General File Structure

Example of a sample classifications data file and detailed information about its structure.

The following illustration is a sample data file:

A data file must adhere to the following structure rules:

  • Classifications cannot have a value of 0 (zero).
  • Adobe recommends that you limit the number of import and export columns to 30.
  • Uploaded files should use UTF-8 without BOM character encoding.
  • Special characters, such as a tabs, newlines, and quotes can be embedded within a cell provided the v2.1 file format is specified and the cell is properly escaped. Special characters include:
    \t     tab character
    \r     form feed character
    \n    newline character
    "       double quote

    The comma is not a special character.

  • Classifications cannot contain a caret (^) since this character is used to denote a sub-classification.
  • Use care when using a hyphen. For example, if you use a hyphen (-) in a Social term, Social recognizes the hyphen as a Not operator (the minus sign). For example, if you specify fragrance-free as a term using the import, Social recognizes the term as fragrance minus free and collects posts that mention fragrance, but not free.
  • Character limits are enforced to classify report data.

    For example, if you upload a classifications text file for products (s.products) with product names longer than 100 characters (bytes), the products will not display in reporting. Tracking Codes and all custom conversion variables (eVars) allow 255 bytes.

  • Tab-delimited data file (create the template file using any spreadsheet application or text editor).
  • Either a .tab or .txt file extension.
  • A pound sign (#) identifies the line as a user comment. Adobe ignores any line that begins with #.
  • A double-pound sign followed by SC (## SC) identifies the line as a pre-processing header comment used by reporting. Do not delete these lines.
  • Classification exports can have duplicate keys due to newline characters in the key. In an FTP or browser export, this can be resolved by turning on quoting for the FTP account. This will place quotes surrounding each key with newline characters.
  • Cell C1 in the first line of the import file contains a version identifier that determines how classifications handle the use of quotes throughout the remainder of the file.
    • v2.0 ignores quotes and assumes they are all part of the keys and values specified. For example, consider this value: "This is ""some value""". v2.0 would interpret this literally as: "This is ""some value""".
    • v2.1 tells classifications to assume that quotes are part of the file formatting used in Excel files. So v2.1 would format the above example to: This is "some value".
    • Problems can arise when v2.1 is specified in the file, but what is actually wanted is v2.0 - namely, when quotes are used in ways that is illegal under Excel formatting. For example, if you have a value: "VP NO REPS" S/l Dress w/ Overlay. With v2.1, this is incorrect formatting (the value should be surrounded by opening and closing quotes and quotes that are part of the actual value should be escaped by quotes) and classifications will not work beyond this point.
    • Make sure you do one of the following: change your file format to v2.0 by changing the header (cell C1) in the files you upload, OR properly implement Excel quoting throughout your files.
  • The first (non-comment) row of the data file contains the column headings used to identify the classification data in that column. The importer requires a specific format for column headings. For more information, see Column Heading Format.
  • Immediately following the header row in a data file are the data rows. Each line of data should contain a data field for each column heading.
  • The data file supports the following control codes, which Adobe uses to provide structure to the file, and correctly import classifications data:
    CONTROL CODE DESCRIPTION

    <New Line>

    A new line character is the only supported delimiter between data lines/records in the data file. Typically, you only need to specifically insert these characters when writing a program to automatically generate data files.

    ~autogen~

    Requests that Adobe automatically generate a unique id for this element.

    In the campaign context, this control value instructs Adobe to assign an identifier to each creative element. See Key.

    ~period~

    Designates that the data column represents the date range associated with the item. See Date.

    Empty field

    Represents a NULL value for the current field. Use this if a particular data column does not apply to the current record.

    PER Modifiers

    Designates that the data column represents a PER Modifier field. See PER Modifier Headings.