ci(veans): add fast veans-test job for unit tests

.github/workflows/test.yml

@@ -91,6 +91,20 @@ jobs:
           version: v2.10.1
           working-directory: veans
 
+  veans-test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@8e8c483db84b4bee98b60c0593521ed34d9990e8 # v6
+      - uses: actions/setup-go@7a3fe6cf4cb3a834922a1244abfce67bcef6a0c5 # v6
+        with:
+          go-version: stable
+      - name: Run unit tests
+        # The e2e package self-skips when VEANS_E2E_API_URL isn't set, so
+        # this job runs only the (fast) unit tests, independent of the
+        # heavier test-veans-e2e job that needs the API artifact.
+        working-directory: veans
+        run: go test -count=1 ./...
+
   check-translations:
     runs-on: ubuntu-latest
     needs: mage

veans/internal/client/types.go

@@ -58,9 +58,9 @@ type Project struct {
 // "table" / "kanban"), not an int — Vikunja's ProjectViewKind has a
 // custom MarshalJSON.
 type ProjectView struct {
-	ID             int64  `json:"id"`
-	Title          string `json:"title"`
-	ProjectID      int64  `json:"project_id"`
+	ID        int64  `json:"id"`
+	Title     string `json:"title"`
+	ProjectID int64  `json:"project_id"`
 	// view_kind / bucket_configuration_mode are serialized as strings on
 	// the wire (custom MarshalJSON on the parent enums), not ints.
 	ViewKind       string `json:"view_kind"`
@@ -86,18 +86,18 @@ type Bucket struct {
 // Task mirrors the on-the-wire task representation. Many fields are omitted —
 // veans only consumes what its commands print or filter on.
 type Task struct {
-	ID          int64       `json:"id"`
-	Title       string      `json:"title"`
-	Description string      `json:"description,omitempty"`
-	Done        bool        `json:"done"`
-	DoneAt      *time.Time  `json:"done_at,omitempty"`
-	Priority    int64       `json:"priority,omitempty"`
-	ProjectID   int64       `json:"project_id"`
-	Index       int64       `json:"index,omitempty"`
-	Identifier  string      `json:"identifier,omitempty"`
-	Position    float64     `json:"position,omitempty"`
-	Created     time.Time   `json:"created,omitempty"`
-	Updated     time.Time   `json:"updated,omitempty"`
+	ID          int64      `json:"id"`
+	Title       string     `json:"title"`
+	Description string     `json:"description,omitempty"`
+	Done        bool       `json:"done"`
+	DoneAt      *time.Time `json:"done_at,omitempty"`
+	Priority    int64      `json:"priority,omitempty"`
+	ProjectID   int64      `json:"project_id"`
+	Index       int64      `json:"index,omitempty"`
+	Identifier  string     `json:"identifier,omitempty"`
+	Position    float64    `json:"position,omitempty"`
+	Created     time.Time  `json:"created,omitempty"`
+	Updated     time.Time  `json:"updated,omitempty"`
 	// BucketID is only set by Vikunja when sending a task to a server-
 	// side endpoint (e.g. the bucket-move POST); reads return it as 0.
 	// The current bucket(s) — one per Kanban view — are exposed via

veans/internal/commands/update.go

@@ -94,6 +94,13 @@ func newUpdateCmd() *cobra.Command {
 	return cmd
 }
 
+// runUpdate is intentionally a single linear flow — the steps it performs
+// (concurrency check → status → field changes → comments → field POST →
+// bucket move → label add/remove → refetch) all share the same task,
+// flag set, and error-handling shape. Splitting them produces five tiny
+// functions that each take the same five arguments.
+//
+//nolint:gocyclo // single-pass orchestration; each branch is one short stanza
 func runUpdate(ctx context.Context, rt *runtime, id int64, f *updateFlags) (*client.Task, error) {
 	current, err := rt.client.GetTask(ctx, id)
 	if err != nil {