You would expect very little possibility to see the coordinative や and the case particle に next to each other.

や factors out the shortest different sections, so you will find [particle]+や in a `XやY` composition that _X_ and _Y_ ends in different particles.

> {東京**や**仙台]*から*来た人 (diff. nouns w/ same particle)
>
> 在籍期間が[2012年*まで*‌**や**2018年*から*]の人 (diff. nouns w/ diff. particles)

That means the theoretical order would be にや, but in reality, you will be not likely to encounter such a situation, because に is one of central case marker in Japanese, and や stricter about syntactic leanness than its kinds* (such as か, とか and やら), which means, you need another particle, that is different than に but compared to に, to make something like にや appear. That would result in a very terrible type of [zeugma][1], like conjoining subject and object together.

---
\* cf.

> [じぶんがいじめられたら先生*に*‌**とか**ともだち*に*そうだんする。][2]  
> × 先生*に*‌**や**ともだち*に*そうだんする。


  [1]: https://en.wikipedia.org/wiki/Zeugma_and_syllepsis
  [2]: http://www.nhk.or.jp/ijimezero/sengen_school/kantou/11_91.html