Question

I want to parse TestComplele 7 logs programmatically. Specifically, I want to get names and statuses (success\failure\warning) of each test.

According to manual, I use the following command to get test run logs:

Log["SaveResultsAs"] ("PathToMyFolder", 1);

This creates a bunch of files and folders, including XMLs which can be parsed by my script.
Here is an example of such an XML:

<ProjectLog>
  <ProjectLogItem id="0" pid="-1" nety="False">
    <Status>805039D.gif</Status>
    <No>1</No>
    <Name><![CDATA[Script Test Log [cTaStoring\cTaStoring_Start]]]></Name>
    <StartTime>13.04.2012 12:26:15</StartTime>
    <EndTime>13.04.2012 12:27:47</EndTime>
    <RunTime>0:01:32</RunTime>
    <Details></Details>
    <Messages/>
   </ProjectLogItem>
   <ProjectLogItem id="1" pid="-1" nety="False">
    <Status>905039C.gif</Status>
    <No>2</No>
    <Name><![CDATA[AutomaticStoring]]></Name>
    <StartTime>13.04.2012 12:27:47</StartTime>
    <EndTime>13.04.2012 12:30:42</EndTime>
    <RunTime>0:02:55</RunTime>
    <Details></Details>
    <Messages/>
   </ProjectLogItem>
</ProjectLog>

As you can see, I can get almost everything I want, but the most valuable piece of data, the status of a test is written as a file name of a picture showing the corresponding icon! E.g. 805039D.gif contains a checkmark and 905039C.gif contains an exclamation mark.

It would be easy to hardcode these names in my scripts, but the problem is that they change constantly!

Questions:

1. Why have they decided to use image names instead of some enumeration
   to output statuses?
2. Why on earth do icons change their names?
3. How to get the statuses of tests?

Answer 1

I was most annoyed by the same issue. So I wrote this in C# to solve it:

I would get the Length of the file that the image refers to, then

            if (b.Length == 610) status = "Pass";
            if (b.Length == 1016) status = "Fail";
            if (b.Length == 353) status = "Inconclusive";

So if the file it refers to is 610 in length, then pass, etc…