Skip to content

Changelog

Notable changes to rn-forge-data, newest first, in the Keep a Changelog format.

[Unreleased]

[0.1.6] - 2026-11-20

Changed

  • Requires tablib[xlsx]>=3.10.0.

[0.1.5] - 2026-11-19

Changed

  • write_xlsx writes large tables faster, and its files declare their dimension, so they load faster.

[0.1.4] - 2026-11-17

Changed

  • The rn-forge-commons requirement admits 0.10 (<0.11).
  • write_xlsx removes a redundant row list to reduce internal memory use; output is unchanged. The usage guide clarifies that the caller's whole dataset and the returned workbook bytes remain in memory.

Fixed

  • PandasUtils.get_string_field preserves present 0, 0.0 and False as "0", "0.0" and "False". Missing values still follow required; get_decimal_field now raises decimal.InvalidOperation for a False cell. Callers relying on empty strings for zero/False or decimal zero for False must adjust to these corrected results.

[0.1.3] - 2026-11-13

Changed

  • The rn-forge-commons requirement admits 0.9 (>=0.8.0,<0.10).

[0.1.2] - 2026-10-29

Changed

  • The package declares the MIT licence and is published to PyPI.

[0.1.1] - 2026-10-25

Changed

  • Requires its rn-forge dependencies by version range from pykit's index (ADR-0004), not by git URL.

[0.1.0] - 2026-10-16

Added

  • The rn-forge-data package: Excel and tabular data helpers. The core, rn_forge.data, holds ExcelUtils, CellFormat, WorkbookTemplate and write_xlsx, and needs only openpyxl and tablib[xlsx]. The pandas adapter, rn_forge.data.pandas, behind the pandas extra, holds PandasUtils, ExcelAdapter, SeriesRow, series_value and series_columns. These moved from rn-forge-commons.