posted on November 15, 2010 02:25:22 PM new
I understand that I cannot convert my old listings to have variation. I would like to do an import. Is it available for listing with variation?
posted on November 16, 2010 10:10:45 AM new
thanks. I was able to import most of the items. But there are a lot of things that need to be tweaked. I really don't want to have to change one by one. So is there no csv file I can upload that can do a bulk change?
posted on November 16, 2010 03:00:38 PM new
Hello,
You can import from a CSV file with attributes in place. The template you can download from the import page already has these columns and were currently working out the documentation, but I will include what we have available thus far:
The attributes will be added to the template spreadsheet as new columns, up to 5, labeled "Variation_1", "Variation_2", ... "Variation_5. The meaning of these columns is interpreted differently, depending on item being a parent or a child:
parent items should specify, in the Variation_x columns, the name of each variation (e.g. "Color" or "Size". The name will be handled case insensitive and capitalized at import.
child items should specify, in the Variation_x columns, the value of each variation (e.g. "Red" or "XL". These will also be handled case insensitive and capitalized at import, unless they are all in capitals.
In addition to these columns, the spreadsheet will contain one more column called "ParentSKU". This is mandatory for a variation: if an item misses it, we cannot identify the parent item and therefore the item will be considered a standalone one and not a variation.
An item is considered to be "parent item" if and only if:
ParentSKU is null or 0
it has at least one not null "Variation_x" column
An item is considered to be "child item" and associated to parent having SKU = N if and only if:
ParentSKU is not null and != 0
it has at least one not null "Variation_x" column
Due to the technical constraints, we will assume that a parent item and its variations are contiguous spreadsheet rows. The parent item MUST appear first then the child items must appear in the subsequent, contiguous rows. As an example, let there be parent item P, child items C1 and C2, and another item A:
Accepted:
row i P
row i+1 C1
row i+2 C2
row i+3 A
Accepted:
row i A
row i+1 P
row i+2 C2
row i+3 C1
Not accepted:
row i P
row i+1 C1
row i+2 A
row i+3 C2 <- error: child occurring after the sequence ended
Not accepted:
row i A
row i+1 C1 <- error: child occurring before parent
row i+2 P
row i+3 C2
If a parent has provided values for a subset ["C1", ... "Cn] of Variation_1, ... Variation_n columns, then all these item's children MUST provide values for all the columns in the subset. If they provide additional values, those will be silently ignored (that is, logged as warnings but not prevent the child from being imported and attached to the parent).
The variation (child) items MUST provide values for:
SKU
ParentSKU
all the subset of Variation_x columns which are not null in the parent
If a variation does not specify a SKU, will be silently rejected from the import.
The variation (child) items MAY provide values for:
quantity
images
which will OVERRIDE the values specified for the parent item. If quantity is not specified, it will be considered 1. If images are not specified, the parent images will be used.
The variation (child) items need NOT to specify any other values. In fact, these will be ignored by the importer and pulled from the parent item.
Obviously, the parent-item relationship will be extracted based on the ParentSKU value.
If you check back later we do hope to have this written up for the import documentation available on the website.
Regards,
Chris
posted on November 23, 2010 12:38:30 PM new
I know in the online item interface, you cannot change a listing from one item to variation. But if I use csv file, can I add a parent sku to existing listings to convert them to a variation listing?
. The name will be handled case insensitive and capitalized at import.