sast-unicode-check: add version 0.1 task

Resolves: https://issues.redhat.com/browse/OSH-739

Signed-off-by: Chuntao Han <[email protected]>

.tekton/pull-request.yaml

@@ -5,7 +5,7 @@ metadata:
   name: build-definitions-pull-request
   annotations:
     pipelinesascode.tekton.dev/on-cel-expression: (event == "pull_request" && target_branch == "main" && ( !has(body.pull_request) || !body.pull_request.draft) ) || (event == "push" && target_branch.startsWith("gh-readonly-queue/main/"))
-    pipelinesascode.tekton.dev/task: "[task/git-clone/0.1/git-clone.yaml, .tekton/tasks/buildah.yaml, .tekton/tasks/task-lint.yaml, .tekton/tasks/e2e-test.yaml, task/sast-snyk-check/0.2/sast-snyk-check.yaml]"
+    pipelinesascode.tekton.dev/task: "[task/git-clone/0.1/git-clone.yaml, .tekton/tasks/buildah.yaml, .tekton/tasks/task-lint.yaml, .tekton/tasks/e2e-test.yaml, task/sast-snyk-check/0.2/sast-snyk-check.yaml, task/sast-unicode-check.yaml/0.1/sast-unicode-check.yaml]"
     pipelinesascode.tekton.dev/max-keep-runs: "5"
 spec:
   params:
@@ -52,6 +52,14 @@ spec:
         workspaces:
           - name: workspace
             workspace: workspace
+      - name: sast-unicode-check
+        runAfter:
+          - build-image-index
+        taskRef:
+          name: sast-unicode-check
+        workspaces:
+          - name: workspace
+            workspace: workspace
       - name: build-container
         runAfter:
           - task-lint-check

task/sast-unicode-check/0.1/README.md

@@ -0,0 +1,30 @@
+# sast-unicode-check task
+
+## Description:
+
+The sast-unicode-check task uses [find-unicode-control](https://github.com/siddhesh/find-unicode-control.git) tool to perform Static Application Security Testing (SAST) to look for non-printable unicode characters in all text files in a source tree.
+
+## Parameters:
+
+| name                         | description                                                                                                                                   |
+|------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------|
+| FIND_UNICODE_CONTROL_GIT_URL | URL from repository to find unicode control.                                                                                                  |
+| FIND_UNICODE_CONTROL_ARGS    | arguments for find-unicode-control command.                                                                                                   |
+| KFP_GIT_URL                  | Known False Positives git URL, optionally taking a revision delimited by #; If empty, filtering of known false positives is disabled.         |
+| PROJECT_NVR                  | Name-Version-Release (NVR) of the scanned project. It is used to find path exclusions (it is optional).                                       |
+| RECORD_EXCLUDED              | Whether to record the excluded findings (defaults to false). If `true`, the the excluded findings will be stored in `excluded-findings.json`. |
+
+## Results:
+
+| name          | description                              |
+|---------------|------------------------------------------|
+| TEST_OUTPUT   | Tekton task test output.                 |
+
+## Source repository for image:
+
+https://github.com/konflux-ci/konflux-test
+
+## Additional links:
+
+* https://github.com/siddhesh/find-unicode-control.git
+* https://gitlab.cee.redhat.com/osh/known-false-positives.git

task/sast-unicode-check/0.1/sast-unicode-check.yaml

@@ -0,0 +1,282 @@
+apiVersion: tekton.dev/v1
+kind: Task
+metadata:
+  labels:
+    app.kubernetes.io/version: "0.1"
+  annotations:
+    tekton.dev/pipelines.minVersion: "0.12.1"
+    tekton.dev/tags: "konflux"
+  name: sast-unicode-check
+spec:
+  description: >-
+    Scans source code for non-printable unicode characters in all text files.
+  results:
+    - description: Tekton task test output.
+      name: TEST_OUTPUT
+  params:
+    - name: image-url
+      type: string
+      description: Image URL.
+      default: ""
+    - name: image-digest
+      type: string
+      description: Image digest to report findings for.
+      default: ""
+    - name: FIND_UNICODE_CONTROL_GIT_URL
+      type: string
+      description: URL from repository to find unicode control.
+      default: "https://github.com/siddhesh/find-unicode-control.git#c2accbfbba7553a8bc1ebd97089ae08ad8347e58"
+    - name: FIND_UNICODE_CONTROL_ARGS
+      type: string
+      description: arguments for find-unicode-control command.
+      default: "-p bidi -v -d -t"
+    - name: KFP_GIT_URL
+      type: string
+      description: URL from repository to download known false positives files.
+      # FIXME: Red Hat internal projects will default to https://gitlab.cee.redhat.com/osh/known-false-positives.git when KONFLUX-4530 is resolved
+      default: ""
+    - name: PROJECT_NVR
+      type: string
+      default: ""
+      description: |
+        Name-Version-Release (NVR) of the scanned project.
+        It is used to find path exclusions (it is optional).
+    - name: RECORD_EXCLUDED
+      type: string
+      default: "false"
+      description: |
+        Whether to record the excluded findings (defaults to false).
+        If `true`, the the excluded findings will be stored in `excluded-findings.json`.
+    - name: caTrustConfigMapName
+      type: string
+      description: The name of the ConfigMap to read CA bundle data from.
+      default: trusted-ca
+    - name: caTrustConfigMapKey
+      type: string
+      description: The name of the key in the ConfigMap that contains the CA bundle data.
+      default: ca-bundle.crt
+  volumes:
+    - name: trusted-ca
+      configMap:
+        name: $(params.caTrustConfigMapName)
+        items:
+          - key: $(params.caTrustConfigMapKey)
+            path: ca-bundle.crt
+        optional: true
+  steps:
+    - name: sast-unicode-check
+      image: quay.io/redhat-appstudio/konflux-test:v1.4.6@sha256:5f298d8d990dfa82023e50029b71b08e19c3c9cedb181dfc4bc86c9ecad8700c
+      # per https://kubernetes.io/docs/concepts/containers/images/#imagepullpolicy-defaulting
+      # the cluster will set imagePullPolicy to IfNotPresent
+      workingDir: $(workspaces.workspace.path)/hacbs/$(context.task.name)
+      volumeMounts:
+      env:
+        - name: KFP_GIT_URL
+          value: $(params.KFP_GIT_URL)
+        - name: FIND_UNICODE_CONTROL_GIT_URL
+          value: $(params.FIND_UNICODE_CONTROL_GIT_URL)
+        - name: FIND_UNICODE_CONTROL_ARGS
+          value: $(params.FIND_UNICODE_CONTROL_ARGS)
+        - name: PROJECT_NVR
+          value: $(params.PROJECT_NVR)
+        - name: RECORD_EXCLUDED
+          value: $(params.RECORD_EXCLUDED)
+        - name: SOURCE_CODE_DIR
+          value: $(workspaces.workspace.path)
+      script: |
+        #!/usr/bin/env bash
+        set -exuo pipefail
+        . /utils.sh
+        trap 'handle_error $(results.TEST_OUTPUT.path)' EXIT
+
+        SCAN_PROP=""
+
+        ca_bundle=/mnt/trusted-ca/ca-bundle.crt
+        if [ -f "$ca_bundle" ]; then
+          echo "INFO: Using mounted CA bundle: $ca_bundle"
+          cp -vf $ca_bundle /etc/pki/ca-trust/source/anchors
+          update-ca-trust
+        fi
+
+        # Clone the source code from upstream repo
+        clone_unicontrol()
+        {
+            GIT_URL=$(echo "${FIND_UNICODE_CONTROL_GIT_URL}" | awk -F'#' '{print $1}')
+            REV=$(echo "${FIND_UNICODE_CONTROL_GIT_URL}" | awk -F'#' '{print $2}')
+
+            # Clone find-unicode-control repository
+            git clone "${GIT_URL}" find-unicode-control >&2
+            status=$?
+            if [ "${status}" -ne 0 ]; then
+                echo "Error: Failed to clone the repository: ${GIT_URL}" >&2
+                return 1
+            else
+                echo "Message: Succeed to clone the repository: ${GIT_URL}" >&2
+            fi
+
+            if [[ -n "${REV}" ]]; then
+                # Get git url suffix
+                git_url_suffix=$(git -C ./find-unicode-control/ rev-parse HEAD)
+                if [[ "${git_url_suffix}" != "${REV}" ]]; then
+                    git -C ./find-unicode-control/ checkout --quiet "${REV}"
+                fi
+                SCAN_PROP="find-unicode-control-git-url:${FIND_UNICODE_CONTROL_GIT_URL}"
+            else
+                SCAN_PROP="find-unicode-control-git-url:${FIND_UNICODE_CONTROL_GIT_URL}#${git_url_suffix}"
+            fi
+
+            # Check usage of find-unicode-control to confirm the tool can work
+            ./find-unicode-control/find_unicode_control.py --help >&2
+            status=$?
+            if [ "${status}" -ne 0 ]; then
+                echo "Error: Failed to check usage of find-unicode-control" >&2
+                return 1
+            else
+                echo "Message: Succeed to check usage of find-unicode-control" >&2
+            fi
+        }
+
+        # Find unicode control
+        unicontrol_scan()
+        {
+            # Split the arguments into an array
+            args=("${FIND_UNICODE_CONTROL_ARGS}")
+
+            LANG=en_US.utf8 ./find-unicode-control/find_unicode_control.py "${args[@]}" "${SOURCE_CODE_DIR}/source" \
+                >raw_sast_unicode_check_out.txt \
+                2>raw_sast_unicode_check_out.log \
+                || FUC_EXIT_CODE=$?
+
+            if [[ -z "${FUC_EXIT_CODE}" || "${FUC_EXIT_CODE}" -eq 1 ]]; then
+                # No non-printable unicode characters found, or successfully found
+                echo "Message: Succeed to run find-unicode-control command" >&2
+            else
+                # Failed to run the command
+                echo "Error: failed to run find unicode control command" >&2
+                return 1
+            fi
+
+            # Translate the output format
+            sed -i raw_sast_unicode_check_out.txt -E -e 's|(.*:[0-9]+)(.*)|\1: warning:\2|' -e 's|^|Error: UNICONTROL_WARNING:\n|'
+            status=$?
+            if [ "${status}" -ne 0 ] && [ "${status}" -ne 1 ]; then
+                echo "Error: failed to translate the unicontrol output format" >&2
+                return 1
+            else
+                echo "Message: Succeed to translate the unicontrol output format" >&2
+            fi
+        }
+
+        # Process all results as configured with CSGERP_OPTS
+        process_all_results()
+        {
+            CSGERP_OPTS=(
+                --mode=json
+                --remove-duplicates
+                --embed-context=3
+                --set-scan-prop="${SCAN_PROP}"
+                --strip-path-prefix="${SOURCE_CODE_DIR}"/source/
+            )
+            # In order to generate csdiff/v1, we need to add the whole path of the source code as sast-unicode-check only provides an URI to embed the context
+            csgrep "${CSGERP_OPTS[@]}" raw_sast_unicode_check_out.txt > processed_sast_unicode_check_out.json
+
+            csgrep --mode=evtstat processed_sast_unicode_check_out.json
+        }
+
+        # Filter known false positive
+        filter_kfp()
+        {
+            echo "Message: Filtering false positives in results files using csfilter-kfp..." >&2
+            CMD=(
+                csfilter-kfp
+                --verbose 
+                --kfp-git-url="${KFP_GIT_URL}"
+            )
+
+            if [[ -n "${PROJECT_NVR}" ]] && [[ "${RECORD_EXCLUDED}" == "true" ]]; then
+                CMD+=(
+                    --project-nvr="${PROJECT_NVR}"
+                    --record-excluded=raw-excluded-findings.json
+                )
+            fi
+
+            "${CMD[@]}" processed_sast_unicode_check_out.json  > sast_unicode_check_out.json
+            status=$?
+            if [ "$status" -ne 0 ]; then
+                echo "Error: failed to filter known false positives" >&2
+                return 1
+            else
+                echo "Message: Succeed to filter known false positives" >&2
+            fi
+
+            if [ -f "raw-excluded-findings.json" ]; then
+                csgrep --mode=json --strip-path-prefix="${SOURCE_CODE_DIR}"/source/ raw-excluded-findings.json > excluded-findings.json
+            fi
+        }
+
+        generate_sarif_report()
+        {
+            csgrep --mode=sarif sast_unicode_check_out.json > sast_unicode_check_out.sarif
+            if [[ -z "${FUC_EXIT_CODE}" ]]; then
+                note="Task $(context.task.name) success: No finding was detected"
+                ERROR_OUTPUT=$(make_result_json -r SUCCESS -t "$note")
+            elif [[ "${FUC_EXIT_CODE}" -eq 1 ]] && [[ ! -s  sast_unicode_check_out.sarif ]]; then
+                note="Task $(context.task.name) success: Some findings were detected, but filtered by known false positive"
+                ERROR_OUTPUT=$(make_result_json -r SUCCESS -t "$note")
+            else
+                echo "sast-unicode-check test failed because of the following issues:"
+                cat processed_sast_unicode_check_out.json
+                TEST_OUTPUT=
+                parse_test_output $(context.task.name) sarif sast_unicode_check_out.sarif  || true
+                note="Task $(context.task.name) failed: For details, check Tekton task log."
+                ERROR_OUTPUT=$(make_result_json -r ERROR -t "$note")
+            fi
+            echo "${TEST_OUTPUT:-${ERROR_OUTPUT}}" | tee $(results.TEST_OUTPUT.path)
+        }
+
+        main() 
+        {
+            clone_unicontrol
+            unicontrol_scan
+            process_all_results
+            filter_kfp
+            generate_sarif_report
+        }
+
+        main
+    - name: upload
+      image: quay.io/konflux-ci/oras:latest@sha256:f4b891ee3038a5f13cd92ff4f473faad5601c2434d1c6b9bccdfc134d9d5f820
+      workingDir: $(workspaces.workspace.path)/hacbs/$(context.task.name)
+      env:
+        - name: IMAGE_URL
+          value: $(params.image-url)
+        - name: IMAGE_DIGEST
+          value: $(params.image-digest)
+      script: |
+        #!/usr/bin/env bash
+
+        if [ -z "${IMAGE_URL}" ] || [ -z "${IMAGE_DIGEST}"  ]; then
+          echo 'No image-url or image-digest param provided. Skipping upload.'
+          exit 0;
+        fi
+
+        UPLOAD_FILES="sast_unicode_check_out.sarif excluded-findings.json"
+        for UPLOAD_FILE in ${UPLOAD_FILES}; do
+            if [ ! -f "${UPLOAD_FILE}" ]; then
+              echo "No ${UPLOAD_FILE} exists. Skipping upload."
+              continue;
+            fi
+
+            if [ "${UPLOAD_FILES}" == "excluded-findings.json" ]; then
+                MEDIA_TYPE=application/json
+            else
+                MEDIA_TYPE=application/sarif+json
+            fi
+
+            echo "Selecting auth"
+            select-oci-auth "${IMAGE_URL}" > "${HOME}/auth.json"
+            echo "Attaching to ${IMAGE_URL}"
+            oras attach --no-tty --registry-config "$HOME/auth.json" --artifact-type "${MEDIA_TYPE}" "${IMAGE_URL}" "${UPLOAD_FILE}:${MEDIA_TYPE}"
+        done
+  workspaces:
+  - name: workspace

task/sast-unicode-check/OWNERS

@@ -0,0 +1,5 @@
+# See the OWNERS docs: https://go.k8s.io/owners
+approvers:
+  - integration-team
+reviewers:
+  - integration-team