# USAGE
#
# To run the benchmarks on a certain branch of Zeitgeist make sure you are
# running an instance of the branch and that it includes the benchmark
# extension.
#
# To benchmark and output timing to 'output.json' run:
#  ./query_timings.py -o output.json
#
# If you already have data in the output file and want to merge both
# data sets (build avg.) run:
#  ./query_timings.py -m -o output.json
#
# To plot the data use the '--plot' argument. Multiple '--plot' arguments
# will define multiple series.
#
# When plotting one needs to choose between several types of plots using
# --type with the following arguments: marsh_time, get_events_time,
# find_ids_time, find_events or overall.
#
# In short, a run always looks like:
#  tools/development/query_timings.py --name "lp:zeitgeist" -o trunk.json \
#       --queries tools/development/query_sets/timerange_always.txt
#  tools/development/query_timings.py --name "lp:some-branch" -o somebranch.json \
#       --queries tools/development/query_sets/timerange_always.txt
#  tools/development/query_timings.py --plot trunk.json --type overall \
#       -o benchmark.svg
