TXGenWeb Landmarks Bulk Converter
=================================

WHAT THIS DOES
This package downloads the public legacy TXGenWeb Landmarks county pages directly from:
https://www.txgenweb.org/research/landmarks.htm

It discovers the county links automatically, downloads each legacy .htm page, extracts the
five-column landmark table, builds a redesigned static .html page, validates the result,
and creates a report.

HOW TO RUN IT ON WINDOWS
1. Extract this ZIP into a normal folder.
2. Double-click RUN-LANDMARKS-CONVERTER.bat
3. Leave the window open while it works.
4. When finished, open the new "output" folder.

THE OUTPUT FOLDER WILL CONTAIN
- landmarks\                 redesigned county .html pages
- _legacy_source\            downloaded original .htm pages
- converted-landmarks.zip    all generated landmark pages
- conversion-report.csv      PASS / REVIEW / FAILED results and record counts

NEXT STEP
Upload BOTH of these files back to ChatGPT:
- converted-landmarks.zip
- conversion-report.csv

The REVIEW/FAILED list lets us concentrate manual work only on pages whose legacy HTML
is unusual instead of rebuilding hundreds of normal pages by hand.

NOTES
- The converter does not use JavaScript fetch() in generated county pages.
- Blank values become em dashes.
- Text from source cells is HTML-decoded and preserved as text.
- Duplicate source records remain duplicates.
- Odd coordinates and spellings are not normalized.
- A copy of every successfully downloaded source page is retained for final verification.