Support nullable base types for (*Build[T]) #638
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
gernest
force-pushed
the
nullable
branch
4 times, most recently
from
a76e291 to
2d1127e
Compare
Before, we had `null` tag that set schema column to nullable. However the base types allowed were `int64|float64|bool|string`. This begged the question, what is `null`? for a `nullable` field of type `int64` do we interpret `0` as `null`?, probably not. We were offering option for something that was impossible to represent. Fortunately, we can safely represent null values with out type system. For base types `*int64|*float64|*bool|*string` defines nullable base types. This commit add support for both `int64|float64|bool|string` and `*int64|*float64|*bool|*string` for base types . This works with dynamic columns too. With this change `null` tag is redundant so It was removed. Fields of type `*int64|*float64|*bool|*string` will automatically generate nullable schema column. NOTE: I also discovered a bug where when `T` has dynamic column and you append multiple `T` without any dynamic column followed by T with dynamic column caused a panic. The issue was how we adjusted dynamic columns to match current `.Append` rows state. The fix is included in this commit because I needed this behavior in tests.
gernest
added a commit
to gernest/frostdb
that referenced
this pull request
Dec 23, 2023
* main: Optimize dynparquet..deserializeDynamicColumns (polarsignals#652) Remove bufutils.Dedupe (polarsignals#650) Performance improvement on pqarrow.RecordDynamicCols (polarsignals#644) Remove dynparquet.ValuesToBuffer (polarsignals#645) parts list no longer used (polarsignals#649) dynparquet/samples: Make ToRecord consistent (polarsignals#648) parquet/converter: New Record call ResetFull on dictionary builders (polarsignals#646) parquet-tool: check in cmd col (polarsignals#643) Remove unused function (polarsignals#641) Support nullable base types for (*Build[T]) (polarsignals#638)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Before, we had
nulltag that set schema column to nullable. However the base types allowed wereint64|float64|bool|string.This begged the question, what is
null? for anullablefield of typeint64do we interpret0asnull?, probably not. We were offering option for something that was impossible to represent.Fortunately, we can safely represent null values with out type system. For
base types
*int64|*float64|*bool|*stringdefines nullable base types.This commit add support for both
int64|float64|bool|stringand*int64|*float64|*bool|*stringfor base types . This works with dynamic columns too.With this change
nulltag is redundant so It was removed. Fields of type*int64|*float64|*bool|*stringwill automatically generate nullable schemacolumn.
NOTE: I also discovered a bug where when
Thas dynamic column and you append multipleTwithout any dynamic column followed by T with dynamic column caused a panic. The issue was how we adjusted dynamic columns to match current.Appendrows state. The fix is included in this commit because I needed this behavior in tests.