test: expand real-world JSON corpus with production API samples

[membphis]

Background

qjson's current real-world test data is limited to:

• twitter.json (simdjson corpus) — unicode-heavy social media
• small_api.json / medium_resp.json — synthetic chat completion payloads
• amazon_cellphones.ndjson — product data

Production JSON parsers encounter diverse API response shapes. V8 and simdjson maintain broader corpora to catch compatibility edge cases.

Goal

Expand fixture coverage to include common production API response patterns, improving confidence in real-world compatibility.

Proposed Additions

Source	Payload Type	Why It Matters
GitHub API	Issue/PR responses	Deeply nested, many optional fields, markdown in strings
Stripe API	Payment objects	Decimal-sensitive numbers, nested metadata
Kubernetes	Resource manifests	Deep nesting, large arrays, YAML-originated quirks
AWS API	EC2/S3 responses	XML-to-JSON conversion artifacts, verbose structures
OpenAPI/Swagger	Schema definitions	$ref heavy, recursive structures

Acceptance Criteria

• Add 3-5 new fixture files representing distinct API patterns
• Each fixture added to tests/fixtures/manifest.json with appropriate checks
• Fixtures parse correctly in both EAGER and LAZY modes
• At least one fixture >100KB for large-payload coverage
• Document fixture sources and licensing in tests/fixtures/README.md

Technical Notes

• Prefer fixtures with permissive licenses (MIT, Apache, public domain)
• Sanitize any PII from real API responses before committing
• Consider adding a ci: ["pr"] entry for smaller fixtures to catch regressions early

References

• Current manifest: tests/fixtures/manifest.json
• simdjson jsonexamples: https://github.com/simdjson/simdjson/tree/master/jsonexamples

[coderabbitai]

🔗 Related PRs

#43 - test: add cjson and simdjson fixture coverage [merged]
#78 - test(fuzz): add nesting-depth stack-safety target [merged]
#99 - ci: test qjson against OpenResty runtimes [merged]
#141 - test: add FFI grammar corpus [merged]
#144 - test: strengthen lazy phase2 fuzz replay [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!