Repository navigation
validation: PostgreSQL backend + Linux re-run of the Foundry suite (SQLite 212/327, PostgreSQL 265/327) - #87
Merged
Conversation
… of both backends (SQLite 212/327, PostgreSQL 265/327) - quirks.py: OdbcPostgresQuirks (psqlodbc, compat compose service), backend selected by ADBCBRIDGE_VALIDATION_BACKEND / --vendor-version - queries/odbc_postgres: BYTEA and decode() for the binary tests; skips for nanosecond precision, negative decimal scale and the rounded 24:00:00 time - RESULTS.md: 2026-09-06 runs of both backends against f262fef with suite b949612; D1-D9, D11, D14 fixed since the 2026-08-21 baseline, no regressions; D10/D12/D13 on SQLite are SQLiteODBC's missing precision; PostgreSQL isolates the driver's own remaining items P1-P4, P7, P8 - README.md: PostgreSQL backend, macOS/Windows notes; ROADMAP row updated Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01YG6wFApZpg2xe61EmKhaSZ
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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-runs the ADBC Driver Foundry validation suite on Linux against the current driver (f262fef, suite b949612 of 2026-09-04) and adds a PostgreSQL 16 backend.
SQLite (the 2026-08-21 baseline backend): 212 pass / 69 fail / 46 skip, up from 192 / 88 / 46, no regressions. Baseline driver findings D1–D9, D11 and D14 are fixed. Every remaining D10/D12/D13 failure has one cause: SQLiteODBC reports no fractional-seconds precision for any TIME or TIMESTAMP column, so the driver keeps its microsecond default. On SQLite that is a backend limit.
PostgreSQL (new, psqlodbc, compat compose service): 265 pass / 28 fail / 34 skip. psqlodbc reports precision truthfully, so the 28 failures are the driver's own, classified in RESULTS.md as P1–P4 (TIME/TIMESTAMP precision and zone fidelity, including zone-less ingest landing in
timestamptz), P7 (adbc.connection.db_schemaGetOption missing) and P8 (schema rows with a NULL catalog). All are smallsrc/changes and are listed in the roadmap row. Backend limits (nanosecond precision, negative decimal scale, the rounded24:00:00time) are declared as query overrides, and the binary tests get BYTEA/decode()overrides.No
src/change in this PR; results only, plus the backend wiring and docs.