You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
F2: cells beyond the header width are silently dropped in both loaders #82
Found in ART Milestone 0 (#53). Code read at main032b522; not reproduced.
Problem
Both loaders iterate only over the header's columns:
dynamic_loader.py:263-264: cell = raw[c] if c < len(raw) else ""
te_loader.py:221: the same pattern.
A data row with more cells than the header loses the extra values, and they appear in no count or error. This contradicts the data-preservation constraint ("never silently discard columns") and the intent of BR-10 (reject rows that don't match the header).
Fix
Treat a row wider than the header as a row error ("row has N cells, header has M"), the same way a cast failure is reported. Add tests for both modes.
Activity
docs(art): Milestone 0 assessment, VCRM draft, ADR-0002 and ADR-0003 (#…
Merge pull request #88 from amar-python/fix/issue-82-ragged-rows