public interface TestData
Test data can come in a variety of flavours based on your personal preference.
The test file to use with your DataSource is specified in the testFileName
DataSource configuration property and will have the .data.csv
extension.
The CSV file can contain an optional header line as the example below will show you. The headline names can be either DataSource title or field name and are not case sensitive. If the header line is missing it will be assumed that the data/columns are the same order as the field declaration in the DataSource file. This sample represents some moon landing data and you can also use double quotes around more complex values.
mission, duration, launchDate, commander, rover "Apollo 11", "195.31", 1969-07-16, "Neil Armstrong", false "Apollo 12", "244.61", 1969-11-14, "Charles Conard", false "Apollo 13", 142.91, "1970-04-11", "Jim Lovell", false "Apollo 14", 216.03, "1971-01-31", "Alan Shepard", false "Apollo 15", 295.20, "1971-07-26", "David Scott", true "Apollo 16", 265.85, "1972-04-16", "John W. Young", true "Apollo 17", "301.87", 1972-12-07, "Eugene Cernan", true
The test file to use with your DataSource is specified in the testFileName
DataSource configuration property and will have the .data.js
extension.
The JSON file needs to contain an array of objects. Each object represents a new record and will contain the properties in a JSON fashion. The names can be either DataSource title or field name and are not case senstive.
[ { category: "Adding Machine/calculator Roll", itemName: "Adding Machine Roll 57x57mm Lint Free", sku: "226500", unitCost: 0.52, units: "Roll" }, { category: "Pastes and Gum", itemName: "Glue UHU Clear Gum 250ml", sku: "724800", unitCost: 2.26, description: "Ideal for paper and card. Dries clear. Easy to use. Washable & Non-Toxic." }, ... ]
The test file to use with your DataSource is specified in the testFileName
DataSource configuration property and will have the .data.xml
extension.
The test data file should consist of a top-level element containing a series of XML elements, each of which creates one DataSource record. Values for each field are given within tags named after the field name or in the parent tag as an attribute.
For example, the following XML represents a list of supply items.
<supplyItems> <supplyItem> <description>A revolutionary cushion-grip ballpoint pen that reduces required gripping power, relieving stress and alleviating writing fatigue. Excellent for people who suffer from arthritis or carpal tunnel syndrome. Medium point, black ink. Refillable.</description> <category>1</category> <itemRef>ODC 204-502-153</itemRef> <maxQuantity>5</maxQuantity> <requiresJustification>0</requiresJustification> <itemName>Dr. Grip Pens -- Blue Barrel</itemName> <itemID>1</itemID> <unitCost>4.99</unitCost> </supplyItem> <supplyItem> <description>A revolutionary cushion-grip ballpoint pen that reduces required gripping power, relieving stress and alleviating writing fatigue. Excellent for people who suffer from arthritis or carpal tunnel syndrome. Medium point, black ink. Refillable.</description> <category>1</category> <itemRef>ODC 204-708-834</itemRef> <maxQuantity>5</maxQuantity> <requiresJustification>0</requiresJustification> <itemName>Dr. Grip Pens -- Black Barrel</itemName> <itemID>2</itemID> <unitCost>4.99</unitCost> </supplyItem> <supplyItem> <description>Personalized business cards for all your networking needs.</description> <category>2</category> <itemRef></itemRef> <maxQuantity>500</maxQuantity> <requiresJustification>1</requiresJustification> <itemName>Personalized business cards -- 500 count</itemName> <itemID>3</itemID> <unitCost>25.00</unitCost> </supplyItem> ... <supplyItems/>
You can name the root element anything you like when you describe its location using
DataSource.recordXPath
. The same
sample could be written this way:
<records> <record> <description>A revolutionary cushion-grip ballpoint pen that reduces required gripping power, relieving stress and alleviating writing fatigue. Excellent for people who suffer from arthritis or carpal tunnel syndrome. Medium point, black ink. Refillable.</description> <category>1</category> <itemRef>ODC 204-502-153</itemRef> <maxQuantity>5</maxQuantity> <requiresJustification>0</requiresJustification> <itemName>Dr. Grip Pens -- Blue Barrel</itemName> <itemID>1</itemID> <unitCost>4.99</unitCost> </record> <record> <description>A revolutionary cushion-grip ballpoint pen that reduces required gripping power, relieving stress and alleviating writing fatigue. Excellent for people who suffer from arthritis or carpal tunnel syndrome. Medium point, black ink. Refillable.</description> <category>1</category> <itemRef>ODC 204-708-834</itemRef> <maxQuantity>5</maxQuantity> <requiresJustification>0</requiresJustification> <itemName>Dr. Grip Pens -- Black Barrel</itemName> <itemID>2</itemID> <unitCost>4.99</unitCost> </record> <record> <description>Personalized business cards for all your networking needs.</description> <category>2</category> <itemRef></itemRef> <maxQuantity>500</maxQuantity> <requiresJustification>1</requiresJustification> <itemName>Personalized business cards -- 500 count</itemName> <itemID>3</itemID> <unitCost>25.00</unitCost> </record> ... <records/>Data for a tree-like DataSource can be specified with the same format. The following code example is for supply categories. This will make use of attribute values instead of separate tags for each property on the supply category. The attribues are not case sensitive and can be named after either the DataSource title or field name.
<categories> <category itemName="Office Paper Products" parentID="root" /> <category itemName="Calculator Rolls" parentID="Office Paper Products" /> <category itemName="Adding Machine/calculator Roll" parentID="Calculator Rolls" /> ... </categories>Note that all records must define values for the itemName primary key field and for the parentID field that establishes the tree relationship.
When you have date, time or datetime fields these need to be formatted the same way they are expected to be formatted in REST responses. If data for a field does not match the expected format it will be parsed using JDK built-in DataFormat parsers, in lenient mode, in the server's default locale. Examples of date, time and datetime formats.
// date - YYYY-MM-DD 2014-12-06 // time - hh:mm:ss 22:01:45 // datetime - YYYY-MM-DDThh:mm:ssZ 2014-12-06T22:01:45-04:00
DateFormatAndStorage