Data Extraction

Data Extraction API v2.0 Release Notes

julia.pond@webtrends.com — Mon, 21 Dec 2009 17:33:37 GMT

<fanfare>Version 2 of the Data Extraction API is now available.</fanfare>

As always, we strongly encourage upgrading to version 2.0, so that you can put the new functionality to work and give us your feedback and ideas for future development.

If you can’t migrate immediately, version 1.1 is still available for use (documentation at http://product.webtrends.com/dxapi/v1/index.html). Note: Using some v1.1 URIs in v2 will produce different results, because of changes in data. In XML format, there will be new values (for example, isSearchable); in HTML, layout is altered.

To use v2 features, specify v2 in your URIs. For example:

https://ws.webtrends.com/v2/ReportService/profiles/YcL4a5dufF6/reports/RZ2YgglpHk5/info?format=xml

This URI uses version 1.1:

https://ws.webtrends.com/v1_1/ReportService/profiles/YcL4a5dufF6/reports/RZ2YgglpHk5/info?format=xml

Additions and Changes

Alerts and Notes API: Use the syndication feeds method to create a feed of alerts and notes.
Key Metrics: Use the keymetrics method to summarize important data for profiles of interest.
Internationalization: The language parameter enables you to specify a language preference for report data.
More reports available: Browsers by Versions, Onsite Ad Impressions, Onsite Ad Clickthrough Rates, Onsite Ad Clickthroughs, Cities, Countries, Mobile Devices, Mobile Browsers (these have been removed from the list of legacy reports not supported).

Enhancements and Bug Fixes

Report API
- Measure ordering: Measures in report data appear in the order in which you specified them.·
- Multidimensional search: The get report data for period method now includes the attribute IsSearchable, so you can search all dimensions (like using the report filter in Analytics).
Trending for two-dimensional reports, including many built-in (legacy) reports: The Data Exchange API now provides the same trending functionality on two-dimensional reports as the Analytics user interface (where the trend is shown for the first dimension). For example, the Content Groups and Sub-groups report is not available in Version 1 of the API; in Version 2, you can get trend data on the Content Groups dimension. Note: On built-in (legacy) reports, interval data is only available for Visits and Page Views.
ENG367660: Reports available to user not accurate
ENG367545: Token names for Mobile Browser and Mobile Device reports not current
ENG367783: API error in Insight

Update to Excel/VBA Sample Application

michael.love@webtrends.com — Wed, 18 Nov 2009 16:00:17 GMT

I have seen a bit of an interest in the original sample application that I put together and since then there have been a number of changes to to DX API. With the latest release of the Webtrends Analytics Insight user interface we have now update this API to V2. Along with that I have had a chance to do additional testing of the Excel VBA application. These two events have now led to version v6 of the Sample Application.

So, what’s in this latest version? Well, first of all, there are bug fixes. I won’t detail them here, but there were a few and there are probably a few still in there. Next I have updated all of the REST calls to V2 of the API. And, finally, I have added new functionality that will allow you to see both the Profile and Report Meta data items for any selected report. It’s all in the documentation, so please read it for details.

One of the interesting updates in the V2 DX API is the inclusion of two new reports: an Account Key Metrics report and a Profile Key Metrics report. These have been added as new reports in this Sample Application as well.

So, as always, enjoy this new version of the application and any feedback you have is welcome. The new version can be found as an embedded object in the attached documentation.

Michael Love

Senior Solutions Engineer

SDK, ADO.NET DataTable/DataSet, and Code Bits

adron.hall@webtrends.com — Fri, 23 Oct 2009 12:31:12 GMT

Recently I have been working a lot with the Webtrends DX web services here internally. Our consulting & services group has been working with various technologies to connect, retrieve data, and provide that data in informational reports to clients and partners. In the course of working on these various projects I have added some functionality to the Webtrends DX web services SDK and also made some fairly solid in roads on the practices used to develop applications for Excel, Outlook, and other Microsoft Office Applications. Here is a break down of the progress I have made over the last few weeks. I hope it is helpful in everyone's efforts to extend, report, and provide analytics data & knowledge to their respective organizations!

Webtrends DX SDK - The SDK provides dramatically simplified standard programmatic access to reports, profile listings, report listings, and a mapper object to provide the reports in various object formats. Recently I have added the ability for the mapper object to take a standard dimensional report object and provide the report as a standard ADO.NET DataTable. This dramatically simplifies a lot of the functions of sorting, finding, or specifying various parts of the report data or manipulating it for presentation. In addition this is a more familiar object format for ETL using SSIS or even other ETL tools out on the market. The following is an example of the code used to retrieve a report with the SDK and then do a conversion into a standard DataTable Object.

User webtrendsUser =
    new User
        {
            AccountName = Resources.Options.Default.Account,
            UserName = Resources.Options.Default.Password,
            Password = EncryptionHelper.Decrypt(Resources.Options.Default.Password)
        };
 
var report = ReportFactory.CreateDimensionalReport("reportIdGoesHere", "profileIdGoesHere", null, null, null,
                                                   null, null, true, webtrendsUser);
Mapper mapper = new Mapper(report);
DataTable dataTable = mapper.MapReportDefinitionToDataTable();

In the code above, I have created a Webtrends User object and assigned the appropriate parameters. In this case I am using the Resources (Options.Settings) available to Windows Application to store and retrieve those settings. This way the user will not have to enter them every time they want to run the report.

I then have the report object being pulled from the ReportFactory Object via the CreateDimensionalReport factory method. The parameters that are actually needed are the report ID, the profile ID, and the Webtrends User Object. The null values are other various parameters that can be set but do not particularly need to be set.

Then I create a Mapper Object and pass it the report object via the constructor. This object requires a report object upon instantiation. I setup the object this way since there always needs to be a report object before any manipulations or mappings can be made.

The last step is to declare the DataTable Object, then pull the data table from the builder method MapReportDefinitionToDataTable().

For the SDK assembly and code check out the Document Outlining the offering. In addition over the next couple of weeks I will be posting more information regarding how to use the SDK, best practices, and code snippets for accessing the Webtrends DX Web Services. There will be additional blog entries and definitely some documentation coming soon (I'm working on it!). But in the meantime, if you have any questions, comments, or ideas on how we should extend or modify the SDK please let us know here on the Webtrends Developer Forums!

I'll have some more entries real soon about best practices, how to code against Outlook or Excel, and more. So subscribe and stay up to speed on everything Webtrends! Thanks for reading.

Dissecting the Sample Excel/VBA Application - Part 4

michael.love@webtrends.com — Thu, 22 Oct 2009 13:50:56 GMT

This is the final installment of the postings on the sample application. Up to this point we have dealt with the process of getting information about profiles and report definitions. We have yet to get any real report data. So, this post will deal with that process and what we do with the report results once we get them.

The first step in the process is, of course that of creating a REST URL that requests report data. The application takes care of all of this based on the selections made by the user for such things as time periods to cover, search criteria, limiting rows, how to deal with totals, and the selection of specific measures. There is a fair amount of work involved in making sure everything is valid, but the end result is a REST URL that looks something the one shown below.

https://ws.webtrends.com/v1_1/ReportService/profiles/YcL4a5dufF6/reports/95df19b6d9f2/?totals=all&period=2009m01d16&format=html&suppress_error_codes=true

The major difference between this URL and those we have seen in the past is that there are multiple parameters specified which control what and how much data we get back. A second difference here is that we are also using a different data type for our results. In this case we have chosen the “html” format rather than JSON. When getting report results for Excel the “html” format makes it extremely easy to populate the result data into Excel. The result data is returned as a table definition which when pasted into a spreadsheet automatically places our data into rows and columns in a tight report fashion. In fact the result data from the REST URL above produces the results shown below.

<table border=1>

    <tr>

        <th>Time</th>

        <th>Browsers</th>

        <th>Views</th>

        <th>Visits</th>

    </tr>

    <tr>

        <td>1/16/2009</td>

        <td>  </td>

        <td>83752</td>

        <td>8941</td>

    </tr>

    <tr>

        <td>1/16/2009</td>

        <td>Microsoft Internet Explorer</td>

        <td>48488</td>

        <td>5193</td>

    </tr>

    <tr>

        <td>1/16/2009</td>

        <td>Mozilla</td>

        <td>13547</td>

        <td>1445</td>

    </tr>

    <tr>

        <td>1/16/2009</td>

        <td>Netscape</td>

        <td>5942</td>

        <td>636</td>

    </tr>

</table>

If you look closely at this result, you will see that we have a single table with a border (width of 1 pixel). The table is divided into several table rows. Each table row has a number of columns. These hold the dimension and measure data for the report. Also important to note are the first two rows in the table data. The first row holds the names of the report columns. In every report I have seen the first column has always been time. The second row holds the “total” data. Even when a report has no totals this row of data will be present and will contain the date value. The remaining rows contain the report detail as specified in the REST URL request.

It is these two top rows that may need some special handling. Suppose I have a series of queries that return report data for individual periods, say one week’s worth of data but one day at a time. What I really want to get to here is a single report that shows each result for each day, but I only want one set of header data and I don’t want any totals since I can create those myself. If I were to paste in the results from two successive days without making any changes to the results, I would get a spreadsheet that looks like the following.

Time	Browsers	Views	Visits
1/16/2009		83752	8941
1/16/2009	Microsoft Internet Explorer	48488	5193
1/16/2009	Mozilla	13547	1445
1/16/2009	Netscape	5942	636
Time	Browsers	Views	Visits
1/15/2009		93520	9992
1/15/2009	Microsoft Internet Explorer	54057	5821
1/15/2009	Mozilla	15601	1666
1/15/2009	Netscape	6120	635

Note that this table contains two sets of headers and two total columns. What we really want is a result that has only one header row and no totals. It should look like the following.

Time	Browsers	Views	Visits
1/16/2009	Microsoft Internet Explorer	48488	5193
1/16/2009	Mozilla	13547	1445
1/16/2009	Netscape	5942	636
1/15/2009	Microsoft Internet Explorer	54057	5821
1/15/2009	Mozilla	15601	1666
1/15/2009	Netscape	6120	635

This gives me a table that I can work with better for creating pivot tables, charts and appropriate totals.

So, how do we do this? Well, first of all, the application allows you to define that you want this data, headers or totals, stripped off before inserting into the spreadsheet of choice. Since we know exactly where these are in the result set, we really only need to identify which row of data to remove: headers are always in row one and totals are in row two. The code in the application that does this is shown below.

' Check to see if we need to strip off the Date/Totals row

If UCase(m_strIgnoreTotalDate) = "Y" Then

m_intCheck = InStr(1, g_strResult, "</tr>") ' Find the first end of table row - that's the header row

If m_intCheck > 1 Then

m_intStart = m_intCheck + 1 ' Start of date/total row

m_intEnd = InStr(m_intStart, g_strResult, "</tr>", vbTextCompare) + 4 ' End of date/total row

' rebuild the result without the date/total stuff

g_strResulta = Left(g_strResult, m_intStart + 3)

g_strResultb = Right(g_strResult, Len(g_strResult) - m_intEnd)

g_strResult = g_strResulta & g_strResultb

End If

' Get our data and put it on the Clipboard

' Check to see if we need to strip off the header row

If UCase(m_strIgnoreHeader) = "Y" Then

m_intCheck = InStr(1, g_strResult, "<table border=1>")

If m_intCheck = 1 Then

m_intStart = InStr(1, g_strResult, "<tr>") ' Start of the header row

m_intEnd = InStr(1, g_strResult, "</tr>", vbTextCompare) + 4 ' End of the header row

' Rebuild our result without that stuff

g_strResulta = Left(g_strResult, 16)

g_strResultb = Right(g_strResult, Len(g_strResult) - m_intEnd)

g_strResult = g_strResulta & g_strResultb

End If

The right approach is to first look to see if we want to remove the “Total” row and do that. The reason for this is that we know this data is in row two. If we were to delete the header row first, the total row would move to the first row and would complicate the logic. So, the easy way to do this once we have decided that removing it is the thing to do, we simply look for where that row starts and ends. The start position is simply the first character after the end of the first row and the end is the last position of the “</tr>” string. Once we calculate those two values we simply recreate our result to copy the data up to the start and after the end of that row.

The same basic approach is taken for the header row. We know exactly where the start of that row is because we know the number of characters leading up to it. We then calculate the end in the same way we did for the totals row except that we are only looking for the end of this first row. Once we have these values calculated, we can again recreate the result set to include only the data prior to and after the first row of data.

Once we have the result set that is needed for our spreadsheet the only real task left is to decide where to put it. By that I mean what worksheet and which cell. The cell we need is the top left cell. Then we paste our data in there and we are done.

I hope these posts have provided useful information. Whether you choose to use this sample application or to write your own my goal here has been to provide some insight into how one can approach using the REST URLs with Excel and VBA. I will continue to work on this application with bug fixes (I am sure there are a few) and new functionality. Eventually I will add the trend reports and will support future versions of the API.

As always, I welcome your comments.

Michael Love

Senior Solutions Engineer

Dissecting the Sample Excel/VBA Application - Part 3

michael.love@webtrends.com — Wed, 21 Oct 2009 14:23:46 GMT

The previous posts on this application have provided some insight on how to get general profile and report data. We have also parsed the results that were returned assuming that we used the JSON data format. So now we are just about ready to create a REST URL that will allow us to fetch some actual report data. But before we do there is one last task to be performed and that is getting a list of measures that are available for our desired report. We need to do that just in case we want to limit the measures return to specific measures rather than all. One other element we are interested in knowing is whether or not we can use the “search” functionality to limit the results based on contents in the dimension data. Whether or not we can do that is determined by whether or not the report is multi-dimensional or hierarchical.

The REST URL to make this request to get what is referred to as the report meta data is shown below. This URL clearly defines both the Profile and the Report desired via their ID values. The “info” field defines this as a request for the report meta data.

https://ws.webtrends.com/v1_1/ReportService/profiles/YcL4a5dufF6/reports/95df19b6d9f2/info/?format=json

This request is made in the same way that previous requests are made and were discussed in my prior posts. The resulting data in JSON format will look something like the data below. Remember, however, that there are no spaces or other extraneous characters returned. These are shown here to make this easy to read.

    "accountID" : 18079,

    "profileID" : "YcL4a5dufF6",

    "ID" : "95df19b6d9f2",

    "name" : "Browsers",

    "language" : null,

    "type" : "dimensional",

    "properties" : {

        "isHierarchy" : false,

        "intervalsEnabled" : false,

        "internalID" : "topbrowsers_v",

        "IsRealTimeCompatible" : true,

        "ProfileCategory" : null

},

    "dimension" : {

        "ID" : "Desc",

        "name" : "Browsers",

        "type" : "data",

        "Range" : null,

        "Properties" : null,

        "SubDimension" : null

},

    "measures" : [

            "name" : "Views",

            "accumulationType" : null,

            "ID" : "Hits-0",

            "columnID" : 0,

            "measureFormatType" : "numeric",

            "AllowTotals" : true

},

            "name" : "Visits",

            "accumulationType" : null,

            "ID" : "Users-0",

            "columnID" : 0,

            "measureFormatType" : "numeric",

            "AllowTotals" : true

The entire result set is enclosed in curly brackets – “{ }”. There are within this set of data three (3) main segments of data. The first segment is general report information. Here we have such information as the Profile ID, the Report ID, the Report Name, and other general properties. It is here that we find a couple of critical pieces of data. These are both in the properties area: “isHierarchy”, and "intervalsEnabled". The “isHierarchy” value tells us whether or not we can use the search functionality. A value of “false” indicates this is a single dimension report and that we can do this. On the other hand, a value of “true” indicates that the report contains more than a single dimension and that we cannot use the search functionality. The second property item that may be of interest to us is the "intervalsEnabled" item. The value of this item tells whether or not trends data is available and we can use those time periods to retrieve data. Again, a value of “false” means interval/trend data is not available and “true” means that interval/trend data is available. This sample application has not implemented or enabled this feature, so for our discussion here the point is moot.

The second segment of data in our result is a list of dimensions. In a single dimension report this list is obviously short and appears as shown above. In a multi-dimensional report the sub dimensions are listed as shown below.

"dimension" : {

"ID" : "contentgroup",

"name" : "Content Group",

"type" : "data",

"Range" : null,

"Properties" : null,

"SubDimension" : {

"ID" : "xIfaV44n7l5",

"name" : "Content Sub-Group",

"type" : "data",

"Range" : null,

"Properties" : null,

"SubDimension" : null

}

Note that each sub dimension is listed as a hierarchical item of the dimension above it.

The key set of data that we want out of this result data is the list of measures. The measures are all in a third segment bounded by square brackets – “[ ]”. Each individual measure is defined within its own space bounded by curly brackets – “{ }”. For each measure there are a number of definition items including, "name", "accumulationType", "ID","columnID", "measureFormatType", and "AllowTotals". For purposes of our application we only need two of these values: “name” and “ID”. The name filed is used for display to the user in the interface and the ID is used when we create the REST URL to request our data. In the application we store this data in a custom data type. This data type is shown below.

Public Type MEASUREDATA

    name As String

    id As String

    accumulationType As String

    columnID As String

    measureFormatType As String

    allowTotals As String

End Type

Public g_Measures() As MEASUREDATA

The function below is used to extract the data we need from the JSON result. Once extracted the list of measures is available for the user to select. A check is also made to either enable or disable the use of the “search” functionality based on this being a single or multi-dimensional report.

Public Sub GetReportMeta()

' Get Report Meta so we can have list of measures

Dim m_intWhere As Integer

Dim m_strIsHier As String

frmProfileList.lstMeasures.Clear ' Clear form we will display list on

g_intMeasureCnt = 0 ' Initialize the profile count to 0

' Check to see if the user has selected the Profile Summary

If frmProfileList.lstReports.ListIndex = frmProfileList.lstReports.ListCount - 1 Then

frmProfileList.optMultiple.Enabled = False

frmProfileList.optSingle.Enabled = False

frmProfileList.optSEPeriods.Value = True

Exit Sub ' Exit this sub because there are no choices for measures

Else

frmProfileList.optMultiple.Enabled = True

frmProfileList.optSingle.Enabled = True

frmProfileList.optSEPeriods.Enabled = True

End If

' Set up URL for fetching Report Meta

g_strURL = g_strBaseREST & g_strSelectedProfileID & g_strReportsREST & g_Reports(frmProfileList.lstReports.ListIndex).id & g_strReportsMetaREST

' Call function to execute REST URL

If Not RunREST(g_strURL, g_strUserName, g_strPassword) Then

Exit Sub

End If

' Check to see if the report is hierarchical - no search on those

m_intWhere = InStr(1, g_strResult, "isHierarchy") ' Find start point for measures list

m_strIsHier = Mid(g_strResult, m_intWhere + 13, 4)

If m_strIsHier = "true" Then

frmProfileList.txtSearch.Enabled = False ' disable text entry in search box

Else

frmProfileList.txtSearch.Enabled = True ' enable text entry in search box

End If

' Display the JSON URL Result

m_intWhere = InStr(1, g_strResult, "[") ' Find start point for measures list

g_strResulta = Mid(g_strResult, m_intWhere + 1, Len(g_strResult) - (m_intWhere + 2))

g_blnMore = True

Do While g_blnMore

g_intSplit = InStr(1, g_strResulta, "},") ' See if there is more than one entry

If g_intSplit = 0 Then

g_strResultb = Mid(g_strResulta, 2, Len(g_strResulta) - 2)

Else

g_strResultb = Mid(g_strResulta, 2, g_intSplit - 2) ' Grab the first set of data

End If

g_strResultb = Replace(g_strResultb, """", "") ' Get rid of the extra " characters

g_strArr = Split(g_strResultb, ",") ' Create an array with this set of data

ReDim Preserve g_Measures(g_intMeasureCnt) ' Make room for new measure

g_Measures(g_intMeasureCnt).name = Mid(g_strArr(0), 6) ' Store the measure name

frmProfileList.lstMeasures.AddItem g_Measures(g_intMeasureCnt).name ' Add name for selection from form

g_Measures(g_intMeasureCnt).id = Mid(g_strArr(2), 4) ' Store the Measure ID

g_Measures(g_intMeasureCnt).accumulationType = Mid(g_strArr(1), 18) ' Store the Accumulation Type

g_Measures(g_intMeasureCnt).allowTotals = Mid(g_strArr(5), 13) ' Store the Allow Totals value

g_Measures(g_intMeasureCnt).columnID = Mid(g_strArr(3), 10) ' Store the Column ID

' Check to see if we are done parsing

If g_intSplit = 0 Then

g_blnMore = False ' We're done

End If

g_strResulta = Right(g_strResulta, Len(g_strResulta) - g_intSplit - 1) ' Set up for next profile

g_intMeasureCnt = g_intMeasureCnt + 1 ' Bump the profile count

Loop

Application.Cursor = xlNorthwestArrow ' Set the application cursor back to normal (arrow)

End Sub

So, now we have all of our data available in the user interface for the user to select what data will be included in the requested report. The three keys areas of selection are the time periods, the search data (if available), and the selected measures. Other selections include limiting what rows to return and how to deal with report totals. I won’t go into all of the logic on these selections and the interactions, but there are certain checks that are made to ensure consistency and validity of the selected items. But once these selections are made the user can create the URL and then copy it to the worksheet for eventual execution.

That’s all for this post. The next and last post on this application will deal with processing the results of the report data requests.

Michael Love

Senior Solutions Engineer

Dissecting the Sample Excel/VBA Application - Part 2

michael.love@webtrends.com — Tue, 20 Oct 2009 19:21:05 GMT

This second post on dissecting the Sample Application deals with the process of requesting two separate, but related, sets of report data: the List of Available Reports for a Profile and the List of Report Periods for which data may be available. I say “may be available” because it is possible that a report may not have existed when the Profile initially began analyzing. A custom report may have been added some time after that initial set up process. In that case the report would not have any data until when it had been enabled for the profile and the report template(s) available to the user. This point really has nothing to do with the application. I provided this comment only as a point of report data clarification.

The process in place within the application is that these two data requests will take place when the user clicks on a Profile Name in the combo list. The request for Report List is made first followed by the List of Periods. There is nothing that requires that one be done before the other; it’s just the sequence in which I chose to do them. The REST URL to request a Report List looks like the following:

https://ws.webtrends.com/v1_1/ReportService/profiles/YcL4a5dufF6/reports/?format=json

Note that again I use the JSON data format to keep down the amount of data returned. The process used to create this URL and send it is much like that used for the Profile List. The difference, of course, is in the format of the REST URL where we need to include the Profile ID and then add the fields to identify the request as one for the Report List.

So, the data format that I get back looks very similar to the format used for the Profile List.

[

{"accountID":18079,

"profileID":null,

"name":"Creatives",

"ID":"WcTGeY1bNj5",

"language":null,

"type":null,

"Category":"Marketing",

"IsHierarchy":false,

"IntervalsEnabled":false,

"IsRealtimeCompatible":false,

"properties":null

{

"accountID":18079,

"profileID":null,

"name":"Creative Types",

"ID":"uZK6QjyBNj5",

"language":null,

"type":null,

"Category":"Marketing",

"IsHierarchy":false,

"IntervalsEnabled":false,

"IsRealtimeCompatible":false,

"properties":null

}

]

The formatting here has been added only to make reading the data easier. The JSON format returns one long string bounded by square brackets – “[ ]”. Within these brackets each report is detailed within a set of curly brackets – “{ }”. Within those brackets are the individual meta items that define the report in a general format. The two items that are really important to us here are the “name” and the “ID”. The application parses this data using the following code.

' Display the JSON URL Result

' We need to parse out the results to get the data we need

frmProfileList.lstReports.Clear ' Clear form we will display list on

g_intReportCnt = 0 ' Initialize the profile count to 0

g_strResulta = Mid(g_strResult, 2, Len(g_strResult) - 2) ' Strip the [] characters from each end

g_blnMore = True

Do While g_blnMore

g_intSplit = InStr(1, g_strResulta, "},") ' See if there is more than one entry

If g_intSplit = 0 Then

g_strResultb = Mid(g_strResulta, 2, Len(g_strResulta) - 2) ' Grab the enry when only one left

Else

g_strResultb = Mid(g_strResulta, 2, g_intSplit - 2) ' Grab the first set of data

End If

g_strResultb = Replace(g_strResultb, """", "") ' Get rid of the extra " characters

g_strArr = Split(g_strResultb, ",") ' Create an array with this set of data

ReDim Preserve g_Reports(g_intReportCnt)

g_Reports(g_intReportCnt).id = Mid(g_strArr(3), 4) ' Store the Report ID

g_Reports(g_intReportCnt).name = Mid(g_strArr(2), 6) ' Store the Report Name

g_Reports(g_intReportCnt).accountid = Mid(g_strArr(0), 11) ' Store the Account ID

g_Reports(g_intReportCnt).profileid = Mid(g_strArr(1), 11) ' Store the Profile ID

g_Reports(g_intReportCnt).language = Mid(g_strArr(4), 10) ' Store the Language

g_Reports(g_intReportCnt).type = Mid(g_strArr(5), 6) ' Store the Report Type

g_Reports(g_intReportCnt).category = Mid(g_strArr(6), 10) ' Store the Report Category

g_Reports(g_intReportCnt).ishierarchy = Mid(g_strArr(7), 13) ' Store Hierarchy vlaue

g_Reports(g_intReportCnt).intervalsenabled = Mid(g_strArr(8), 18) ' Store Interval enablement

g_Reports(g_intReportCnt).isrealtimecompatible = Mid(g_strArr(9), 22) ' Store Realtime

g_Reports(g_intReportCnt).properties = Mid(g_strArr(10), 12) ' Store Properties

' Check to see if we are done parsing

If g_intSplit = 0 Then

g_blnMore = False ' We're done

End If

g_strResulta = Right(g_strResulta, Len(g_strResulta) - g_intSplit - 1) ' Set up for next report

g_intReportCnt = g_intReportCnt + 1 ' Bump the report count

Loop

' Sort the report list arrays

SortReports

' Load Report Names into form

For i = LBound(g_Reports) To UBound(g_Reports)

frmProfileList.lstReports.AddItem g_Reports(i).name ' Add name for selection from form

The parsing process here is almost identical to what was done for the List of Profiles, the exception being the actual fields of data that are extracted and stored. To store the data I have again set up a specific data type for the report data. It looks like the following:

Public Type REPORTDATA

accountid As String

profileid As String

name As String

id As String

language As String

type As String

category As String

ishierarchy As String

intervalsenabled As String

isrealtimecompatible As String

properties As String

End Type

Public g_Reports() As REPORTDATA

While this application does not yet use this, the “intervalsenabled” value is used to determine whether or not the trends option for report is valid.

Prior to loading the report names into the combo list on the form the application will sort that list into alphabetical sequence. Unfortunately, the list returned from the request does not do that for you. Using the custom data type here makes that sorting process easier because all of the report information I kept together.

The next step in the process here is to get the List of Periods that are available. This is yet another REST URL format for this set of data. This REST URL has the following format:

https://ws.webtrends.com/v1_1/ReportService/profiles/YcL4a5dufF6/periods/?format=json

This REST URL is almost identical to the Report List request with the one exception that it has the “periods” data request defined. As before, we still use the JSON format for our data type. This request is submitted just like all of the others we have done. The format of the result data is shown in the following:

{"Report":["2009m01d04","2009m01d05","2009m01d06","2009m01d07","2009m01d08","2009m01d09","2009m01d10","2009w02","2009m01d11","2009m01d12","2009m01d13","2009m01d14","2009m01d15","2009m01d16","2009m01d17","2009w03","2009m01","2009q01","2009"],"Realtime":[“2009m01d18”]}

This format is a bit different than what we have seen with Profiles and Reports. The outside delimiters in this case are the curly brackets – “{ }” – and the inside delimiters are the square brackets – “[ ]”. In parsing this data all we care about are the individual periods. The application, at this time also does not deal with the “Realtime” or express data. That data will be ignored in this version of the application. So, all we need to parse are the comma-delimited values, without the quote marks that are contained inside the first set of square brackets. The code to do that looks like the following:

' Parse out the URL Result - set into individual time periods

If InStr(1, g_strResult, "Realtime") Then

g_strResultb = Mid(g_strResult, 12, (InStr(1, g_strResult, "Realtime") - 15)) ' Need to account for profiles with Express Data

Else

g_strResultb = Mid(g_strResult, 12, Len(g_strResult) - 29) ' No express data

End If

g_strResultb = Replace(g_strResultb, """", "") ' Get rid of the extra " characters

g_strPeriods() = Split(g_strResultb, ",") ' Create an array with this set of data

' Load data onto Report Periods Form

frmProfileList.lstPeriods.Clear ' Start with empty list

For i = 0 To UBound(g_strPeriods)

ReDim Preserve g_strReportPeriod(i)

frmProfileList.lstPeriods.AddItem g_strPeriods(i) ' Save in the form

g_strReportPeriod(frmProfileList.lstPeriods.ListCount - 1) = g_strPeriods(i) ' Save in our array for use later

This is fairly straightforward code to do this. However, there are a few other items that we need to add to the list for the user to pick for time periods. These are the relative starting point of “current day”, “current month”, and “current year”. So these are manually added to our pick list since they have no representation in the list of periods returned. The code to do that is as follow:

' Add relative time periods at end of form and array

ReDim Preserve g_strReportPeriod(i + 3)

frmProfileList.lstPeriods.AddItem "Today"

g_strReportPeriod(frmProfileList.lstPeriods.ListCount - 1) = "current_day"

frmProfileList.lstPeriods.AddItem "This Month"

g_strReportPeriod(frmProfileList.lstPeriods.ListCount - 1) = "current_month"

frmProfileList.lstPeriods.AddItem "This Year"

g_strReportPeriod(frmProfileList.lstPeriods.ListCount - 1) = "current_year"

So, that ends this portion of the blog post. In the next blog post we will be getting additional report meta information and dealing with creating a REST URL to fetch actual report data.

So, until the next posting,

Michael Love

Senior Solutions Engineer

Dissecting the Sample Excel/VBA Application

michael.love@webtrends.com — Mon, 19 Oct 2009 19:23:06 GMT

Recently I posted a blog entry with a Sample Excel/VBA Application. The idea behind this application was to provide any developer who is interested a baseline set of code for using the REST API for extracting data from Webtrends Analytics. The latest version of that application is v4 and it supports the v1.1 REST URL API. This new version is attached to this blog. Please note that the attachment is a Microsoft Word document. The application which is and Excel spreadsheet with VBA driving the action is included in that document as an embedded file.

I have decided to follow up that posting with more detail on how that application works. This follow up work will come in a series of postings that sort of start at the beginning of the application and works its way through to the end. The focus will be on the process of building the REST URLs and then parsing the resultant data from each call.

The sample application assumes that there are no REST URLs to begin with and that the user of the application will need to create them interactively. This aspect of the application builds off the idea of the REST Generator application at http://generator.webtrends.com. To make this work requires a series of steps and calls to Webtrends to fetch specific data. I am assuming for purposes of this discussion that the reader has a fair knowledge of VBA with Excel and I will, therefore, not go into detail on every line of code in this application. The code is documented on a line by line basis.

So, let’s start with the sequence in which the application gets its data and what is does with it. To dynamically, interactively, generate a REST URL that requests report data we need to do the following:

1. Get a list of the profiles available to us.

2. Get a list of the reports that exist in the desired profile and the time periods that might exist for the reports.

3. Get the report data.

There’s a bit more detail involved in these and I will cover that as I cover each of the mail points in the list above. This posting will cover just the process involved in requesting a list of profiles.

I will warn you now that I am a brute force programmer and you may have better ideas than I do about specific functions and how they work. That’s good and I welcome any feedback you have.

Let’s begin by looking at the general format for a valid REST URL. For more detail on this I refer you to the Webtrends developer site at http://product.webtrends.com/dxapi/api_description.html. The general REST URL has the following format:

https://ws.webtrends.com/v1.3.1/ReportService.svc/profiles/faj38jaFH/reports/J893kIL34/?period=2009&format=xml

where,

https: = protocol

ws.webtrends.com = site where data request is sent

v1.3.1 = API version number

ReportService.svc = web service called

Faj38jaFH = profile id

J389kIL34 = report id

period=2009 = report period

format=xml = result data format

The actual format of the REST URL you will use will depend heavily on what the request is targeted to return. For a list of profiles the REST URL will look like the following:

https://ws.webtrends.com/v1_1/ReportService.svc/profiles/?format=xml

or,

https://ws.webtrends.com/v1_1/ReportService.svc/profiles/?format=json

The difference here is the format in which the result information is returned. In the sample application I have used JSON. So, this is the format I will reference in this post.

Using VBA to request web data in this form is a simple process. In fact, in the sample application I have included a standard function to make this retrieval, no matter what data is being requested. This function looks like this:

Private Function RunREST(url As String, ByRef username As String, ByRef password As String)

' Function will execute a REST url along with the UserName and Password passed as parameters

Application.Cursor = xlWait ' Set the Wait cursor while executing the REST Call

' Set up the object needed for the web service request

Set g_objHttp = CreateObject("MSXML2.XMLHTTP")

' Open our connection and pass the user credentials

g_objHttp.Open "GET", url, False, username, password ' False indicates the call is synchronous - wait for URL Result

g_objHttp.send ' Send the request

g_strResult = g_objHttp.responseText ' Get the URL Result

' Check to see if our credentials passed - display error when they don't

If g_strResult = "Must authenticate" Then

Application.Cursor = xlNorthwestArrow ' Reset the application cursor to normal (arrow)

MsgBox "Authentication Failed", vbCritical, "Authentication Error"

username = "" ' Reset User Name

password = "" ' Reset Password

RunREST = False ' Set return value for failure

Exit Function ' Exit - we're done here

End If

RunREST = True ' Set return value for success

End Function

The key statements in this function really are those that do the actual call of the REST URL. They are:

g_objHttp.Open "GET", url, False, username, password

g_objHttp.send

g_strResult = g_objHttp.responseText

The g_objHttp.Open statement creates an open connection for the HTTP request. Asscociated with that statement we pass the type of call, “GET”, the actual URL to be used with the call, and the user name and password. The parameter shown as “False” indicates how this call is to be handled by the application. The value “False” indicates that it is a synchronous call, meaning that the application will wait for the response before continuing. Passing the user name and password prevents the user from being challenged when the call is made. The g_objHttp.send statement does just what is implied – it sends the URL request via the web. The last statement – g_strResult = g_objHttp.responseText – assigns whatever the result of the request is to our internal variable for further analysis.

The remaining statements in this function simply check to determine whether or not the authentication credentials had been accepted or rejected. If either the user name or password were not valid, the result returned will simply state “Must authenticate”. That’s what this code checks for and displays an appropriate error and returns ‘False” as the function value.

Now that we have a result set it is time to do something with the data. Remember that the data is returned in JSON format. The JSON format for a list of profiles looks something like this:

[{"ID":"YcL4a5dufF6","name":"...sample: Zedesco 2009","AccountID":18079},

{"ID":"CxoSsHxKlF6","name":"..Sample: ExactTarget 2009","AccountID":18079},

{"ID":"1ZnDMawXD16","name":".sample: Z-2005","AccountID":18079},

{"ID":"1To2KI7wD16","name":"Sample: ET-2005","AccountID":18079}]

The paragraph formatting has been added to make this more readable. The actual result is returned as one log string. The entire string is bounded by square brackets – []. Each individual set of profile data is bounded by curly brackets – {}. Eand each of these is separated by a single comma. Each set of profile data has three name/value pairs. The names are: “ID”, “name”, and “AccountID”. Not that these are all surrounded by double quotes and also separated by commas. So parsing this data out should not be any serious challenge. Using my brute force coding style I accomplish this with this set of statements:

' Display the JSON URL Result

' We need to parse all of the results to get what we want

frmProfileList.lstProfiles.Clear ' Clear form we will display list on

g_intProfileCnt = 0 ' Initialize the profile count to 0

g_strResulta = Mid(g_strResult, 2, Len(g_strResult) - 2) ' Strip the [] characters from each end

g_blnMore = True

Do While g_blnMore

g_intSplit = InStr(1, g_strResulta, "},") ' See if there is more than one entry

If g_intSplit = 0 Then

g_strResultb = Mid(g_strResulta, 2, Len(g_strResulta) - 2) ' Only one entry - grab it

Else

g_strResultb = Mid(g_strResulta, 2, g_intSplit - 2) ' Grab the first set of data

End If

g_strResultb = Replace(g_strResultb, """", "") ' Get rid of the extra " characters

g_strArr = Split(g_strResultb, ",") ' Create an array with this set of data

ReDim Preserve g_Profiles(g_intProfileCnt)

g_Profiles(g_intProfileCnt).id = Mid(g_strArr(0), 4) ' Save the Profile ID

g_Profiles(g_intProfileCnt).name = Mid(g_strArr(1), 6) ' Save the Profile Name

g_Profiles(g_intProfileCnt).accountid = Mid(g_strArr(2), 11) ' Save the Profile Account ID

' Check to see if we are done parsing

If g_intSplit = 0 Then

g_blnMore = False ' We're done

End If

g_strResulta = Right(g_strResulta, Len(g_strResulta) - g_intSplit - 1) ' Set up for next profile

g_intProfileCnt = g_intProfileCnt + 1 ' Bump the profile count

Loop

' Sort the resulting array into sequence

SortProfiles

' Now load the form with the Profile Names

For i = LBound(g_Profiles) To UBound(g_Profiles)

frmProfileList.lstProfiles.AddItem g_Profiles(i).name ' Add name for selection from form

If you look for this code in the application, you will find it in the “GetProfiles” subroutine. One of the methods I used to help store the profile data was to set up a specific data type for it. That data type, found in the Data_Module module looks like this:

Public Type PROFILEDATA

name As String

id As String

accountid As String

End Type

Public g_Profiles() As PROFILEDATA

Creating this data type makes it easy to keep all of the profile data for one profile together and then be able to sort the data into “name” sequence. When we present the profile data to the end user we use the Profile Name. When we use the profile data to later request a list of reports we will need the corresponding Profile ID.

The next blog post will cover creating a REST URL to retrieve a List of Reports and what do with that result.

Michael Love

Senior Solutions Engineer

Data Extraction API v1.1 Release Notes

julia.pond@webtrends.com — Thu, 24 Sep 2009 22:27:39 GMT

To use v1.1 features, specify v1_1 in your requests. This URI uses version 1.1:

https://ws.webtrends.com/v1_1/ReportService/profiles/YcL4a5dufF6/reports/RZ2YgglpHk5/info?format=xml

The prior version is still available*, but we encourage you to move quickly to adopt the new version, so that our enhancement efforts are driven by your feedback.

This URI uses version 1:

https://ws.webtrends.com/v1/ReportService/profiles/YcL4a5dufF6/reports/RZ2YgglpHk5/info?format=xml

Changes made in version 1.1 are flagged in the documentation.

* Note: URIs produced with Generator v1.1 are not compatible with the Data Extraction API v1.0. If you need to produce URIs for version 1.0, then you can access Generator v1.0 here:

https://generator.webtrends.com/v1

Additions and Changes

XML parsing: Additional metadata is defined for some views in the XML format (such as the profile summary method), so data collected with XML parser code that uses position, rather than the XML tag name, may not be positioned as before when using v1.1. Webtrends recommends as a best practice to use the XML tag name, rather than position, in XML parsing, so this is a good time to change position-dependent parser code to use tag names.
Combining data from different accounts: When you log in to a parent account, you can query profiles from any sub-accounts that you have permission to access. Visibility into profiles is no longer restricted to profiles in the account to which you authenticate in your initial session. You can now combine data from profiles in different accounts, specifying an account with the account name parameter.
Templates and report access: The Data Exchange API now restricts access to data based on permissions and on whether a report is included in the templates available to you.
New XML2 format: “XPath-friendly” output is now available (useful in dynamic applications).
Measure totals: Additional choices for returning totals.
Changes to the Generator utility: Added XML2 and CSV formats and new totals functionality. Removed charting and mobile routing functionality.

Bug Fixes and Enhancements

ESC365229: The profile summary report has one changed measure and one new measure:
Changed "Average per Day" to "Average Page Views per Day"
Added "Average Visits per Day"
ENG363413: Profile and report permissions error in Generator.
Data Extraction API numbers are returned directly from the analytics engine without manipulation, so they are the same as those in Webtrends reports. Discrepancies that still exist are due to calculations performed by the Webtrends user interface (rounding, averages, and calculated measures).
Search string filters on reports are now applied to both dimensions.
Reports including the Pages dimension now return page titles as a data column, along with the URI (formerly, the page title was omitted).

Known Issues

ENG365353: The list templates method may not produce expected results.
ENG365139: The Languages and Java Versions reports may not produce expected results when using the search parameter, since actual values differ from display values (“en” in the data, but “English” is displayed).
ENG364956: Comparison of certain measures (particularly averages and subtotals) may reveal differences between v1 and v1.1 of the API. v1.1 is considered more accurate, because the data comes directly from analysis databases.
ENG365121: Incorrect or partial data is returned (values contain percent symbols).
ENG365341: Column aggregates on the geography drilldown in Excel are incorrectly displayed.
ENG365425: Report and template listings do not reflect the user’s view permissions.

Excel, VBA, and REST URLs - A Sample Application

michael.love@webtrends.com — Wed, 26 Aug 2009 13:07:40 GMT

It has often been said that Excel is the one application most often used when analyzing data on a computer. Its use far outnumbers any other tool including those tools that fall within the category of Business Intelligence. It is also our experience at Webtrends that our customers also use Excel to a great extent for the further analysis, integration , and presentation of data extracted or exported from Webtrends Analytics reports. So, it is important that Webtrends make available to its customers the tools needed to get that data from the Webtrends Analytics reporting tool and into Excel in an easy and straightforward way. This is what the new REST URL DX API is all about.

There are multiple ways that you get data into Excel using the REST URL APIs. The most important part of the process is to create/generate a valid REST URL in the first place. Once that is done, it a fairly straightforward process to use it to get web data and populate rows of data where you want them. Two very simple ways to do this, of course, are: 1) use the REST URL Generator application available at https://generator.webtrends.com; and 2) to use REST URLs that are generated with the new Webtrends Analytics 9 Insight interface (http://insight.webtrends.com). In each of these all you need to know is your log in information and what data (report) you want to extract. Using either of these methods will get you well on your way to populating Excel workbooks and then using that information in the way you want.

But, suppose you want to do more. Suppose I want to extract data from Webtrends and then use the power to of Excel to store it, update it, and present it in the ways you need. And, I want all of this in a single tool. I don't want to go back and forth between a Webtrends application and Excel. I want it all in Excel.

Well, this is possible. Using tools such as VBA (Visual Basic for Applications) within Excel, you can access the Webtrends data, build REST URLs and populate the data into worksheets where you need the data. The interesting part of this process is really how to you build and manage the REST URLs that extract your data. If you read the developer notes on this site about how to build REST URLs, you will quickly learn that there is more than one format based on the data you need to get and that data can be returned in different formats. These formats go by the names of XML, JSON, and HTML. Somewhere imbedded in these formats hides the data I want and need to get the job done. And why do you need to know this? Because the various REST URL methods are needed to find out what data is available, such as profile, reports, time periods and the like. And you will need this information if you are going to build a working REST URL that ultimately gets report data.

So, we have developed this Excel/VBA Sample Application that will make this easy for you. The application has embedded a REST URL Generator functionally equivalent to the REST URL Generator mentioned above. The VBA code that is part of the application uses all of the appropriate methods for extracting configuration information and parses it out from the data streams that come from using the REST URLs methods for doing things like getting a list of profiles and so forth. The application also stores the REST URLs that you generate and lets you define where you want this data to go.

So, what does this mean for you? Well, you can use the application as is. It is quite functional and has complete usage documentation associated with it. Or, as it is more intended, you can use this as a base for exploration or expansion on its functionality. Add to it to do more formatting, analysis, data comparisons and the like. Then feed what you did back into this developer network so that others may use your experience as a launching pad for their own application.

Whatever you do with it, enjoy!

Michael Love

Senior Solutions Engineer

Open Campaign and our DX API

eric.butler@webtrends.com — Sat, 08 Aug 2009 21:29:57 GMT

The very clever Webtrends Open Campaign is all about transparency. We are showcasing our technologies, along with creative solutions from some of our partners, to demonstrate what's possible using data of all shapes and sizes. From our site:

Have a look around the web site. Get to know the people and partners involved. And come back often. We'll explore our results and introduce new marketing strategies every week.

One post early on in the blog was from our own Robin Balmer, who has put together some visualizations around metrics from the campaign. He's pulling data from our DX API, and put a flash visualization together to present a trend of views and visits over time. Simple, elegant and relatively easy to do as he notes in his blog post:

I set this task of visualization-building as a fun programming challenge for myself and as a means to get the visualizations to work stylistically with the design of the web site. But you don’t have be an experienced coder to get at Webtrends data - you can use the API to pull data right into Excel and analyze or visualize it from there. I saw some great examples of this during Webtrends’ last Developer Day.

Nice work Robin!

Webtrends Data Extraction is Live!

eric.butler@webtrends.com — Mon, 27 Jul 2009 14:41:50 GMT

We are pleased to announce the general availability of our Webtrends Data Extraction API. This v1 release is the culmination of a ton of effort of many individuals inside Webtrends, as well as the terrific support from our partners and customers who have provided feedback and suggestions during our beta process. Thank you all for helping us reach this milestone.

This open, innovative API service, provides you with elegant access to your powerful Webtrends analytics data. This is the first time we have leveraged a full beta cycle as part of our iterative development process. It is also the first time we've reached out to our community, via this Developer Network, to solicit feedback and validate our approach. It's been a huge win for us.

What's new?

Recently, we've been working on some behind the scenes governance (security and usage controls), and focusing on fixing bugs/issues. We've also modified the HTML and CSV formats slightly to return data back in a more easy-to-consume structure, per your suggestions. These changes, along with some additional functionality has made this API much easier to use directly within Excel, making it the only API in the industry that is powerful enough for sophisticated applications, while easy enough to also be used by everyone via Excel.

We've launched the formal "/v1/" version structure into the URL. If you are using the "/beta/" request in your URL, you should switch over to "/v1/". We'll continue to use the "/beta/" structure for future updates, so please don't use it for any production services or applications. Please consult our documentation for the most current information on the API.

What's next?

We are going to continue to iterate rapidly on this Data Extraction platform. It's already become very popular with our customers and partners as it is so easy to use. Look for a regular maintenance release on a monthly basis, along with a 3-4 month cycle of releases that will include awesome new features.

We will continue to use this Developer Network for communications, ideas, and general sharing of information. As always, please feel free to login and post questions, respond to other questions, or post ideas you have. Your creativity and innovation is contagious. It's all good.

Data Extraction API Updates

eric.butler@webtrends.com — Tue, 07 Jul 2009 18:59:13 GMT

Hi everyone, here are a couple of updates for you on our Data Extraction (DE) API rollout. Your comments and suggestions have been extremely helpful...keep the feedback coming...

Getting Close to v1

We are currently finalizing some additional functionality and fixing some issues found during this Beta process. Our plan is to release RC1 (release candidate) this week.

As we release RC1, we will start our versioning process. We'll keep the /beta/ path open for a few weeks still to allow everyone ample time to migrate over, but we'll be introducing the /v1/ path into the request URL as the first supported version of the service. We'll be updating our documentation accordingly.

New functionality

We've added some new functionality to our Data Extraction API (our reporting REST-based web services) in the past couple of weeks. Here's a quick synopsis:

Express Report Support - Based on very popular demand, Webtrends reporting data that is Express compliant and enabled will now show realtime data via the DX API. There are various method additions to support this:
- The list report time period method shows the additional time period as "list name="Realtime"".
- The list reports for profile method now indicates whether the report "IsRealtimeCompatible", so you know if that particular report will give you Express data.
- And to make the request for Express data, simply use "period=current_day" in the get report data for period method. Also now included in that response is the "LastUpdate" time stamp so you know how recently your data was updated.
Support for custom time ranges - In addition to the standard "periods" of time available (eg, current_month, etc.), you can now select relative and specific custom time ranges in your request. For example, if you'd like to select the past 7 days, starting with yesterday, you would include "start_period=current_day-7&end_period=current_day-1", or "start_period=2009m6d30&end_period=2009m7d6"
New response format - HTML - We've morphed our "excel" format into a new "html" format response structure. You can still select "format=excel" as an option, but the responses are all coming back in the new HTML format. This gives us better control of the response, and more consistency with the structure.

More coming soon

Look for more information coming from us over the next couple of weeks. It's all good!