---
url: >-
  /statelake/administering-statelake/designer/data-definitions/file-definitions/xml-file-definitions/design-xml-file-definitions/xml-dataview-fields-tab-xml-file-definitions.md
---

# XML Dataview - Fields Tab - XML File Definitions

The **Fields** tab window is split into 2 distinct panes - **Available Fields** and **Selected Fields**.

Upon opening, the upper **Available Fields** pane echoes the items that were open on the **Records**  tab, and shows all previously **Selected Fields** (if any) in the lower pane.

*Here, only the wrapper M10B2BORDER had been open before the Fields tab was accessed*

*The wrapper and the first complex element POheader were open on the Records tab.*

*In this example, POheader had been expanded on the Records tab.*

So potentially, the **Fields** tab can appear to be different every time you access it.

If no **Fields** have yet been selected, then the **Fields** tab will have no entries populating the lower **Selected Fields** pane, as illustrated below. And while the complex elements themselves (in this case **M10B2BORDER** and **POheader**) appear in the **Available Fields** list, they are simply the name or identifying label or wrapper for the group of fields within them, and are not able to be selected as fields.

## Fields

| Field Name | Description |
|---|---|
| Selected Fields *Field Name* | The **Name** of the field that has been selected. |
| *Data Type* | The **Data Type** of the selected field. |
| *Size* | The maximum **size** of the field. |
| *Display format* | The **format** specified for the field. |
| *Field XPath* | Shows the **path** for the field - such as \POheader\HeaderCode.::: warningMake sure that this field looks OK for each field. It is relative to the **Root Node** so a mistake here can lead to incorrect references.::: |
| *Namespace* | The **Namespace** for the field.A **namespace** is a logical declared container space which groups together and stores the names associated with one particular event or process. |
| *Default Value* | The **Default Value** specified for the field. |
| *In Memory* | Will be either **True**  or **False**. The default is **False**. |

## Buttons

| Button Name | Description |
|---|---|
| Add Custom XML Field | Click to **Add Custom XML Field**. A permanent additional field. Unlike an **In-Memory Field**, these fields will be written to the group o which they have been added and will be saved with the dataview. |
| Add In-Memory Field | Click to **Add In-Memory Field**. These fields are temporary additional fields that are available in memory only while the dataview is active. As soon as the dataview is closed, these temporary fields no longer exist. The **In Memory** column for this type of field will be set to **True**, |
| Copy Create SQL Statement to the Clipboard |  |
| OK | Click to save any changes you have made. All changes will be lost if you do not click the **OK** button. |
| Cancel | Click to **Cancel** any changes you have made. |

## Fields

To expand any group, click on the ( + ) symbol beside it, or double-click on the name of the element - this will show all of the fields within the selected record grouping.

The complex group header or wrapper elements cannot be selected as fields - only the simple elements. For instance - if a **Schema** includes \<Buyer> \<Address1> \<Address2> \</Buyer>, then only the two (2) **Address** elements within the **Buyer** wrapper would be selected as they would be the items containing the read data. \<Buyer> \</Buyer> is simply the wrapper that groups the other elements together.

Complex grouping is not the same as repeating. For example, \<Buyer> \<Address1> \<Address2> \</Buyer> \<Seller> \<Address1> \<Address2> \</Seller> represents four (4) simple **Address** elements at the same level, and therefore within the same dataview. \<Buyer> \</Buyer> and \<Seller> \</Seller> are simply the group wrappers.

Elements at the same level may have a repeating element listed within them. For example, \<Buyer> \<Address1> \<Address2> \</Buyer> \<Item\[\*]> \<Seller> \<Address1> \<Address2> \</Seller> is still four (4) **Address** elements at the same level, with the additional repeating dataview for **Item** at the next level down.

::: info
It is usually better to include an element for simple elements only, instead of not including it - even if it is not required for **Mapping** e.t.c., in the rest of your configuration. You never know when it may come is useful.
:::

To select fields for this dataview query that are to be selected from the **Available Fields** list, double‑click on each field required in turn. Once selected they will appear in the **Selected Fields** pane, showing the data related to each field. As you select the fields, they do not disappear from the **Available Fields** list.

In the **Selected Fields** pane, all fields will be prefixed with the element group header followed by an underscore, to identify the original record.

::: info
While the next example illustrates how the **XML Dataview** works (in terms of selecting fields/records and specifying the **Field Name** and **XPath**), it is intentionally structurally incorrect.

The dataview pictured where there are items from both **POheader** and **POdetail** selected would only identify and extract the selected information from the first record from **POdetail** - all subsequent detail lines that pertained to the current header would be ignored.

So it makes no sense to try and design a dataview in the way illustrated. This design structure might only be successful if **POdetail** was not a repeating complex element, depending on the structure.
:::

::: info
You will notice that the intentionally inserted but incorrect fields from **POdetail** have been removed from the following examples.
:::

::: info
If you want to select two instances of the same field, click on the field name in the **Available Fields** pane for each instance. The second and subsequent selections will appear in the **Selected Fields** pane with a sequential number appended to the end of the field name, as shown below.
:::

::: info
If you make a mistake and click on one of the fields in the **Available Fields** pane that you do not actually require, or click on a field more than once in error, to de-select it simply double-click on the field name in the **Selected Fields** pane that you do not want.
:::

Once all your selection is complete, click the **OK** button to save and close the window. You will return to the designer canvas, and the dataview you have just created will be displayed.

Using **Test File Load** at this stage will produce a **Preview** with one entry.

So what about the last set of records/fields containing the details for each header - **POdetail**?

The folder icon that was visible beside **POdetail** on the **Fields** tab showed that this item was repeating within the header record – meaning that there could potentially be multiple detail records that belong to each header record.

But with only fields selected from **POheader** as per this example, then no data will be pulled from **POdetail** to flesh out each **POheader** record.

To extract data from this lower information tier of repeating elements, it must be treated as the start of a node - and therefore it is treated as a node in its own right. In **Statelake**, it will be considered to be an “**expanded**” **Root Node**. A node within a node.

To do this we must make another **XML Dataview** for **POdetail**, and specify it as a **Root Node**, just as with **M10B2BORDER**. So repeat the process as detailed on [Design - XML File Definitions](/statelake/administering-statelake/designer/data-definitions/file-definitions/xml-file-definitions/design-xml-file-definitions/) using this next lower level complex element.