Implementation Guide › Implementation › Data Gathering (Data Source Expert) › How to Configure Adapters and Translations › File Adapter Case Study › Configuration File DataSourceInterface Section

Configuration File DataSourceInterface Section

The DatasourceInterface section consists of attributes specifying the connection and the connection type between the Adapter and data source. There are two types of interface, file and SQL. The main difference between the two is that for files, the Files collection is required, and for SQL, the queries collection is required.

The DataSourceInterface section also defines the way in which the Adapter manages the source file (whether it deletes the original file if it was created only for the Adapter, or whether it leaves the data if it is needed for other uses, and so on).

For file Adapters, to read and parse ASCII files, the files interface is used as shown in the following figure. Choose the following values for the settings as follows:

The Files section under the DataSource Interface node relates to the connection to the data source. Configure the following attributes.

Note: This section looks completely different for an SQL adapter.

• DeleteFileAfterProcessing: Sets the way in which the Adapter handles the source file and determines how the Adapter controls the reading in order to read only new records. In this case, the source files are left in place on the server and this value is set to 'no'.

  In cases where a file is created only for the Adapter and it can be deleted after processing, it should be set it to "yes". The file is then renamed, processed, and deleted.

  When set to "no", the file is copied and the processing takes place using the copied file. If new records are appended to the end of this file, the Adapter copies these new records to the work file during the next cycle. If new records are not appended to the file, the Adapter searches for the next file with the same pattern and name (in lexicographic order) as the current file. If the Adapter finds such a file, work proceeds on this file. The Adapter does not revert to the previous file even if new records are appended.

  Set to "no" when you need to keep the integrity of the source file and when the file should be appended.

• InputFormat: Refers to the name given to the InputFormat element in the next InputFormatCollection which handles the data from this data source connection. The Input format is the fields structure of the input data coming from the data source after being parsed by the Adapter. The parsing method is specified in the delimiters attribute as explained below. When handling more than one connection using different interface formats this field becomes more important and determines which input format structure handles each data sources data.
• Path: Physical location of the data source files. For example : C:\Program Files\CA\Cloud Insight\Adapters\ServersAdapter\data\.
• NamePattern: Specifies the data source file name. Wildcards may be used if more than one file uses the same input format. If a file name is specified with no wildcards, the Adapter searches only for a file with this name. Where using wildcards, the Adapter searches for all files that correspond to the pattern, sorts them lexicographically, then reads them one by one. During the next run, it searches for new records in the last file before proceeding to the next.

  In this example, if the wildcard character '*' is used, then the attribute values is "ServerData*.csv". (The Adapter reads all files with names that begin with ServerData and have the extension, .csv.)

  Important! It is recommended that a date and time are added to the end of the filenames using the following format YYYYMMDD-HHMISS to ensure that the files are sorted correctly, read in the correct order, and that none of the files are missed. The Time portion can also be added if there are multiple files being produced each day.

• Delimiters: Method of determining how the file is parsed. One or more characters can be specified to serve as delimiters, according to the data rows that are to be parsed into fields. If not specified, the default value is "/t".

The data source file in this example is a CSV (comma separated) file. The simplest way of parsing such files is to specify the comma as a delimiter.

Other methods available for parsing are:

• RegExForParser: Uses a regular expression to set the parsing rule.
• LogicLineDefinition: Used when a line in the file is built out of several lines.
• TitleExists (optional): Specifies whether the first non-blank line in the data source file is a title line. The title can be used by the Adapter when parsing the data. In this example, each data file contains the title row, so 'yes' should be specified for this attribute.