Int64, true),
        Field::new("c2", DataType::Utf8, false),
        Field::new("c3", DataType::Utf8, false),
    ]),
);

source

pub const fn fields(&self) -> &Fields

Returns an immutable reference of the vector of Field instances.

source

pub fn flattened_fields(&self) -> Vec<&Field>

Returns a vector with references to all fields (including nested fields)

§Example

use std::sync::Arc;
use arrow_schema::{DataType, Field, Fields, Schema};

let f1 = Arc::new(Field::new("a", DataType::Boolean, false));

let f2_inner = Arc::new(Field::new("b_inner", DataType::Int8, false));
let f2 = Arc::new(Field::new("b", DataType::List(f2_inner.clone()), false));

let f3_inner1 = Arc::new(Field::new("c_inner1", DataType::Int8, false));
let f3_inner2 = Arc::new(Field::new("c_inner2", DataType::Int8, false));
let f3 = Arc::new(Field::new(
    "c",
    DataType::Struct(vec![f3_inner1.clone(), f3_inner2.clone()].into()),
    false
));

let mut schema = Schema::new(vec![
  f1.clone(), f2.clone(), f3.clone()
]);
assert_eq!(
    schema.flattened_fields(),
    vec![
        f1.as_ref(),
        f2.as_ref(),
        f2_inner.as_ref(),
        f3.as_ref(),
        f3_inner1.as_ref(),
        f3_inner2.as_ref()
   ]
);