Struct DataFile 

pub struct DataFile {

    pub file_path: String,
    pub file_format: FileFormat,
    pub content: FileContent,
    pub partition: HashMap<String, Value>,
    pub record_count: i64,
    pub file_size_in_bytes: i64,
    pub column_sizes: HashMap<i32, i64>,
    pub value_counts: HashMap<i32, i64>,
    pub null_value_counts: HashMap<i32, i64>,
    pub nan_value_counts: HashMap<i32, i64>,
    pub lower_bounds: HashMap<i32, Vec<u8>>,
    pub upper_bounds: HashMap<i32, Vec<u8>>,
    pub statistics: HashMap<i32, ColumnStats>,
    pub sort_order_id: Option<i32>,
    pub key_metadata: Option<Vec<u8>>,
    pub split_offsets: Option<Vec<i64>>,
    pub equality_ids: Option<Vec<i32>>,
}

Metadata about a data or delete file.

Fields

file_path: String

Path to the file.

file_format: FileFormat

The file format.

content: FileContent

The type of content in this file.

partition: HashMap<String, Value>

The partition data for this file.

record_count: i64

Number of records in this file.

file_size_in_bytes: i64

File size in bytes.

column_sizes: HashMap<i32, i64>

Column sizes (column_id -> size in bytes).

value_counts: HashMap<i32, i64>

Value counts (column_id -> non-null value count).

null_value_counts: HashMap<i32, i64>

Null value counts (column_id -> null value count).

nan_value_counts: HashMap<i32, i64>

NaN value counts (column_id -> NaN count).

lower_bounds: HashMap<i32, Vec<u8>>

Lower bounds (column_id -> serialized lower bound).

upper_bounds: HashMap<i32, Vec<u8>>

Upper bounds (column_id -> serialized upper bound).

statistics: HashMap<i32, ColumnStats>

Column-level statistics.

sort_order_id: Option<i32>

Sort order ID, if the file is sorted.

key_metadata: Option<Vec<u8>>

Key metadata for encryption.

split_offsets: Option<Vec<i64>>

Split offsets for the file.

equality_ids: Option<Vec<i32>>

Field IDs used for equality comparison in equality delete files. This is required for content=EqualityDeletes and should be null otherwise.

Implementations

impl DataFile

pub fn new( file_path: impl Into<String>, file_format: FileFormat, record_count: i64, file_size_in_bytes: i64, ) -> Self

Creates a new data file with minimal required fields.

pub fn with_partition(self, partition: HashMap<String, Value>) -> Self

Sets the partition data for this file.

pub fn with_column_stats( self, column_id: i32, size: i64, value_count: i64, null_count: i64, ) -> Self

Adds column statistics.

pub fn with_content(self, content: FileContent) -> Self

Sets the content type (e.g., Data, PositionDeletes, EqualityDeletes).

pub fn with_equality_ids(self, ids: Vec<i32>) -> Self

Sets the equality IDs for equality delete files.

pub fn with_key_metadata(self, metadata: Vec<u8>) -> Self

Sets the key metadata.

pub fn with_split_offsets(self, offsets: Vec<i64>) -> Self

Sets the split offsets.

Trait Implementations

impl Clone for DataFile

fn clone(&self) -> DataFile

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for DataFile

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for DataFile

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for DataFile

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.