Great Expectations 1.18.1: Data Docs HTML Escape Fix

Great Expectations 1.18.1 shipped on June 11, 2026 as a patch release for the open source data validation library. The functional change is a Data Docs rendering fix: regex patterns with angle brackets now get HTML escaped instead of being interpreted as markup. Documentation cleanup and temporary warehouse test skips round out a release with no validation logic changes.

The full release notes and downloads are on the GitHub release page.

PR #11909 addresses a display bug in Data Docs, the static HTML site Great Expectations generates from an Expectation Suite. When an expectation carried a regex containing < or > characters, the renderer passed those symbols through without escaping. Browsers treated them as HTML tags, which broke the surrounding layout and hid the actual pattern text.

Data engineers use Data Docs as an audit trail. After a checkpoint run against a batch or streaming slice, the generated site shows which expectations passed, which failed, and the parameters each rule evaluated. Regex based expectations on column values, URL fields, or log line formats are common in pipelines that validate semi structured input. A corrupted regex display makes post run review harder and can mislead reviewers who skim the HTML instead of reading raw checkpoint JSON.

The fix applies standard HTML entity escaping to regex content before it reaches the Data Docs template layer. Expectation evaluation, checkpoint execution, validator batch handling, and result serialization are unchanged between 1.18.0 and 1.18.1. Pipelines that run headless and never call build_data_docs() see no behavioral delta. Teams that publish Data Docs to object storage or serve them behind an internal reverse proxy should upgrade if any suite includes angle bracket regex patterns.

Two documentation commits ship in the same tag. PR #11900 syncs the docs version label to the released 1.18.0 baseline, aligning the published doc set with the prior minor release numbering. PR #11906 removes the GX Cloud docs site from the documentation tree.

The GX Cloud docs removal is an operator visible signal. Teams that bookmarked cloud hosted setup guides or linked internal runbooks to GX Cloud pages should verify those references still resolve. The open source library docs remain the primary source for checkpoint configuration, datasource wiring, and Expectation Suite management. Neither docs pull changes runtime imports, CLI entry points, or Python package layout. They matter for organizations that pin internal wiki versions to specific library releases and need the doc label to match the installed wheel.

PR #11908 temporarily skips BigQuery integration tests in the upstream CI suite. PR #11911 does the same for Snowflake integration tests. Both are tagged [MAINTENANCE] in the release notes.

Temporary skips in a data validation library usually mean flaky credentials, quota limits, or endpoint drift on shared test infrastructure rather than connector removal. Great Expectations ships SQLAlchemy and native connectors for both warehouses, and this release does not list any connector API changes. Operators running checkpoints against BigQuery or Snowflake tables should not treat these skips as a signal to change datasource configuration.

The practical impact falls on contributors and vendors who mirror upstream CI. If your fork runs the same integration matrix, warehouse specific assertions may not execute on every commit until the skips revert. Watch the 1.18.1 release page and subsequent tags for maintenance PRs that restore coverage. Production users who only consume published wheels are unaffected by CI scheduling decisions.