exonware-xwjson 0.9.0.9

Extended binary JSON format - single version of truth for all format conversions

xwjson

Binary JSON plus local data-engine behavior for real workloads. MessagePack-based .xwjson is built for high-throughput reads/writes, lazy and path-scoped access, schema-aware validation, encrypted payload support, and atomic-first ACID-style transaction flows with WAL. In this repo, performance/stress suites include large-file scenarios validated up to 20GB, stdlib JSON comparisons, and conversion stress runs.

Longer tour, examples, and troubleshooting: README_LONG.md.

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

---

Install

pip install exonware-xwjson
# Full (optional)
pip install exonware-xwjson[full]

---

Quick start

from exonware.xwjson import XWJSONSerializer

serializer = XWJSONSerializer()
# Async (recommended)
data = await serializer.load_file_async("data.xwjson")
await serializer.save_file_async(data, "output.xwjson")
# Sync
data = serializer.load_file("data.xwjson")
serializer.save_file(data, "output.xwjson")

See docs/ for REF_*, guides, and examples.

---

What you get

Area	What's in it
Binary engine	MessagePack-based encoding with fast serialization paths and metadata-aware payloads.
Lazy + path I/O	Parse on access, stream/paging support, and path-level operations without full materialization.
References	$ref, @href, *anchor preserved and resolved.
Security	Password-based encrypted file support and encrypted payload detection helpers.
Atomic operations	Atomic read/write/update/delete paths for safe state transitions under concurrent workloads.
Transactions	ACID-style flows with WAL-backed transaction components, rollback support, and dependency-aware batch work.
Index/cache acceleration	xwnode strategy integrations for faster lookups and cache-assisted repeated access.
Integration	xwnode, xwschema, xwquery, xwdata, xwstorage; format metadata carried through.

---

Atomic operations - core strength

• xwjson is designed around atomic state changes, not best-effort updates.
• Single-file and path-level operations are built to commit safely or fail safely.
• Transaction flows combine atomic steps with WAL-backed durability mechanics.
• For local data-engine workloads, this is a headline capability: atomic operations are first-class, not optional.

---

Performance and benchmark posture

• Built-in performance guidance and benchmark methodology: docs/_archive/GUIDE_PERFORMANCE.md.
• Repo test artifacts include conversion/performance stress suites and large-file scenarios in tests/2.integration/scenarios/.
• File size note: no hard file-size cap is enforced by xwjson; largest documented test scenario in this repo is 20GB.
• Test summaries in tests/FINAL_TEST_SUMMARY.md and tests/TEST_SUMMARY.md document speed-oriented targets (including stdlib JSON comparison paths and cache/index speedups).
• Positioning: xwjson can act like a database-like local data layer for many workloads; external database claims should stay workload-specific and benchmark-backed.

---

Docs and tests

• Start: docs/INDEX.md or docs/.
• Guides: Basic/advanced usage, performance, conversion, schema validation under docs/ when present.
• Tests: From repo root, follow this package's runner or pytest layout.

---

License and links

MIT - see LICENSE. Homepage: https://exonware.com · Repository: https://github.com/exonware/xwjson
Version: 0.9.0.9 | Updated: 31-Mar-2026

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