odsslicer

Python reader and writer for .ods files (OpenDocument Spreadsheet — LibreOffice / OpenOffice Calc), with a numpy-inspired indexing API:

from odsslicer import ODSReader

table = ODSReader("workbook.ods")
sheet = table.sheet("Sheet1")

sheet["A1"].value          # typed value: str / float / bool / date / time / None
sheet["A1:B3"].to_numpy()  # any block as a numpy array
sheet[:, 0]                 # entire column A

sheet["C1"].formula = "SUM(A1:A10)"
sheet["C1"].style.bold = True
table.save("out.ods")                     # add recalculate=True to have LibreOffice compute formulas/pivots

odsslicer works directly on the ODF XML (via BeautifulSoup/lxml), so it preserves what other tools tend to drop — cell formats (currency, percentage, date, time), formulas, merged and repeated cells, styles, comments — and writes them back faithfully. It has no calculation engine of its own: like ODF itself, it describes what to compute — and can hand the file to a local LibreOffice (save(..., recalculate=True)) to compute formulas and pivot tables for you.

Full documentation with an example for every feature: DOCS.md.

Installation

pip install odsslicer

Requires Python ≥ 3.10. Dependencies (beautifulsoup4, lxml, numpy) are installed automatically. For local development:

git clone https://github.com/antnardo/odsslicer.git
cd odsslicer
pip install -e ".[test]"

What it does

Each line links to the detailed section (with examples) in DOCS.md.

See Known limitations for what's deliberately out of scope (no calculation engine, no charts/images, no partial rich text…).

Are there already equivalent PyPI modules?

(Versions and dates as of August 2026.)

Where odsslicer sits: a spreadsheet-shaped API (sheet["A1:B3"], numpy arrays, fill handles, copy/sort/merge) rather than a generic ODF DOM, with read and write of the things data-oriented tools usually lose — formats, formulas (in ordinary syntax), styles, merged cells, comments, pivot definitions — and every write verified against a real LibreOffice. If you only need to read values fast, use python-calamine; if you need to manipulate arbitrary ODF documents (text, presentations) at the XML level, use odfdo.

Tests

pip install -e ".[test]"
pytest

The main suite (tests/test_odsslicer.py) covers every feature above plus regression tests for the bugs fixed along the way; it runs on every push/PR via GitHub Actions across Python 3.10 to 3.13.

tests/test_wild_files.py confronts the API with real-world files written by other generators — Excel 16, a 2012-era LibreOffice 3.5, recent LibreOffice on Linux and Windows, a Google Sheets export (see tests/wild/README.md for provenance) — reading, writing and round-tripping each one.

tests/test_libreoffice_consistency.py is an opt-in suite that hands files odsslicer wrote to a real, local LibreOffice (soffice --headless --convert-to fods) and inspects what LibreOffice itself made of them — the strongest available signal that a write is genuinely valid ODF, not just something our own reader happens to parse back. It skips automatically if no soffice/libreoffice binary is on PATH.

Versions

Version numbers are derived automatically from git tags (via setuptools-scm) and follow Semantic Versioning — while the major version stays 0, the API can still change between minor versions. See CHANGELOG.md (and the Releases page) for what changed in each version.

License

MIT — reuse with essentially no restriction, just keep the copyright notice.

Project name

odsslicer, to reflect the module's real differentiator — numpy-style indexing/slicing by cell address — rather than a generic "ods reader".