Are source files mandatory for clover-report

[Abhinav Sahu]

Hi,

I am getting error when running clover-report, like:

[clover-report] Failed to render syntax highlights for /home/ta230208/TeamCity_Agent_Space/sdvsue800/TeamCity-Agents/ta230208/work/47becb009e2f5761/Commons/src/java/com/teradata/datamovement/model/DDLComparable.java

[clover-report] java.io.FileNotFoundException: /home/ta230208/TeamCity_Agent_Space/sdvsue800/TeamCity-Agents/ta230208/work/47becb009e2f5761/Commons/src/java/com/teradata/datamovement/model/DDLComparable.java (No such file or directory)java.io.FileNotFoundException: /home/ta230208/TeamCity_Agent_Space/sdvsue800/TeamCity-Agents/ta230208/work/47becb009e2f5761/Commons/src/java/com/teradata/datamovement/model/DDLComparable.java (No such file or directory)

[clover-report]         at java.io.FileInputStream.open(Native Method)

[clover-report]         at java.io.FileInputStream.<init>(FileInputStream.java:120)

[clover-report]         at java.io.FileReader.<init>(FileReader.java:55)

[clover-report]         at com.cenqua.clover.registry.FileInfo.getSourceReader(FileInfo.java:442)

[clover-report]         at com.cenqua.clover.reporters.html.source.SourceRenderHelper.getChecksummingReader(SourceRenderHelper.java:413)

[clover-report]         at com.cenqua.clover.reporters.html.source.SourceRenderHelper.render(SourceRenderHelper.java:389)

[clover-report]         at com.cenqua.clover.reporters.html.source.SourceRenderHelper.gatherSrcRenderInfo(SourceRenderHelper.java:154)

[clover-report]         at com.cenqua.clover.reporters.html.source.SourceRenderHelper.insertLineInfosForFile(SourceRenderHelper.java:83)

[clover-report]         at com.cenqua.clover.reporters.html.RenderFileAction.insertLineInfos(RenderFileAction.java:284)

[clover-report]         at com.cenqua.clover.reporters.html.RenderFileAction.render(RenderFileAction.java:131)

[clover-report]         at com.cenqua.clover.reporters.html.RenderFileAction.call(RenderFileAction.java:115)

[clover-report]         at com.cenqua.clover.util.CloverExecutors$LoggingCallable.call(CloverExecutors.java:86)

[clover-report]         at clover.retrotranslator.edu.emory.mathcs.backport.java.util.concurrent.FutureTask.run(FutureTask.java:178)

[clover-report]         at clover.retrotranslator.edu.emory.mathcs.backport.java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1061)

[clover-report]         at clover.retrotranslator.edu.emory.mathcs.backport.java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:575)

[clover-report]         at java.lang.Thread.run(Thread.java:662)

[clover-report] /home/ta230208/TeamCity_Agent_Space/sdvsue800/TeamCity-Agents/ta230208/work/47becb009e2f5761/Commons/src/java/com/teradata/datamovement/model/DDLComparable.java (No such file or directory)

I believe this is due instrumented source files not found, right?

My question is whether is it manadatory to have source files when running clover-report? Will my coverage percentage increase if I include source files? Right now getting only 4% for a test but as told by developer it should be above 30%.

My build server is different than test server, still can I give source path? Packages (java) directory structure would be same but the location will be different.

Thanks in advance.

Regards.

[Marek Parfianowicz · Accepted solution]

Right now getting only 4% for a test but as told by developer it should be above 30%. My build server is different than test server [...]?

If coverage is lower that expected then it must be something wrong in your environment/build setup. There are many possibilities how to configure Clover for builds on mutiple machines. See:

https://confluence.atlassian.com/display/CLOVER/Using+Clover+in+various+environment+configurations

Questions worth answering:

• which configuration would fit best to your environment as per the Decision Matrix?
• do you use relative or absolute initstring for clover database(s)?
• have you copied clover database(s) from a build server to a test server or are they accessible through a shared network drive?
• is there a need to use -Dclover.initstring or -Dclover.initstring.basedir property on a test sever (see manual above)?
• do you have clover.jar deployed on test server (for instance in an application sever's /lib directory; depends on what kinds of tests do you run)?
• do you see coverage recording files produced on the test server in the directory where clover.db is located?

[Abhinav Sahu]

Actually my product has three components, one can be said as client and other two are servers. They all communicate through a message queue (MQ). Now the issue is I am getting coverage for only client code, I am not able to see any coverage for servers code and that's why my coverage is very low. I have instrumented code for all three components and a coverage.db file.

Can you please let me know how can I get coverage info of those codes as well?

[Marek Parfianowicz]

I suggest a following lecture:

* https://confluence.atlassian.com/display/CLOVER/Working+with+Distributed+Applications

Shortly speaking, you have to ensure that:

* Clover database (clover.db) is available from your application server - either use an absolute path for initstring (assuming that build server and test server is the same machine) or provide a -Dclover.inistring JVM option for server's JVM

* Clover runtime JAR (clover.jar) is available in classpath - the easiest way is to put it into <server_dir>/lib directory, so that Clover becomes available for all applications deployed on a server

I suggest checking server's logs whether they contain any Clover-related messages - they shall be present in case of problems mentioned above. You can also run an application server with -Dclover.logging.level=debug to get even more details.

Cheers
Marek

[Abhinav Sahu]

Hi Marek,

That's the problem build server and test server can't be same. :(

Can we give any URL kind of option in clover ant task so that .db can be accessed via other two components?

Thanks again.

[Marek Parfianowicz]

You can't use URL. Instead of this, just copy the Clove database from a build server to a test server and point to a new db location on that machine by passing the -Dclover.initstring=/path/to/clover.db JVM option.

[Deleted user]

Marek, just to be clear, the clover.license file must be in all the same places that the clover.jar is. So, if you put clover.jar in your application server's lib directory, a copy of the license file has to go there as well.

[Marek Parfianowicz]

Not really ... The clover.license is required for code instrumentation and for report generation. License key is not required to run instrumented application and record coverage files.

Excerpt from https://www.atlassian.com/licensing/clover:

A Clover machine is defined as a single physical computer that Clover is used on to instrument source code or generate coverage reports [...]

See also:

https://answers.atlassian.com/questions/52021/usage-of-10-machine-clover-license-via-25-remote-agents-of-bamboo

[Marek Parfianowicz]

It's not mandatory to include sources in the HTML report. Lack of sources does not decrease code coverage either.

You can use <clover-report> with srcLevel=false to skip source page rendering:

<clover-report>
  <current outfile="clover_html">
    <format type="html" srcLevel="false"/>
  </current>
</clover-report>

You can also use <clover-report> with <sourcepath> option to point to a different location of sources:

<clover-report>
  <current outfile="clover_html">
    <format type="html"/>
    <sourcepath>
      <pathelement path="/some/other/location"/>
    </sourcepath>
  </current>
</clover-report>

[Abhinav Sahu]

Yes part of the problem solved, thanks.