Created by potrace 1.16, written by Peter Selinger 2001-2019

Miva Developer Docs

CSV#

  • The first line is the output names, unless the feed uses Use Value as Feed. Put the column line in the Header text then.
  • Values are separated by the Delimiter: comma, tab or pipe.
  • Values are quoted and escaped as CSV needs.
  • Each line ends with the Line Ending: LF, CRLF or CR.
  • A list, such as Additional Images, is one cell with the values separated by commas.
  • A field left out by conditional output is an empty cell, so columns stay lined up.

Use comma for .csv files and tab for Google’s and Microsoft’s .txt format.

JSON#

  • One array of objects, written on one line: [{...},{...}].
  • Each object has a property per field, named by the output name.
  • A field left out by conditional output has no property.
  • A list is an array of strings.
  • Line Ending doesn’t apply. The Preview tab indents JSON to make it readable; the feed doesn’t.

The output is a single array, not JSON Lines.

XML#

Each row is one <item> element with an element per field:

<item>
  <g:id>TEE-RED-S</g:id>
  <title>Feed Test T-Shirt Red S</title>
  <g:price>20.00 USD</g:price>
</item>
  • There’s no root element, namespace or XML declaration. Add them in the Header and Footer text.
  • Element names are the output names, lowercased. Letters, digits, _, - and : are kept; anything else becomes _. A name starting with a digit gets f_ in front.
  • Include a prefix such as g: in the output name itself. The module doesn’t add one.
  • Values are entity-encoded.
  • A field left out by conditional output has no element.
  • A list repeats the element once per value, for example one <g:additional_image_link> per image.

Data Builder’s stored XML rows use <record> instead of <item>.

Line endings#

Setting Characters
LF Line feed. The default.
CRLF Carriage return and line feed (Windows)
CR Carriage return

Line Ending applies to CSV and XML rows, the line ending added after the Header text, and the {{newline}} token.