Tips for PureCoverage UNIX Users

Related Topics

You can use PureCoverage for Windows to view data that was collected using PureCoverage for UNIX. For information, click images\shortcut.gif

If you are accustomed to using PureCoverage for UNIX, you will notice that PureCoverage for Windows is substantially different. The most important differences are listed here:

§          PureCoverage for Windows instruments complete Visual C++ executables, not individual object files at link time.

§          You use the PureCoverage user interface to perform most operations. (You can also specify PureCoverage commands in a script or from the command line.)

§          You instrument and run a Visual C++ executable in a single step. For more information, click images\shortcut.gif

§          You display the results of each run individually, using a series of data coverage windows: the Run Summary window, the Coverage Browser window, the Function List window, and the Annotated Source window. You can run more than one program at a time and display the results in these windows.

§          You can use the Navigator to view the run status of programs. For more information, click images\shortcut.gif

§          PureCoverage uses a pure_api.c file, instead of purecov_stubs.a, to call PureCoverage API functions from Visual C++ programs. You also add #include "pure.h" to any source file that uses API functions, rather than #include "purecov.h".

§          PureCoverage saves data in a PureCoverage data file (.cfy), instead of a .pcv file. For more information, click images\shortcut.gif

§          You can compare differences between runs directly in PureCoverage, using the images\diffic.gif tool button. You don't need to use the pc_diff script.

§          PureCoverage does not support any report scripts. However, you can save coverage data to a PureCoverage data file (.cfy), to use for further analysis or to share with other PureCoverage users. You can also save data to a tab-delimited ASCII text file (.txt) to use outside of PureCoverage, for example, in test scripts or in Microsoft Excel. You can also copy data from the Coverage Browser window to use in a text editor or from the Function List window to perform further analysis in Excel.

§          By default, PureCoverage assigns names to instrumented Visual C++ programs and saves them in a cache directory. However, you can still choose to assign an alternative name to the instrumented program and store it wherever you want. For more information, click images\shortcut.gif

§          PureCoverage does not save cumulative coverage data into a .pcv file, like the UNIX version. Instead, PureCoverage creates an Auto Merge run, by default, when you run a program, and merges the data from subsequent runs of the same program in the current session into this Auto Merge. You can also merge program runs manually. You can use the user interface to save these merge runs into PureCoverage data files (.cfy), or use the command line option /SaveMergeData.

§          PureCoverage does not use environment variables such as PURECOVOPTIONS to control preferences and settings at build time and run time.To change a run-time option, use the PureCoverage dialog boxes or command-line options.

§          PureCoverage does not provide adjustments to mark source code that is difficult or impossible to reach, or to keep it from being displayed as untested and counted as unused code.

§          PureCoverage does not support a -merge command-line option. If you want to merge PureCoverage data files (.cfy), open the saved .cfy files in PureCoverage, merge them manually, and save them as a new .cfy file or an ASCII text file (.txt).

(C) Copyright IBM Corporation 1993, 2010.