Struct polars::prelude::JsonReader

pub struct JsonReader<'a, R>
where
    R: MmapBytesReader,
{ /* private fields */ }

Available on crate feature polars-io only.

Reads JSON in one of the formats in JsonFormat into a DataFrame.

Implementations

impl<'a, R> JsonReader<'a, R>

where R: MmapBytesReader,

pub fn with_schema(self, schema: Arc<Schema>) -> JsonReader<'a, R>

Set the JSON file’s schema

pub fn with_schema_overwrite(self, schema: &'a Schema) -> JsonReader<'a, R>

Overwrite parts of the inferred schema.

pub fn infer_schema_len(self, max_records: Option<usize>) -> JsonReader<'a, R>

Set the JSON reader to infer the schema of the file. Currently, this is only used when reading from JsonFormat::JsonLines, as JsonFormat::Json reads in the entire array anyway.

When using JsonFormat::JsonLines, max_records = None will read the entire buffer in order to infer the schema, Some(1) would look only at the first record, Some(2) the first two records, etc.

It is an error to pass max_records = Some(0), as a schema cannot be inferred from 0 records when deserializing from JSON (unlike CSVs, there is no header row to inspect for column names).

pub fn with_batch_size(self, batch_size: NonZero<usize>) -> JsonReader<'a, R>

Set the batch size (number of records to load at one time)

This heavily influences loading time.

pub fn with_projection( self, projection: Option<Vec<String>> ) -> JsonReader<'a, R>

Set the reader’s column projection: the names of the columns to keep after deserialization. If None, all columns are kept.

Setting projection to the columns you want to keep is more efficient than deserializing all of the columns and then dropping the ones you don’t want.

pub fn with_json_format(self, format: JsonFormat) -> JsonReader<'a, R>

pub fn with_ignore_errors(self, ignore: bool) -> JsonReader<'a, R>

Return a null if an error occurs during parsing.

Trait Implementations

impl<'a, R> SerReader<R> for JsonReader<'a, R>

where R: MmapBytesReader,

fn finish(self) -> Result<DataFrame, PolarsError>

Take the SerReader and return a parsed DataFrame.

Because JSON values specify their types (number, string, etc), no upcasting or conversion is performed between incompatible types in the input. In the event that a column contains mixed dtypes, is it unspecified whether an error is returned or whether elements of incompatible dtypes are replaced with null.

fn new(reader: R) -> JsonReader<'a, R>

Create a new instance of the [SerReader]

fn set_rechunk(self, rechunk: bool) -> JsonReader<'a, R>

Make sure that all columns are contiguous in memory by aggregating the chunks into a single array.