feat(cli): add static dir file discovery depth flag (#940)

Co-authored-by: Connor Clark <[email protected]>
GoogleChrome · Dec 14, 2023 · de1614e · de1614e
1 parent 2cf4dc0
commit de1614e
Show file tree

Hide file tree

Showing 5 changed files with 48 additions and 7 deletions.
diff --git a/docs/configuration.md b/docs/configuration.md
@@ -232,6 +232,9 @@ Options:
   --maxAutodiscoverUrls      The maximum number of pages to collect when using the staticDistDir
                              option with no specified URL. Disable this limit by setting to 0.
                                                                                [number] [default: 5]
+  --staticDirFileDiscoveryDepth The maximum depth of nested folders Lighthouse will look into to discover 
+                                URLs on a static file folder.
+                                                                               [number] [default: 2]
 ```
 
 #### `method`
@@ -422,6 +425,25 @@ lhci collect --url=https://example-1.com --url=https://example-2.com
 lhci collect --start-server-command="yarn serve" --url=http://localhost:8080/ --puppeteer-script=./path/to/login-with-puppeteer.js
 ```
 
+### `staticDirFileDiscoveryDepth`
+
+The maximum depth level of nested folders that Lighthouse will look into to discover URLs. If not set, this will default to 2.
+
+### Example
+```text
+
+public/
+├── index.html               #level 0                      
+├── contact/
+│   └── index.html           #level 1
+├── projects/
+│   ├──index.html            #level 1
+│   └── crisis/
+│       ├──index.html        #level 2
+│       └── earthquake/
+│           └── index.html   #level 3
+```
+
 ---
 
 ### `upload`

diff --git a/packages/cli/src/collect/collect.js b/packages/cli/src/collect/collect.js
@@ -94,6 +94,12 @@ function buildCommand(yargs) {
       default: 5,
       type: 'number',
     },
+    staticDirFileDiscoveryDepth: {
+      description:
+        'The maximum depth level of nested folders that Lighthouse will look into to discover URLs. If not set, this will default to 2.',
+      default: 2,
+      type: 'number',
+    },
   });
 }
 
@@ -187,7 +193,8 @@ async function startServerAndDetermineUrls(options) {
       : options.autodiscoverUrlBlocklist
       ? [options.autodiscoverUrlBlocklist]
       : [];
-    const availableUrls = server.getAvailableUrls();
+    const maxStaticDirFileDiscoveryDepth = options.staticDirFileDiscoveryDepth || 2;
+    const availableUrls = server.getAvailableUrls(maxStaticDirFileDiscoveryDepth);
     const normalizedBlocklist = autodiscoverUrlBlocklistAsArray.map(rawUrl => {
       const url = new URL(rawUrl, 'http://localhost');
       url.port = server.port.toString();

diff --git a/packages/cli/src/collect/fallback-server.js b/packages/cli/src/collect/fallback-server.js
@@ -72,9 +72,20 @@ class FallbackServer {
     );
   }
 
-  /** @return {string[]} */
-  getAvailableUrls() {
-    const htmlFiles = FallbackServer.readHtmlFilesInDirectory(this._pathToBuildDir, 2);
+  /**
+   * @param {number} maxDepth
+   * @return  {string[]}
+   */
+  getAvailableUrls(maxDepth) {
+    if (maxDepth >= 0) {
+      maxDepth = Math.floor(maxDepth);
+    } else {
+      process.stderr.write(
+        `WARNING: staticDirFileDiscoveryDepth must be greater than 0. Defaulting to a discovery depth of 2\n`
+      );
+      maxDepth = 2;
+    }
+    const htmlFiles = FallbackServer.readHtmlFilesInDirectory(this._pathToBuildDir, maxDepth);
     return htmlFiles.map(({file}) => `http://localhost:${this._port}/${file}`);
   }
 

diff --git a/packages/cli/test/collect-puppeteer.test.js b/packages/cli/test/collect-puppeteer.test.js
@@ -91,9 +91,9 @@ describe('Lighthouse CI collect CLI with puppeteer', () => {
     const chromePathHelp = stdout.match(/--chromePath.*\n.*\n.*/);
     expect(chromePathHelp).toMatchInlineSnapshot(`
       [
-        "--chromePath                The path to the Chrome or Chromium executable to use for collection.
-        --puppeteerScript           The path to a script that manipulates the browser with puppeteer before running Lighthouse, used for auth.
-        --puppeteerLaunchOptions    The object of puppeteer launch options",
+        "--chromePath                   The path to the Chrome or Chromium executable to use for collection.
+        --puppeteerScript              The path to a script that manipulates the browser with puppeteer before running Lighthouse, used for auth.
+        --puppeteerLaunchOptions       The object of puppeteer launch options",
       ]
     `);
     expect(stderr).toMatchInlineSnapshot(`""`);

diff --git a/types/collect.d.ts b/types/collect.d.ts
@@ -63,6 +63,7 @@ declare global {
         additive: boolean;
         settings?: LighthouseSettings;
         maxAutodiscoverUrls?: number;
+        staticDirFileDiscoveryDepth?: number;
       }
     }
   }