exonware-xwformats 0.9.0.27

Heavyweight serialization formats for xwsystem (schema, scientific, database, binary, RON). CSV/JSON/TOML/YAML/XML, BSON, and other core xwsystem codecs are not registered by this package

xwformats

Many serialization formats, one package. Schema (Protobuf, Parquet, Thrift, …), scientific (HDF5, Feather, Zarr, …), database (LMDB, LevelDB, …), plus binary and text. Same converter style as xwsystem; formats register on import. Pick lite, lazy, full, or dev extras to match how you deploy.

Company: eXonware.com · Author: eXonware Backend Team · Email: [e-mail]

---

Install

Install	What you get	When to use
pip install exonware-xwformats	Base — exonware-xwsystem[full] (shared text/JSON stack, etc.) + lazy hooks	Heavier than core-only xwsystem; add [full] for extra heavyweight codecs.
pip install exonware-xwformats[lazy]	Lazy - xwsystem[lazy] + lazy extras	Matches lazy-optional stack in docs.
pip install exonware-xwformats[full]	Full - common format deps installed up front	Production or CI; on Windows, LevelDB/RocksDB wheels may be missing (see docs/REF_54_BENCH.md).
pip install exonware-xwformats[dev]	Dev - pytest, black, mypy, …	Contributors

Base install pulls exonware-xwsystem[full] (CSV, JSON, TOML, YAML, XML, and other xwsystem optionals) and enables xwlazy when available. [lazy], [dev], and [full] are extras on top.

---

Quick start

from exonware.xwformats import XWFormats

xf = XWFormats()
print(xf.list_formats())                       # All registered formats
data = xf.convert(json_bytes, "json", "yaml")  # Bidirectional conversion
ser = xf.get_serializer("parquet")
ser.encode(value)
ser.decode(data)

Importing xwformats registers its formats with xwsystem; use xwsystem's codec/facade for detection by file extension. See REF_14_DX and REF_15_API.

---

What you get

Area	What's in it
Schema	Protobuf, Parquet, Thrift, ORC, Cap'n Proto, FlatBuffers, Arrow. Avro is optional (may be off default import on some platforms - see __init__.py and docs/_archive/KNOWN_ISSUES.md).
Scientific	HDF5, Feather, Zarr, NetCDF, MAT.
Database	LMDB, LevelDB, RocksDB, GraphDB.
Binary / text	UBJSON, Bincode, Dill, Postcard, and RON are implemented in xwformats. BSON, CSV, JSON, TOML, YAML, XML, and other core xwsystem binaries are xwsystem-only — not registered or exported by xwformats.
Integration	Same converter/facade as xwsystem; codec registry updated on import.

Lite = few deps. Lazy = optional format wheels on first use. Full = common optionals pre-installed. Platform notes (e.g. RocksDB on Windows): docs/logs/setup/. Known issues: docs/_archive/KNOWN_ISSUES.md and REF_22_PROJECT.md.

---

Ecosystem functional contributions

xwformats extends format coverage; sibling XW libs provide runtime loading strategy, registration, and downstream consumption layers. You can use xwformats standalone for broad serializer coverage and conversion tasks. Integrating with additional XW libraries is optional and mainly useful for enterprise and mission-critical pipelines that need unified runtime, storage, and query infrastructure.

Supporting XW lib	What it provides to xwformats	Functional requirement it satisfies
XWSystem	Core codec registry/facade and serializer contracts that xwformats registers into.	One unified serialization API despite many optional format backends.
XWLazy	On-demand dependency loading/install behavior for optional format stacks.	Practical deployment of many formats without forcing heavy installs.
XWData	Higher-level data transformation workflows that consume xwformats codecs.	Real pipeline usage beyond encode/decode-only primitives.
XWStorage	Persistence engine integrations that use specialized format serializers.	Storage interoperability for scientific/schema/database-oriented payloads.
XWJSON	Binary JSON and local data-engine flows that can interoperate with format conversion paths.	Efficient local format conversion and transition workflows.
XWQuery	Query pipelines over data loaded via format serializers.	End-to-end analytics/transform use cases after format normalization.

Competitive edge: instead of isolated serializer plugins, xwformats plugs into a shared runtime and data stack so advanced formats are immediately usable in storage/query/application workflows.

---

Docs and tests

Aligned with project REFs and docs/GUIDE_01_USAGE.md (see GUIDE_63_README in the monorepo).

• Start: docs/INDEX.md - index and quick links.
• Use: docs/GUIDE_01_USAGE.md - usage and format list.
• Requirements / status: docs/REF_01_REQ.md, docs/REF_22_PROJECT.md.
• API / design: docs/REF_15_API.md, docs/REF_13_ARCH.md, docs/REF_14_DX.md.
• Tests: docs/REF_51_TEST.md. Summaries: docs/logs/tests/. Run: python tests/runner.py or pytest from repo root.

---

License and links

Apache-2.0 - see LICENSE.

• Homepage: https://exonware.com
• Repository: https://github.com/exonware/xwformats

Async Support

• xwformats is primarily synchronous in its current implementation.
• Source validation: 0 async def definitions and 0 await usages under src/.
• This module still composes with async-capable xw libraries at integration boundaries when needed.

Version: 0.9.0.27 | Updated: 08-Apr-2026

Built with ❤️ by eXonware.com - Revolutionizing Python Development Since 2025