feat(js): Add shared dataCollection prompt and preset (v11 only)

CHANGELOG.md

@@ -8,6 +8,7 @@
 - feat(sveltekit): Import `sentrySvelteKit` from `@sentry/sveltekit/vite` on SDK v11 and write `sentryUrl` for self-hosted ([#1348](https://github.com/getsentry/sentry-wizard/pull/1348))
 - feat(cloudflare): Write `nodejs_compat` instead of `nodejs_als` to the wrangler config and drop a stale `nodejs_als` ([#1353](https://github.com/getsentry/sentry-wizard/pull/1353))
 - feat(js): Warn when the wizard runs on a Node.js version that SDK v11 does not support ([#1354](https://github.com/getsentry/sentry-wizard/pull/1354))
+- feat(js): Add shared dataCollection prompt and preset (v11 only) ([#1358](https://github.com/getsentry/sentry-wizard/pull/1358))
 - fix(sourcemaps): Treat prerelease SDK versions as up to date ([#1345](https://github.com/getsentry/sentry-wizard/pull/1345))
 
 ## 7.0.3

src/react-router/sdk-version.ts

@@ -1,4 +1,4 @@
-import { major, minVersion } from 'semver';
+import { getMajorVersion } from '../utils/semver';
 
 /**
  * Version range of `@sentry/react-router` (and `@sentry/profiling-node`) the
@@ -27,23 +27,11 @@ export function getSentryReactRouterVitePluginImportPath(
   installedSdkVersion: string | undefined,
 ): string {
   const sdkMajor =
-    getMajor(installedSdkVersion) ??
-    getMajor(SENTRY_REACT_ROUTER_SDK_RANGE) ??
+    getMajorVersion(installedSdkVersion) ??
+    getMajorVersion(SENTRY_REACT_ROUTER_SDK_RANGE) ??
     0;
 
   return sdkMajor >= 11
     ? SENTRY_REACT_ROUTER_VITE_IMPORT_PATH
     : SENTRY_REACT_ROUTER_ROOT_IMPORT_PATH;
 }
-
-function getMajor(version: string | undefined): number | undefined {
-  if (!version) {
-    return undefined;
-  }
-  try {
-    const minVer = minVersion(version);
-    return minVer ? major(minVer) : undefined;
-  } catch {
-    return undefined;
-  }
-}

src/sveltekit/utils.ts

@@ -1,4 +1,6 @@
-import { lt, major, minVersion } from 'semver';
+import { lt, minVersion } from 'semver';
+
+import { getMajorVersion } from '../utils/semver';
 
 /**
  * Version range of `@sentry/sveltekit` the wizard installs. Also decides the
@@ -22,25 +24,15 @@ export function getSentrySvelteKitVitePluginImportPath(
   installedSdkVersion: string | undefined,
 ): string {
   const sdkMajor =
-    getMajor(installedSdkVersion) ?? getMajor(SENTRY_SVELTEKIT_SDK_RANGE) ?? 0;
+    getMajorVersion(installedSdkVersion) ??
+    getMajorVersion(SENTRY_SVELTEKIT_SDK_RANGE) ??
+    0;
 
   return sdkMajor >= 11
     ? SENTRY_SVELTEKIT_VITE_IMPORT_PATH
     : SENTRY_SVELTEKIT_ROOT_IMPORT_PATH;
 }
 
-function getMajor(version: string | undefined): number | undefined {
-  if (!version) {
-    return undefined;
-  }
-  try {
-    const minVer = minVersion(version);
-    return minVer ? major(minVer) : undefined;
-  } catch {
-    return undefined;
-  }
-}
-
 export type KitVersionBucket =
   | 'none'
   | 'invalid'

src/utils/data-collection.ts

@@ -0,0 +1,116 @@
+// @ts-expect-error - clack is ESM and TS complains about that. It works though
+import * as clack from '@clack/prompts';
+import * as Sentry from '@sentry/node';
+
+import { traceStep } from '../telemetry';
+import { abortIfCancelled } from './clack';
+import { getMajorVersion } from './semver';
+
+export const DATA_COLLECTION_DOCS_URL =
+  'https://docs.sentry.io/platforms/javascript/configuration/options/#dataCollection';
+
+/**
+ * Keys that commonly reveal user identity in HTTP headers, cookies, and URL
+ * query parameters (forwarding chains, IP addresses, user hints).
+ */
+const PII_KEY_DENYLIST = ['forwarded', '-ip', 'remote-', 'via', '-user'];
+
+const PII_KEY_DENYLIST_SNIPPET = `[${PII_KEY_DENYLIST.map(
+  (key) => `"${key}"`,
+).join(', ')}]`;
+
+/**
+ * `dataCollection` value the wizard writes when the user opts to reduce data
+ * collection. Kept as an object so magicast-based wizards can insert it as an
+ * AST node; `getDataCollectionSnippet` renders the same preset for
+ * template-string-based wizards.
+ */
+export const DATA_COLLECTION_PII_PRESET = {
+  userInfo: false,
+  graphQL: { document: false, variables: false },
+  genAI: { inputs: false, outputs: false },
+  databaseQueryData: false,
+  queues: false,
+  httpBodies: [],
+  httpHeaders: { deny: [...PII_KEY_DENYLIST] },
+  cookies: { deny: [...PII_KEY_DENYLIST] },
+  urlQueryParams: { deny: [...PII_KEY_DENYLIST] },
+};
+
+export const DATA_COLLECTION_COMMENT_LINES = [
+  'Turns off collection of data that could identify users. Adjust per category:',
+  DATA_COLLECTION_DOCS_URL,
+];
+
+/**
+ * Renders the PII preset (with its explanatory comment) as code to embed in a
+ * generated `Sentry.init` options object. Every line is prefixed with
+ * `indent`; the result ends without a trailing newline.
+ */
+export function getDataCollectionSnippet(indent = '  '): string {
+  const lines = [
+    ...DATA_COLLECTION_COMMENT_LINES.map((line) => `// ${line}`),
+    'dataCollection: {',
+    '  userInfo: false,',
+    '  graphQL: { document: false, variables: false },',
+    '  genAI: { inputs: false, outputs: false },',
+    '  databaseQueryData: false,',
+    '  queues: false,',
+    '  httpBodies: [],',
+    `  httpHeaders: { deny: ${PII_KEY_DENYLIST_SNIPPET} },`,
+    `  cookies: { deny: ${PII_KEY_DENYLIST_SNIPPET} },`,
+    `  urlQueryParams: { deny: ${PII_KEY_DENYLIST_SNIPPET} },`,
+    '},',
+  ];
+
+  return lines.map((line) => `${indent}${line}`).join('\n');
+}
+
+/**
+ * The `dataCollection` option only makes sense to offer from SDK v11 on,
+ * where the SDK collects rich context by default. Unknown or unparseable
+ * installed versions fall back to the major of the range the wizard installs.
+ */
+export function sdkSupportsDataCollection(
+  installedSdkVersion: string | undefined,
+  wizardInstallRange: string,
+): boolean {
+  const sdkMajor =
+    getMajorVersion(installedSdkVersion) ??
+    getMajorVersion(wizardInstallRange) ??
+    0;
+
+  return sdkMajor >= 11;
+}
+
+/**
+ * Asks whether the wizard should write the PII-reducing `dataCollection`
+ * preset. Call this only when `sdkSupportsDataCollection` returned `true`.
+ */
+export async function askShouldReduceDataCollection(): Promise<boolean> {
+  return traceStep('data-collection', async () => {
+    const reduceDataCollection: boolean = await abortIfCancelled(
+      clack.select({
+        message:
+          'Sentry collects request data, user info, and other context by default. Do you want to reduce this to avoid sending personally identifiable information (PII)?',
+        initialValue: false,
+        options: [
+          {
+            value: true,
+            label: 'Yes',
+            hint: 'Add a dataCollection config to Sentry.init() that turns off PII-heavy categories',
+          },
+          {
+            value: false,
+            label: 'No',
+            hint: 'Keep the SDK defaults',
+          },
+        ],
+      }),
+    );
+
+    Sentry.setTag('data-collection-reduced', reduceDataCollection);
+
+    return reduceDataCollection;
+  });
+}

src/utils/semver.ts

@@ -1,4 +1,30 @@
-import { satisfies, subset, valid, validRange } from 'semver';
+import {
+  major,
+  minVersion,
+  satisfies,
+  subset,
+  valid,
+  validRange,
+} from 'semver';
+
+/**
+ * Returns the major version of an exact version or the lowest version that a
+ * range permits (e.g. `^10` -> 10, `~11.2.0` -> 11). Returns `undefined` for
+ * missing or unparseable input.
+ */
+export function getMajorVersion(
+  version: string | undefined,
+): number | undefined {
+  if (!version) {
+    return undefined;
+  }
+  try {
+    const minVer = minVersion(version);
+    return minVer ? major(minVer) : undefined;
+  } catch {
+    return undefined;
+  }
+}
 
 export function fulfillsVersionRange({
   version,

test/utils/data-collection.test.ts

@@ -0,0 +1,92 @@
+import { describe, expect, it } from 'vitest';
+
+import {
+  getDataCollectionSnippet,
+  sdkSupportsDataCollection,
+} from '../../src/utils/data-collection';
+import { getMajorVersion } from '../../src/utils/semver';
+
+describe('getMajorVersion', () => {
+  it.each([
+    ['^10', 10],
+    ['^10.73.0', 10],
+    ['~11.2.0', 11],
+    ['11.0.0', 11],
+    // Prerelease versions keep their major, so v11 rc/beta installs pass the gate
+    ['11.0.0-rc.1', 11],
+    ['^11.0.0-rc.0', 11],
+    ['~11.0.0-beta.2', 11],
+    ['>=10.57.0 <12', 10],
+  ])('returns the lowest permitted major for %s', (version, expected) => {
+    expect(getMajorVersion(version)).toBe(expected);
+  });
+
+  it.each([undefined, '', 'not-a-version', 'workspace:*'])(
+    'returns undefined for %s',
+    (version) => {
+      expect(getMajorVersion(version)).toBeUndefined();
+    },
+  );
+});
+
+describe('sdkSupportsDataCollection', () => {
+  it.each([
+    ['^11.0.0', '^10'],
+    ['11.2.3', '^10'],
+    ['^12', '^10'],
+    ['11.0.0-rc.1', '^10'],
+    ['^11.0.0-beta.0', '^10'],
+  ])('returns true for installed version %s', (installed, range) => {
+    expect(sdkSupportsDataCollection(installed, range)).toBe(true);
+  });
+
+  it.each([
+    ['^10.73.0', '^10'],
+    ['10.99.0', '^10'],
+    ['^9', '^10'],
+    ['10.99.0-rc.1', '^10'],
+  ])('returns false for installed version %s', (installed, range) => {
+    expect(sdkSupportsDataCollection(installed, range)).toBe(false);
+  });
+
+  it('falls back to the wizard install range when the installed version is missing', () => {
+    expect(sdkSupportsDataCollection(undefined, '^11')).toBe(true);
+    expect(sdkSupportsDataCollection(undefined, '^10')).toBe(false);
+  });
+
+  it('falls back to the wizard install range when the installed version is unparseable', () => {
+    expect(sdkSupportsDataCollection('workspace:*', '^11')).toBe(true);
+    expect(sdkSupportsDataCollection('workspace:*', '^10')).toBe(false);
+  });
+
+  it('returns false when neither version is parseable', () => {
+    expect(sdkSupportsDataCollection(undefined, 'not-a-version')).toBe(false);
+  });
+});
+
+describe('getDataCollectionSnippet', () => {
+  it('renders the preset with the default indentation', () => {
+    expect(getDataCollectionSnippet()).toMatchInlineSnapshot(`
+      "  // Turns off collection of data that could identify users. Adjust per category:
+        // https://docs.sentry.io/platforms/javascript/configuration/options/#dataCollection
+        dataCollection: {
+          userInfo: false,
+          graphQL: { document: false, variables: false },
+          genAI: { inputs: false, outputs: false },
+          databaseQueryData: false,
+          queues: false,
+          httpBodies: [],
+          httpHeaders: { deny: ["forwarded", "-ip", "remote-", "via", "-user"] },
+          cookies: { deny: ["forwarded", "-ip", "remote-", "via", "-user"] },
+          urlQueryParams: { deny: ["forwarded", "-ip", "remote-", "via", "-user"] },
+        },"
+    `);
+  });
+
+  it('prefixes every line with the given indent', () => {
+    const snippet = getDataCollectionSnippet('    ');
+    for (const line of snippet.split('\n')) {
+      expect(line.startsWith('    ')).toBe(true);
+    }
+  });
+});