JSON SSIS Source

Using the JSON Source Component

The JSON Source Component is a source component used to retrieve JSON documents from HTTP requests or files and extract them into column data. Configure your outputs by importing an example JSON, or manually adding/removing nodes. From there just configure the columns for each output and the component is ready to extract data from JSON documents from the Input.

Data Source Page

The data source page determines where your JSON documents are coming from. The connection manager property can either be set to an HTTP Connection Manager, or a local file.

Local File

• Local File Path - The path on the file system to the JSON document that will be extracted.

• Encoding - The encoding of the document that will be retrieved.

HTTP Connection Manager

If an HTTP connection manager is selected the UI will display request settings that are used to determine the content of HTTP requests.

HTTP Method - The HTTP method that will be used to send requests.

Relative Path - The relative path is the second part of the URL that will be used to make the HTTP request.  It is combined with the base path from the connection manager to form the full path.  The relative path can contain query string parameters, but they can also be added individually in the grid below.

Query String Parameters - Query string parameters that will be appended to the full path when sending HTTP requests.

Request Headers - Request headers that will be combined with the headers in the connection manager and included with HTTP requests.

Request Body - The body of the HTTP request.

Request Again Rule - The request again rules allows paged data to be retrieved.  After each request the request again rule is evaluated and if it evaluates to 'True' another request will be sent.  If the rule evaluates to false or is left blank no more requests will be sent.

Max Identical Requests - The max identical requests property will raise an error if a certain number of identical requests are sent.  The purpose of this feature is to prevent accidentally sending the same request in a never ending loop.  If 0 is selected the feature will be disabled and an unlimited of identical requests can be sent.

Preview Button - The preview button launch the Preview dialog which will send the first request based on the current settings of the component, and show a preview of the response.  If there are more requests based on the value of the request again rule they can be sent and the response previewed by clicking the 'Next' button.  The Preview dialog is a good way to identify HTTP configuration problems without running the package.

Expression Editor Preceding Request Tokens

Many of the properties on this page can be set as either a text merged field or an expression. Properties default to text merge mode, but can be changed to expression mode in the Expression Editor. The expression editor for a property can be launched by clicking the fx button. The expression editor has some predefined tokens:

RequestCount - The number of requests that have been sent.  This will be 0 on the first request and incremented by 1 after each request.

Response Body - This is the body of the response from the previous request.  On the first request is will be empty.

GetValueFromPreviousResponseHeaders - This is a function that is used to the value of HTTP headers from the previous response. This function is only available in expression mode.

Expression Editor Text Merge Mode

In text merge mode, property values be will whatever text was entered, but with any variables or special tokens are replaced with their actual values before each HTTP request is sent.

Expression Editor Expression Mode

In expression mode the property is set to an expression which is evaluated before each request to determine the value of the field. Expressions can contain variables and special tokens, but also functions and conditional logic. The functions in the Value Lookup Functions category may be particularly useful for this component, as they can be used to look up values and JSON documents and HTTP headers.

Design Page

This page allows you to import an example JSON document, either from the file system or from the web using the http settings specified on the datasource page.  The imported document is used to generate the expected hierarchy of incoming JSON documents. From there you can add and remove nodes manually to refine the structure.

Once you've added some nodes by either importing or manually adding, you may need to refine the structure by editing the node properties. The following properties will have an effect on the extract process:

• JSON Name - This is the name of the node as it is in the expected JSON. Notice when this property is changed the JPath and Data Name properties are automatically updated.
• Data Name - This is the name of the Output (green rows) or Column (yellow rows). Notice that some rows do not have a Data Name property. This is because these are nodes that define hierarchy and have no immediate value to extract. These nodes are important however because it affects the JPath of child nodes.
• Type - There are 3 different JSON node types to choose from:
  • Array - This node contains a list of nodes. Notice the JPath of the child elements will contain a '[n]' at the end. This means this node repeats.
  • Object - This node often just defines hierarchy and is simply a node that contains nodes.
  • Value - This is a basic node containing a value available for extraction.
• JPath - This property is not editable but instead generated based on the above properties.

General

Configure Input and Output settings.

JSON Column 

Specifies the input column that contains the JSON to extract.

Null Mode

Specifies what represents a NULL value in the JSON. There are 2 options:

• Null Value - If a node cannot be found or is null then the value is NULL.
• Empty String - If a node cannot be found or is empty then the value is NULL.

Outputs

In the Outputs data grid, you can check and uncheck outputs to specify if you wish to have these outputs in your component. You can also configure the following properties:

• Output Name - The name of the output
• Key Field - This specifies the identifier column for this output. This is important for linking outputs with each other. By default, the Key Field is set to '_RowIndex' which is a special field that contains the current count of this output node. This is useful because many times a JSON object will have no key field because relationships are defined hierarchically. During runtime, the value of this field will go into the _ParentKeyField field of all of this output's children. 

Columns

Configure column settings for each output.

Select the output whose columns you wish to configure in the top left combo box. Select a column in the data grid on the left to populate its properties in the property grid on the right. There are a couple of special columns to take note of:

• _RowIndex - This column contains the current count of this output node. 
• _ParentKeyField - This column contains the value of this records parent key field.