Skosify 2.3.0

This is a minor version that makes some operations (prefLabel selection, logging) more deterministic, helping to ensure that successive runs of Skosify on the same input produce the same result every time. A new facility for performing SPARQL Update operations just before serializing the output has been added.

On the more technical side, Skosify now works with rdflib 6 as well, Python 2.7 support has been dropped, support for Python 3.9 introduced, and there are some changes to the unit tests and CI environment.

• #79/#80 Deterministic skosify log
• #85 add post_update_query option
• #87 Fix compatibility with rdflib 6.0.0
• #88 Drop Python 2 compatibility, add Python 3.9 support
• #89 Switch from Travis CI to GitHub Actions
• #90 Upgrade pytest and switch to flake8
• #91 Make prefLabel policies deterministic by using label string sort order to break ties

Skosify 2.2.1

This is a patch release that works around a bug in the rdflib N-Triples parser which prevented using N-Triples files as input under Python 3.

• #78 Work around N-Triples parser issue by using N3 parser instead

Skosify 2.2.0

This is a minor release which adds new prefLabel selection policies.

• #77 New preflabel policies "uppercase" and "lowercase" plus combinations

Skosify 2.1.2

This is a patch release that fixes a Python 3 compatibility problem.

• #76 Use open() instead of file() to read SPARQL query files
• update author/contributor list in top level README.rst file

Skosify 2.1.1

This is a patch release that fixes a couple of problems in the 2.1.0 release.

• #75 Fix further instances of rdflib 5 compatibility problems
• update Trove classifiers in setup.py to match the supported Python versions (2.7, 3.6, 3.7, 3.8)

Skosify 2.1.0

This is a long-awaited release that fixes many compatibility issues with newer Python versions and rdflib 5.0.0.
Supported Python versions are now 2.7, 3.6, 3.7 and 3.8.

Includes these fixes and improvements: (not necessarily a complete list)

• #56 Extract more functionality into the public api (credit: @danmichaelo)
• #50 Allow sections to be skipped in skosify.cfg (credit: @danmichaelo)
• #63 Fix: Read to_format from config (credit: @danmichaelo)
• #64 Move label checks into public api (credit: @danmichaelo)
• #65/#66 Drop Python 3.3 support and use Travis CI Xenial environment
• #68 Force UTF-8 encoding for input files
• #70 Config parser fix (credit: @kouralex)
• #71 Fix RDFlib 5 compatibility issue (credit: @danmichaelo)
• #72 loop hierarchically redundant relations in a sorted fashion (credit: @kouralex)
• #73 Write encoded output to stdout.buffer, as required by Python 3

Skosify 1.0

This is Skosify 1.0. It was originally released on Google Code in January 2014. This GitHub release was created for historical continuity. Original release notes:

ANN: Skosify 1.0, SKOS validation, conversion and correction tool

I'm happy to announce the release Skosify 1.0, a tool for validating and
correcting SKOS vocabularies and converting from legacy formats to SKOS.
Skosify can be used to find and, in many cases, correct many common
problems in SKOS files. The program takes a SKOS or RDF file as input
and outputs a processing report as well as a corrected vocabulary file.

Version 1.0 contains various enhancements, including richer mapping and
transformation facilities. The code is now sufficiently stable to
warrant a 1.0 version number.

The code is available on Google Code [2] and is open source, under the
MIT license. The tool is implemented in Python, using rdflib. The
current code is compatible with both Python 2.x and 3.x and follows the
PEP8 coding style guidelines.

Skosify has been used to convert, validate and correct many ontologies
developed and published in the FinnONTO projects as well as many other
SKOS vocabularies found on the Web of Data. More information about the
project, including two academic papers, can be found at the Skosify
project page [3].

Skosify is currently being maintained at the National Library of Finland
[4] where it is used to prepare vocabularies for publication in the
upcoming Finto [5] thesaurus and ontology service, a spin-off of the
ONKI system developed in the FinnONTO projects.

As always, I'm happy to hear any comments. Bug reports or requests for
features can be submitted to the issue tracker at the Google Code site.

Thanks,
Osma

[1] http://demo.seco.tkk.fi/skosify
[2] http://code.google.com/p/skosify
[3] http://www.seco.tkk.fi/tools/skosify
[4] http://www.nationallibrary.fi
[5] http://finto.fi/