github/kubernetes
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #122250 from thockin/fix_verify-licenses

Browse Source 
Fix verify-licenses.sh
This commit is contained in:
Kubernetes Prow Robot
2023-12-14 07:27:39 +01:00
committed by GitHub
parent ac425531b1 7dfe3d866b
commit 485e57c395
1 changed files with 58 additions and 77 deletions
Download Patch File Download Diff File Expand all files Collapse all files

93
hack/verify-licenses.sh
View File

@@ -26,26 +26,25 @@ KUBE_ROOT=$(dirname "${BASH_SOURCE[0]}")/..
source "${KUBE_ROOT}/hack/lib/init.sh"
source "${KUBE_ROOT}/hack/lib/util.sh"
# This sets up the environment, like GOCACHE, which keeps the worktree cleaner.
kube::golang::setup_env
kube::util::ensure-temp-dir
# Creating a new repository tree
# Deleting vendor directory to make go-licenses fetch license URLs from go-packages source repository
git worktree add -f "${KUBE_TEMP}"/tmp_test_licenses/kubernetes HEAD >/dev/null 2>&1 || true
cd "${KUBE_TEMP}"/tmp_test_licenses/kubernetes && rm -rf vendor
# Ensure that we find the binaries we build before anything else.
export GOBIN="${KUBE_OUTPUT_BINPATH}"
PATH="${GOBIN}:${PATH}"
# Explicitly opt into go modules, even though we're inside a GOPATH directory
export GO111MODULE=on
function http_code() {
curl -I -s -o /dev/null -w "%{http_code}" "$1"
}
allowed_licenses=()
packages_flagged=()
@@ -60,77 +59,59 @@ go install github.com/google/go-licenses@latest
# Refer: https://github.com/cncf/foundation/blob/main/allowed-third-party-license-policy.md
curl -s 'https://spdx.org/licenses/licenses.json' -o "${KUBE_TEMP}"/licenses.json
number_of_licenses=$(jq '.licenses | length' "${KUBE_TEMP}"/licenses.json)
loop_index_length=$(( number_of_licenses - 1 ))
echo '[INFO] Fetching current list of CNCF approved licenses...'
for index in $(seq 0 $loop_index_length);
do
licenseID=$(jq ".licenses[$index] .licenseId" "${KUBE_TEMP}"/licenses.json)
if [[ $(jq ".licenses[$index] .isDeprecatedLicenseId" "${KUBE_TEMP}"/licenses.json) == false ]]
then
allowed_licenses+=("${licenseID}")
fi
done
while read -r L; do
allowed_licenses+=("${L}")
done < <(jq -r '.licenses[] | select(.isDeprecatedLicenseId==false) .licenseId' "${KUBE_TEMP}"/licenses.json)
# Scanning go-packages under the project & verifying against the CNCF approved list of licenses
echo '[INFO] Starting license scan on go-packages...'
go-licenses report ./... >> "${KUBE_TEMP}"/licenses.csv
echo -e 'PACKAGE_NAME LICENSE_NAME LICENSE_URL\n' >> "${KUBE_TEMP}"/approved_licenses.dump
while IFS=, read -r GO_PACKAGE LICENSE_URL LICENSE_NAME
do
FORMATTED_LICENSE_URL=
if [[ " ${allowed_licenses[*]} " == *"${LICENSE_NAME}"* ]];
then
if [[ "${LICENSE_URL}" == 'Unknown' ]];
then
if [[ "${GO_PACKAGE}" != k8s.io/* ]];
then
while IFS=, read -r GO_PACKAGE LICENSE_URL LICENSE_NAME; do
if ! printf -- "%s\n" "${allowed_licenses[@]}" | grep -q "^${LICENSE_NAME}$"; then
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/notapproved_licenses.dump
packages_flagged+=("${GO_PACKAGE}")
continue
fi
if [[ "${LICENSE_URL}" == 'Unknown' ]]; then
if [[ "${GO_PACKAGE}" != k8s.io/* ]]; then
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/approved_licenses_with_missing_urls.dump
packages_url_missing+=("${GO_PACKAGE}")
else
LICENSE_URL='https://github.com/kubernetes/kubernetes/blob/master/LICENSE'
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/approved_licenses.dump
fi
elif curl -Is "${LICENSE_URL}" | head -1 | grep -q 404;
then
# Check whether the License URL is incorrectly formed
# TODO: Remove this workaround check once PR https://github.com/google/go-licenses/pull/110 is merged
IFS='/' read -r -a split_license_url <<< ${LICENSE_URL}
for part_of_url in "${split_license_url[@]}"
do
if [[ ${part_of_url} == '' ]]
then
continue
elif [[ ${part_of_url} == 'https:' ]]
then
FORMATTED_LICENSE_URL+='https://'
else
if [[ ${part_of_url} =~ ^v[0-9]+\.[0-9]+\.[0-9]+ ]]
then
FORMATTED_LICENSE_URL+="${part_of_url}/${split_license_url[-1]}"
fi
if [[ "$(http_code "${LICENSE_URL}")" != 404 ]]; then
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/approved_licenses.dump
continue
fi
# The URL 404'ed. Try parent-paths.
#echo -e "DBG: err 404 ${LICENSE_URL}"
dir="$(dirname "${LICENSE_URL}")"
file="$(basename "${LICENSE_URL}")"
while [[ "${dir}" != "." ]]; do
dir="$(dirname "${dir}")"
#echo "DBG: try ${dir}/${file}"
if [[ "$(http_code "${dir}/${file}")" != 404 ]]; then
#echo "DBG: it worked"
echo "${GO_PACKAGE} ${LICENSE_NAME} ${dir}/${file}" >> "${KUBE_TEMP}"/approved_licenses.dump
break
else
FORMATTED_LICENSE_URL+="${part_of_url}/"
fi
fi
#echo "DBG: still 404"
done
if curl -Is "${FORMATTED_LICENSE_URL}" | head -1 | grep -q 404;
then
if [[ "${dir}" == "." ]];then
#echo "DBG: failed to find a license"
packages_url_missing+=("${GO_PACKAGE}")
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/approved_licenses_with_missing_urls.dump
else
echo "${GO_PACKAGE} ${LICENSE_NAME} ${FORMATTED_LICENSE_URL}" >> "${KUBE_TEMP}"/approved_licenses.dump
fi
else
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/approved_licenses.dump
fi
else
echo "${GO_PACKAGE} ${LICENSE_NAME} ${LICENSE_URL}" >> "${KUBE_TEMP}"/notapproved_licenses.dump
packages_flagged+=("${GO_PACKAGE}")
fi
done < "${KUBE_TEMP}"/licenses.csv
awk '{ printf "%-100s : %-20s : %s\n", $1, $2, $3 }' "${KUBE_TEMP}"/approved_licenses.dump