Skip to content

feat(dw-data-access): Invoke-DwMojibakeCensus.ps1 - #98

Open
Justinvolved wants to merge 1 commit into
feat/dw-debrandfrom
feat/dw-mojibake-census
Open

feat(dw-data-access): Invoke-DwMojibakeCensus.ps1#98
Justinvolved wants to merge 1 commit into
feat/dw-debrandfrom
feat/dw-mojibake-census

Conversation

@Justinvolved

Copy link
Copy Markdown
Collaborator

Wave-1 script 7 of 7 (4.30.0), stacked on the debrand PR. READ-ONLY double-encoded-UTF-8 census per table.column across every string column (or -Table subset): broken markers vs healthy typography as the partial-vs-wholesale contrast, U+FFFD context spans with every non-ASCII character escaped. Markers built from code points, so the file's own encoding cannot corrupt the needles and the repo validator does not trip on its own detector. Comparisons run COLLATE Latin1_General_BIN2 — the default CI collation case-folds the stray-Â marker onto healthy â and gives U+FFFD an undefined weight so CHARINDEX never finds it (caught live in smoke). Repair stays a separate deliberate step; U+FFFD is called out as unrecoverable in place. Consolidates the three scattered mojibake notes (visual-qa.md hard pointer, modelling-discipline.md soft pointer, the repo-file/ftfy side untouched).

Verification: validator 0 errors, parses. Live smoke: clean verdict with healthy-contrast on a repaired demo DB; synthetic damaged table detected both marker classes and produced escaped FFFD context spans; test table dropped after.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant