Examples

A dynamic document

examples/encode_value.mojo builds nothing by hand. It decodes a short document and encodes it again.

from toml import decode_toml, encode_toml

def main() raises:
    var doc = decode_toml(String("name = \"Tom\"\nage = 42\n"))
    print(encode_toml(doc), end="")

The printed text is a TOML document. Key order follows the order the keys were defined. Nested tables become [header] blocks after the scalar keys of their parent.

Inline tables and compact arrays

from runtime.options import EncodeOptions
from toml import decode_toml, encode_toml

def main() raises:
    var doc = decode_toml(String("[point]\nx = 1\ny = 2\n"))
    print(encode_toml(doc, EncodeOptions(True, False)))

EncodeOptions(True, False) writes tables as inline tables. EncodeOptions(False, True) writes an array of tables as one array of inline tables. Scalar arrays are one line in every mode.

A generated struct

testdata/schema/benchmark_v2.json is the schema for Message. After pixi run generate:

from Message import Message
from runtime.datum import decode_text, encode_text

def main() raises:
    var m = Message()
    m.f_bool = True
    m.f_int32 = Int64(42)
    m.f_string = String("hi")
    var back = decode_text[Message](encode_text(m))
    print(back.f_int32)

Document in the same schema has a nested meta table and an items array of tables. The encoder writes those as [meta] and [[items]].