Refactor Extend to preserve declaration links

[Andarist]

This targets v4 and it's basically the same change as in #4150 (but that targets v3).

The proposed version is better because it preserves declaration links and that helps in-IDE navigation. To get proper declaration links you need to iterate over a single "modifier object" and [P in keyof T | keyof U] breaks that (it sources keys from 2 objects, so the compiler can't decide from which one the declaration, and modifiers, should be picked from). See this TS playground

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

🪧 Tips

Chat

There are 3 ways to chat with CodeRabbit:

• Review comments: Directly reply to a review comment made by CodeRabbit. Example:
  • I pushed a fix in commit <commit_id>, please review it.
  • Generate unit testing code for this file.
  • Open a follow-up GitHub issue for this discussion.
• Files and specific lines of code (under the "Files changed" tab): Tag @coderabbitai in a new review comment at the desired location with your query. Examples:
  • @coderabbitai generate unit testing code for this file.
  • @coderabbitai modularize this function.
• PR comments: Tag @coderabbitai in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples:
  • @coderabbitai gather interesting stats about this repository and render them as a table. Additionally, render a pie chart showing the language distribution in the codebase.
  • @coderabbitai read src/utils.ts and generate unit testing code.
  • @coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.
  • @coderabbitai help me debug CodeRabbit configuration file.

Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments.

CodeRabbit Commands (Invoked using PR comments)

• @coderabbitai pause to pause the reviews on a PR.
• @coderabbitai resume to resume the paused reviews.
• @coderabbitai review to trigger an incremental review. This is useful when automatic reviews are disabled for the repository.
• @coderabbitai full review to do a full review from scratch and review all the files again.
• @coderabbitai summary to regenerate the summary of the PR.
• @coderabbitai generate docstrings to generate docstrings for this PR.
• @coderabbitai resolve resolve all the CodeRabbit review comments.
• @coderabbitai configuration to show the current CodeRabbit configuration for the repository.
• @coderabbitai help to get help.

Other keywords and placeholders

• Add @coderabbitai ignore anywhere in the PR description to prevent this PR from being reviewed.
• Add @coderabbitai summary to generate the high-level summary at a specific location in the PR description.
• Add @coderabbitai anywhere in the PR title to generate the title automatically.

CodeRabbit Configuration File (.coderabbit.yaml)

• You can programmatically configure CodeRabbit by adding a .coderabbit.yaml file to the root of your repository.
• Please see the configuration documentation for more information.
• If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: # yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json

Documentation and Community

• Visit our Documentation for detailed information on how to use CodeRabbit.
• Join our Discord Community to get help, request features, and share feedback.
• Follow us on X/Twitter for updates and announcements.

[Andarist]

@colinhacks this should pass CI now, could you re-approve the workflow run?

[Andarist]

I'm not sure why Flatten was previously within ExtendObject, what was the deal with Identity here or why do you have both ExtendShape and ExtendObject in the first place. If you share more context behind it then I could try to figure out a different solution.

I guess that maybe ExtendObject is here just to use different constraints. If that's the case then I think moving Flatten into ExtendShape makes sense. Let just all its instantiations "benefit" from it instead of keeping it solely within ExtendObject.

I also thought about using an extra branch here, like: keyof A | keyof B extends keyof A ? B : ..., but that would mess up correct modifiers preservation (for example, I ran into issues in partial-interface.test.ts)

[colinhacks]

Love this!!

Rebased to fix CI.

Cleaned up the utils — the Identity/Flatten were in there to force TS to resolve the the type. With the old implementation of ExtendShape (without Identity) I was seeing unresolved types in my return types, e.g.:

But it seems Identity/Flatten aren't necessary with the new and improved version. Very nice 😊

[colinhacks]

New benchmarks for Zod 3 (pre and post this change) and Zod 4 (pre and post)

pnpm bench omit-extend

> @zod/tsc-perftest@1.0.0 bench /Users/colinmcd94/Documents/projects/zod/packages/tsc
> tsx ./bench/index.ts omit-extend

Generating 40 chained calls to .extend() w/ 6 keys...
╔═════════════════════╗
║     Zod v3.24.2     ║
╚═════════════════════╝
Generating 14 chained calls to .extend() w/ 6 keys...

> @zod/tsc-perftest@1.0.0 build:bench /Users/colinmcd94/Documents/projects/zod/packages/tsc
> tsc -p tsconfig.bench.json

Files:                         163
Lines of Library:            39498
Lines of Definitions:        54891
Lines of TypeScript:            93
Lines of JavaScript:             0
Lines of JSON:                   0
Lines of Other:                  0
Identifiers:                 87859
Symbols:                     84027
Types:                       11099
Instantiations:            6971996
Memory used:               127755K
Assignability cache size:     3207
Identity cache size:             6
Subtype cache size:              0
Strict subtype cache size:       0
I/O Read time:               0.01s
Parse time:                  0.16s
ResolveModule time:          0.02s
ResolveTypeReference time:   0.00s
ResolveLibrary time:         0.01s
Program time:                0.23s
Bind time:                   0.09s
Check time:                  1.27s
transformTime time:          0.15s
commentTime time:            0.00s
I/O Write time:              0.00s
printTime time:              0.16s
Emit time:                   0.16s
Total time:                  1.75s
╔═════════════════════╗
║     Zod v3.24.3     ║
╚═════════════════════╝
Generating 14 chained calls to .extend() w/ 6 keys...

> @zod/tsc-perftest@1.0.0 build:bench /Users/colinmcd94/Documents/projects/zod/packages/tsc
> tsc -p tsconfig.bench.json

Files:                         163
Lines of Library:            39498
Lines of Definitions:        54891
Lines of TypeScript:            93
Lines of JavaScript:             0
Lines of JSON:                   0
Lines of Other:                  0
Identifiers:                 87867
Symbols:                     84081
Types:                       10814
Instantiations:             206486
Memory used:               116895K
Assignability cache size:     3168
Identity cache size:             6
Subtype cache size:              0
Strict subtype cache size:       0
I/O Read time:               0.01s
Parse time:                  0.16s
ResolveModule time:          0.02s
ResolveTypeReference time:   0.00s
ResolveLibrary time:         0.01s
Program time:                0.23s
Bind time:                   0.08s
Check time:                  0.17s
transformTime time:          0.02s
commentTime time:            0.00s
I/O Write time:              0.00s
printTime time:              0.03s
Emit time:                   0.03s
Total time:                  0.51s
╔══════════════════════╗
║     Zod v4 (pre)     ║
╚══════════════════════╝
Generating 14 chained calls to .extend() w/ 6 keys...

> @zod/tsc-perftest@1.0.0 build:bench /Users/colinmcd94/Documents/projects/zod/packages/tsc
> tsc -p tsconfig.bench.json

Files:                         177
Lines of Library:            39498
Lines of Definitions:        56673
Lines of TypeScript:            93
Lines of JavaScript:             0
Lines of JSON:                   0
Lines of Other:                  0
Identifiers:                 94094
Symbols:                     65480
Types:                        1454
Instantiations:               8385
Memory used:               111503K
Assignability cache size:      225
Identity cache size:             0
Subtype cache size:              6
Strict subtype cache size:       0
I/O Read time:               0.01s
Parse time:                  0.17s
ResolveModule time:          0.03s
ResolveTypeReference time:   0.00s
ResolveLibrary time:         0.01s
Program time:                0.25s
Bind time:                   0.09s
Check time:                  0.06s
transformTime time:          0.01s
commentTime time:            0.00s
I/O Write time:              0.00s
printTime time:              0.02s
Emit time:                   0.02s
Total time:                  0.41s
╔═══════════════════════╗
║     Zod v4 (curr)     ║
╚═══════════════════════╝
Generating 14 chained calls to .extend() w/ 6 keys...

> @zod/tsc-perftest@1.0.0 build:bench /Users/colinmcd94/Documents/projects/zod/packages/tsc
> tsc -p tsconfig.bench.json

Files:                         178
Lines of Library:            39498
Lines of Definitions:        56682
Lines of TypeScript:            93
Lines of JavaScript:             0
Lines of JSON:                   0
Lines of Other:                  0
Identifiers:                 94093
Symbols:                     65488
Types:                        1479
Instantiations:               9341
Memory used:               111275K
Assignability cache size:      225
Identity cache size:             0
Subtype cache size:              6
Strict subtype cache size:       0
I/O Read time:               0.01s
Parse time:                  0.17s
ResolveModule time:          0.03s
ResolveTypeReference time:   0.00s
ResolveLibrary time:         0.01s
Program time:                0.24s
Bind time:                   0.09s
Check time:                  0.06s
transformTime time:          0.01s
commentTime time:            0.00s
I/O Write time:              0.00s
printTime time:              0.02s
Emit time:                   0.02s
Total time:                  0.40s

Unfortunately there seems to be no escaping some kind of exponential explosion with chained called to .extend() that I can find. (Mateusz, this comment is mostly for posterity, not trying to nerd snipe you here. 🙃)

To replicate:

pnpm i
cd packages/tsc
pnpm bisect

This runs a script that increases the number of chained .extend() calls by 2x until tsc fails. Instantiations are very superlinear. Valibot has found a way to keep it linear...by avoiding a .extend() method/function altogether and exclusively recommending spread syntax. This works for Zod too.

const A = z.object({ ... });

const B = z.object({ 
  ...A.shape, 
  // new fields...
})

Thinking an .extends() API is a just footgun that I should discourage over spreading .shape. Not sure why a method formulation explodes, but not the spread.

[Andarist]

Unfortunately there seems to be no escaping some kind of exponential explosion with chained called to .extend() that I can find. (Mateusz, this comment is mostly for posterity, not trying to nerd snipe you here. 🙃)

I think this is what I was trying to reproduce here. Maybe v3 and v4 behave differently for some reason here? Or maybe there is some crucial difference between what you are generating and the stuff I prepared?

I have checked this bench briefly and from what I see:

• the current answer for Zod 4 is 38
• with the build from here it's 46

So it's higher - but not by a lot. But it gives up based on instantiationDepth's limit (100) and not on the instantiationCount's count limit (5000000). Despite being able to check more "layers" it's also substantially faster (total time 2.90s -> total time 0.52s).