Struct CsvReader 

pub struct CsvReader { /* private fields */ }

Reads CSV or TSV text into Rows. See the module docs.

Implementations

impl CsvReader

pub fn new() -> Self

Sniff the dialect from DELIMITERS and the header, failing when the sniffer finds no delimiter (set a fallback to read such input anyway).

pub fn tsv() -> Self

Tab-separated, with excel’s quoting (csv.get_dialect("excel-tab")).

pub fn delimiter(self, delimiter: char) -> Self

Use this delimiter, with excel’s quoting, instead of sniffing.

pub fn dialect(self, dialect: Dialect) -> Self

Use this dialect instead of sniffing.

pub fn fallback(self, delimiter: char) -> Self

The delimiter to fall back to when the sniffer finds none, as rich --csv falls back to , for a .csv file and a tab for a .tsv one.

pub fn header(self, header: bool) -> Self

Whether the first row is a header (true/false), instead of sniffing it.

pub fn header_mode(self, header: Header) -> Self

Set the header decision, Header::Sniff included.

pub fn detect(&self, content: &str) -> Option<(Dialect, bool)>

The dialect and header decision for content, as read makes them; None when the sniffer finds no delimiter and there is no fallback.

pub fn read(&self, content: &str) -> Result<Rows, DataError>

Read content. Blank lines are skipped; a row longer than the header adds columns named by position, and a shorter one is padded with nulls, so no field is lost.

impl CsvReader

pub fn source<R: BufRead>(&self, reader: R) -> Result<CsvSource<R>, DataError>

Stream rows from reader instead of reading the text whole, so an input larger than memory can be read (a profile samples it).

The dialect and header are decided from the first SAMPLE_CHARS characters, as read decides them, and records are parsed one at a time: a quoted field may span lines, up to MAX_RECORD bytes. The columns are decided from the rows the sniffer read, as read decides them from every row (a row longer than the header adds columns named by position); a later row is padded with nulls or cut to fit, and longer_rows counts the cut ones. Bytes that are not UTF-8 are replaced.

use rich_data::csv::CsvReader;
use rich_data::{RowSource, Value};

let text = "name,note\nweb,\"two\nlines\"\napi,ok\n";
let mut source = CsvReader::new().source(text.as_bytes()).unwrap();
assert_eq!(source.columns(), ["name", "note"]);
let first = source.next_row().unwrap().unwrap();
assert_eq!(first[1], Value::from("two\nlines"));
source.next_row().unwrap().unwrap();
assert!(source.next_row().is_none());

Trait Implementations

impl Clone for CsvReader

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for CsvReader

impl Debug for CsvReader

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

Formats the value using the given formatter.

impl Default for CsvReader

fn default() -> Self

Returns the “default value” for a type.