rdf_core 0.9.3

A type-safe, modular Dart library for modeling, encoding, and decoding RDF data.

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.

[0.9.3] 2025-06-24

Fixed

• TurtleDecoder: Fixed parsing of prefixed names with colons in local parts

  • Corrected prefixed name splitting to only split on the first colon, allowing colons in local names per W3C Turtle specification
  • Updated parser to properly handle IRIs like prefix:local:name which are valid according to PN_LOCAL grammar

• TurtleTokenizer: Enhanced local name validation according to PN_LOCAL specification

  • Implemented proper PN_LOCAL grammar validation that prevents invalid patterns like dots at the end of local names
  • Added validation to reject local names ending with dots (e.g., author.me → invalid)
  • Added validation to reject local names starting with dots or hyphens
  • Added validation to reject double dots and hyphen-dot patterns
  • Improved tokenizer to properly handle dots in the middle of valid local names
  • Fixed tokenizer backtracking when invalid trailing dots are detected

• RdfNamespaceMappings: Enhanced IRI validation for prefix generation

  • Added comprehensive _isValidPnLocal validation function that enforces W3C Turtle PN_LOCAL rules
  • Improved extractNamespaceAndLocalPart to reject IRIs with invalid local name patterns
  • Added validation for domain-like suffixes and percent encoding in local names
  • Performance optimization: pre-compiled regex patterns as static final fields to avoid repeated compilation

Enhanced

• Performance: Optimized regex usage in TurtleTokenizer by using static final compiled patterns
  • Replaced inline RegExp(r'[0-9]') calls with pre-compiled _isDigitRegExp
  • Replaced inline character class regexes with static compiled patterns for better performance

Added

• Testing: Added comprehensive test suite for Turtle local name validation
  • Tests for invalid PN_LOCAL patterns (dots at end, double dots, hyphen-dot combinations)
  • Tests for valid PN_LOCAL patterns with dots in the middle
  • Validation that invalid local names are serialized as full IRIs instead of using prefix notation

[0.9.2] - 2025-05-15

Fixed

• Turtle encoding of IRIs that contain url escape (e.g. %20) must not use prefixes

[0.9.1] - 2025-05-15

Changed

• TurtleEncoder: Fixed prefix and relative URI handling with base URI

  • Improved handling of IRIs with baseUri to correctly prioritize relative IRIs for subjects/objects
  • Always use prefixes for predicates, even when they fall under the base URI
  • Apply prefixes only to subjects/objects when their namespace is longer than the base URI
  • Fixed issue where base URI namespaces were sometimes incorrectly used as prefixes

• TurtleEncoder: Improved handling of local names that start with a digit

  • Local names that start with a digit are now consistently serialized as full IRIs rather than using prefix notation by default
  • This behavior can be configured through the useNumericLocalNames option in TurtleEncoderOptions
  • Fixed internal parameter consistency to ensure proper serialization of IRIs with numeric local names regardless of prefix availability

• RdfNamespaceMappings: Improved prefix generation for URIs with hyphens

  • Changed prefix generation strategy to use initials for components with hyphens (e.g., "test-complex-ontology" → "tco")
  • Ensures compliance with RDF specifications that prohibit hyphens in prefixes
  • Makes prefix generation consistent between domain names and path components with hyphens
  • Added comprehensive tests for hyphen handling in prefix generation

[0.9.0] - 2025-05-14

Breaking Changes

• RdfCore API: Changed signature for codec retrieval

  • Modified how codecs are accessed and configured in the RdfCore class
  • Users who directly access or modify codecs will need to update their code

• Codec Architecture: Enhanced options support

  • RdfGraphCodec, RdfGraphEncoder, and RdfGraphDecoder now support distinct RdfGraphEncoderOptions and RdfGraphDecoderOptions
  • All codec implementations must now properly handle these options

Added

• TurtleEncoder: Added support for automatic prefix generation
  • New generateMissingPrefixes option in TurtleEncoderOptions (enabled by default)
  • The encoder now automatically generates meaningful prefixes for IRIs that don't match existing prefixes
  • Reuses generated prefixes consistently within the same serialization

Enhanced

• Documentation was greatly enhanced for all exported classes

• RdfNamespaceMappings: Added static utility method for namespace manipulation

  • New public static method extractNamespaceAndLocalPart for splitting IRIs into namespace and local part
  • Can be used by clients to implement custom namespace-aware functionality

• TurtleEncoder: Improved handling of base URIs and relative IRIs

  • Properly writes base directive in Turtle output when baseUri is provided
  • Correctly converts absolute IRIs to relative IRIs when they start with the base URI
  • Ensures no prefix generation for IRIs that should be serialized as relative paths

[0.8.2] - 2025-05-13

Fixed

• AutoDetectingGraphCodec: Improved handling of codec detection failures
  • Fixed implementation to properly delegate to registry.detectGraphCodec
  • Modified AutoDetectingGraphDecoder to try all registered codecs in sequence when auto-detection fails
  • Corrected supportedMimeTypes to use the default codec's supported types
• Turtle Encoder: Enhanced relative URI support
  • Added proper handling of relative URIs when baseUri is provided
  • Improved base URI handling in the Turtle output format
• Exception Handling: Updated tests to handle both FormatException and custom exception types:
  • Modified tests to accept RdfSyntaxException, RdfDecoderException in addition to FormatException
  • Fixed encoder comparison to check for runtimeType equality instead of instance identity

[0.8.1] - 2025-05-13

Breaking Changes - DO NOT USE 0.8.0 please

• API Refactoring: In preparation for RDF 1.1 Dataset support:
  • Renamed RdfCodec → RdfGraphCodec
  • Renamed RdfEncoder → RdfGraphEncoder
  • Renamed RdfDecoder → RdfGraphDecoder

[0.8.0] - 2025-05-12

Breaking Changes

• API Refactoring: Changed terminology to align with dart:convert standards
  • Renamed parse → decode
  • Renamed serialize → encode
  • Renamed RdfFormat → RdfCodec
  • Renamed withStandardFormats → withStandardCodecs
  • Renamed withFormats → withCodecs

Added

• Added additionalCodecs parameter to withStandardCodecs factory constructor
• Added global convenience variables (turtle, jsonld, ntriples, rdf) for easier API usage
• Improved documentation for registry access through the registry property

[0.7.6] - 2025-05-08

Added

• RdfNamespaceMappings can now generate a prefix automatically if there was no matching prefix found

Fixed

• IriTerm used to be case insensitive in equals, but that is actually wrong - for example: https://schema.org/Datatype != https://schema.org/datatype !

[0.7.5] - 2025-05-07

Changed

• Another turtle tweak: suppress duplicate objects in multivalue predicates

[0.7.4] - 2025-05-07

Changed

• Better sorting of turtle output

[0.7.3] - 2025-05-07

Changed

• Removed unused code

[0.7.2] - 2025-05-07

Added

• Added additional well-known namespaces to standard mappings:
  • Added geo (http://www.w3.org/2003/01/geo/wgs84_pos#)
  • Added contact (http://www.w3.org/2000/10/swap/pim/contact#)
  • Added time (http://www.w3.org/2006/time#)
  • Added vs (http://www.w3.org/2003/06/sw-vocab-status/ns#)
  • Added dcmitype (http://purl.org/dc/dcmitype/)
  • Added void (http://rdfs.org/ns/void#)
  • Added prov (http://www.w3.org/ns/prov#)
  • Added gr (http://purl.org/goodrelations/v1#)

Changed

• Enhanced Turtle serialization for improved readability:
  • Implemented inline blank node serialization for nodes referenced exactly once
  • Added support for nested inline blank nodes in square bracket notation
  • Added proper whitespace formatting with double line breaks between subject groups
  • Optimized collection serialization to work with inline blank nodes
• Added linting configuration using package:lints/core.yaml
• Improved code quality with lint fixes and optimized Dart idioms

[0.7.1] - 2025-05-06

Changed

• Triple, IriTerm, BlankNodeTerm and LiteralTerm classes now output in n-triple style in .toString() for better readability of debug output.
• Included information about json-ld in README and homepage.

[0.7.0] - 2025-05-05

Added

• Added support for N-Triples format, including parser and serializer implementations
  • Implemented N-Triples as per W3C RDF 1.1 N-Triples specification
  • Added proper escaping and validation for N-Triples syntax
  • Integrated N-Triples format into the standard formats registry
  • Available via MIME type 'application/n-triples' and file extension '.nt'

[0.6.10] - 2025-05-05

Added

• Added new factory methods LiteralTerm.integer(), LiteralTerm.decimal(), and LiteralTerm.boolean() for more concise and type-safe creation of common literal types.
• Updated TurtleParser to use these specialized factory methods for literal values.
• Enhanced TurtleSerializer to use native syntax for integer, decimal and boolean literals, resulting in more idiomatic and concise Turtle output.
• Enhanced TurtleSerializer to use native syntax for lists (aka rdf collections: ex:Person1 ex:someProperty ("foo" "bar" "blub");) and sets (aka multiple triple per subject/predicate pair: ex:Person1 ex:someProperty "foo", "bar", "blub";).

[0.6.9] - 2025-05-02

Fixed

• Export exceptions

[0.6.8] - 2025-05-02

Fixed

• Added back path and yaml to dev dependencies

[0.6.7] - 2025-05-02

Fixed

• More dependency cleanups

[0.6.6] - 2025-05-02

Fixed

• Fixed imports

[0.6.5] - 2025-05-02

Fixed

• TurtleFormat and JsonLdFormat should have been exported
• unused http dependency was removed

[0.6.4] - 2025-05-02

Fixed

• Fixed tests to use relative paths instead of absolute paths for better portability
• Improved examples for handling broken Turtle syntax

[0.6.3] - 2025-05-02

Added

• Enhanced Turtle parsing flexibility with configurable parsing flags in TurtleTokenizer
  • Added TurtleParsingFlag enum with options for handling non-standard Turtle syntax
  • Implemented support for digits in local names (allowDigitInLocalName)
  • Added support for missing dots after prefix declarations (allowMissingDotAfterPrefix)
  • Implemented auto-addition of common prefixes (autoAddCommonPrefixes)
  • Added support for prefix declarations without @ symbol (allowPrefixWithoutAtSign)
  • Improved handling of missing final dots (allowMissingFinalDot)
  • Implemented support for identifiers without colons (allowIdentifiersWithoutColon)
• Improved error reporting with detailed position information and diagnostics
• Added comprehensive logging to help diagnose parsing issues in real-world Turtle files

Changed

• Refactored TurtleTokenizer for better maintainability and modularity
• Improved documentation with practical examples of non-standard Turtle syntax handling

Fixed

• Made Turtle parser more robust when processing real-world datasets with syntax variations
• Enhanced error messages to be more informative for debugging parsing issues

[0.6.2] - 2025-05-01

Fixed

• Important fix: build.yaml should not have been committed, it breaks the build of user projects

[0.6.1] - 2025-05-01

Changed

• Removed a deprecated annotation that was no longer needed after the vocabulary cleanup

[0.6.0] - 2025-05-01

Removed

• BREAKING CHANGE: Removed all deprecated vocabulary classes from the vocab directory
  • Removed all previously deprecated vocabulary classes (SchemaProperties, SchemaTypes, SchemaPersonProperties, etc.)
  • Users should now use direct IriTerm instances instead of vocabulary classes
  • Example: replace SchemaProperties.name with IriTerm('https://schema.org/name')
• Updated examples to use direct IriTerm instances rather than vocabulary constants

Added

• Improved documentation around IriTerm usage patterns
• Added more comprehensive test coverage for core RDF functionality
• Support for multiline strings in turtle files

[0.5.1] - 2025-04-30

Fixed

• Fixed release script to correctly parse the changelog format with square brackets
• Added missing development dependencies required for the release tooling

[0.5.0] - 2025-04-30

• BREAKING CHANGE: Marked all vocabulary classes as deprecated
  • Added @deprecated annotation to all classes in vocabulary modules (acl, dc, dc_terms, foaf, ldp, etc.)
  • Classes will be removed in a future release due to API design concerns
  • Users should migrate to the upcoming new vocabulary API
• Improved code documentation
• Enhanced developer warnings for approaching breaking changes

[0.4.0] - 2025-04-29

• BREAKING CHANGE: Reorganized API structure for better modularity and usability
• Added RdfNamespaceMappings for improved namespace handling
• Enhanced serializers to support custom namespace mappings
• Improved documentation across the codebase
• Updated API documentation to reflect new structure

0.3.1

• Extended Schema.org vocabulary with postal address support
  • Added PostalAddress, ContactPoint, and Country classes
  • Added SchemaAddressProperties for complete address modeling
  • Added properties for street address, locality, region, postal code, country, etc.
• Added comprehensive tests for new postal address components
• Added example demonstrating postal address usage with Schema.org vocabulary

0.3.0

• Enhanced blank node handling in RDF serialization and parsing
• Refactored IriTerm usage with prevalidated constructor and improved validation logic
• Refactored RDF Core Vocabulary: replaced constants with vocabulary modules
• Added RDF vocabularies for RDFS, Schema.org, SKOS, Solid, and vCard
• Improved documentation and consistency across RDF vocabularies
• Added comprehensive tests for RDF exception handling and parsing
• Enhanced test coverage for TurtleTokenizer and JsonLdFormat functionality
• Fixed edge cases in blank node handling
• Improved test assertions and documentation for various RDF vocabularies

0.2.0

• Numerous improvements and new features since 0.1.2.
• See commit history for details on all enhancements and fixes.

0.1.2

• Cleaned up all library names for consistency (removed rdf_core. prefix, now use simple names like exceptions.base).
• Minor formatting and style improvements in core files.
• No breaking changes; all public APIs remain backward compatible.

0.1.1

• Example and documentation now consistently use canonical RDF vocabularies (e.g., http://xmlns.com/foaf/0.1/) with http://, not https://.
• Turtle and JSON-LD serializers automatically warn if non-canonical (https) namespaces are used when canonical (http) is available.
• Added documentation and comments on best practices for prefixes and vocabularies.
• Improved prefix handling and static analysis compliance.
• Added and refined tests for prefix usage and serialization correctness.

0.1.0

• Initial release of rdf_core: type-safe, extensible Dart library for RDF data manipulation.
• Implements RDF graph, triple, term, and serialization/parsing for Turtle and JSON-LD.
• Plugin architecture for formats and adapters.
• Comprehensive test coverage.