remove special-casing of default in hook.js #6401

bengl · 2025-09-08T18:09:23Z

What does this PR do?

See title. Plus:

appropriate testing
modifying instrumentations when needed

Motivation

The previous model of assuming default === exports was very incorrect for ESM since they're never the exact same object. We also can't assume they should have the same properties, or even that they're both objects/functions, at least not in the general case.

Therefore, we need to be specific about what's being instrumented in every single instrumentation.

Additional Notes

This is obviously in a draft state. The only instrumentation edited so far is http2.

This should fix these issues:
Refs: #3500
Refs: #5479

github-actions · 2025-09-08T18:09:57Z

Overall package size

Self size: 12.42 MB
Deduped: 112.51 MB
No deduping: 112.91 MB

Dependency sizes

| name | version | self size | total size | |------|---------|-----------|------------| | @datadog/libdatadog | 0.7.0 | 35.02 MB | 35.02 MB | | @datadog/native-appsec | 10.2.1 | 20.64 MB | 20.65 MB | | @datadog/native-iast-taint-tracking | 4.0.0 | 11.72 MB | 11.73 MB | | @datadog/pprof | 5.10.0 | 9.91 MB | 10.3 MB | | @opentelemetry/core | 1.30.1 | 908.66 kB | 7.16 MB | | protobufjs | 7.5.4 | 2.95 MB | 5.6 MB | | @datadog/wasm-js-rewriter | 4.0.1 | 2.85 MB | 3.58 MB | | @datadog/native-metrics | 3.1.1 | 1.02 MB | 1.43 MB | | @opentelemetry/api | 1.9.0 | 1.22 MB | 1.22 MB | | jsonpath-plus | 10.3.0 | 617.18 kB | 1.08 MB | | import-in-the-middle | 1.14.2 | 122.36 kB | 850.93 kB | | lru-cache | 10.4.3 | 804.3 kB | 804.3 kB | | opentracing | 0.14.7 | 194.81 kB | 194.81 kB | | source-map | 0.7.6 | 185.63 kB | 185.63 kB | | pprof-format | 2.2.1 | 163.06 kB | 163.06 kB | | @datadog/sketches-js | 2.1.1 | 109.9 kB | 109.9 kB | | lodash.sortby | 4.7.0 | 75.76 kB | 75.76 kB | | ignore | 7.0.5 | 63.38 kB | 63.38 kB | | istanbul-lib-coverage | 3.2.2 | 34.37 kB | 34.37 kB | | rfdc | 1.4.1 | 27.15 kB | 27.15 kB | | dc-polyfill | 0.1.10 | 26.73 kB | 26.73 kB | | @isaacs/ttlcache | 1.4.1 | 25.2 kB | 25.2 kB | | tlhunter-sorted-set | 0.1.0 | 24.94 kB | 24.94 kB | | shell-quote | 1.8.3 | 23.74 kB | 23.74 kB | | limiter | 1.1.5 | 23.17 kB | 23.17 kB | | retry | 0.13.1 | 18.85 kB | 18.85 kB | | semifies | 1.0.0 | 15.84 kB | 15.84 kB | | jest-docblock | 29.7.0 | 8.99 kB | 12.76 kB | | crypto-randomuuid | 1.0.0 | 11.18 kB | 11.18 kB | | ttl-set | 1.0.0 | 4.61 kB | 9.69 kB | | mutexify | 1.4.0 | 5.71 kB | 8.74 kB | | path-to-regexp | 0.1.12 | 6.6 kB | 6.6 kB | | koalas | 1.0.2 | 6.47 kB | 6.47 kB | | module-details-from-path | 1.0.4 | 3.96 kB | 3.96 kB |

_{🤖 This report was automatically generated by heaviest-objects-in-the-universe}

codecov · 2025-09-08T18:10:21Z

Codecov Report

❌ Patch coverage is 85.29412% with 5 lines in your changes missing coverage. Please review.
✅ Project coverage is 84.07%. Comparing base (964886d) to head (a6fe881).
⚠️ Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
...kages/datadog-instrumentations/src/helpers/hook.js	78.57%	3 Missing ⚠️
...s/datadog-instrumentations/src/helpers/register.js	71.42%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #6401   +/-   ##
=======================================
  Coverage   84.07%   84.07%           
=======================================
  Files         484      484           
  Lines       20246    20257   +11     
=======================================
+ Hits        17021    17031   +10     
- Misses       3225     3226    +1

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

BridgeAR

LGTM, just the deactivation of the warnings is not ideal in my opinion. Could we limit that a version, if that's only applying in a particular version?

integration-tests/helpers/index.js

pr-commenter · 2025-09-08T20:29:17Z

Benchmarks

Benchmark execution time: 2025-09-24 15:02:41

Comparing candidate commit a6fe881 in PR branch bengl/no-default-special-case with baseline commit 964886d in branch master.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 1612 metrics, 68 unstable metrics.

rochdev · 2025-09-11T16:39:46Z

packages/datadog-instrumentations/src/fastify.js


-addHook({ name: 'fastify', versions: ['>=3'] }, fastify => {
-  const wrapped = shimmer.wrapFunction(fastify, fastify => wrapFastify(fastify, true))
+addHook({ name: 'fastify', versions: ['>=3'] }, (fastify, _1, _2, isIitm) => {


As discussed offline, the fact that this is needed is probably due to a bug or something not properly handled in the new implementation. Let's fix it before landing.

bengl · 2025-09-18T18:16:16Z

packages/datadog-instrumentations/src/helpers/hook.js

    if (newExports &&
      (typeof newExports === 'object' || typeof newExports === 'function') &&
-      (moduleName === 'protobufjs' || !moduleName.includes('protobuf'))) {
+      (moduleName === 'protobufjs' || !moduleName.includes('protobuf')) && !moduleName.includes('aws-sdk')) {


@perhardhernandez Lets put a comment here indicating why these are special-cased. We ought to also be careful with includes, and instead specify as precisely as we can.

bengl · 2025-09-18T18:17:39Z

packages/datadog-instrumentations/src/helpers/hook.js

    ) {
-      newExports.default = onrequire(moduleExports.default, moduleName, moduleBaseDir, moduleVersion, isIitm)
+      defaultWrapResult = onrequire(moduleExports.default, moduleName, moduleBaseDir, moduleVersion, isIitm)
+      if (moduleName.includes('express')) {


@perhardhernandez Lets put a comment here indicating why this special-cased. We ought to also be careful with includes, and instead specify as precisely as we can.

packages/datadog-plugin-pino/test/integration-test/client.spec.js

BridgeAR · 2025-09-21T15:08:37Z

packages/datadog-plugin-pino/test/integration-test/client.spec.js

+      variants = varySandbox(sandbox, 'server.mjs', {
+        default: 'import pino from \'pino\'',
+        star: 'import * as modPino from \'pino\'; const pino = (...a) => modPino.default(...a)',
+        destructure: 'import { default as pino } from \'pino\'',


I think we miss one more case that we should also test:

import { pino } from 'pino'

I believe that might not be the case for all modules, but definitely for some.

That export is not possible as 'pino' is not directly exposed.

I did not check if it's pino in new versions or Node.js that allows it, but it's definitely possible to load pino like that with new versions :)

You are right, it does seem like it's pino version issue. If we want to add into the tests we would need to something like in the pg integration test file.

BridgeAR · 2025-09-21T15:15:44Z

packages/datadog-plugin-connect/test/integration-test/client.spec.js

+        default: 'import connect from \'connect\'',
+        star: 'import * as starConnect from \'connect\'; const connect = starConnect.default;',
+        destructure: 'import { default as connect } from \'connect\';'


Since most imports always look like these, should we just have shortcut that accepts the name of the module and that creates these three possible test cases automatically?
That way we only have to define them in more detail in case the default case does not work as expected, which is less frequent.

BridgeAR · 2025-09-21T15:16:57Z

packages/datadog-plugin-azure-functions/test/integration-test/client.spec.js

+    // TODO(bengl): The `varySandbox` helper function isn't well set-up for dealing
+    // with Azure Functions and the way the `func` command expects to find files. I
+    // have manually tested that all the usual import variants work, but really we ought
+    // to figure out a way of automating this.


We have to memorize to do that again before landing with the latest changes.

BridgeAR · 2025-09-21T15:19:23Z

packages/datadog-instrumentations/src/helpers/hook.js

    const filename = path.join(...parts)

-    if (this._patched[filename]) return moduleExports
+    if (this._patched[filename] && patched.has(moduleExports)) return patched.get(moduleExports)


I am uncertain why we use both cases. Would it be possible to get a comment explaining this? :)

I'm pretty sure using both doesn't even do anything? What happens if the set is removed?

The idea of keeping both was to ensure that if we ever encounter the same filename twice, we return the patched export, if that's not possible, is this._patched even needed (considering onrequire takes case of filename verification)? And if it is possible, my understanding is that moduleExports always enters our process as the original export object.

BridgeAR · 2025-09-21T15:20:07Z

packages/datadog-instrumentations/src/cassandra-driver.js

 })

-addHook({ name: 'cassandra-driver', versions: ['>=4.4'] }, cassandra => {
+addHook({ name: 'cassandra-driver', versions: ['>=4.4'] }, (cassandra, _1, _2, isIitm) => {


Before landing, it is probable good to rename _1 and _2 to the actual variable names, just with a leading underscore. That applies to all cases.

Since this same check is in multiple places, could it be an option instead? For example:

// Could be another name too, I'm not 100% what this even does. { name: '@azure/functions', versions: ['>=4'], patchDefault: false }

Then each individual hook doesn't need to know how to handle it and it fixes the weird signature of the function.

The above, was to prevent double wrapping when moduleExports.default and moduleExports share a same prototype, handling it as Roch suggested would work.

I am fine either way

rochdev · 2025-09-22T13:25:01Z

packages/datadog-instrumentations/src/helpers/hook.js

+    if (
+      isIitm &&
+      moduleExports.default &&
+      (typeof moduleExports.default === 'object' ||


Why is this check needed? If a hook is requesting the export it should always be called.

It coming from isIitm does not ensure that .default exist, hence the condition.

Sure, but why the explicit type check? Wouldn't checking just for the presence enough?

Express and next have some weird exports that are not of those types, hence the check.

rochdev · 2025-09-22T13:27:28Z

packages/datadog-instrumentations/src/helpers/hook.js

-    this._patched[filename] = true
+    if (newExports &&
+      (typeof newExports === 'object' || typeof newExports === 'function')) {
+      patched.set(moduleExports, newExports)


What does this do at this point since this code should never run more than once for an export now (since we switched back to using the filename)

Did we confirm that an export never runs more than once? If so why did we have the following

if (this._patched[filename]) return moduleExports

The idea of keeping both was to make sure we were always returning the patched export.

rochdev · 2025-09-22T20:13:55Z

packages/datadog-instrumentations/src/pino.js

 }

-addHook({ name: 'pino', versions: ['2 - 3', '4', '>=5 <5.14.0'] }, pino => {
+addHook({ name: 'pino', versions: ['2 - 3', '4'], patchDefault: false }, (pino, _1, _2, isIitm) => {


Is isIitm still needed here? The hook shouldn't have to know how to get the right export.

Yes, with out it pino tests fail. I'm guessing this is related to some of the weird things that pino do with their exports.

Can you look into it a bit more? I'd love to be able to just remove that parameter entirely if we can, and the hook should already be receiving both the export and the default. I wonder if it's because Pino needs the opposite of patchDefault?

Fixed ESM instrumentation, the previous model of assuming default === exports was very incorrect for ESM since they're never the exact same object. We also can't assume they should have the same properties, or even that they're both objects/functions, at least not in the general case. Therefore, we need to be specific about what's being instrumented in every single instrumentation. Added appropriate testing and modifications to ensure any known broken instrumentation is properly instrumented with all import scenarios. --------- Co-authored-by: perhardhernandez <[email protected]>

bengl requested review from a team as code owners September 8, 2025 18:09

BridgeAR reviewed Sep 8, 2025

View reviewed changes

integration-tests/helpers/index.js Outdated Show resolved Hide resolved

This comment has been minimized.

Sign in to view

bengl added the semver-patch label Sep 9, 2025

rochdev requested changes Sep 11, 2025

View reviewed changes

perhardhernandez force-pushed the bengl/no-default-special-case branch 2 times, most recently from b290d9c to ae9e749 Compare September 12, 2025 15:01

perhardhernandez closed this Sep 12, 2025

perhardhernandez reopened this Sep 12, 2025

perhardhernandez closed this Sep 12, 2025

perhardhernandez reopened this Sep 12, 2025

perhardhernandez mentioned this pull request Sep 12, 2025

Perhardhernandez/no default special case #6430

Closed

7 tasks

perhardhernandez force-pushed the bengl/no-default-special-case branch from 6eed43b to e9fc4f0 Compare September 15, 2025 13:18

perhardhernandez mentioned this pull request Sep 18, 2025

Perhardhernandez/no default special case updated (testing purposes) #6471

Closed

7 tasks

bengl commented Sep 18, 2025

View reviewed changes

perhardhernandez force-pushed the bengl/no-default-special-case branch from 65db538 to 607fb1a Compare September 19, 2025 20:12

BridgeAR reviewed Sep 21, 2025

View reviewed changes

rochdev reviewed Sep 22, 2025

View reviewed changes

perhardhernandez force-pushed the bengl/no-default-special-case branch 2 times, most recently from 0b4a374 to 1acd2c7 Compare September 22, 2025 16:15

rochdev reviewed Sep 22, 2025

View reviewed changes

perhardhernandez requested a review from rochdev September 23, 2025 14:16

rochdev previously approved these changes Sep 23, 2025

View reviewed changes

Sign commit

3ec4b08

perhardhernandez force-pushed the bengl/no-default-special-case branch from fe6096d to 3ec4b08 Compare September 24, 2025 13:42

readded missing unhook property

a6fe881

perhardhernandez dismissed rochdev’s stale review via a6fe881 September 24, 2025 14:55

perhardhernandez requested a review from rochdev September 24, 2025 15:59

rochdev approved these changes Sep 24, 2025

View reviewed changes

perhardhernandez merged commit a38a294 into master Sep 24, 2025
774 of 776 checks passed

perhardhernandez deleted the bengl/no-default-special-case branch September 24, 2025 18:18

This was referenced Sep 25, 2025

v5.68.0 proposal #6426

Closed

v5.68.0 proposal #6576

Merged

remove special-casing of default in hook.js #6401

remove special-casing of default in hook.js #6401

Uh oh!

Conversation

bengl commented Sep 8, 2025 • edited by BridgeAR Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What does this PR do?

Motivation

Additional Notes

Uh oh!

github-actions bot commented Sep 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Overall package size

Uh oh!

codecov bot commented Sep 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

BridgeAR left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

pr-commenter bot commented Sep 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmarks

Uh oh!

This comment has been minimized.

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

perhardhernandez Sep 22, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

perhardhernandez Sep 22, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

rochdev Sep 22, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

bengl commented Sep 8, 2025 •

edited by BridgeAR

Loading

github-actions bot commented Sep 8, 2025 •

edited

Loading

codecov bot commented Sep 8, 2025 •

edited

Loading

pr-commenter bot commented Sep 8, 2025 •

edited

Loading

perhardhernandez Sep 22, 2025 •

edited

Loading

perhardhernandez Sep 22, 2025 •

edited

Loading

rochdev Sep 22, 2025 •

edited

Loading