ci: Add a CI check for REUSE-compliant licensing/copyright headers

This doesn’t enforce licensing/copyright headers to be present on all
files, but does check that at least a minimum number of files are
correct.

This should help avoid new files being added without appropriate
licensing information in future.

The baseline is set at what `reuse lint` outputs for me at the moment.

See https://reuse.software/tutorial/#step-2 for information about how to
add REUSE-compliant licensing/copyright to files.

Signed-off-by: Philip Withnall <pwithnall@endlessos.org>

Helps: #1415

.gitlab-ci.yml

@@ -13,7 +13,7 @@ cache:
 variables:
   FEDORA_IMAGE: "registry.gitlab.gnome.org/gnome/glib/fedora:v19"
   COVERITY_IMAGE: "registry.gitlab.gnome.org/gnome/glib/coverity:v7"
-  DEBIAN_IMAGE: "registry.gitlab.gnome.org/gnome/glib/debian-stable:v13"
+  DEBIAN_IMAGE: "registry.gitlab.gnome.org/gnome/glib/debian-stable:v14"
   MINGW_IMAGE: "registry.gitlab.gnome.org/gnome/glib/mingw:v9"
   MESON_TEST_TIMEOUT_MULTIPLIER: 4
   G_MESSAGES_DEBUG: all
@@ -72,6 +72,17 @@ sh-and-py-check:
       - "**/*.py"
       - "**/*.sh"
 
+style-check-mandatory:
+  extends: .only-default
+  image: $DEBIAN_IMAGE
+  stage: style-check
+  allow_failure: false
+  script:
+    - .gitlab-ci/run-reuse.sh
+  variables:
+    # The submodules are needed for `reuse lint` to work properly
+    GIT_SUBMODULE_STRATEGY: "normal"
+
 fedora-x86_64:
   extends:
     - .build-linux

.gitlab-ci/debian-stable.Dockerfile

@@ -68,6 +68,9 @@ ENV LANG=C.UTF-8 LANGUAGE=C.UTF-8 LC_ALL=C.UTF-8
 
 RUN pip3 install meson==0.60.3
 
+# FIXME: Once we use Debian Bookworm, we can just install the `reuse` package
+RUN pip3 install reuse==1.0.0
+
 ARG HOST_USER_ID=5555
 ENV HOST_USER_ID ${HOST_USER_ID}
 RUN useradd -u $HOST_USER_ID -ms /bin/bash user

.gitlab-ci/run-reuse.sh

@@ -0,0 +1,62 @@
+#!/bin/bash
+#
+# Copyright 2022 Endless OS Foundation, LLC
+#
+# SPDX-License-Identifier: LGPL-2.1-or-later
+#
+# Original author: Philip Withnall
+
+set -e
+
+# We need to make sure the submodules are up to date, or `reuse lint` will fail
+# when it tries to run `git status` internally
+git submodule update --init
+
+# Run `reuse lint` on the code base and see if the number of files without
+# suitable copyright/licensing information has increased from a baseline
+# FIXME: Eventually this script can check whether *any* files are missing
+# information. But for now, let’s slowly improve the baseline.
+files_without_copyright_information_max=407
+files_without_license_information_max=559
+
+# The || true is because `reuse lint` will exit with status 1 if the project is not compliant
+# FIXME: Once https://github.com/fsfe/reuse-tool/issues/512 or
+# https://github.com/fsfe/reuse-tool/issues/183 land, we can check only files
+# which have changed in this merge request, and confidently parse structured
+# output rather than the current human-readable output.
+lint_output="$(reuse lint || true)"
+
+files_with_copyright_information="$(echo "${lint_output}" | awk '/^\* Files with copyright information: / { print $6 }')"
+files_with_license_information="$(echo "${lint_output}" | awk '/^\* Files with license information: / { print $6 }')"
+total_files="$(echo "${lint_output}" | awk '/^\* Files with copyright information: / { print $8 }')"
+error=0
+
+files_without_copyright_information="$(( total_files - files_with_copyright_information ))"
+files_without_license_information="$(( total_files - files_with_license_information ))"
+
+if [ "${files_without_copyright_information}" -gt "${files_without_copyright_information_max}" ] || \
+   [ "${files_without_license_information}" -gt "${files_without_license_information_max}" ]; then
+  echo "${lint_output}"
+fi
+
+if [ "${files_without_copyright_information}" -gt "${files_without_copyright_information_max}" ]; then
+  echo ""
+  echo "Error: New files added without REUSE-compliant copyright information"
+  echo "Please make sure that all files added in this branch/merge request have correct copyright information"
+  error=1
+fi
+
+if [ "${files_without_license_information}" -gt "${files_without_license_information_max}" ]; then
+  echo ""
+  echo "Error: New files added without REUSE-compliant licensing information"
+  echo "Please make sure that all files added in this branch/merge request have correct license information"
+  error=1
+fi
+
+if [ "${error}" -eq "1" ]; then
+  echo ""
+  echo "See https://reuse.software/tutorial/#step-2 for information on how to add REUSE information"
+  echo "Also see https://gitlab.gnome.org/GNOME/glib/-/issues/1415"
+fi
+
+exit "${error}"