fixup! test(@angular/build): add library builder scaling benchmark suite

Implemented the style-flavor matrix (`--style-flavor`, see "Style-flavor matrix" above):
`inline`, `inline-css`, `inline-scss`, `external`, `external-css`, `external-scss`, as a sweep
axis independent of `--layout`/`--count`/`--depth`. Defaults to `inline` only (matching the
prior behavior) since the full matrix multiplies build count by up to 6x; opt in with
`--style-flavor=<list>`.

Author: dherges

scripts/benchmark.mts

@@ -9,6 +9,7 @@
 import fs from 'node:fs';
 import path from 'node:path';
 import { type BenchmarkCliOptions, runI18nBenchmarks } from './benchmarks/i18n/index.mts';
+import { STYLE_FLAVORS, type StyleFlavor } from './benchmarks/library-builder/fixtures.mts';
 import {
   type LibraryBuilderBenchmarkOptions,
   runLibraryBuilderBenchmarks,
@@ -44,16 +45,26 @@ function parseIntList(raw: string | number | undefined): number[] | undefined {
   if (raw === undefined) {
     return undefined;
   }
+
   return String(raw)
     .split(',')
     .map((s) => Number(s.trim()));
 }
 
+function parseStringList(raw: string | undefined): string[] | undefined {
+  if (raw === undefined) {
+    return undefined;
+  }
+
+  return raw.split(',').map((s) => s.trim());
+}
+
 async function runLibraryBuilderSubsystem(
   options: {
     layout?: string;
     count?: string;
     depth?: string;
+    'style-flavor'?: string;
     iterations?: string | number;
     json?: boolean;
     csv?: string | boolean;
@@ -108,10 +119,26 @@ async function runLibraryBuilderSubsystem(
     return 1;
   }
 
+  const styleFlavor = parseStringList(options['style-flavor']);
+  const validStyleFlavors: readonly string[] = STYLE_FLAVORS;
+  if (styleFlavor !== undefined && styleFlavor.some((f) => !validStyleFlavors.includes(f))) {
+    // eslint-disable-next-line no-console
+    console.error(
+      'Error: --style-flavor must be a comma-separated list of: ' +
+        STYLE_FLAVORS.join(', ') +
+        '. Got "' +
+        options['style-flavor'] +
+        '".',
+    );
+
+    return 1;
+  }
+
   const cliOptions: LibraryBuilderBenchmarkOptions = {
     layout: options.layout as LibraryBuilderBenchmarkOptions['layout'],
     count,
     depth,
+    styleFlavor: styleFlavor as StyleFlavor[] | undefined,
     iterations,
     json: Boolean(options.json),
     csv: options.csv !== undefined ? path.resolve(cwd, options.csv as string) : undefined,
@@ -140,6 +167,7 @@ export default async function (
     layout?: string;
     count?: string;
     depth?: string;
+    'style-flavor'?: string;
     help?: boolean;
     [key: string]: unknown;
   },
@@ -178,6 +206,9 @@ Options (library-builder):
   --depth=<n,n,...>          Explicit tree depths, layout=deep only. Overrides the depths
                              that would otherwise be derived from --count (the smallest
                              depth with 2^depth >= each target count).
+  --style-flavor=<f,f,...>   Style-flavor targets, one or more of: inline, inline-css,
+                             inline-scss, external, external-css, external-scss
+                             (default: inline)
   --iterations=<n>           Measured iterations (default: 3)
   --json                     Output raw measurements as JSON to stdout
   --csv=<path>               Write raw per-iteration measurements as CSV to <path>

scripts/benchmarks/library-builder/README.md

@@ -56,9 +56,31 @@ This suite supports two primary layouts to test different filesystem and entry-p
   - Design: Creates a nested directory structure (`sub-a`/`sub-b` at each level).
   - Depth $N$ is not an independent sweep parameter -- see "Sweep rounds" below.
 
-Each entry point consists of a `public-api.ts` and a component file with an inline template
-containing a bit of per-entry content derived from the entry's id (e.g. `<h2>Component no.
-0042</h2>`), so entries aren't byte-identical copies of each other.
+Each entry point consists of a `public-api.ts` and a component file with a bit of per-entry
+content derived from the entry's id (e.g. `<h2>Component no. 0042</h2>`), so entries aren't
+byte-identical copies of each other.
+
+### Style-flavor matrix
+
+`--style-flavor` is a second, independent sweep axis (crossed with `--layout` and the
+count/depth sweep) controlling each entry point's template location and stylesheet
+presence/language:
+
+| Flavor          | Template               | Stylesheet                                        |
+| :-------------- | :--------------------- | :------------------------------------------------ |
+| `inline`        | inline `template`      | none                                              |
+| `inline-css`    | inline `template`      | inline `styles`, plain CSS                        |
+| `inline-scss`   | inline `template`      | inline `styles`, SCSS (via `inlineStyleLanguage`) |
+| `external`      | external `templateUrl` | none                                              |
+| `external-css`  | external `templateUrl` | external `styleUrl`, `.css` file                  |
+| `external-scss` | external `templateUrl` | external `styleUrl`, `.scss` file                 |
+
+`external` isolates the cost of the extra per-entry template file read from the cost of
+stylesheet processing (it has no stylesheet, like `inline`, but reads its template from disk
+like the other `external-*` flavors). SCSS content uses a per-entry variable
+(`$ref-color-NNNN`) so the preprocessor does real work rather than passing through literal CSS.
+Defaults to `inline` only -- the full 6-flavor matrix is opt-in via `--style-flavor=<list>`
+since it multiplies the number of builds run by up to 6x.
 
 ### Sweep rounds
 
@@ -88,14 +110,15 @@ pnpm admin benchmark library-builder [options]
 
 ### Options
 
-| Option                  | Description                                                                                          | Default                                  |
-| :---------------------- | :--------------------------------------------------------------------------------------------------- | :--------------------------------------- |
-| `--layout=<flat\|deep>` | Fixture layout                                                                                       | `flat`                                   |
-| `--count=<n,n,...>`     | Entry-point count targets, drives both layouts                                                       | `10,20,50,100,200,300,400,500,1000,2000` |
-| `--depth=<n,n,...>`     | **`layout=deep` only.** Explicit tree depths, overriding the depths otherwise derived from `--count` | derived from `--count`                   |
-| `--iterations=<n>`      | Measured iterations per size                                                                         | `3`                                      |
-| `--json`                | Output raw per-iteration measurements as JSON to stdout                                              | N/A                                      |
-| `--csv=<path>`          | Write raw per-iteration measurements as CSV to `<path>`                                              | N/A (opt-in only)                        |
+| Option                     | Description                                                                                                                                               | Default                                  |
+| :------------------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------- | :--------------------------------------- |
+| `--layout=<flat\|deep>`    | Fixture layout                                                                                                                                            | `flat`                                   |
+| `--count=<n,n,...>`        | Entry-point count targets, drives both layouts                                                                                                            | `10,20,50,100,200,300,400,500,1000,2000` |
+| `--depth=<n,n,...>`        | **`layout=deep` only.** Explicit tree depths, overriding the depths otherwise derived from `--count`                                                      | derived from `--count`                   |
+| `--style-flavor=<f,f,...>` | Style-flavor targets (see "Style-flavor matrix" above), one or more of `inline`, `inline-css`, `inline-scss`, `external`, `external-css`, `external-scss` | `inline`                                 |
+| `--iterations=<n>`         | Measured iterations per size                                                                                                                              | `3`                                      |
+| `--json`                   | Output raw per-iteration measurements as JSON to stdout                                                                                                   | N/A                                      |
+| `--csv=<path>`             | Write raw per-iteration measurements as CSV to `<path>`                                                                                                   | N/A (opt-in only)                        |
 
 > `--count` and `--depth` are not independent sweep axes for `layout=deep`: `--count` is translated
 > into depths (smallest `depth` with `2^depth >= n` per target, de-duplicated); `--depth` only
@@ -110,6 +133,15 @@ Sweep the flat layout at a single size:
 pnpm admin benchmark library-builder --layout=flat --count=500
 ```
 
+Sweep the full style-flavor matrix at a single size, both layouts:
+
+```sh
+pnpm admin benchmark library-builder --layout=flat --count=500 \
+  --style-flavor=inline,inline-css,inline-scss,external,external-css,external-scss
+pnpm admin benchmark library-builder --layout=deep --count=500 \
+  --style-flavor=inline,inline-css,inline-scss,external,external-css,external-scss
+```
+
 ## Benchmark Results
 
 Each entry below is one benchmark run, with its own summary and conclusions.
@@ -161,10 +193,36 @@ but not a regression, and not yet investigated further.
 
 All 60 builds (10 sizes x 3 iterations x 2 layouts) completed with `status: "ok"`; zero failures.
 
+### 2026-10-16: Style-flavor matrix implemented, fixture-correctness verification
+
+Implemented the style-flavor matrix (`--style-flavor`, see "Style-flavor matrix" above):
+`inline`, `inline-css`, `inline-scss`, `external`, `external-css`, `external-scss`, as a sweep
+axis independent of `--layout`/`--count`/`--depth`. Defaults to `inline` only (matching the
+prior behavior) since the full matrix multiplies build count by up to 6x; opt in with
+`--style-flavor=<list>`.
+
+This entry only verifies fixture correctness at small scale, not full-sweep performance
+numbers -- a full style-flavor sweep is still Future Work.
+
+Verification performed:
+
+- `--layout=flat --count=5` and `--layout=deep --count=5` (8 entries, depth 3), all 6 flavors,
+  `--iterations=1`: all 12 builds completed with `status: "ok"`.
+- Manually inspected a generated `inline-scss` fixture's `angular.json` to confirm
+  `inlineStyleLanguage: "scss"` is set only for that flavor, and compiled it standalone to
+  confirm the per-entry SCSS variable (`$ref-color-NNNN`) resolves to a literal color in the
+  built `.mjs` output (not left as unresolved `$...` syntax) -- i.e. SCSS is actually being
+  preprocessed, not passed through as plain CSS.
+- Repeated the same compiled-output check for `external-scss` with the `deep` layout --
+  confirms external `.scss` stylesheets referenced via `styleUrl` also compile correctly.
+- Verified `--csv` and `--json` output both include the new `styleFlavor` field.
+- Verified `--style-flavor` rejects an unrecognized flavor name with a clear error listing the
+  valid set, consistent with existing `--layout`/`--count`/`--depth` validation.
+
 ## Future Work
 
-- Style-flavor matrix: inline / inline-css / inline-scss / external / external-css / external-scss, swept across both layouts.
 - A/B comparison against `ng-packagr` (both are already registered Architect builders — `@angular/build:ng-packagr` and `@angular/build:library` — so this is a same-builder-package comparison, not a separate tool invocation).
+- Full style-flavor matrix sweep (default `--count` sweep x 6 flavors x 2 layouts) to quantify the actual performance impact of stylesheet processing and external file reads, now that the matrix is implemented.
 - Dependencies between entry points.
 - Shared stylesheet tokens imported by multiple components.
 - Watch mode.