fix(schema): fix OASv3 schema generation fails when model used in @OneOf has kind property

[Romakita]

Closes: #2413

[github-actions]

Benchmarks

• Machine: linux x64 | 2 vCPUs | 6.8GB Mem
• Node: v16.20.2
• Run: Sun Sep 03 2023 08:39:21 GMT+0000 (Coordinated Universal Time)
• Method: autocannon -c 100 -d 10 -p 10 localhost:3000 (two rounds; one to warm-up, one to measure)

	Version	Router	Requests/s	Latency	Throughput/Mb
fastify	3.29.4	✓	45097.6	21.73	8.04
nest-fastify	8.4.3	✓	40092.0	24.47	7.15
koa	2.13.4	✗	32915.2	29.83	5.87
fastify-injector	3.29.4	✓	23022.9	42.90	5.25
express	4.18.1	✓	8800.6	112.66	1.57
tsed-koa	7.34.7	✓	8219.8	120.67	7.18
nest	8.4.3	✓	7621.2	130.19	1.83
fastify-big-json	3.29.4	✓	7499.2	132.15	86.26
express-injector	4.18.1	✗	6916.9	143.23	1.58
tsed-express	7.34.7	✓	6382.8	155.31	1.17
express-morgan	4.18.1	✗	5117.5	193.47	0.91

Explanation

The benchmark shows a performance difference between the frameworks. We note that Ts.ED is often last. In fact, Ts.ED uses features useful to a production application which reduce its performance.

For example, Ts.ED initializes a sandbox (async_hook) for each request in order to work in an isolated context if necessary.
It also initializes the elements necessary for monitoring requests in a log manager.

All this at a necessary cost that reflects the reality of a production application ;)

[github-actions]

Benchmarks

• Machine: linux x64 | 2 vCPUs | 6.8GB Mem
• Node: v16.20.2
• Run: Sun Sep 03 2023 08:43:54 GMT+0000 (Coordinated Universal Time)
• Method: autocannon -c 100 -d 10 -p 10 localhost:3000 (two rounds; one to warm-up, one to measure)

	Version	Router	Requests/s	Latency	Throughput/Mb
fastify	3.29.4	✓	45033.6	21.73	8.03
nest-fastify	8.4.3	✓	41356.8	23.78	7.37
koa	2.13.4	✗	36112.8	27.21	6.44
fastify-injector	3.29.4	✓	23753.6	41.60	5.41
express	4.18.1	✓	9060.6	109.45	1.62
tsed-koa	7.34.7	✓	7987.1	124.18	6.98
fastify-big-json	3.29.4	✓	7960.4	124.67	91.59
nest	8.4.3	✓	7685.0	129.04	1.85
express-injector	4.18.1	✗	7189.2	137.99	1.64
tsed-express	7.34.7	✓	6630.4	149.46	1.21
express-morgan	4.18.1	✗	5171.4	191.66	0.92

Explanation

The benchmark shows a performance difference between the frameworks. We note that Ts.ED is often last. In fact, Ts.ED uses features useful to a production application which reduce its performance.

For example, Ts.ED initializes a sandbox (async_hook) for each request in order to work in an isolated context if necessary.
It also initializes the elements necessary for monitoring requests in a log manager.

All this at a necessary cost that reflects the reality of a production application ;)

[Romakita]

🎉 This PR is included in version 7.34.8 🎉

The release is available on:

• v7.34.8
• GitHub release

Your semantic-release bot 📦🚀

[Romakita]

🎉 This PR is included in version 7.35.0-beta.5 🎉

The release is available on:

• v7.35.0-beta.5
• GitHub release

Your semantic-release bot 📦🚀