Bugfix #895: UPPERCASE area group headers in VSCode sidebar

codev/projects/bugfix-895-vscode-uppercase-area-group-he/status.yaml

@@ -0,0 +1,16 @@
+id: bugfix-895
+title: vscode-uppercase-area-group-he
+protocol: bugfix
+phase: pr
+plan_phases: []
+current_plan_phase: null
+gates:
+  pr:
+    status: pending
+    requested_at: '2026-05-28T01:02:24.638Z'
+iteration: 1
+build_complete: false
+history: []
+started_at: '2026-05-28T00:56:37.908Z'
+updated_at: '2026-05-28T01:02:24.638Z'
+pr_ready_for_human: true

codev/state/bugfix-895_thread.md

@@ -0,0 +1,15 @@
+# bugfix-895 thread
+
+## Investigate
+- Issue #895: Replace per-word capitalization with full uppercase for area group headers in VSCode sidebar.
+- Current `formatAreaForDisplay` in `packages/core/src/area-grouping.ts` splits on `[-_\s]+`, capitalizes first char of each word, joins with space. Produces `Cross Cutting` for `cross-cutting`.
+- Issue's before/after table shows `Cross-cutting (1)` → `CROSS-CUTTING (1)` (preserves hyphen). The simplest reading that matches the table: replace the entire pipeline with `area.toUpperCase()`. That preserves `-` and `_` and matches every example in the table.
+- Plan: rename `formatAreaForDisplay` → `uppercaseAreaName`, implementation becomes `area.toUpperCase()`. Update test + one caller in `area-group-tree-item.ts`.
+- Callers: `packages/vscode/src/views/area-group-tree-item.ts` (one site) + tests in `packages/vscode/src/test/area-grouping.test.ts`.
+
+## Fix
+- `packages/core/src/area-grouping.ts`: replaced multi-step pipeline with `area.toUpperCase()`; renamed `formatAreaForDisplay` → `uppercaseAreaName`; updated docstring.
+- `packages/vscode/src/views/area-group-tree-item.ts`: import + call site renamed; updated docstring comment.
+- `packages/vscode/src/test/area-grouping.test.ts`: suite renamed + assertions updated to uppercase expectations. Separator behavior changed: hyphens/underscores are now preserved (e.g. `cross-cutting` → `CROSS-CUTTING`, not `CROSS CUTTING`) — matches the issue's before/after table. The old "consecutive-separator collapsing" test was repurposed to lock in verbatim preservation.
+- Verified: `pnpm --filter @cluesmith/codev-core build` (tsc), `pnpm --filter codev-vscode check-types`, `pnpm --filter codev-vscode lint`, `pnpm --filter codev-vscode test:unit` (49 pass), `pnpm --filter codev-vscode compile-tests` all clean.
+

packages/core/src/area-grouping.ts

@@ -8,24 +8,21 @@ import { UNCATEGORIZED_AREA } from './constants.js';
  * `UNCATEGORIZED_AREA` sentinel) MUST continue to be used for
  * bucketing (`groupByArea`), id stability, and `===` matchers.
  *
- * Rule: split on `-`, `_`, and whitespace; capitalize the first
- * character of each word; rejoin with a single space. Examples:
- *   'vscode'        -> 'Vscode'
- *   'cross-cutting' -> 'Cross Cutting'
- *   'front_end'     -> 'Front End'
- *   'Uncategorized' -> 'Uncategorized'  (no-op; sentinel is already
- *                                        single-word, first-char-upper)
+ * Rule: uppercase the entire string. Separators (`-`, `_`) are
+ * preserved as-is so the header reads as a single visual token
+ * matching VSCode's own container-label convention (EXPLORER,
+ * SOURCE CONTROL, etc.). Examples:
+ *   'vscode'        -> 'VSCODE'
+ *   'cross-cutting' -> 'CROSS-CUTTING'
+ *   'front_end'     -> 'FRONT_END'
+ *   'Uncategorized' -> 'UNCATEGORIZED'
  *
  * Purely structural: no hardcoded list of "known acronyms". Teams
  * using Codev decide their own labeling semantics; a per-repo
  * override map can be layered on top as a non-breaking extension.
  */
-export function formatAreaForDisplay(area: string): string {
-  return area
-    .split(/[-_\s]+/)
-    .filter(w => w.length > 0)
-    .map(w => w.charAt(0).toUpperCase() + w.slice(1))
-    .join(' ');
+export function uppercaseAreaName(area: string): string {
+  return area.toUpperCase();
 }
 
 /**

packages/vscode/src/test/area-grouping.test.ts

@@ -1,5 +1,5 @@
 import * as assert from 'assert';
-import { groupByArea, formatAreaForDisplay } from '@cluesmith/codev-core/area-grouping';
+import { groupByArea, uppercaseAreaName } from '@cluesmith/codev-core/area-grouping';
 
 interface AreaItem {
 	id: string;
@@ -81,36 +81,35 @@ suite('groupByArea', () => {
 	});
 });
 
-suite('formatAreaForDisplay', () => {
-	test('lowercase single word -> capitalized first char', () => {
-		assert.strictEqual(formatAreaForDisplay('vscode'), 'Vscode');
-		assert.strictEqual(formatAreaForDisplay('tower'), 'Tower');
-		assert.strictEqual(formatAreaForDisplay('porch'), 'Porch');
+suite('uppercaseAreaName', () => {
+	test('lowercase single word -> full uppercase', () => {
+		assert.strictEqual(uppercaseAreaName('vscode'), 'VSCODE');
+		assert.strictEqual(uppercaseAreaName('tower'), 'TOWER');
+		assert.strictEqual(uppercaseAreaName('porch'), 'PORCH');
 	});
 
-	test('hyphenated -> separator replaced with space, each word capitalized', () => {
-		assert.strictEqual(formatAreaForDisplay('cross-cutting'), 'Cross Cutting');
+	test('hyphenated -> uppercase with hyphen preserved', () => {
+		assert.strictEqual(uppercaseAreaName('cross-cutting'), 'CROSS-CUTTING');
 	});
 
-	test('underscored -> separator replaced with space, each word capitalized', () => {
-		assert.strictEqual(formatAreaForDisplay('front_end'), 'Front End');
+	test('underscored -> uppercase with underscore preserved', () => {
+		assert.strictEqual(uppercaseAreaName('front_end'), 'FRONT_END');
 	});
 
-	test('mixed separators and >2 words', () => {
-		assert.strictEqual(formatAreaForDisplay('front-end_ui'), 'Front End Ui');
+	test('mixed separators preserved verbatim', () => {
+		assert.strictEqual(uppercaseAreaName('front-end_ui'), 'FRONT-END_UI');
 	});
 
-	test('Uncategorized sentinel is a no-op (single word, first char already upper)', () => {
-		assert.strictEqual(formatAreaForDisplay('Uncategorized'), 'Uncategorized');
+	test('Uncategorized sentinel uppercases to UNCATEGORIZED', () => {
+		assert.strictEqual(uppercaseAreaName('Uncategorized'), 'UNCATEGORIZED');
 	});
 
 	test('empty string -> empty string (defensive)', () => {
-		assert.strictEqual(formatAreaForDisplay(''), '');
+		assert.strictEqual(uppercaseAreaName(''), '');
 	});
 
-	test('collapses multiple consecutive separators', () => {
-		// Pathological but cheap to lock — no accidental double-spacing.
-		assert.strictEqual(formatAreaForDisplay('a--b'), 'A B');
-		assert.strictEqual(formatAreaForDisplay('a__b'), 'A B');
+	test('consecutive separators preserved verbatim (no collapsing)', () => {
+		assert.strictEqual(uppercaseAreaName('a--b'), 'A--B');
+		assert.strictEqual(uppercaseAreaName('a__b'), 'A__B');
 	});
 });

packages/vscode/src/views/area-group-tree-item.ts

@@ -1,5 +1,5 @@
 import * as vscode from 'vscode';
-import { formatAreaForDisplay } from '@cluesmith/codev-core/area-grouping';
+import { uppercaseAreaName } from '@cluesmith/codev-core/area-grouping';
 
 export type AreaGroupKind = 'backlog' | 'builder';
 
@@ -17,8 +17,8 @@ export type AreaGroupKind = 'backlog' | 'builder';
  * `areaName`, `id`, and `contextValue` all use the raw wire value so
  * expansion-state persistence and `areaName === ...` matchers in the
  * per-view providers keep working. Only the human-visible label is
- * passed through `formatAreaForDisplay` (title-case + separator-to-
- * space, see #885).
+ * passed through `uppercaseAreaName`, matching VSCode's own
+ * container-label convention (EXPLORER, SOURCE CONTROL, etc.).
  */
 export class AreaGroupTreeItem extends vscode.TreeItem {
   constructor(
@@ -27,7 +27,7 @@ export class AreaGroupTreeItem extends vscode.TreeItem {
     count: number,
     collapsibleState: vscode.TreeItemCollapsibleState,
   ) {
-    super(`${formatAreaForDisplay(areaName)} (${count})`, collapsibleState);
+    super(`${uppercaseAreaName(areaName)} (${count})`, collapsibleState);
     this.id = `${kind}-group:${areaName}`;
     this.contextValue = `${kind}-group`;
   }