Data mapping

Your store calls a column article; another calls it sku. Scalarcut does not require either. You map each column in your file to a Scalarcut field once, and the mapping is remembered.

How mapping works

When you upload a file, Scalarcut reads the header row and fingerprints it. If you have imported this header layout before, the saved mapping is applied automatically — you only map a new column layout once.

Each mapped column carries a transform that tells the importer how to read the value:

Transform	Used for	Behaviour
trim	SKU, name, collection, ids	Whitespace trimmed.
integer	Width, height, thickness, pack qty	Parsed as a whole number.
parse_size_pair	Combined size	Accepts 600×300, 600x300, 600*300, with mm, cm, or m.
url	Image URL	Must be public HTTPS or it is dropped with a warning.
enum_map	Material type	Matches values like porcelain, gres, or керамогранит to a known type.

The import flow

1. Upload the file and choose a mode (below).
2. Map the columns — or let a saved profile apply automatically.
3. Validate — every row is checked; errors and warnings are reported.
4. Confirm — you review the created / updated / rejected counts before anything is published.
5. Commit — the rows are written to the catalog and the revision is bumped.

If any row fails, you can download an errors CSV listing the row number, SKU, and the reason, fix the source file, and re-upload.

Validation rules

• A row needs a SKU and a size. Without either, it is rejected.
• A tile size below 10 mm is rejected.
• A missing or non-HTTPS image URL is a warning, not an error — the tile still imports.

Keeping the catalog up to date

Your catalog changes constantly. You do not need to re-embed the widget or touch your snippet — the widget always serves the latest committed revision on each load. The workflow for everyday changes:

Add new products

Add the new rows to your usual export and re-upload. The same headers are recognised automatically, so the new SKUs are added and the existing ones are left untouched.

Edit a product

Update that product's row (a new size, a corrected pack quantity, a new image URL) and re-import. The existing SKU is updated in place; the other rows are unchanged.

Remove products

Removal is an explicit choice, so nothing disappears by accident. Choose the mode that fits:

Mode	What it does
Upsert	Adds or updates the rows in the file and leaves everything else alone. Use for adding and editing.
Replace a collection	Replaces one collection entirely — any SKU in that collection that is no longer in the file is removed.
Full catalog sync	Makes the catalog match the file exactly — any tile SKU missing from the file is removed.

For a one-off product that is out of stock or discontinued, edit the file and re-import rather than leaving a stale SKU in the planner.

Images and prices stay yours

The mapping only records where things come from. Images stay on your servers (referenced by URL), and prices are display-only in the widget — the checkout remains the source of truth.

Next steps

• CSV structure — the columns and file formats.
• Catalog — what the catalog stores and how the widget uses it.