git pages bug fix

git-pages/files/retention-cleanup.sh

@@ -176,46 +176,90 @@ if [ "${#TO_DELETE[@]}" -eq 0 ]; then
 fi
 
 echo ""
-echo "=== Phase 4: whiteout deletion ==="
-echo "Creating whiteout tar for ${#TO_DELETE[@]} report(s)..."
+echo "=== Phase 4: full site rebuild ==="
+echo "Rebuilding site (${#TO_DELETE[@]} report(s) to delete)..."
 
-WHITEOUT_TAR=$(mktemp)
-trap 'rm -f "$WHITEOUT_TAR"' EXIT
+ARCHIVE_FILE=$(mktemp)
+SITE_DIR=$(mktemp -d)
+NEW_TAR=$(mktemp)
+cleanup_phase4() {
+  rm -f "$ARCHIVE_FILE" "$NEW_TAR"
+  rm -rf "$SITE_DIR"
+}
+trap cleanup_phase4 EXIT
 
-python3 -c "
-import tarfile, sys
+# Try archive.tar first
+echo "Downloading archive.tar..."
+HTTP_CODE=$(curl_with_host -o "$ARCHIVE_FILE" -w "%{http_code}" -sS "${PAGES_URL}/.git-pages/archive.tar")
 
-tar = tarfile.open(name='${WHITEOUT_TAR}', mode='w')
+if [ "$HTTP_CODE" = "200" ] && tar -tf "$ARCHIVE_FILE" >/dev/null 2>&1; then
+  echo "Extracting archive..."
+  tar -xf "$ARCHIVE_FILE" -C "$SITE_DIR"
 
-dirs = set()
-for d in sys.argv[1:]:
-    dirs.add(d.strip())
+  for dir in "${TO_DELETE[@]}"; do
+    if [ -d "$SITE_DIR/$dir" ]; then
+      echo "  Removing: $dir"
+      rm -rf "$SITE_DIR/$dir"
+    fi
+  done
+else
+  echo "archive.tar failed (HTTP ${HTTP_CODE}) - falling back to manifest-based rebuild"
 
-tarinfo = tarfile.TarInfo()
-tarinfo.type = tarfile.CHRTYPE
-tarinfo.devmajor = 0
-tarinfo.devminor = 0
+  ALL_PATHS=$(echo "$MANIFEST" | jq -r '.contents | keys[]' 2>/dev/null || true)
 
-for d in sorted(dirs, key=len, reverse=True):
-    info = tarinfo
-    info.name = d
-    tar.addfile(info)
+  if [ -z "$ALL_PATHS" ]; then
+    echo "ERROR: no files in manifest - cannot rebuild" >&2
+    exit 1
+  fi
 
-tar.close()
-" "${TO_DELETE[@]}"
+  EXCLUDE_GREP=""
+  for dir in "${TO_DELETE[@]}"; do
+    EXCLUDE_GREP="${EXCLUDE_GREP}${EXCLUDE_GREP:+|}^${dir}/"
+  done
 
-echo "Patching ${PAGES_URL}/ with whiteout tar..."
-HTTP_CODE=$(curl_with_host -X PATCH "${PAGES_URL}/" \
+  if [ -n "$EXCLUDE_GREP" ]; then
+    KEEP_PATHS=$(echo "$ALL_PATHS" | grep -v -E "$EXCLUDE_GREP" || true)
+  else
+    KEEP_PATHS="$ALL_PATHS"
+  fi
+
+  if [ -z "$KEEP_PATHS" ]; then
+    echo "No files to keep - site will be empty"
+    mkdir -p "$SITE_DIR/__placeholder__"
+    echo "placeholder" > "$SITE_DIR/__placeholder__/index.html"
+  else
+    FILE_COUNT=$(echo "$KEEP_PATHS" | wc -l | tr -d ' ')
+    echo "Downloading ${FILE_COUNT} file(s)..."
+    while IFS= read -r path; do
+      [ -z "$path" ] && continue
+      dir=$(dirname "$SITE_DIR/$path")
+      mkdir -p "$dir"
+      curl_with_host -o "$SITE_DIR/$path" -sS "${PAGES_URL}/${path}" || {
+        echo "  WARN: failed to download ${path}"
+      }
+    done <<< "$KEEP_PATHS"
+  fi
+fi
+
+if [ -z "$(ls -A "$SITE_DIR" 2>/dev/null)" ]; then
+  echo "Site is empty - creating placeholder"
+  mkdir -p "$SITE_DIR/__placeholder__"
+  echo "placeholder" > "$SITE_DIR/__placeholder__/index.html"
+fi
+
+tar -cf "$NEW_TAR" -C "$SITE_DIR" .
+
+echo "PUT: replacing site contents..."
+HTTP_CODE=$(curl_with_host -X PUT "${PAGES_URL}/" \
   -H "Content-Type: application/x-tar" \
-  -H "Atomic: no" \
-  --data-binary @"${WHITEOUT_TAR}" \
+  --data-binary @"${NEW_TAR}" \
   -w "%{http_code}" \
   -o /dev/null)
 
-echo "HTTP $HTTP_CODE"
-if [ "$HTTP_CODE" = "200" ] || [ "$HTTP_CODE" = "204" ]; then
-  echo "Retention cleanup finished."
+echo "HTTP ${HTTP_CODE}"
+if [ "$HTTP_CODE" = "200" ] || [ "$HTTP_CODE" = "201" ] || [ "$HTTP_CODE" = "204" ]; then
+  echo "Site rebuild completed."
 else
-  echo "ERROR: retention HTTP ${HTTP_CODE}" >&2
+  echo "ERROR: PUT HTTP ${HTTP_CODE}" >&2
   exit 1
 fi

git-pages/templates/init-job.yaml

@@ -6,7 +6,7 @@ metadata:
   labels:
     {{- include "git-pages.componentLabels" . | nindent 4 }}
   annotations:
-    "helm.sh/hook": post-install, post-upgrade
+    "helm.sh/hook": post-install
     "helm.sh/hook-delete-policy": hook-succeeded,before-hook-creation
 spec:
   backoffLimit: 5
@@ -32,6 +32,16 @@ spec:
                 -H "Host: {{ .Values.ingress.host }}" \
                 -o /dev/null "http://git-pages:3000/.git-pages/health"
               do sleep 2; done
+              echo "Init: checking if site already exists..."
+              MANIFEST=$(curl -sf \
+                -H "Host: {{ .Values.ingress.host }}" \
+                "http://git-pages:3000/.git-pages/manifest.json" 2>/dev/null || echo "")
+
+              if echo "$MANIFEST" | grep -q '"contents"'; then
+                echo "Init: site already initialized, skipping"
+                exit 0
+              fi
+
               echo "Init: creating placeholder site..."
               WORK=$(mktemp -d)
               mkdir -p "$WORK/__init__"