We continue our discussion on external data and data warehouse. Today we look at the role of metadata for external data. The metadata includes document ID, date of entry into the warehouse, description of the document, source, source date , classification , index words, purge date, physical location, length, and references. Associated with data is another type of data called notification data.
This is merely a convenience for the users to subscribe to alerts when certain type of external data is available.
External data is not directly stored in the warehouse. Its entries are made in the metadata of the data warehouse and the external data resides elsewhere.
One of the important design considerations of external data is that it often contains many different components some of which are more useful than others, so they have to be divided into manageable units.
Since external data cannot be reshaped, it has no place in the data model. The best that can be done here seems to be to keep references to the external data.
Summarization and Trends of external data are often more useful and are stored with the external data. Take Dow Jones index for example, the daily report is important but the trends and the monthly average can be calculated and put together with the external data.
 Every piece of information external or otherwise has a lifetime at the end of which it could be discarded or archived. On the other hand, the cost of the metadata is pretty low and that could be kept in tact.
One of the advantages of using external data is that it can be used to compare with internal data.  The comparison allows management a unique perspective. This comparison is usually done on a common key but that common key is sometimes difficult to find.  For example, the key-structure of the external data may need to be converted to that of the internal data. Or the data may not be cleaned such as segregating products to compare apples to apples and not apples with oranges.
In short, we have seen that the external data holds a lot of information but has many issues to deal with. Hence it is captured, processed, stored on different media, setup with notifications and referenced within the warehouse. The warehouse manages and tracks the external data without the latter pervasively invading the system.

---------------
In addition to our usual post today, I want to show sample code for using the UI automation library with iOS:
The following code tries to add a sentence six times to the Notes app.

var target = UIATarget.localTarget();
var app = target.frontMostApp();
var testName = 'Take Notes'
UIALogger.logStart(testName);
for( var i = 0; i < 6; i++)
{
if ( i == 0) app.navigationBar().buttons()["Add"].tap();
else app.mainWindow().tableViews()[0].cells()[0].tap();
var page = app.mainWindow().textFields()[0];
var currentContent = page.getValue();
currentContent +=  "The big brown bear ate the brownies in the big paper bag." ;
page.setValue(currentContent);
UIALogger.logMessage("Added sentence : " + i );
app.navigationBar().buttons()["Notes"].tap();
}
app.mainWindow().tableViews()[0].cells()[0].tap();

if (app.mainWindow().textFields()[0].getValue.match('/bear/g') ) {    UIALogger.logPass(testName);}
else {    UIALogger.logFail(testName);}