Standard library

nox.csv

Read and write CSV text following RFC 4180: comma-separated fields, optional double-quoted fields that may contain commas, quotes and line breaks, and "" as an escaped quote.

Text

import nox.csv
from nox.csv import CsvError

Capability: none. The module works on text; read files with nox.fs.

Functions#

Function	Description
`parse(text)`	parses the whole text into rows of fields (`list[list[str]]`); raises `CsvError` for an unterminated quoted field
`parse_dicts(text)`	treats the first row as the header and returns one `dict[str, str]` per following row
`dump_row(fields)`	formats one row, quoting fields that need it (no trailing newline)
`dump(rows)`	formats all rows, one per line, each followed by `\n`

Input may use \n or \r\n line endings; output always uses \n. All values are strings — convert with int(...) / float(...) yourself.

Nox

import nox.csv
from nox.csv import CsvError

rows: list[list[str]] = nox.csv.parse("a,b,c\n1,\"x,y\",3\n")
print(rows, len(rows))
dicts: list[dict[str, str]] = nox.csv.parse_dicts("name,age\nada,36\nalan,41\n")
print(dicts[1]["name"], len(dicts))
print(nox.csv.dump_row(["a", "b,c", "d\"e"]))
print(nox.csv.dump([["x", "y"], ["1", "2"]]), end="")
try:
    nox.csv.parse("a,\"unterminated")
except CsvError as e:
    print("bad csv")

Output

[['a', 'b', 'c'], ['1', 'x,y', '3']] 2
alan 2
a,"b,c","d""e"
x,y
1,2
bad csv

Out of scope#

Custom delimiters (;, tab), a byte-order mark, and reading files directly (pass the text from nox.fs.read_to_string).

The names write and write_row were removed in 2.0; use dump and dump_row (Migrating).

Edit this page on GitHub