Elizabeth Thompson fb410b39e6 fix(reports): retrofit execution-id/cache-key tracing onto existing screenshot-capture logs ...

Extends the log_context plumbing landed for the new tile-wait-budget log
lines to the rest of the report-screenshot capture path, so any log line
in this flow can be traced back to the run that produced it without
cross-referencing: superset/utils/screenshot_utils.py (combine_screenshot_tiles
and the non-budget log lines in take_tiled_screenshot), superset/utils/webdriver.py
(WebDriverPlaywright and WebDriverSelenium get_screenshot/find_unexpected_errors),
and superset/utils/screenshots.py (BaseScreenshot.driver/compute_and_cache/resize_image,
using the thumbnail cache_key where there is no report execution id).

superset/commands/report/execute.py needed no changes: every log line in
its screenshot-acquisition path already carries execution_id.

Message-text/plumbing only -- no behavior or log-level changes here beyond
what already landed. Left untouched, to avoid conflicting with the
concurrent fix-tile-readiness-check PR: the per-tile spinner-wait log
lines inside take_tiled_screenshot, and the two non-tiled ".loading" wait
blocks in webdriver.py (Playwright's wait_for_function-based checks and
Selenium's WebDriverWait-based check).

Co-Authored-By: Claude <noreply@anthropic.com>

2026-07-23 01:00:06 +00:00

49 KiB

Raw Blame History

View Raw