fix: sync MEOS API drift (meosType→MeosType + spatial-rel arg drop)#118
Closed
estebanzimanyi wants to merge 1 commit intomainfrom
Closed
fix: sync MEOS API drift (meosType→MeosType + spatial-rel arg drop)#118estebanzimanyi wants to merge 1 commit intomainfrom
estebanzimanyi wants to merge 1 commit intomainfrom
Conversation
Two upstream MEOS changes that the source had not caught up to:
1. MEOS catalog header now exposes the type as MeosType (PascalCase).
MobilityDuck source historically used meosType (lowercase). Sweep
rename across temporal/{set,span,spanset}.hpp,
temporal/temporal_functions.hpp, geo/tgeompoint_functions.cpp,
index/rtree_module.hpp, etc.
2. The MEOS exports tcontains_geo_tgeo, tdisjoint_{geo_tgeo,tgeo_geo,
tgeo_tgeo}, tintersects_{geo_tgeo,tgeo_geo,tgeo_tgeo},
ttouches_{geo_tgeo,tgeo_geo}, tdwithin_{tgeo_tgeo,tgeo_geo,geo_tgeo}
lost their trailing bool restr, bool at_value parameters. The result
is now always the full tbool; restriction is composed at the call
site:
Temporal *ret = tcontains_geo_tgeo(gs, tgeom);
if (ret && restr) {
Temporal *r2 = tbool_at_value(ret, at_value);
free(ret);
ret = r2;
}
The 3-arg form for tdwithin_* keeps the dist parameter; only
restr/at_value are dropped.
Currently make release on main fails with 'meosType does not name a
type' and 'too many arguments to function tcontains_geo_tgeo' (× 13
call sites). Without this, no parity work can build locally and only
branches that already happen to carry this fix get green CI.
Cherry-picked from perf/duck-stack-attempt2 (commits 7a8cc22 and
c37b9e2).
estebanzimanyi
force-pushed
the
fix/api-drift-sync
branch
from
cd1d5d2 to
f4b1679
Compare
3 tasks
Member
Author
|
Superseded by #121 |
estebanzimanyi
added a commit
that referenced
this pull request
May 10, 2026
#116 + #118) Two cleanups landing together so the rest of the open PR queue can build and merge. Part 1 — revert main's single-timezone commits (was #116): - 39921f1 fix(tz): single-timezone model — drops the AutoLoadExtension(icu) + SetOptionByName("TimeZone", ...) block from LoadInternal. - 08a5598 docs(tz): clarify two-timezone reality in comments. - 47df9ca fix(tz): set MEOS timezone to Europe/Brussels — drops meos_initialize_timezone("Europe/Brussels"). - d3d9810 fix(tz): force MEOS UTC timezone — drops the earlier meos_initialize_timezone("UTC") so LoadInternal matches PR #111's parent state exactly. - Shift remaining +01 expectations to +00 in 23 test files (calendar dates unchanged; only displayed offset shifts to match TZ=UTC). Part 2 — sync MEOS API drift (was #118): - meosType → MeosType sweep (16 files): MEOS catalog header now exposes the type as MeosType (PascalCase); MobilityDuck source historically used meosType (lowercase). - t{contains,disjoint,intersects,touches,dwithin}_* arg drop (12 call sites): MEOS exports lost their trailing bool restr, bool at_value parameters; restriction is now composed at the call site via tbool_at_value when needed. Result: LoadInternal collapses to ConfigureMeosSridCsvOnce() + meos_initialize() + error handler, exactly matching PR #111's parent state. All MobilityDuck source compiles against the current vcpkg MEOS. PR #111 then replaces the call_once block with EnsureMeosInitializedOnThread() — clean diff, no policy fight. The +00 shift is interim shape per the project's timezone-neutral policy (PR #111 / commit 9dd765a); the longer-term direction is to rewrite each assertion to value-equality / accessor / asText round-trip form, caught by scripts/lint-tz-pinned-tests.py (PR #120). ## Closes - #116 (consolidate: revert main's tz commits and align tests with TZ=UTC) - #118 (fix: sync MEOS API drift)
Member
Author
|
Folded into #121 (same) per the ecosystem PR consolidation policy. The consolidation PR explicitly declares this PR as one of its inputs in the |
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.
Summary
Two cherry-picked commits unblocking local compilation against the current vcpkg MEOS:
fe4a0da refactor: meosType -> MeosType (sync to upstream MEOS PR #785)— sweeps every reference of the lowercasemeosTypetypedef in MobilityDuck source to the PascalCaseMeosTypethat the current MEOS catalog header exposes.cd1d5d2 fix(geo): adapt 12 t{contains,disjoint,intersects,touches,dwithin}_* call sites to 3-arg MEOS API— thetcontains_geo_tgeo,tdisjoint_*,tintersects_*,ttouches_*,tdwithin_*exports lost theirrestr/at_valuetrailing arguments (the result is now always the full tbool — restriction is composed at the call site viatbool_at_valuewhen needed). Updates the 12 call sites intgeompoint_functions.cpp.Both commits originate from the parallel
perf/duck-stack-attempt2work (7a8cc22,c37b9e2). Cherry-picked here ontomainso they land independently and other in-flight branches stop compiling against a stale API.Why now
Currently
make releaseonmainfails with'meosType' does not name a type; did you mean 'MeosType'?andtoo many arguments to function 'Temporal* tcontains_geo_tgeo(...)'(× 13 call sites). Without these, no parity work can build locally, and only branches that already happen to carry this fix get green CI.Test plan
linux_amd64andlinux_arm64.restr=truewas used are now wrapped in atbool_at_value(ret, at_value)post-restriction that produces the same result the dropped MEOS args used to compute internally.