Usage

The overview documents some basic usage on how to use this tool. It is highly recommended to read the overview first, and you might not even need to read this chapter. This chapter covers details that might not be immediately obvious when using the tool. This chapter does not cover everything, assuming that the user is helped enough by reuse --help and reuse <subcommand> --help.

Implementation details

This section covers implementation details that are true for the entire tool.

When searching for copyright and licensing tags inside of files, the tool does not strictly limit itself to the header comment as prescribed by the specification. It searches the first 4 kibibytes of the file. This makes sure that the tool can parse any type of plain-text file, even if the comment style is not recognised.

If a file is found to have an unparseable tag, that file is not parsed at all. This is a bug.

The tool does not verify the correctness of copyright notices. If it finds any line containing ‘©’, ‘Copyright’, or ‘SPDX-FileCopyrightText:’, then the tag and everything following it is considered a valid copyright notice, even if the copyright notice is not compliant with the specification.

When running the tool, the root of the project is automatically found if the working directory is inside a VCS repository. Otherwise, it treats the working directory as the root of the project. You can override the root of the project with the --root optional argument.

Git submodules are automatically ignored unless --include-submodules is passed as optional argument.

Symbolically links and files that are zero-sized are automatically ignored.

addheader

addheader makes it possible to semi-automatically add copyright and licensing information into the header of a file. This is useful especially in scenarios where you want to add a copyright holder or license to a lot of files without having to manually edit the header of each file.

Warning

You should be cautious with using addheader in automated processes. While nothing is stopping you from using it in your release script, you should make sure that the information it adds is actually reflective of reality. This is best verified manually.

The basic usage is reuse addheader --copyright="Jane Doe" --license=MIT my_file.py. This will add the following header to the file (assuming that the current year is 2019):

# SPDX-FileCopyrightText: 2019 Jane Doe
#
# SPDX-License-Identifier: MIT

You can use as many --copyright and --license arguments, so long as there is at least one such argument.

The REUSE header is placed at the very top of the file. If a different REUSE header already existed—at the top or elsewhere—its tags are copied, and the header is replaced in-place.

With the argument --copyright-style it is posible to change the default SPDX-FileCopyrightText to one of the following style:

spdx:           SPDX-FileCopyrightText: <year> <statement>
spdx-symbol:    SPDX-FileCopyrightText: © <year> <statement>
string:         Copyright <year> <statement>
string-c:       Copyright (C) <year> <statement>
string-symbol:  Copyright © <year> <statement>
symbol:         © <year> <statement>

Shebangs are always preserved at the top of the file.

Comment styles

The tool normally tries to auto-detect the comment style to use from the file extension of a file, and use that comment style. If the tool is unable to detect the comment style, or if it detects the wrong style, you can override the style using --style. The supported styles are:

  • AppleScript

  • ASPX

  • BibTex

  • C

  • CSS

  • Haskell

  • HTML

  • Jinja

  • JSX

  • ML

  • Python

  • TeX

If your comment style is not supported or a file extension is not correctly detected, please open an issue.

Normally, the tool uses a single-line comment style when it is available (e.g., // is used instead of /* */ for C comment styles). If no single-line comment style is available, a multi-line style is used. You can force a certain comment style using –single-line or –multi-line.

Templates

When the tool adds a header to a file, it normally first lists all copyright statements alphabetically, adds a single empty line, and then lists all SPDX License Expressions alphabetically. That is all that the header contains. It is possible to change this behaviour, and use a custom type of header that contains extra text. This is done through Jinja2 templates.

The default template is:

{% for copyright_line in copyright_lines %}
{{ copyright_line }}
{% endfor %}

{% for expression in spdx_expressions %}
SPDX-License-Identifier: {{ expression }}
{% endfor %}

Templates are automatically commented by the tool, depending on the detected or specified comment style.

You can create your own Jinja2 templates and place them in .reuse/templates/. If you create the template mytemplate.jinja2, you can use it with reuse addheader --copyright="Jane Doe" --template=mytemplate foo.py.

Inside of the template, you have access to the following variables:

  • copyright_lines — a list of copyright notices (string).

  • spdx_expressions — a list of SPDX License Expressions (string).

In the future, more variables will be added.

In some cases, you might want to do custom comment formatting. In those cases, you can pre-format your header as a comment. When doing so, suffix your template with .commented.jinja2.

An example of a custom template with manual commenting is:

/*
{% for copyright_line in copyright_lines %}
 * {{ copyright_line }}
{% endfor %}
{% if copyright_lines and spdx_expressions %}
 *
{% endif %}
{% for expression in spdx_expressions %}
 * SPDX-License-Identifier: {{ expression }}
{% endfor %}
{% if "GPL-3.0-or-later" in spdx_expressions %}
 *
 * This program is free software: you can redistribute it and/or modify it under
 * the terms of the GNU General Public License as published by the Free Software
 * Foundation, either version 3 of the License, or (at your option) any later
 * version.
 *
 * This program is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
 * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License along with
 * this program. If not, see <https://www.gnu.org/licenses/>.
{% endif %}
 */

lint

lint is the main component of the tool. Summarily, it verifies whether the project is compliant with the REUSE Specification. Its main goal is to find all files that do not have copyright and licensing information in their headers, but it also checks a few other things.

The STDOUT output of reuse lint is valid Markdown. Occasionally some logging will be printed to STDERR, which is not valid Markdown.

This is some example output of reuse lint:

# BAD LICENSES

'bad-license' found in:
* LICENSES/bad-license.txt


# UNUSED LICENSES

The following licenses are not used:
* bad-license


# MISSING COPYRIGHT AND LICENSING INFORMATION

The following files have no copyright and licensing information:
* no-information.txt


# SUMMARY

* Bad licenses: bad-license
* Deprecated licenses:
* Licenses without file extension:
* Missing licenses:
* Unused licenses: bad-license
* Used licenses: Apache-2.0, CC-BY-SA-4.0, CC0-1.0, GPL-3.0-or-later
* Read errors: 0
* Files with copyright information: 57 / 58
* Files with license information: 57 / 58

Unfortunately, your project is not compliant with version 3.0 of the REUSE Specification :-(

Criteria

These are the criteria that the linter checks against:

Bad licenses

Licenses that are found in LICENSES/ that are not found in the SPDX License List or do not start with LicenseRef- are bad licenses.

Deprecated licenses

If a license has an SPDX License Identifier that has been deprecated by SPDX, the license will show up here.

Licenses without file extension

These are licenses whose file names are a valid SPDX License Identifier, but which do not have a file extension.

Missing licenses

If a license is referred to in a comment header, but the license is not found in the LICENSES/ directory, then that license is missing.

Unused licenses

Conversely, if a license is found in the LICENSES/ directory but is not referred to in any comment header, then that license is unused.

Read errors

Not technically a criterion, but files that cannot be read by the operating system are read errors, and need to be fixed.