This change log follows the Keep a Changelog spec. Every release contains the following sections:
Added
for new features.Changed
for changes in existing functionality.Deprecated
for soon-to-be removed features.Removed
for now removed features.Fixed
for any bug fixes.Security
in case of vulnerabilities.
The versions follow semantic versioning.
- Bumped versions of requirements (#288).
This release was delayed due to the absence of the lead developer (Carmen, me, the person writing these release notes). Many thanks to Max Mehl for coordinating the project in my absence. Many thanks also to the contributors who sent in pull requests, in reverse chronological order: Olaf Meeuwissen, Mikko Piuola, Wolfgang Traylor, Paul Spooren, Robert Cohn, ethulhu, pukkamustard, and Diego Elio Pettenò.
-
Separate Docker image with additional executables installed (
fsfe/reuse:latest-extra
) (#238) -
Allow different styles of copyright lines: SPDX (default), String, String (C), String ©, and © (#248)
-
Convenience function to update resources (SPDX license list and exceptions) (#268)
-
More file types are recognised:
- ClojureScript (
.cljc
,.cljs
) - Fortran (
.F
,.F90
,.f90
,.f95
,.f03
,.f
,.for
) - Makefile (
.mk
) - PlantUML (
.iuml
,.plantuml
,.pu
,.puml
) - R (
.R
,.Renviron
,.Rprofile
) - ReStructured Text (
.rst
) - RMarkdown (
.Rmd
) - Scheme (
.scm
) - TypeScript (
.ts
) - TypeScript JSX (
.tsx
) - Windows Batch (
.bat
)
- ClojureScript (
-
More file names are recognised:
- .dockerignore
- Gemfile
- go.mod
- meson.build
- Rakefile
- Use UTF-8 explicitly when reading files (#242)
- Updated license list to 3.11.
- Similar to CAL-1.0 and CAL-1.0-Combined-Work-Exception, SHL-2.1 is now ignored because it contains an SPDX tag within itself.
- Added
--skip-unrecognised
flag toaddheader
in order to skip files with unrecognised comment styles instead of aborting without processing any file.
-
Always write the output files encoded in UTF-8, explicitly. This is already the default on most Unix systems, but it was not on Windows.
-
All symlinks and 0-sized files in projects are now ignored.
-
The licenses CAL-1.0 and CAL-1.0-Combined-Work-Exception contain an SPDX tag within themselves. Files that are named after these licenses are now ignored.
-
Fixed a bug where
addheader
wouldn't properly apply the template on.license
files if the.license
file was non-empty, but did not contain valid SPDX tags.
- Updated license list to 3.8-106-g4cfec76.
-
Add support for autoconf comment style (listed as m4).
-
More file types are recognised:
- Cython (
.pyx
,.pxd
) - Sass and SCSS (
.sass
,.scss
) - XSL (
.xsl
) - Mailmap (
.mailmap
)
- Cython (
-
Added
--single-line
and--multi-line
flags toaddheader
. These flags force a certain comment style.
- The Docker image has an entrypoint now. In effect, this means running:
docker run -v $(pwd):/data fsfe/reuse lint
instead of
docker run -v $(pwd):/data fsfe/reuse reuse lint
.
-
Added support for Mercurial 4.3+.
-
A pre-commit hook has been added.
-
When an incorrect SPDX identifier is forwarded to
download
orinit
, the tool now suggests what you might have meant.
-
Under the hood, a lot of code that has to do with Git and Mercurial was moved into its own module.
-
The Docker image has been changed such that it now automagically runs
reuse lint
on the/data
directory unless something else is specified by the user.
-
Fixed a bug with
addheader --explicit-license
that would result infile.license.license
iffile.license
already existed. -
Fixed a Windows-only bug to do with calling subprocesses.
-
Fixed a rare bug that would trigger when a directory is both ignored and contains a
.git
file.
-
Support Jinja (Jinja2) comment style.
-
Support all multi-line comment endings when parsing for SPDX information.
-
Improvements to German translation by Thomas Doczkal.
-
No longer remove newlines at the end of files when using
addheader
. -
There can now be a tab as whitespace after
SPDX-License-Identifier
andSPDX-FileCopyrightText
.
-
Implemented
--root
argument to specify the root of the project without heuristics. -
The linter will complain about licenses without file extensions.
-
Deprecated licenses are now recognised.
lint
will complain about deprecated licenses. -
ProjectReport generation (
lint
,spdx
) now uses Python multiprocessing, more commonly called multi-threading outside of Python. This has a significant speedup of approximately 300% in testing. Because of overhead, performance increase is not exactly linear. -
For setups where multiprocessing is unsupported or unwanted,
--no-multiprocessing
is added as flag. -
addheader
now recognises many more extensions. Too many to list here. -
addheader
now also recognises full filenames such asMakefile
and.gitignore
. -
Added BibTex comment style.
-
Updated translations:
- Dutch (André Ockers, Carmen Bianca Bakker)
- French (OliBug, Vincent Lequertier)
- Galician (pd)
- German (Max Mehl)
- Esperanto (Carmen Bianca Bakker)
- Portuguese (José Vieira)
- Spanish (Roberto Bauglir)
- Turkish (T. E. Kalayci)
-
The linter output has been very slightly re-ordered to be more internally consistent.
-
reuse --version
now prints a version with a Git hash on development versions. Towards that end, the tool now depends onsetuptools-scm
during setup. It is not a runtime dependency.
-
lint
no longer accepts path arguments. Where previously one could doreuse lint SUBDIRECTORY
, this is no longer possible. When linting, you must always lint the entire project. To change the project's root, use--root
. -
FileReportInfo
has been removed.FileReport
is used instead.
-
A license that does not have a file extension, but whose full name is a valid SPDX License Identifier, is now correctly identified as such. The linter will complain about them, however.
-
If the linter detects a license as being a bad license, that license can now also be detected as being missing.
-
Performance of
project.all_files()
has been improved by quite a lot. -
Files with CRLF line endings are now better supported.
-
The program's package name on PyPI has been changed from
fsfe-reuse
toreuse
.fsfe-reuse==1.0.0
has been created as an alias that depends onreuse
.fsfe-reuse
will not receive any more updates, but will still host the old versions. -
For users of
fsfe-reuse
, this means:-
If you depend on
fsfe-reuse
orfsfe-reuse>=0.X.Y
in your requirements.txt, you will get the latest version ofreuse
when you installfsfe-reuse
. You may like to change the name toreuse
explicitly, but this is not strictly necessary. -
If you depend on
fsfe-reuse==0.X.Y
, then you will keep getting that version. When you bump the version you depend on, you will need to change the name toreuse
. -
If you depend on
fsfe-reuse>=0.X.Y<1.0.0
, then 0.6.0 will be the latest version you receive. In order to get a later version, you will need to change the name toreuse
.
-
-
--include-submodules
is added to also include submodules when linting et cetera. -
addheader
now also recognises the following extensions:- .kt
- .xml
- .yaml
- .yml
-
Made the workaround for
MachineReadableFormatError
introduced in 0.5.2 more generic. -
Improved shebang detection in
addheader
. -
For
addheader
, the SPDX comment block now need not be the first thing in the file. It will find the SPDX comment block and deal with it in-place. -
Git submodules are now ignored by default.
-
addheader --explicit-license
now no longer breaks on unsupported filetypes.
python3 -m reuse
now works.
- Updated license list to 3.6-2-g2a14810.
-
Performance of
reuse lint
improved by at least a factor of 2. It no longer does any checksums on files behind the scenes. -
Also handle
MachineReadableFormatError
when parsing DEP5 files. Tries to import that error. If the import is unsuccessful, it is handled.
This release was replaced by 0.5.2 due to importing
MachineReadableFormatError
, which is not a backwards-compatible change.
-
TeX and ML comment styles added.
-
Added
--year
and--exclude-year
toreuse addheader
. -
Added
--template
toreuse addheader
. -
Added
--explicit-license
toreuse addheader
. -
binaryornot
added as new dependency. -
Greatly improved the usage documentation.
-
reuse addheader
now automatically adds the current year to the copyright notice. -
reuse addheader
preserves the original header below the new header if it did not contain any SPDX information. -
reuse addheader
now correctly handles.license
files. -
Bad licenses are no longer resolved to LicenseRef-Unknown. They are instead resolved to the stem of the path. This reduces the magic in the code base.
-
.gitkeep
files are now ignored by the tool. -
Changed Lisp's comment character from ';;' to ';'.
--all
argument help toreuse download
, which downloads all detected missing licenses.
-
When using
reuse addheader
on a file that contains a shebang, the shebang is preserved. -
Copyright lines in
reuse spdx
are now sorted. -
Some publicly visible TODOs were patched away.
This release is a major overhaul and refactoring of the tool. Its primary focus is improved usability and speed, as well as adhering to version 3.0 of the REUSE Specification.
-
reuse addheader
has been added as a way to automatically add copyright statements and license identifiers to the headers of files. It is currently not complete. -
reuse init
has been added as a way to initialise a REUSE project. Its functionality is currently scarce, but should improve in the future.
-
reuse lint
now provides a helpful summary instead of merely spitting out non-compliant files. -
reuse compile
is nowreuse spdx
. -
In addition to
Copyright
and©
, copyright lines can be marked with the tagSPDX-FileCopyrightText:
. This is the new recommended default. -
Project no longer depends on pygit2.
-
The list of SPDX licenses has been updated.
-
Valid-License-Identifier
is no longer used, and licenses and exceptions can now only live inside of the LICENSES/ directory.
-
Removed
--ignore-debian
. -
Removed
--spdx-mandatory
,--copyright-mandatory
,--ignore-missing
arguments fromreuse lint
. -
Remove
reuse license
. -
GPL-3.0 and GPL-3.0+ (and all other similar GPL licenses) are no longer detected as SPDX identifiers. Use GPL-3.0-only and GPL-3.0-or-later instead.
-
Scanning a Git directory is a lot faster now.
-
Scanning binary files is a lot faster now.
This release should be a short-lived one. A new (slightly backwards-incompatible) version is in the works.
- Copyrights can now start with
©
in addition toCopyright
. The former is now recommended, but they are functionally similar.
- The source code of reuse is now formatted with black.
- The repository has been moved from https://git.fsfe.org/reuse/reuse to https://gitlab.com/reuse/reuse.
- Any files with the suffix
.spdx
are no longer considered licenses.
- The documentation now builds under Python 3.7.
- When using reuse from a child directory using pygit2, correctly find the root.
- The output of
reuse compile
is now deterministic. The files, copyright lines and SPDX expressions are sorted alphabetically.
- When a GPL license could not be found, the correct
-only
or-or-later
extension is now used in the warning message, rather than a bareGPL-3.0
. - If you have a license listed as
SPDX-Valid-License: GPL-3.0-or-later
, this now correctly matches corresponding SPDX identifiers. Still it is recommended to useSPDX-Valid-License: GPL-3.0
instead.
- Internationalisation support added. Initial support for:
- English.
- Dutch.
- Esperanto.
- Spanish.
- The license list of SPDX 3.0 has deprecated
GPL-3.0
andGPL-3.0+
et al in favour ofGPL-3.0-only
andGPL-3.0-or-later
. The program has been amended to accommodate sufficiently for those licenses.
Project.reuse_info_of
now extracts, combines and returns information both from the file itself and from debian/copyright.ReuseInfo
now holds sets instead of lists.- As a result of this,
ReuseInfo
will not hold duplicates of copyright lines or SPDX expressions.
- As a result of this,
- click removed as dependency. Good old argparse from the library is used instead.
- The
reuse --help
text has been tidied up a little bit.
- Release date in change log fixed.
- The PyPI homepage now gets reStructuredText instead of Markdown.
- Successfully parse old-style C and HTML comments now.
- Added
reuse compile
, which creates an SPDX bill of materials. - Added
--ignore-missing
toreuse lint
. - Allow to specify multiple paths to
reuse lint
. chardet
added as dependency.pygit2
added as soft dependency. reuse remains usable without it, but the performance withpygit2
is significantly better. Becausepygit2
has a non-Python dependency (libgit2
), it must be installed independently by the user. In the future, when reuse is packaged natively, this will not be an issue.
- Updated to version 2.0 of the REUSE recommendations. The
most important change is that
License-Filename
is no longer used. Instead, the filename is deducted fromSPDX-License-Identifier
. This change is NOT backwards compatible. - The conditions for linting have changed. A file is now non-compliant
when:
- The license associated with the file could not be found.
- There is no SPDX expression associated with the file.
- There is no copyright notice associated with the file.
- Only read the first 4 KiB (by default) from code files rather than the entire file when searching for SPDX tags. This speeds up the tool a bit.
Project.reuse_info_of
no longer raises an exception. Instead, it returns an emptyReuseInfo
object when no reuse information is found.- Logging is a lot prettier now. Only output entries from the
reuse
module.
reuse --ignore-debian compile
now works as expected.- The tool no longer breaks when reading a file that has a non-UTF-8
encoding. Instead,
chardet
is used to detect the encoding before reading the file. If a file still has errors during decoding, those errors are silently ignored and replaced.
- Removed dependency on
os.PathLike
so that Python 3.5 is actually supported
- Fixed the link to PyPI in the README.
This is a very early development release aimed at distributing the program as soon as possible. Because this is the first release, the changelog is a little empty beyond "created the program".
The program can do roughly the following:
- Detect the license of a given file through one of three methods (in
order of precedence):
- Information embedded in the .license file.
- Information embedded in its header.
- Information from the global debian/copyright file.
- Find and report all files in a project tree of which the license could not be found.
- Ignore files ignored by Git.
- Do some logging into STDERR.