def export(self, destination):
"""
.. todo:: Break into several smaller methods.
"""
records = TimingsRecorder(self.env).loadExistingRecords()
#
# http://openbook.galileocomputing.de/python/python_kapitel_19_005.htm
# has some helpful examples how to work with the csv module.
#
with open(destination, 'w') as fh:
# Create a list 'operations' containing all tests from all categories.
# Each test is prefixed with its category to avoid name clashes.
# E.g. the following two (stripped) timing records
# 'catA': {'test1', 'testfoo'}
# 'catB': {'test1', 'gui'}
# are transformed into the list
# ['catA::test1', 'catA::testfoo', 'catB::test1', 'catB::gui']
operations = []
categories = records.keys()
for cat in categories:
for time in records[cat]:
ops = records[cat][time].keys()
operations += map(lambda op: self._format(cat, op), ops)
operations.sort()
# Create a list of csv columns by adding 'Timestamp' as the first column
# name to the list of operations and initiate the csv writer.
columns = ['Timestamp'] + operations
writer = csv.DictWriter(fh, columns, dialect=csv.excel)
# Assemble the header row, i.e. create a data row such as this:
# {'Timestamp': 'Timestamp', 'catA::opA': 'catA::opA', 'catB::opB': 'catB::opB'}
header = {}
for key in columns:
header[key] = key
writer.writerow(header)
# Collect all timestamps present in any of the records.
timestamps = set()
for cat in categories:
timestamps = timestamps.union(records[cat].keys())
# Create an ordered list of collected timestamps.
timestamps = list(timestamps)
timestamps.sort()
# Assemble the csv rows that contain the elapsed times per operation.
# Each row corresponds to a single timestamp. Operations that have not
# been timed at a given timestamp are represented as empty cells in the
# final csv file.
rows = []
for time in timestamps:
row = {'Timestamp': time}
# Iterate over categories and the operations within.
for cat in categories:
for op in records[cat]:
# Ensure that the current category contains the current timestamp.
if time in records[cat]:
record = records[cat][time]
# Create a cell per operation.
for op in record:
column = self._format(cat, op)
row[column] = record[op]
rows.append(row)
# Write the data to the csv file.
writer.writerows(rows)
self.note('Exported %s csv rows to %s' % (len(rows), destination))
def run(self):
if len(self.env.data['timings']['timings']) > 0:
worker = TimingsRecorder(self.env)
worker.add(self.env.data['timings'])
def run(self):
if len(self.env.data['timings']['timings']) > 0:
worker = TimingsRecorder(self.env)
worker.add(self.env.data['timings'])