Jax Liu 4a4f72af5d fix(core-py): drop orphan cubes when their base model is pruned ...

Manifest extraction prunes models down to those a query reaches, but kept
all cubes unconditionally. A cube whose baseObject model got pruned then
failed wren-core MDL analysis with "baseObject is not a defined Model or
View" — so any query touching only tables unrelated to a cube's base, or a
scalar query referencing no model, broke once a cube existed (issue #2360).

Filter cubes to those whose baseObject still resolves to a kept model or
view. Cube queries emit `SELECT ... FROM <baseObject>`, so a queried cube's
base model is always pulled in — dropping orphans is therefore safe.

Backport of Canner/wren-engine-saas #892 (+ #895 view-cube coverage).

Closes #2360

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-06-12 11:18:28 +08:00

..

wren

docs(wren): fix cube quickstart and align YAML/CLI examples with implementation (#2359)

2026-06-12 10:28:15 +08:00

wren-core

feat(wren-core): support composite (multi-column) primary keys (#2345)

2026-06-09 15:58:21 +08:00

wren-core-base

feat(wren-core): support composite (multi-column) primary keys (#2345)

2026-06-09 15:58:21 +08:00

wren-core-py

fix(core-py): drop orphan cubes when their base model is pruned

2026-06-12 11:18:28 +08:00

wren-core-wasm

deps: clear 5 high-severity Dependabot alerts on main (#2330)

2026-05-28 15:18:02 +08:00

wren-mdl

feat(wren-core): support composite (multi-column) primary keys (#2345)

2026-06-09 15:58:21 +08:00