Options

The options provided by the RiveScript Coverage Plugin are designed to help troubleshoot the plugin itself and should not be needed by normal users. For completeness, they are documented here. The options go in your .coveragerc file and are specified as follows:

[rivescript_coverage_plugin]
show_startup = True
show_parsing = True
show_tracing = True
clean_rs_objects = False
capture_streams = False

The default values are the opposite of what I show above. The options are defined as follows:

show_startup

Show information about the plugin’s startup sequence is shown, including version information and command line arguments. This is False by default.

show_parsing

Show information about which lines of the RiveScript files are executable and which are not. This is False by default.

show_tracing

Generate a complete trace of the execution of the RiveScript interpreter, including which lines are being marked as executed as we interpret the Debug output of the RiveScript interpreter. This is False by default.

clean_rs_objects

Preserve the _rs_objects_ temporary directory, where the RiveScript Coverage Plug creates files representing each < object NAME python in the RiveScript in order to trace it’s execution. This directory will contain a rs_obj_NAME.py file for each object named NAME in your RiveScript. This is True by default, which means this directory will be removed after analysis.

capture_streams

Capture coverage when RiveScript is dynamically created using the rs.stream method. In order to preserve these streams of RiveScript, a directory _rs_streams_ is created and each stream is identified in a file named s-SOURCEFILE1_LINENO1-SOURCEFILE2_LINENO2.rive. Here, SOURCEFILE1 identifies the python source file containing the direct caller of rs.stream, and SOURCEFILE2 identifies a parent caller that is not located in the same source file. The LINENOs identify the line numbers where the call(s) occur. If multiple calls occur at the same place in the source code with different RiveScript strings being passed, then a 3-digit occurence count will be appended to the filename. This directory of streams is left in place after coverage analysis exits, so that when you run coverage html, it will be able to annotate the files with coverage information. If this is specified as False, then no coverage is captured for streams. (Since v1.1.0)