Result Parser¶

A result parser is a function that receives the parameters:

• root The root path of the run
• runspec An instance of Runspec, containing various information about the run as defined in the runscript.
• instance An instance of Instance, containing the information regarding the benchmark instance used for this run, such as location and instance name.

and returns a list of triples (key, type, value) for later use by the summarizer.

Example¶

A full example can be seen in benchmarks/examples/**/resultparser.py.

def myresultparser(root, runspec, instance):
    # ...
    return [
        ('wall', 'float', 1.0242),
        ('mem', 'float', 82519554),
    ]