Merge pull request #288 from T145/develop

[FIX] More workflow tweaks

Author: T145

.github/workflows/publish_docker.yml

@@ -19,7 +19,7 @@ jobs:
       - name: Checkout repo
         uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8
       - name: Set up QEMU
-        uses: docker/setup-qemu-action@49b3bc8e6bdd4a60e6116a5414239cba5943d3cf
+        uses: docker/setup-qemu-action@c7c53464625b32c7a7e944ae62b3e17d2b600130
       - name: Set up Docker Buildx
         id: buildx
         uses: docker/setup-buildx-action@6524bf65af31da8d45b59e8c27de4bd072b392f5

.github/workflows/publish_lists.yml

@@ -66,6 +66,7 @@ jobs:
         if: always()
         run: rm -rf build/ && rm -f lists.tar
   upload_release:
+    if: github.ref == 'refs/heads/master'
     runs-on: ubuntu-22.04
     needs: build_lists
     continue-on-error: true
@@ -87,7 +88,7 @@ jobs:
           cd ..
       # https://github.com/softprops/action-gh-release
       - name: Create release
-        uses: softprops/action-gh-release@7b4da11513bf3f43f9999e90eabced41ab8bb048
+        uses: softprops/action-gh-release@a06a81a03ee405af7f2048a818ed3f03bbf83c7b
         with:
           files: build/*
           generate_release_notes: true
@@ -102,6 +103,7 @@ jobs:
         with:
           name: lists
   update_docs:
+    if: github.ref == 'refs/heads/master'
     runs-on: ubuntu-22.04
     needs: upload_release
     permissions: write-all
@@ -123,7 +125,7 @@ jobs:
       - name: Make the working tree safe
         run: git config --global --add safe.directory /__w/black-mirror/black-mirror
       - name: Update documents
-        uses: stefanzweifel/git-auto-commit-action@0b492c0d951b87f3cd12523a542dbd156c1dbc80
+        uses: stefanzweifel/git-auto-commit-action@fc84150d7c7500bbbaedb7ec522588fcc41f4d4d
         with:
           # homage to the python linting utility "black"
           commit_message: "ci(build): ✨🍰✨"

.github/workflows/snyk_container.yml

@@ -44,7 +44,7 @@ jobs:
             jq '.runs = [.runs[0]]' snyk.sarif > snyk.sarif.tmp && mv snyk.sarif.tmp snyk.sarif
           fi
       - name: Upload result to GitHub Code Scanning
-        uses: github/codeql-action/upload-sarif@v3
+        uses: github/codeql-action/upload-sarif@v4
         with:
           sarif_file: snyk.sarif
           category: snyk-container

.github/workflows/update_contributions.yml

@@ -3,7 +3,9 @@ name: Update Contributions
 
 on:
   push:
-    paths: ["data/contrib/*"]
+    branches: ["develop"]
+    paths:
+      - "data/contrib/*"
   workflow_dispatch:
 
 permissions: read-all
@@ -32,6 +34,6 @@ jobs:
       - name: Make commit directory safe
         run: git config --global --add safe.directory /__w/black-mirror/black-mirror
       - name: Commit changes
-        uses: stefanzweifel/git-auto-commit-action@0b492c0d951b87f3cd12523a542dbd156c1dbc80
+        uses: stefanzweifel/git-auto-commit-action@fc84150d7c7500bbbaedb7ec522588fcc41f4d4d
         with:
           commit_message: "ci(contrib): ✨📚✨"

.github/workflows/update_lists.yml

@@ -3,6 +3,7 @@ name: Update Lists
 
 on:
   push:
+    branches: ["develop"]
     paths:
       - "data/v2/manifest.json"
   workflow_dispatch:
@@ -41,6 +42,6 @@ jobs:
       - name: Make commit directory safe
         run: git config --global --add safe.directory /__w/black-mirror/black-mirror
       - name: Commit changes
-        uses: stefanzweifel/git-auto-commit-action@0b492c0d951b87f3cd12523a542dbd156c1dbc80
+        uses: stefanzweifel/git-auto-commit-action@fc84150d7c7500bbbaedb7ec522588fcc41f4d4d
         with:
           commit_message: "ci(lists): ✨📚✨"

.github/workflows/validate_manifest.yml

@@ -34,6 +34,10 @@ SHOW_ELAPSED_TIME: true
 FILEIO_REPORTER: false
 PRINT_ALPACA: false
 FLAVOR_SUGGESTIONS: false
+# schema validation
+PRE_COMMANDS:
+  - command: npx ajv-cli validate -s data/v2/manifest.schema.json -d data/v2/manifest.json
+    cwd: "workspace"
 # remove lingering docker containers when done
 POST_COMMANDS:
   - command: docker system prune -f

.mega-linter.yml

@@ -8378,7 +8378,7 @@
     "checksums": {},
     "content": {
       "filter": "NONE",
-      "retriever": "LYNX",
+      "retriever": "ARIA2",
       "type": "CSV"
     },
     "formats": [

data/v2/manifest.json

@@ -110,10 +110,10 @@ USER root
 WORKDIR /root
 
 LABEL maintainer="T145" \
-      version="7.1.1" \
+      version="7.1.3" \
       description="Runs the \"Black Mirror\" project! Check it out GitHub!" \
       org.opencontainers.image.created="2025-11-01" \
-      org.opencontainers.image.revision="7.1.1" \
+      org.opencontainers.image.revision="7.1.3" \
       org.opencontainers.image.source="https://github.com/T145/black-mirror" \
       org.opencontainers.image.url="https://github.com/T145/black-mirror" \
       org.opencontainers.image.vendor="T145" \
@@ -194,6 +194,7 @@ RUN apk add --no-cache --virtual .perl-build build-base curl gnutar xz perl-util
     make -j"$(nproc)"; \
     TEST_JOBS="$(nproc)" make test_harness || true; \
     make install; \
+    ln -sf /usr/local/bin/perl /usr/bin/perl; \
     apk del .perl-build
 
 WORKDIR /usr/src

docker/Dockerfile

@@ -212,8 +212,8 @@ process_list() {
 			'WATCHLIST_INTERNET') mlr --mmap --csv --ifs ';' -N cut -f 1 ;;
 			'CRUZ_IT') mlr --mmap --csv --headerless-csv-output clean-whitespace then cut -f ip_address ;;
 			# PhishTank often includes large hashes in its URLs, which may lead to parsing issues.
-			# Therefore use regex that ignores paths and queries after the domain.
-			'PHISHTANK') mlr --mmap --csv --headerless-csv-output --lazy-quotes put -S '$url =~ "https?://([^/]+)"; $Domain = "\1"' then cut -f Domain ;;
+			# Therefore use a robust regex that ignores paths and queries after the domain.
+			'PHISHTANK') mlr --mmap --csv --headerless-csv-output --lazy-quotes --skip-comments clean-whitespace then cut -f url | get_domains_from_urls ;;
 			'BLOCKLIST_UA') mlr --mmap --csv --headerless-csv-output --ifs ';' cut -f IP ;;
 			# The C2 feed has malformed CSVs.
 			'THREATVIEW_C2_HOSTS') mawk -F, '/^[^#]/{print $3}' ;;