- Updated
oxrdf
,oxttl
, andthiserror
dependencies
- Updated
oxrdf
andoxttl
dependencies
- Updated
oxrdf
,oxttl
, anditertools
dependencies
- Updated
oxrdf
andoxttl
dependencies
- Added a test feature for generating W3C EARL test report
- Added two tests from w3c/rdf-canon
- Updated test manifest
- Updated
oxrdf
andoxttl
dependency
- Updated README
- Updated
oxrdf
andoxttl
from GitHub source to pre-release version on crates.io - Published pre-release version on crates.io
- Added an explicit lifetime name to an associated constant to align with rust-lang/rust#115010
- Updated
itertools
dependency - Separeted CHANGELOG from README
- enables selection of internal hash function
- update
oxrdf
andoxttl
to the latest ones (as a result, Rust needs to be 1.70 or higher)
- add
sort
andsort_graph
functions to sort canonicalizedDataset
andGraph
to getVec<Quad>
andVec<Triple>
, respectively
- add
*_graph
APIs to allowGraph
as input - rename
*_with_options
APIs with simplified*_with
- re-export
serialize
function to enable direct use by users - update
oxrdf
andoxttl
dependencies - add more detailed explanation about
oxrdf
andoxttl
dependencies to the README
- Make
oxrdf
andoxttl
point to specific commits on GitHub
- add
*_quads
APIs to allow input inVec<quad>
instead ofDataset
- fix debug log indentations
- modify the test code to use all official test cases from w3c/rdf-canon with their test manifest
- extract
api
module fromcanon
- add info-level log message to notify if there are duplicate hashes in
hash path list
- fix debug log indentations
- update algorithm interfaces to support the canonicalized dataset as additional output
- remove
indexmap
dependency - update comments for latest specification changes
- Add HNDQ call limit as a countermeasure against poison dataset attack by restricting the maximum number of calls to the Hash N-Degree Quads (HNDQ) algorithm
- Use
oxttl
as a N-Quads parser to avoid the dependency on the wholeoxigraph
- Make logger and YamlLayer public for external crates to use them when debugging
- import the latest-updated test cases from w3c/rdf-canon
- Turn logger into a feature: the logger can now be optionally included in our builds, depending on the requirements of each specific build
- Fix logger to avoid unexpected panic
- update
.gitignore
- Add
serialize
function to serialize a normalized dataset into a canonical N-Quads document - Add an example into README
- Revise input/output of canonicalization using OxRDF
Dataset
instead ofVec<Quad>
- Some optimizations
- Fix bug related to module scopes
We have moved away from using our ad-hoc N-Quads parser and RDF data structures and have instead adopted the use of Oxigraph (and its internal OxRDF). This change makes it easier to canonicalize Oxigraph's internal data and output. However, the current version of Oxigraph does not provide the latest Canonical N-Triples and Canonical N-Quads representations. Therefore, we are currently relying on our forked Oxigraph and Oxrdf that supports canonical representations.
Initial release. It does not pass test060 since it uses an ad-hoc N-Quads parser and serializer. (See #1)