Labels and records

Prepare CSV Data for Bulk Spine Labels

A clean CSV makes a bulk print job predictable. Each row should describe one intended label, and each column should have one defined role. Keep an untouched source export so you can trace any change introduced during preparation.

LibCatalogTools learning guide · 3 min read

Use the headers the tool expects

For DDC, use this header and fictional example row:

location,classNumber,authorLastName,authorFirstName,title,year,volume,copy
Ref,025.3,Besorio,Darwin,Cataloging practice workbook,2026,,1

The current DDC validator requires location, classNumber, authorLastName, title, and year as both headers and nonblank values on every data row. First name, volume, and copy are optional. Do not invent missing cataloging facts to satisfy validation; resolve the record or use a suitable alternative workflow.

For LC, the header is:

location,lcLetters,lcNumber,authorLastName,authorFirstName,year,volume,copy

LC requires location, lcLetters, lcNumber, authorLastName, and year. Header spelling and capitalization matter. The LC mark is still generated from Cutter-Sanborn data, so review it independently before accepting an LC label.

Preserve text and quote commas

Set class numbers and identifiers to text before pasting or importing into a spreadsheet. Otherwise a value such as 025.3 may become 25.3. Export as comma-separated UTF-8 text and reopen the file to verify leading zeros and names with accents.

Ref,025.3,Besorio,Darwin,"Cataloging, step by step",2026,,1

The quotation marks keep the comma inside the title cell. A literal quote inside a quoted value is doubled, as in "The ""local"" collection". Check that each row has the same number of cells as the header. Do not use semicolon-separated output or add commentary lines above the header.

Run a small preflight batch

Choose the matching DDC or LC mode before importing.
Import three representative rows, including a long number and a title with punctuation.
Generate labels and read any validation message.
Compare author marks and every displayed component with the approved records.
Download the output CSV and confirm the number of generated records.
Test the print layout on plain paper before loading adhesive stock.

The bulk tool processes at most 50 records per run, subject to the selected limit. Split larger jobs into explicitly numbered batches and reconcile the record count after each run. A preview containing labels does not prove that all rows in the original file were processed.

Keep batch identity outside the print data

Name source files clearly, for example cataloging-batch-01-source.csv, and keep reviewed exports separately. Avoid inserting a batch identifier into a required field such as location merely to track the job; that value may then appear on every label.

After printing, count labels and match them to items. If a source row is corrected, regenerate from the corrected input and mark the earlier output as superseded in your work process. This prevents two versions of a batch from being applied to the same books.

Frequently asked questions

Why is a visually correct CSV rejected?

Check exact headers, the selected mode, missing required cells, and whether your export used commas rather than another delimiter.

Can I process more than 50 records at once?

No. Split the job and verify the generated count for each batch.

Is the downloaded CSV a MARC record export?

No. It contains label-workflow data, not a complete MARC bibliographic record.

This guide describes the current application and suggested local workflows. Examples are fictional; adopt procedures appropriate to your institution.

Found a cataloging issue or an example that needs correction? Send feedback with the page title, example, and relevant source. Read our scope and limitations.