Add support for generated identifier format strings (id prefixing). #509

sanny-io · 2025-12-17T04:34:21Z

Sample Usage

model User {
    id Int @id
    uuid7 String @default(uuid(7, 'user_uuid7_%s'))
    cuid String @default(cuid(2, 'user_cuid_%s'))
    nanoid String @default(nanoid(21, 'user_nanoid_%s'))
    ulid String @default(ulid('user_ulid_%s'))
}

There are arguments to be made that @default could be a better place to handle this, which would open up opportunities for default values to be composed of complex expressions.

For now, this seems OK.

Summary by CodeRabbit

New Features
- ID generation functions (UUID, CUID, NanoID, ULID) now accept an optional format string; the format may include a "%s" placeholder to embed the generated value and produce custom identifier shapes.
Tests
- Added end-to-end tests verifying formatted ID generation across multiple strategies and creation scenarios (top-level, nested, deeply nested) to ensure prefixes and formatting are applied consistently.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

coderabbitai · 2025-12-17T04:34:30Z

Walkthrough

Public stdlib ID generator signatures were extended to accept an optional format parameter; ORM ID generation now applies a formatting helper that validates a %s placeholder and substitutes the generated value; and a new E2E test verifies formatted generated IDs across multiple generators and nested create scenarios.

Changes

Cohort / File(s)	Summary
Schema: stdlib model `packages/language/res/stdlib.zmodel`	Extended public function signatures to accept an optional `format: String?` parameter: `uuid`, `cuid`, `nanoid` now accept `format`, and `ulid` now accepts `format: String?`.
Runtime: ID generation formatting `packages/orm/src/client/crud/operations/base.ts`	ID generation for `cuid`, `uuid`, `nanoid`, and `ulid` now routes generated values through a new private `formatGeneratedValue(generated, expr?)` helper which validates the format expression is a string containing `%s` and substitutes the first occurrence with the generated value.
Tests: e2e generated ID formats `tests/e2e/orm/client-api/generated-id-format-strings.test.ts`	Added E2E tests that push a SQLite-backed schema exercising uuid/uuid7, cuid/cuid2, nanoid/nanoid8, and ulid strategies across `User`/`Post`/`Comment` models and assert formatted generated IDs for top-level, nested, and deeply nested creates.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Inspect formatGeneratedValue() error paths and handling of non-literal or non-string expressions.
Verify correct mapping of schema argument indices to generator parameters (length vs format).
Confirm tests cover both presence and absence of format and nested creation permutations.

Possibly related PRs

getting more policy test cases to pass #3 — modifies default-value generation/processing representation and touches the same default-value/generator handling area.

Poem

🐰 I hop through code and stitch a tag,
I find the %s and make it brag.
UUIDs, CUIDs, nanoids in line,
ULIDs sparkle with a formatted sign.
Hooray — new IDs all look divine! ✨

Pre-merge checks and finishing touches

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title directly describes the main change: adding support for format strings in generated identifiers (id prefixing), which matches the core functionality across all modified files and the new test.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing touches

📝 Generate docstrings

🧪 Generate unit tests (beta)

Create PR with unit tests
Post copyable unit tests in a comment

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

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

coderabbitai

Actionable comments posted: 1

🧹 Nitpick comments (2)

tests/e2e/orm/client-api/generated-id-format-strings.test.ts (2)
7-1060: Consider refactoring verbose schema definition.

The schema definition spans over 1000 lines with significant duplication—each field's default value is defined identically in both attributes[0].args[0].value and the default property. While this may be required by the SchemaDef type structure, consider:

Extract a helper function to generate ID field definitions:
function makeIdField(name: string, funcName: string, args: any[], format: string) {
  const callExpr = { kind: 'call', function: funcName, args };
  return {
    name,
    type: 'String',
    attributes: [{ name: '@default', args: [{ name: 'value', value: callExpr }] }],
    default: callExpr
  };
}
This would reduce the ~1050-line schema to ~100-200 lines, improving readability and maintainability.

1062-1167: Solid E2E coverage for core scenarios.

The tests effectively validate ID formatting across top-level, nested, and deeply nested creation scenarios, covering all supported ID generation strategies (uuid, uuid7, cuid, cuid2, nanoid, nanoid8, ulid).

Consider adding test cases for edge cases:

Invalid format strings (missing "%s") to verify validation errors

Format strings with special characters or unusual patterns

Verification that the generated suffix is a valid ID (not just the prefix)

📜 Review details

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between dbe7257 and afe1e7d.

📒 Files selected for processing (3)

packages/language/res/stdlib.zmodel (2 hunks)
packages/orm/src/client/crud/operations/base.ts (2 hunks)
tests/e2e/orm/client-api/generated-id-format-strings.test.ts (1 hunks)

🧰 Additional context used

📓 Path-based instructions (2)

**/*.zmodel

📄 CodeRabbit inference engine (CLAUDE.md)

**/*.zmodel: Always run zenstack generate after modifying ZModel schemas
ZModel schema files should define database structure and policies that compile to TypeScript via zenstack generate

Files:

packages/language/res/stdlib.zmodel

tests/e2e/**/*.{ts,tsx}

📄 CodeRabbit inference engine (CLAUDE.md)

E2E tests should validate real-world schema compatibility with established projects

Files:

tests/e2e/orm/client-api/generated-id-format-strings.test.ts

🧠 Learnings (7)