ingestr v1.1.82 - Spanner Array Handling and a GitHub Commits Table


ingestr v1.1.82 shipped on November 3, 2026, with two changes: Spanner now reads every array element type and preserves NULL elements, and the GitHub source gains a commits table. Small release, but one of the two fixes a silent data correctness issue. The full notes are on the GitHub release page.

The full release notes and downloads are on the GitHub release page. This is a stable release, not a prerelease.

PR #1368 changes how ingestr reads ARRAY columns from Cloud Spanner. Two things happened:

  • All element types in an array are read, not a subset. Previously some element types could slip through and end up as lost data in the destination.
  • NULL elements inside an array are preserved. An array like [1, NULL, 3] no longer gets compacted to [1, 3].

Why this matters: if you replicate Spanner tables into BigQuery or Postgres and a downstream consumer relies on positional array access, silently dropping NULL elements shifts positions. Any aggregation over array contents was also wrong. This is the kind of bug that does not show up in schema checks, only in row level diffs. If you pull Spanner arrays, recheck a few rows after upgrading.

PR #1371 adds a commits table to the GitHub source. It is the first contribution from @kerem-kirici.

Use case is straightforward: ingesting repository commit history alongside issues and pull requests lets you correlate development activity per repo or per org in your warehouse. Keep GitHub API rate limits in mind when backfilling a large org; the commits endpoint is the expensive one.

No breaking changes are listed in the notes. The changelog for this release is short, two PRs, and the full changelog confirms nothing else moved. If you run Spanner sources in production, treat the NULL element fix as a data contract change and spot check historic ingests.