Skip to content

CHANGELOG.md

Latest commit

 

History

History
628 lines (531 loc) · 28.6 KB

CHANGELOG.md

File metadata and controls

628 lines (531 loc) · 28.6 KB
Raw

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

Added

  • Added gRPC enhanced ACK parity for commit ACK codes CA, CE, and CR alongside the existing AA, AE, and AR application ACK codes.
  • Added shared dirty real-world corpus fixture categories and CLI parity coverage for corpus summary, fingerprint, and diff.
  • Added REST and gRPC server dirty-corpus parity coverage using the shared dirty real-world fixture categories.
  • Added REST profile lint, explain, and inline fixture-test endpoints that emit the shared profile evidence report shapes with opt-in v2 provenance.
  • Added local Python wheel dirty-corpus parity coverage using the shared dirty real-world fixture categories.
  • Added xtask check-dirty-corpus-parity to run the shared dirty-corpus acceptance proof across Rust core, CLI, REST, and gRPC surfaces.
  • Added a CLI dirty real-world validate/redact/bundle/replay workflow to the shared dirty-corpus parity runner.
  • Added a REST dirty real-world validate-redacted/bundle/replay workflow to the shared dirty-corpus parity runner.
  • Added a gRPC dirty real-world validate-redacted/bundle/replay workflow to the shared dirty-corpus parity runner.
  • Added xtask check-bundle-replay-parity to run the shared bundle/replay acceptance proof across Rust core, CLI, REST, and gRPC surfaces.
  • Added xtask check-evidence-parity-acceptance as the aggregate local Rust/CLI/REST/gRPC parity acceptance suite, with optional local Python wheel smoke.
  • Added xtask python-local-wheel-proof to build, install, import, and smoke-test the local Python hl7v2 wheel without claiming TestPyPI or PyPI availability.
  • Added xtask python-public-registry-proof to install hl7v2 from TestPyPI or PyPI in a scratch venv and run the same Python smoke/evidence scripts after public registry upload succeeds.
  • Added xtask check-deployment-provenance to reject floating deployment image examples and keep hl7v2 image tags aligned with the workspace version.
  • Added a checked deployment digest example/template under the deployment provenance rail for future production image receipts.
  • Aligned legacy ADR toolchain and package-surface examples with the current Rust 1.95 support boundary and post-demicrocrating product graph.
  • Routed TestPyPI/PyPI install-back workflow jobs through xtask python-public-registry-proof so hosted public-registry proof and local reproduction use the same command boundary.
  • Added local Python dirty real-world validate/redact/bundle/replay workflow proof using the shared Z-segment fixture.
  • Added the dirty real-world Python evidence workflow smoke to the hosted Python Wheels workflow and policy guard.
  • Added a shared schema-version parity fixture and manifest guard for representative CLI, REST, gRPC, and local Python evidence proof.
  • Added a dirty real-world odd MSH metadata fixture and parity assertions for Rust core, CLI, REST, gRPC, and local Python smoke expectations.
  • Added a dirty real-world vendor ORU narrative/null observation fixture and parity assertions for Rust core, CLI, REST, gRPC, and local Python smoke expectations.
  • Added xtask check-first-10-minutes-guide and wired it into check-first-use-guides so the job-first CLI onboarding path is executable from a source checkout.
  • Added xtask check-first-use-by-surface-guide to prove the Rust, CLI, and server routes from the first-use routing guide while keeping Python public registry proof separate.
  • Added xtask check-vendor-upgrade-diff-guide to prove the before/after corpus drift workflow from the vendor upgrade guide.
  • Added xtask check-operator-error-guidance-guide to prove representative REST safe-error fields and CLI validation report guidance from the operator error guide.
  • Added xtask check-schema-version-parity to run the shared schema-version acceptance proof across Rust core, CLI, REST, and gRPC surfaces.
  • Added xtask check-safe-error-phi-parity to run the shared safe-error and PHI sentinel acceptance proof across Rust core, CLI, REST, and gRPC surfaces.
  • Added xtask check-profile-parity to run the shared profile lint/explain/test acceptance proof across Rust core, CLI, REST, and gRPC surfaces.
  • Added xtask check-safe-support-bundle-guide to prove the operator support-bundle guide's redact, bundle, replay, artifact, and PHI-sentinel promises.
  • Added xtask check-evidence-artifacts-guide to prove the operator artifact interpretation guide's doctor, profile, validation, corpus, redaction, bundle, manifest, environment, field-path, replay, and PHI-sentinel promises.
  • Added xtask check-sidecar-guide to start a local HTTP sidecar and prove the deployment guide's readiness, redacted validation, invalid-message quarantine, bundle, replay, ACK policy, metrics, corpus diff, and PHI-sentinel path.
  • Hardened xtask check-sidecar-guide to use an ephemeral loopback port for the executable smoke proof while leaving the manual sidecar guide on its fixed local example port.
  • Added generated SAFE-SHARING.md checklists to CLI, Python/core, and server evidence bundles so support packets carry a reviewed pre-share checklist.
  • Added gRPC evidence replay parity with configured-root bundle replay, fail-closed missing/unsafe bundle handling, shared replay reports, and opt-in v2 replay provenance.
  • Added gRPC quarantine output parity through ValidateRedacted with configured-root writes for failed redacted validation and opt-in v2 quarantine provenance.
  • Added gRPC inline corpus diff parity, including before/after inline corpus deltas and opt-in v2 provenance.
  • Added gRPC profile explain parity with shared profile explain reports and opt-in v2 provenance.
  • Added gRPC profile fixture-test parity with shared profile test reports, inline fixture inputs, expected-report subset comparisons, and opt-in v2 provenance.
  • Added gRPC evidence bundle creation parity with configured-root writes, hashed public output IDs, unsafe bundle ID rejection, and opt-in v2 bundle artifacts.
  • Added gRPC profile lint parity with shared profile lint reports and opt-in v2 provenance.

Fixed

  • Derived the Nix package version and Docker image tag from the workspace version so the flake no longer emits a floating latest image tag.
  • Pinned the Kubernetes deployment image tag and version labels to v1.5.0 instead of the previous floating latest image reference.
  • Removed the remaining Grafana documentation example that defaulted the administrator password to admin.
  • Pinned Python wheel proof workflows to Rust 1.95.0 so public Python package build and install-back evidence uses the declared release toolchain.
  • Hardened Python wheel and public-registry proof commands to fail when the imported hl7v2 module version does not match the expected package version.
  • Hardened public Python registry install-back proof to install wheel artifacts only with pip caching disabled.
  • Aligned the e2e HTTP client test dependency on rustls so all-features workspace checks do not re-enable native TLS through reqwest defaults.
  • Added CPU and memory bounds to the Docker Compose validation sidecar smoke stack so local operator proof cannot consume unlimited host resources.
  • Pinned the API Contracts workflow's OpenAPI, schema, and proto toolchain installs to exact versions for reproducible contract validation.
  • Surfaced nightly mutation and security audit failures instead of masking them, while pinning the nightly audit and mutation tools.
  • Pinned the Security workflow's cargo-audit install to a specific version for reproducible dependency-audit runs.
  • Made routed benchmark lane failures visible while bounding the CI benchmark sample to the internal benchmark harness.
  • Made the nightly workflow summary fail the workflow when required nightly jobs report failures instead of downgrading them to warnings.
  • Fixed the evidence-artifacts guide smoke so corpus diff count assertions use the current dirty real-world fixture directories instead of stale after-counts.
  • Added synchronize-only concurrency controls to the API Contracts and Security workflows so stale PR updates can be canceled without canceling push, scheduled, or manual proof runs.
  • Rejected non-finite float spellings such as INF, Infinity, and NaN from HL7 numeric validation.

Documentation

  • Clarified the non-Nix developer setup path, including just installation, xtask fallbacks, optional local quality tools, and OpenSSL/pkg-config troubleshooting boundaries.
  • Removed stale cargo-tarpaulin coverage workflow references from CI/testing docs so they match the current cargo-llvm-cov coverage lane.
  • Replaced the stale issues and next-steps planning snapshot with a current blocker map for public Python proof, evidence parity, operator workflows, dirty corpus expansion, gRPC hardening, RIPR calibration, and later TypeScript/WASM work.
  • Corrected the documentation index so new planning work no longer defaults to the closed historical plans/1.4.1/ directory.
  • Added crate-local context for the canonical Rust hl7v2 crate and the hl7v2-python binding backend, preserving the Rust API versus Python backend boundary for future maintenance.
  • Corrected the ADR index status for ADR-0006 so it matches the proposed OPA policy-enforcement ADR.
  • Corrected the v1.5.0 readiness receipt so the public Python hl7v2 TestPyPI blocker points at the still-open Trusted Publisher issue #563 instead of describing it as closed.
  • Clarified current-state wording for the public Python hl7v2 TestPyPI/PyPI blocker after the registry proof command, workflow routing, parity boundary, and gap audit landed without a new upload attempt.
  • Clarified the TestPyPI release-proof guide so the latest failed publishing-mode run is no longer described as current main after later registry proof scaffolding landed.
  • Refreshed the manifest-linked cross-surface evidence parity gap audit after guide smokes, dirty-corpus expansion, and Python public-registry proof boundaries landed on main.
  • Recorded a second hosted-traffic calibration audit for the advisory ripr evidence surface after evidence-parity and Python-proof PR traffic, keeping the lane non-blocking.
  • Recorded executable safe support-bundle guide smoke proof for the operator-facing support packet workflow.
  • Recorded executable sidecar guide smoke proof for the HTTP deployment walkthrough, with follow-up hardening so the executable check chooses an ephemeral loopback port instead of depending on the manual example port.
  • Recorded the blocked public Python registry install-back proof commands in the machine-readable evidence parity manifest.
  • Updated the current-main TestPyPI publishing-mode proof retry after the v1.5.0 refactor cleanup readiness refresh; wheel smoke passed, while upload remains blocked by TestPyPI Trusted Publisher setup.
  • Recorded a post-release current-main readiness refresh after the Python wheel dirty evidence smoke, Windows policy guard, server validation-helper refactor, and synthetic value/CLI help cleanup landed through #740/#742/#743/#738.
  • Recorded a post-release current-main readiness refresh after REST and gRPC dirty real-world validate-redacted/bundle/replay workflows landed through #734.
  • Recorded a current-main TestPyPI publishing-mode proof retry for public Python package hl7v2; wheel smoke passed, while upload remains blocked by TestPyPI Trusted Publisher setup.
  • Recorded a post-release current-main readiness refresh after gRPC enhanced ACK parity landed through #705.
  • Recorded a current-main TestPyPI publishing-mode proof attempt for public Python package hl7v2; wheel smoke passed, while upload remains blocked by TestPyPI Trusted Publisher setup.
  • Added a repeatable public crates.io install-back smoke script and receipt for the v1.5.0 Rust library, CLI, and server first-use paths.
  • Recorded a post-release current-main readiness refresh after normalization and CLI ACK parity landed through #698.
  • Recorded a post-release current-main readiness refresh after shared Rust/CLI/server/Python dirty-corpus parity landed through #695.
  • Recorded the shared dirty real-world corpus fixture proof for Rust core and CLI corpus evidence commands.
  • Recorded the server dirty real-world corpus parity proof for REST and gRPC corpus evidence commands.
  • Recorded the local Python wheel dirty real-world corpus parity proof.
  • Recorded a post-release current-main readiness refresh after the focused SRP module split train landed through #691.
  • Recorded a fresh hosted non-publishing Python TestPyPI Proof receipt on current main and confirmed public Python hl7v2 is still absent from TestPyPI and PyPI.
  • Recorded a post-release v1.5.0 objective audit showing that the crates.io, tag, GitHub release, and Rust/CLI/server install-back portions are now complete while public Python TestPyPI/PyPI proof remains blocked or undecided.
  • Recorded a current hosted non-publishing Python TestPyPI Proof receipt for public package hl7v2 while keeping TestPyPI upload/install-back blocked on Trusted Publisher setup.
  • Aligned the cross-surface evidence parity docs with the already-tested Python ACK local binding proof.
  • Aligned the gRPC parity docs with the currently implemented profile, bundle/replay, corpus, ACK, and normalize RPC proof surface.
  • Refreshed the v1.5.0 release-readiness receipt after the Python ACK and gRPC parity documentation syncs landed on main.
  • Recorded a pre-release package registry state audit showing that v1.5.0 was not yet published at that snapshot and that public Python hl7v2 was not on PyPI or TestPyPI.
  • Recorded a prompt-to-artifact objective completion audit for the active v1.5.0 lane and kept the lane open until publish, Python, and npm receipts exist.
  • Recorded the v1.5.0 crates.io publish, tag, GitHub release, registry resolution, and Rust/CLI/server install-back receipt.
  • Recorded the final non-publishing v1.5.0 pre-publish proof for the selected crates.io graph before any upload, tag, or GitHub release action.
  • Refreshed the v1.5.0 release-readiness receipt after the advisory RIPR calibration audit and gRPC profile lint parity landed on main.
  • Refreshed the v1.5.0 release-readiness receipt after gRPC profile explain parity landed on main.
  • Refreshed the v1.5.0 release-readiness receipt after gRPC quarantine output parity landed on main.
  • Refreshed the v1.5.0 release-readiness receipt after gRPC evidence bundle creation parity landed on main.
  • Confirmed the v1.5.0 release graph decision still selects the primary Rust graph plus hl7v2-python as binding backend after the latest readiness refresh.
  • Refreshed the v1.5.0 release-readiness receipt after the gRPC inline corpus fingerprint and diff parity work landed on main.
  • Refreshed the v1.5.0 release-readiness receipt after the nightly property command repair and finite numeric validation fix landed on main.
  • Recorded the first hosted-traffic calibration audit for the advisory ripr evidence surface and kept the lane non-blocking.

[1.5.0] - 2026-05-15

Added

  • Added gRPC inline corpus fingerprint parity, including optional inline profile validation issue-code counts and opt-in v2 provenance.
  • Added an advisory ripr static mutation-exposure workflow and suppression policy, keeping runtime mutation as a targeted backstop rather than a default PR tax.
  • Added targeted mutation lane routing for high-risk parser, MLLP, profile, redaction, bundle/replay, evidence-schema, server, Python, and release surfaces.
  • Added a Rust 1.95 / v1.5.0 release-readiness workflow and receipt home for non-publishing proof before crates.io release.

Changed

  • Raised MSRV from Rust 1.93 to Rust 1.95 and pinned rust-toolchain.toml to Rust 1.95.0 for local and CI consistency.
  • Prepared the Rust package graph as version 1.5.0 for hl7v2, hl7v2-server, and hl7v2-cli.
  • Prepared hl7v2-python as a publishable crates.io binding backend for the public Python hl7v2 package, selected it for the v1.5.0 binding-backend release graph, and kept it outside the primary Rust product graph.
  • Tightened the compiler, Clippy, no-panic, and file-policy rails for high-throughput maintenance without increasing default CI weight.
  • Retargeted the public Python distribution metadata and proof workflows to hl7v2 while keeping the internal hl7v2-python Rust crate outside the primary Rust product graph.

Fixed

  • Burned down selected CLI monitor and server metrics Clippy debt with bounded conversions and narrower test assertions.

[1.4.0] - 2026-05-09

Added

  • Added opt-in v2 provenance contracts and producer paths across validation, profile, corpus, redaction, bundle, replay, quarantine, and doctor evidence artifacts while keeping v1 defaults compatible.
  • Added server sidecar hardening after v1.3.0: inline corpus evidence endpoints, server replay, bundle artifact schema opt-in, readiness path-leak protection, redacted structured evidence logs, evidence metrics, and Docker Compose smoke coverage.
  • Added Python evidence hardening: validation/corpus/redaction/bundle/replay v2 parity, cross-surface PHI sentinel coverage, a manual TestPyPI proof workflow, and a Python evidence workflow guide.
  • Added cargo run -p xtask -- evidence-schema-check as the maintained local evidence fixture/schema validation rail.

Fixed

  • Fixed server evidence bundles so replay reproduces messages whose MSH.9 includes a third message-structure component such as ADT^A01^ADT_A01.

Documentation

  • Prepared the v1.4.0 package line and publish dry-run receipt for the final Rust crates.io graph: hl7v2, hl7v2-server, and hl7v2-cli.
  • Verified and refreshed the validation sidecar guide with live server replay output and the current inline corpus diff response shape.
  • Added current-state and contract-index documentation for the post-v1.3.0 evidence-contract line.

[1.3.0] - 2026-05-09

Added

  • Released the v1.3.0 Evidence Loop around deterministic HL7 interface evidence: first-run diagnostics, typed validation reports, profile lint/test/explain, corpus summarize/fingerprint/diff, safe-analysis redaction, evidence bundle/replay, and Python binding parity.
  • Added schema-backed evidence artifacts and golden fixtures for validation reports, profile reports, corpus reports, redaction receipts, bundle summaries, and replay reports.
  • Added CLI automation semantics for evidence commands: stable exit codes, machine-readable stdout, diagnostic stderr, --output, --quiet, and --no-color support.
  • Added server-side edge-guard workflows: sanitized --print-config, readiness checks, redacted validation, evidence bundle creation, policy-driven ACK/NAK decisions, and quarantine output hooks.
  • Added Python binding APIs for parse, JSON export, normalize, validation reports, corpus summary/fingerprint/diff, safe-analysis redaction, evidence bundle creation, and replay verification.
  • Added workflow guides for first-use evidence, vendor upgrade diffs, safe support bundles, and validation sidecar deployment.

Changed

  • Evidence bundles now include manifest and README artifacts, and replay verifies manifest hashes before comparing regenerated evidence.
  • Current release documentation positions v1.3.0 as the published Evidence Loop release for the final Rust package graph.

[1.2.1] - 2026-05-08

Release

  • Moved the package line to 1.2.1 so the final Rust publish can use a fresh release tag instead of reusing the historical v1.2.0 tag.
  • Published the final Rust package graph to crates.io: hl7v2, hl7v2-server, and hl7v2-cli.

Documentation

  • Clarified the post-collapse public package surface: hl7v2, hl7v2-python, hl7v2-server, and hl7v2-cli.
  • Marked old implementation crate READMEs as private deprecated compatibility shims and pointed new Rust users to hl7v2 module paths.
  • Added the 2026-05-07 workspace-patched publish dry-run receipt for the final four-crate package graph.
  • Recorded the direct crates.io dry-run result: hl7v2 passes before publish, while dependent crates correctly wait on hl7v2 being present in the crates.io index.

Historical pre-v1.2.1 recovery notes - 2026-05-03

Added

Modernization & Integrity

  • Rust 2024 Migration: Entire workspace upgraded to the Rust 2024 edition.
  • MSRV 1.93: Set minimum supported Rust version to 1.93 to leverage modern language features (let chains).
  • Let Chains Adoption: Extensively refactored nested conditional logic into idiomatic Rust 2024 let chains across all core crates.
  • gRPC Service: Fully implemented Hl7Service providing high-performance RPCs for Parse, Validate, GenerateAck, and Normalize.
  • Message Lifecycle: New hl7v2-lifecycle crate for enterprise message retention, archival state machines, and legal hold management.
  • Statistical Guard: Enhanced hl7v2-guard with statistical baseline anomaly detection and automated warmup learning periods.

CI & Stability

  • Deterministic Build: Integrated protoc-bin-vendored for cross-platform gRPC code generation without external dependencies.
  • Network Stabilization: Refactored E2E tests to use OS-assigned ports and oneshot channel synchronization, eliminating CI flakiness.
  • Security Hardening: Corrected TruffleHog scan ranges and refined license regression checks to eliminate false positives in documentation.

Fixed

  • Corrected field indices for MSH metadata extraction in gRPC and Lifecycle components.
  • Resolved type inference ambiguities in examples caused by recent model refactoring.
  • Fixed numerous Clippy warnings including collapsible-if, field-reassign-with-default, and io-other-error.
  • Improved gRPC error reporting for invalid profiles and malformed HL7 messages.

[1.2.0] - 2026-03-04

Added

Production Readiness & Security

  • HTTP Server: Production-ready REST API built with Axum.
  • Observability: Integrated Prometheus metrics (/metrics) and structured JSON tracing.
  • Security: API Key authentication (X-API-Key header) and per-IP rate limiting via tower-governor.
  • Swagger UI: Interactive API documentation served at /api/docs via OpenAPI 3.0 spec.
  • Nix Support: Added flake.nix and .envrc for reproducible development environments.

CLI Enhancements

  • Streaming Parse: High-performance, memory-efficient parsing for large files via --streaming flag.
  • Normalization: Improved --canonical-delims support for standardizing HL7 messages.
  • Profiling: Real-time performance monitoring and system resource tracking in CLI commands.

Architecture (SRP Microcrate Refactoring)

  • New Microcrates: Extracted logic into 28 specialized crates for better maintainability and reduced dependency trees.
  • Network: New hl7v2-network crate for MLLP over TCP/TLS.
  • Stream: New hl7v2-stream crate for event-based parsing.
  • Validation: New hl7v2-validation crate for rule-based engine.
  • Generation: Extracted hl7v2-ack, hl7v2-faker, hl7v2-template, and hl7v2-template-values.

Fixed

  • Fixed critical infinite loop in streaming parser during partial segment reads.
  • Resolved message boundary detection issues in sequential MLLP streams.
  • Fixed race conditions in E2E tests caused by TCP port collisions.
  • Improved error reporting for HL7 query path out-of-bounds access.
  • Corrected numerous clippy lints and formatting issues across the workspace.
  • Refactored server authentication to use isolated state for reliable testing.
  • Fixed benchmark compatibility with newer Rust versions (criterion deprecations).

Documentation

  • Created comprehensive docs/API_GUIDE.md for the REST server.
  • Added detailed README.md and CLAUDE.md to every microcrate.
  • Documented key decisions in ADRs 0011-0014 (Security, Observability, Rules).
  • Created RELEASE_PROCESS.md for project maintainers.

[1.1.0] - 2025-11-13

Added (v1.1.0 Features)

Core Parsing (hl7v2-core)

  • Event-based streaming parser with delimiter switching
  • MLLP frame wrapping/unwrapping
  • Complete escape sequence handling (\F, \S, \R, \E, \T)
  • JSON serialization to canonical format
  • Batch processing (BHS/BTS, FHS/FTS)
  • Field path access API with presence semantics
  • Performance benchmarks

Profile Validation (hl7v2-prof)

  • Profile loading from YAML
  • Profile inheritance with parent resolution
  • Profile merging with conflict resolution
  • Constraint validation (required, length, pattern)
  • HL7 table value set validation
  • Cross-field validation rules
  • Advanced data type validation (ST, ID, CX, PN, TS, DT, TM, NM, SI, FT, TX)
  • Temporal rules (date/time comparisons)
  • Contextual rules (if/then logic)
  • Custom validators (phone, email, SSN, birth date, checksums)

Message Generation (hl7v2-gen)

  • Template-based message generation
  • Deterministic seeding for reproducibility
  • Realistic data generators:
    • Names (gender-aware)
    • Addresses (US format)
    • Phone numbers
    • Social Security Numbers
    • Medical Record Numbers
    • ICD-10 codes
    • LOINC codes
    • Medications
    • Allergens
    • Blood types
    • Ethnicity/Race
  • Value distributions (fixed, lists, ranges, normal)
  • Error injection (invalid segments/fields)
  • Corpus generation with multi-template support
  • Golden hash verification

CLI Interface (hl7v2-cli)

  • Parse command (with JSON output, MLLP support)
  • Normalize command (message normalization)
  • Validate command (profile validation)
  • ACK command (ACK generation with AA/AE/AR codes)
  • Generate command (template-based generation)
  • Interactive REPL mode

Known Limitations (v1.1.0)

  • Zero-copy parsing claims overstated (uses Vec internally)
  • No backpressure/bounded channels in streaming
  • No memory bounds enforcement
  • No resume parsing across chunk boundaries
  • No highlight escapes (\H...\N)
  • No remote profile loading
  • No server mode HTTP/gRPC
  • No configuration file support
  • Network module contains stubs only

See docs/STATUS.md for complete status.

Documentation

[1.0.0] - 2025-01-01 (Hypothetical)

Initial Release

  • Core HL7 v2 parsing
  • Basic MLLP support
  • Message normalization
  • Simple JSON serialization
  • Basic validation rules
  • CLI interface (parse, validate, normalize)

Future Releases

  • Continue server sidecar hardening, Python distribution proof, profile conformance quality, and compatibility-shim policy cleanup as separate release trains.

Contributing

For information about contributing changes, see CONTRIBUTING.md.

Versioning

This project follows Semantic Versioning:

  • PATCH (1.0.x): Bug fixes, documentation, internal improvements
  • MINOR (1.x.0): New features, backward compatible
  • MAJOR (x.0.0): Breaking changes, major redesigns

Compatibility

Rust Version Support

  • MSRV (Minimum Supported Rust Version): 1.95
  • Stable: Latest stable Rust recommended

HL7 Versions Supported

  • HL7 v2.3
  • HL7 v2.4
  • HL7 v2.5
  • HL7 v2.5.1
  • HL7 v2.7
  • HL7 v2.8
  • HL7 v2.9

Release Notes

v1.1.0 Release Notes

Highlights:

  • Complete core parsing implementation
  • Comprehensive profile validation
  • Realistic message generation
  • CLI interface for common operations

Performance:

  • Parse: ≥100k messages/minute
  • Memory: Proportional to message size
  • Latency: Sub-millisecond

Quality:

  • 87%+ test coverage
  • Zero unsafe code in public APIs
  • Comprehensive error handling

See docs/STATUS.md for complete feature list.

Links

License

This project is licensed under the GNU Affero General Public License, version 3 or later (AGPL-3.0-or-later). See LICENSE.