Struct ArrowTable 

pub struct ArrowTable { /* private fields */ }

Arrow record batches shown as one table: rows run through the batches in order.

Formatting is chosen per column from its data type when a batch is added, so writing a cell is a lookup and a format into a stack buffer, without allocating. Batches can be appended while the table is shown (a stream); their arrays are shared, not copied.

Formatted types: booleans, integers, floats, Decimal128, strings (Utf8, LargeUtf8, Utf8View), dictionaries of those, dates, times, timestamps (with a time zone: in UTC, marked Z) and durations. Other types (lists, structs, maps, binary, intervals) show ?.

A batch is accepted when its columns have the table’s data types in order; field names, nullability and metadata may differ. A timestamp’s time zone is part of its data type.

Implementations

impl ArrowTable

pub fn new(schema: SchemaRef) -> Self

An empty table for batches with schema; columns are sized by their titles until the first rows arrive.

pub fn from_batches( batches: impl IntoIterator<Item = RecordBatch>, ) -> Result<Self, TableError>

A table of batches, with the first batch’s schema.

Errors

TableError::NoBatches for none; a column error if a batch’s schema differs.

pub fn push(&mut self, batch: RecordBatch) -> Result<(), TableError>

Appends batch’s rows. The first rows to arrive size the columns (their titles and up to the first 1,000 values, at most 40 characters); later rows widen a number or time column whose values no longer fit (measuring some rows of each batch, the last among them), never narrowing it. Columns sized with set_column_chars keep their width.

Errors

A column error if batch has other columns or data types; the table is unchanged.

pub fn schema(&self) -> &SchemaRef

The table’s schema.

pub fn rows(&self) -> usize

Rows in all batches.

pub fn columns(&self) -> Vec<Column>

Grid columns: titled by field name, sized by content, numbers and times aligned right.

pub fn set_decimals( &mut self, column: usize, decimals: Option<usize>, ) -> Result<(), TableError>

Prints floating-point values of column with decimals fixed decimal places (None: the shortest exact form). A column sized by the first rows is sized again from them, as now printed.

Errors

TableError::NoSuchColumn if there is no such column.

pub fn set_column_chars( &mut self, column: usize, chars: usize, ) -> Result<(), TableError>

Sets column to chars characters wide (1 to CELL_BYTES, the most a cell can hold): columns reports it, longer text is cut to fit ending in …, and the first rows to arrive no longer size it.

Errors

TableError::NoSuchColumn if there is no such column.

pub fn write_cell( &self, row: usize, column: usize, text: &mut CellText, ) -> CellState

Appends the value at row of column to text, or a placeholder (see CellState), never a partial value. Never allocates.

pub fn view<'a>(&'a self, first_row: usize, palette: &Palette) -> ArrowGrid<'a>

A GridSource showing the table from first_row (scrolled), in palette’s text color, with nulls and unsupported values dimmed.

Trait Implementations

impl Clone for ArrowTable

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ArrowTable

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

Formats the value using the given formatter.