spaceSplit and spaceTimeSplit accept LATERAL column inputs#132
Open
estebanzimanyi wants to merge 2 commits into
Open
spaceSplit and spaceTimeSplit accept LATERAL column inputs#132estebanzimanyi wants to merge 2 commits into
estebanzimanyi wants to merge 2 commits into
Conversation
Re-register both splitters as in_out_function table functions so the per-row parameters (tgeompoint, sizing constants, origin, duration, torigin, bitmatrix) are read at exec time from the input DataChunk rather than bound once from literals. Mirrors duckdb/extension/icu/ icu-table-range.cpp. The bind callback now only declares return types; SpaceSplitLocalState carries iteration state so a single input row's bins stream out in STANDARD_VECTOR_SIZE-bounded chunks. When the input chunk is drained, the function leaves output.size()==0 and returns NEED_MORE_INPUT, which matches PhysicalTableScan's FINISHED contract (return value ignored, zero-size chunk terminates) and PhysicalTableInOutFunction's LATERAL contract (NEED_MORE_INPUT advances to the next input chunk). Adds 058c_tpoint_split_lateral.test covering the column-input path so the LATERAL plumbing has dedicated coverage alongside 058b's literal-input regression.
…o feat/spacesplit-lateral
estebanzimanyi
added a commit
that referenced
this pull request
May 14, 2026
The portfile pinned REF f11b7443e (Mar 30) but the SHA512 corresponds to 742c1fb5 (May 11). Vcpkg's SHA512 cache silently masked the discrepancy on some runs (PR #132 succeeded by reusing a cached tarball that matched the SHA512), but a fresh download fetches f11b7443e's content which still spells the catalog enum `meosType` (lowercase). That makes `using meosType = MeosType;` in src/include/tydef.hpp fail to compile because `MeosType` (the Pascal-cased rename introduced by MobilityDB PR #790, merged Apr 28) is not yet defined at REF f11b7443e. Update REF to 742c1fb5 so it matches the existing SHA512. The tarball content vcpkg actually used in PR #132's green run is now the explicit pin, and the forward-compat alias compiles cleanly. Pure portfile fix; no other changes.
estebanzimanyi
added a commit
that referenced
this pull request
May 14, 2026
The portfile pinned REF f11b7443e (Mar 30) but the SHA512 corresponds to 742c1fb5 (May 11). Vcpkg's SHA512 cache silently masked the discrepancy on some runs (PR #132 succeeded by reusing a cached tarball that matched the SHA512), but a fresh download fetches f11b7443e's content which still spells the catalog enum `meosType` (lowercase). That makes `using meosType = MeosType;` in src/include/tydef.hpp fail to compile because `MeosType` (the Pascal-cased rename introduced by MobilityDB PR #790, merged Apr 28) is not yet defined at REF f11b7443e. Update REF to 742c1fb5 so it matches the existing SHA512. The tarball content vcpkg actually used in PR #132's green run is now the explicit pin, and the forward-compat alias compiles cleanly. Pure portfile fix; no other changes.
Member
Author
Reviewer's quickstart — ~2-3 minutesWhat this PR does: spaceSplit and spaceTimeSplit accept LATERAL column inputs. Risk: focused, single-purpose. Spot-check the source diff + matching tests; CI confirms. Cross-link: Linux arm64 CI here needs #161 for the orthogonal |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Re-register both splitters as in_out_function table functions so per-row parameters (tgeompoint, sizing constants, origin, duration, torigin, bitmatrix) are read at exec time from the input DataChunk rather than bound once from literals; pattern mirrors duckdb/extension/icu/icu-table-range.cpp. When the input chunk drains, the function leaves output.size() at zero and returns NEED_MORE_INPUT, matching PhysicalTableScan's FINISHED-by-chunk-size contract and PhysicalTableInOutFunction's NEED_MORE_INPUT advance. test/sql/parity/058c_tpoint_split_lateral.test exercises the LATERAL column-input path alongside 058b's literal-input regression. Stacked on #126, which carries the MEOS pin bump and API adaptations the build depends on.