CHANGELOG.md

3.0.0 (Unreleased)

Schema

The schema changes are preliminary and temporary, since they are developed in another repository. I only included them here to keep track of changes.

1. Added contributors key, which are currently exactly the same type as authors.
2. Added key relation to identifiers. Its value is an enum same as Zenodo's related_identifiers.relation (cites, isNewVersionOf, isCitedBy, etc).
3. Updated ISBN regex pattern
4. Updated the schema with new licenses from SPDX license list

CLI

1. Long outputs, for example those resulting from validation errors, are now truncated by default, but this behavior can be disabled using --verbose flag #278
2. cffconvert exits with non zero code when something went wrong converting or validating #266
3. Added "Star on GitHub" footnote in CLI

Library

1. Added conversion and validation behavior for CFF 1.3
2. Converted metadata now includes information from CFF 1.3 key contributors for target formats that support it #333, #334
3. Zenodo metadata now includes related_identifiers and optionally relation_type such as citedBy, compiles, isSupplementTo etc using CFF 1.3 key relation on elements in identifiers #327, #329
4. Converted metadata now uses author email for target formats that support it #285
5. entity authors get exported as Organization when converting to schema.org or codemeta (was Person) #239
6. CFF key name takes precedence over alias when converting entity authors (was the other way around) #308
7. schema.org and codemeta conversion export an affiliation name as name (was legalName) #272
8. Added "upload_type": "software" when exporting to Zenodo from CFF files that use schema versions 1.0.x or 1.1.x #306
9. Apalike conversion now uses "and" or ", and" to concatenate author names when there are 2 or more authors #226
10. Bibtex conversion now uses braces to protect names of entity authors #156
11. Library now explicitly exports Citation and nothing else
12. Requests to GitHub now use versioned API
13. Requests to GitHub can now be authenticated via environment variable CFFCONVERT_API_TOKEN for higher rate limit #353
14. Requests to GitHub now use the target repo's default branch if user leaves branch unspecified #263

Development

1. Changed the repo name from cff-converter-python to just cffconvert #283
2. Moved to a src/ directory layout #311
3. Clearer separation between cli parts and lib parts #276
4. Replaced setup.cfg and setup.py with pyproject.toml #312
5. Added pytest markers for apalike, bibtex, codemeta, endnote, ris, schemaorg, zenodo tests #330
6. Added pytest marker lib for library tests and cli for command line interface tests
7. Refactored testing directory layout/naming to facilitate testing subparts of the test tree #315
8. Consistent styling now enforced with ruff #339
9. Added linting via pre-commit (isort, pyroma, ruff, prospector) #324
10. Added Markdown link checker pre-commit hook
11. Added cffconvert metadata validation checker via pre-commit
12. Updated jsonschema dependency to a wider range, now includes 4.x #292
13. Updated python versions used in CI #295
14. Added tests to verify that behavior maps such as those for authors or URLs implement all possible keys, that there are no extra keys, and that there are no extra methods. #300
15. Added missing codemeta tests to 1.2.0/author-creators #355
16. Fixed bug with YAML parsing related to testing order #343
17. Fixed invalid CFF in one of the tests #297

Other

1. cffconvert is now available as a pre-commit hook with id validate-cff #269
2. Updated the Dockerfile to use latest version of cffconvert and recent version of Alpine #340

2.0.0

CLI

• added APA output (PR #149; thanks @wleoncio)
• added support for validation and conversion of CITATION.cff files with cff-version: 1.2.0
• argument --outputformat was renamed to --format
• argument -ig, --ignore-suspect-keys was removed
• argument --verbose was removed
• argument --show-trace was added

Library

• added APA output (PR #149; thanks @wleoncio)
• added support for validation and conversion of CITATION.cff files with cff-version: 1.2.0
• simplified the Citation class and its interface
• cli is no longer part of the public interface of the library
• URLs are now constructed from identifiers, repository, repository-artifact, repository-code, or url, with a transparent mechanism to choose what to use given the data that is available from a given CITATION.cff file
• Authors are now constructed from given-names, family-names (including name-particle and name-suffix), alias, name, affiliation and orcid, with a transparent mechanism to choose what to use given the data that is available from a given CITATION.cff file

Other

• switched to static configuration (setup.cfg over setup.py)
• dependencies are now in setup.cfg as opposed to requirements[-dev].txt
• updated version ranges for dependencies
• tests are no longer unittest.TestCase based, but pytest with fixtures
• added jsonschema based validation for CITATION.cff files with cff-version: 1.2.0
• implemented State pattern for Citation to help it deal with multiple behaviors under past and future versions of the Citation File Format.
• switched from TravisCI to GitHub Actions workflows, added linting and publishing workflows
• CI is now testing against Python 3.6, 3.7, 3.8, and 3.9 on Mac, Linux and Windows
• copies of the relevant schemas are now bundled with the package
• organized the tests to be more orthogonal to each other / less overlap between tests

1.3.3

• With recent changes to the release process, the schema will be in a different place than before. This release fixes #119).

1.3.2

• the ruamel.yaml dependency was not specified tightly enough, requirements.txt has been updated as have the notes for maintainers.

1.3.1

• 'cff-version: 1.0.3' is now interpreted as 1.0.3-1 (the latest schema version that implements the spec 1.0.3). This will fix some problems with the list of SPDX license abbreviations. These additional licenses should now work:
  • AGPL-3.0-only
  • AGPL-3.0-or-later
  • BSD-1-Clause
  • BSD-2-Clause-Patent
  • CDLA-Permissive-1.0
  • CDLA-Sharing-1.0
  • EPL-2.0
  • EUPL-1.2
  • GFDL-1.1-only
  • GFDL-1.1-or-later
  • GFDL-1.2-only
  • GFDL-1.2-or-later
  • GFDL-1.3-only
  • GFDL-1.3-or-later
  • GPL-1.0-only
  • GPL-1.0-or-later
  • GPL-2.0-only
  • GPL-2.0-or-later
  • GPL-3.0-only
  • GPL-3.0-or-later
  • LGPL-2.0-only
  • LGPL-2.0-or-later
  • LGPL-2.1-only
  • LGPL-2.1-or-later
  • LGPL-3.0-only
  • LGPL-3.0-or-later

1.3.0

• added schema.org converter method

1.2.2

• added documentation for the Google Cloud Function interface

1.2.1

• setup.py no longer includes test dependencies as install dependencies

1.2.0

• corrected an error where cffconvert could not raise an error during validation (#94).

1.1.0

• replaced pykwalifire with its parent pykwalify
• now works for python 3.7 (refs #80)
• not using PyYAML anymore (but it still comes along with pykwalify for some reason)
• added a function that can be used as Google Cloud function
• hopefully fixed parsing of strings that should have been entered as dates (the new validator does not find that offensive, hence I had to fix it myself)

1.0.4

• replaced PyYAML dependency with ruamel.yaml

1.0.3

• security bugfix by updating requests from 2.18.4 to 2.20.0

1.0.2

• fixed bug #82 (warnings on stdout)

1.0.1

• fixed bug #73 (orcid format in zenodo export)

1.0.0

• first stable release
• solved bug #59 (cffconvert creates local file data.yaml and schema.yaml on validate)

0.0.5

• Minor changes

0.0.4

• added optional validation of CITATION.cff files using pykwalifire (--validate)
• added printing the CITATION.cff contents from the command line
• added unit tests for command line interface
• added integration with sonarcloud code quality monitoring
• removed shorthand command line argument -v (represented both --validate and --verbose)
• added showing its own version (--version)
• command line argument --ignore-suspect-keys no longer needs to be assigned a value, it's simply a flag