<layout b="," h="1" t="&quot;" />

This format definition works if the source data looks like this:

"Joe","Brown","1745 E. New St., Apt. 22","Chicago","Illinois","46702"

Double-quotes are wrapped around each field, while a comma separates each field. By including the quotes, Slate interprets "1745 E. New St., Apt. 22" as one field rather than two.  

<layout b="&#x9;"h="1"/>

<layout b="|" h="1" />

<layout type="convert" h="1" />

<layout width="43">
    <f s="1" w="25" id="Student Last Name" />
    <f s="26" w="16" id="Student First Name" />
    <f s="42" w="1" id="Student Middle Initial" />
</layout>

The <layout> node's width attribute defines how many characters each row of the file has. In the above example, each row would be 43 characters long.

Fixed-width files require <f> nodes so that Slate can determine where each field begins and ends, using the following attributes:

• The s attribute defines which character the field starts at. For example, s="1" means that this field starts with the first character in the file.

• The w attribute defines the width of the field. For example, w="25" means that the field is 25 characters long.

• The id attribute is the source field name that's shown in the Field Mappings stage of Upload Dataset. It should be a user-friendly name that's unique and no longer than 64 characters.

<layout type="xml" node="/persons/person">
  <f s="first" id="First Name" />
  <f s="last" id="Last Name" />
  <f s="email" id="Email" />
  <f s="birthdate" id="Birthdate" />
  <f s="schools/school[1]/ceeb" id="School 1 Code"/>
  <f s="schools/school[2]/ceeb" id="School 2 Code" />
</layout>

This format definition works if your incoming data is structured like this:

<persons>
  <person>
    <first>Joe</first>
    <last>Brown</last>
    <email>[email protected]</email>
    <birthdate>2000-01-01</birthdate>
    <schools>
      <school>
        <ceeb>380880</ceeb>
      </school>
      ...
    </schools>
  </person>
  ...
</persons>

In the <layout> node, the type attribute tells Slate which file type to expect - in this case, XML. The node attribute tells Slate the path to each "row" of the file; in this case, there is a <person> node for each row contained within an overall <persons> node.

Each <f> node in the format definition represents a field you will map.

• The s attribute uses XPath syntax to define where the data exists in the source. For example, s="email" means Slate will use the <email> node in the source data.

• The id attribute is the source field name that's shown in the Field Mappings stage of Upload Dataset. It should be a user-friendly name that's unique and no longer than 64 characters.

The last two <f> nodes demonstrate how to handle multirelational data - in this case, Schools:

• The School 1 Code has a path of schools/school[1]/ceeb. Unlike the previous paths, this path contains square brackets with a number, [1], which indicates that we should pull the <ceeb> node from the first <school> node that we encounter under the <schools> node.

• We get to the second <school> code by incrementing the number contained in the square brackets. You will need to anticipate how many nodes might appear and map them all separately, incrementing the bracketed number each time.

<layout type="json" node="/students">
  <f s="first" id="First Name" />
  <f s="last" id="Last Name" />
  <f s="email" id="Email" />
  <f s="birthdate" id="Birthdate" />
  <f s="schools[1]/ceeb" id="School 1 Code"/>
  <f s="schools[2]/ceeb" id="School 2 Code" />
</layout>

This format definition works if your incoming data is structured like this:

{
  students: [
    {
      first: "Joe",
      last: "Brown",
      email: "[email protected]",
      birthdate: "2000-01-01",
      schools: [
        {
          ceeb: "380880"
        },
        ...
      ]
    },
    ...
  ]
}

In this format, the type attribute is set to JSON, and the node attribute tells Slate the name of the root node. If your data begins on a lower level of nesting, you can use slashes to indicate the path; for example: students/student.

Each <f> node in the format definition represents a field you will map.

• The s attribute uses XPath syntax to define where the data exists once the JSON has been converted to XML. For example, s="email" means Slate will use the email property in the source data.

• The id attribute is the source field name that's shown in the Field Mappings stage of Upload Dataset. It should be a user-friendly name that's unique and no longer than 64 characters.

The last two <f> nodes demonstrate how to handle JSON arrays:

• The School 1 Code has a path of schools[1]/ceeb. Unlike the previous paths, this path contains square brackets with a number, [1], which indicates that we should pull the ceeb property from the first object in the schools array.

• We get to the second CEEB code by incrementing the number contained in the square brackets. You will need to anticipate how many items might be in the array and map them all separately, incrementing the bracketed number each time.