[Docs][Format] Make Variant value field mandatory
- Dominant language
- C++
- Stars
- 17.1k
- Forks
- 4.3k
- Avg merge
- 3d 13h
- Merged PRs (30d)
- 88
Description
### Describe the enhancement requested
Fix wording in the Arrow Variant spec to make the `value` field mandatory.
https://github.com/apache/arrow/blob/f10c93c000775c5fb30ca230f540358611091589/docs/source/format/CanonicalExtensions.rst?plain=1#L448
Per the conversation in https://github.com/apache/parquet-format/pull/591#issuecomment-4897543546
Note the requirement is level-dependent in the Parquet spec, so the Arrow wording should mirror it level by level rather than making `value` mandatory everywhere:
- top-level variant group: `value` must be present
- shredded object fields: the field group contains both `value` and `typed_value`
- list elements: `value` [may be omitted](https://github.com/apache/parquet-format/blob/master/VariantShredding.md#arrays) when shredding elements as a specific type, but at least one of the two fields must be present
### Component(s)
Format
Contributor guide
Research direction
Open docs/source/format/CanonicalExtensions.rst at the referenced line and compare its Variant wording with parquet-format PR #591 and the linked VariantShredding arrays guidance. Update the text to describe the required fields separately for top-level groups, shredded object fields, and list elements, then verify each level matches the referenced Parquet requirements.
Written by the indexing model from the issue text.
Assessment
- Domain
- documentation
- Issue type
- Documentation
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Activity status
- Quiet
- Clarity
- Clearly specified
- Newbie friendliness
- 72/100