Skip to content

Latest commit

 

History

History
65 lines (45 loc) · 2.38 KB

index.rst

File metadata and controls

65 lines (45 loc) · 2.38 KB
Raw

ament_copyright

Checks C / C++ / CMake / Python source files for the existence of a copyright and license notice. Files with the following extensions are being considered: .c, .cc, .cpp, .cxx, .h, .hh, .hpp, .hxx, .cmake, .py.

When being searched for recursively the following directories and files are being excluded:

  • directories starting with . (e.g. .git) or _ (e.g. __pycache__)
  • setup.py when being a sibling of package.xml

Additionally it checks if the root of a repository contains a LICENSE and CONTRIBUTING.md file.

How to run the check from the command line?

ament_copyright [<path> ...]

When using the option --list-copyright-names a list of known copyright holders is shown. The option --list-licenses outputs a list of known licenses.

When using the option --add-missing <KNOWN_COPYRIGHT_NAME | "Copyright holder string"> <KNOWN_LICENSE> a copyright notice and license is added to all files which lack one. The first argument can either be a name from the list returned by --list-copyright-names or a custom string. The second argument must be a name from the list returned by --list-licenses.

When using the option --add-copyright-year existing copyright notices are being updated to include the current year.

How to run the check from within a CMake ament package as part of the tests?

The CMake integration is provided by the package ament_cmake_copyright.

How to add support for more licenses or copyright holders?

The package uses Python entry points to get all list of known licenses and copyright holder. You can implement a custom package and contribute more implementations to these entry points or extend this package with more licenses.

Why are my existing copyright / license notices not detected?

This script currently only checks line comments (lines starting with # / // depending on the language). Block comments / C-style comment (starting with /*) are not being detected to keep the complexity minimal. Also the content must match the templates exactly.