The Directoryz field guide

Imports and cleanup

Prepare data before it becomes a public directory.

13 practical guides1 topic

Find a guide13 guides shown

Map CSV headers to directory fields

Header mapping should connect meanings, not just similar names.

Garbled names can result from an encoding mismatch before the directory sees the data.

A file may use commas, semicolons, or another separator despite its extension.

Spreadsheet exports can include blank rows, notes, or repeated headings.

Equivalent website URLs can appear in several forms, but aggressive normalization can erase meaningful paths.

Duplicate detection needs more than a matching name.

A merge should preserve the best-supported value, not simply the newest row.

A sample should exercise the difficult cases in the dataset.

A bulk update is easier to recover from when the previous values and affected records are known.

CSV files contain exported values rather than the live logic of a workbook.

Categories and tags need an agreed representation in the source file.

An audit log should identify what changed and how to investigate it.

A successful import message confirms processing, not editorial correctness.