Skip to content

[Spark] Fix schema evolution issue with nested struct (within a map) and column renamed #4575

[Spark] Fix schema evolution issue with nested struct (within a map) and column renamed

[Spark] Fix schema evolution issue with nested struct (within a map) and column renamed #4575

Workflow file for this run

name: "Unidoc"
on: [push, pull_request]
jobs:
build:
name: "U: Scala ${{ matrix.scala }}"
runs-on: ubuntu-20.04
strategy:
matrix:
# These Scala versions must match those in the build.sbt
scala: [2.13.13, 2.12.18]
steps:
- name: install java
uses: actions/setup-java@v3
with:
distribution: "zulu"
java-version: "8"
- uses: actions/checkout@v3
- name: generate unidoc
run: build/sbt "++ ${{ matrix.scala }}" unidoc