"""Class for printing reports on profiled python code."""
# Written by James Roskind
# Based on prior profile module by Sjoerd Mullender...
# which was hacked somewhat by: Guido van Rossum
# Copyright Disney Enterprises, Inc. All Rights Reserved.
# Licensed to PSF under a Contributor Agreement
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
# http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND,
# either express or implied. See the License for the specific language
# governing permissions and limitations under the License.
from functools import cmp_to_key
from dataclasses import dataclass
__all__ = ["Stats", "SortKey", "FunctionProfile", "StatsProfile"]
class SortKey(str, Enum):
CALLS = 'calls', 'ncalls'
CUMULATIVE = 'cumulative', 'cumtime'
FILENAME = 'filename', 'module'
def __new__(cls, *values):
obj = str.__new__(cls, value)
for other_value in values[1:]:
cls._value2member_map_[other_value] = obj
@dataclass(unsafe_hash=True)
@dataclass(unsafe_hash=True)
'''Class for keeping track of an item in inventory.'''
func_profiles: Dict[str, FunctionProfile]
"""This class is used for creating reports from data generated by the
Profile class. It is a "friend" of that class, and imports data either
by direct access to members of Profile class, or by reading in a dictionary
that was emitted (via marshal) from the Profile class.
The big change from the previous Profiler (in terms of raw functionality)
is that an "add()" method has been provided to combine Stats from
several distinct profile runs. Both the constructor and the add()
method now take arbitrarily many file names as arguments.
All the print methods now take an argument that indicates how many lines
to print. If the arg is a floating point number between 0 and 1.0, then
it is taken as a decimal percentage of the available lines to be printed
(e.g., .1 means print 10% of all available lines). If it is an integer,
it is taken to mean the number of lines of data that you wish to have
The sort_stats() method now processes some additional options (i.e., in
addition to the old -1, 0, 1, or 2 that are respectively interpreted as
'stdname', 'calls', 'time', and 'cumulative'). It takes either an
arbitrary number of quoted strings or SortKey enum to select the sort
For example sort_stats('time', 'name') or sort_stats(SortKey.TIME,
SortKey.NAME) sorts on the major key of 'internal function time', and on
the minor key of 'the name of the function'. Look at the two tables in
sort_stats() and get_sort_arg_defs(self) for more examples.
All methods return self, so you can string together commands like:
Stats('foo', 'goo').strip_dirs().sort_stats('calls').\
print_stats(5).print_callers(5)
def __init__(self, *args, stream=None):
self.stream = stream or sys.stdout
self.all_callees = None # calc only if needed
self.get_top_level_stats()
print("Invalid timing data %s" %
(self.files[-1] if self.files else ''), file=self.stream)
def load_stats(self, arg):
elif isinstance(arg, str):
with open(arg, 'rb') as f:
self.stats = marshal.load(f)
file_stats = os.stat(arg)
arg = time.ctime(file_stats.st_mtime) + " " + arg
except: # in case this is not unix
elif hasattr(arg, 'create_stats'):
raise TypeError("Cannot create or construct a %r object from %r"
def get_top_level_stats(self):
for func, (cc, nc, tt, ct, callers) in self.stats.items():
if ("jprofile", 0, "profiler") in callers:
if len(func_std_string(func)) > self.max_name_len:
self.max_name_len = len(func_std_string(func))
def add(self, *arg_list):
for item in reversed(arg_list):
if type(self) != type(item):
self.total_calls += item.total_calls
self.prim_calls += item.prim_calls
self.total_tt += item.total_tt
for func in item.top_level:
if self.max_name_len < item.max_name_len:
self.max_name_len = item.max_name_len
for func, stat in item.stats.items():
old_func_stat = self.stats[func]
old_func_stat = (0, 0, 0, 0, {},)
self.stats[func] = add_func_stats(old_func_stat, stat)
def dump_stats(self, filename):
"""Write the profile data to a file we know how to load back."""
with open(filename, 'wb') as f:
marshal.dump(self.stats, f)
# list the tuple indices and directions for sorting,
# along with some printable description
sort_arg_dict_default = {
"calls" : (((1,-1), ), "call count"),
"ncalls" : (((1,-1), ), "call count"),
"cumtime" : (((3,-1), ), "cumulative time"),
"cumulative": (((3,-1), ), "cumulative time"),
"filename" : (((4, 1), ), "file name"),
"line" : (((5, 1), ), "line number"),
"module" : (((4, 1), ), "file name"),
"name" : (((6, 1), ), "function name"),
"nfl" : (((6, 1),(4, 1),(5, 1),), "name/file/line"),
"pcalls" : (((0,-1), ), "primitive call count"),
"stdname" : (((7, 1), ), "standard name"),
"time" : (((2,-1), ), "internal time"),
"tottime" : (((2,-1), ), "internal time"),
def get_sort_arg_defs(self):
"""Expand all abbreviations that are unique."""
if not self.sort_arg_dict:
self.sort_arg_dict = dict = {}
for word, tup in self.sort_arg_dict_default.items():
return self.sort_arg_dict
def sort_stats(self, *field):
if len(field) == 1 and isinstance(field[0], int):
# Be compatible with old profiler
field = [ {-1: "stdname",
2: "cumulative"}[field[0]] ]
if type(arg) != type(field[0]):
raise TypeError("Can't have mixed argument type")
sort_arg_defs = self.get_sort_arg_defs()
if isinstance(word, SortKey):
sort_tuple = sort_tuple + sort_arg_defs[word][0]
self.sort_type += connector + sort_arg_defs[word][1]
for func, (cc, nc, tt, ct, callers) in self.stats.items():
stats_list.append((cc, nc, tt, ct) + func +
(func_std_string(func), func))
stats_list.sort(key=cmp_to_key(TupleComp(sort_tuple).compare))
self.fcn_list = fcn_list = []
fcn_list.append(tuple[-1])
self.stats = newstats = {}
for func, (cc, nc, tt, ct, callers) in oldstats.items():
newfunc = func_strip_path(func)
if len(func_std_string(newfunc)) > max_name_len:
max_name_len = len(func_std_string(newfunc))
for func2, caller in callers.items():
newcallers[func_strip_path(func2)] = caller
newstats[newfunc] = add_func_stats(
(cc, nc, tt, ct, newcallers))
newstats[newfunc] = (cc, nc, tt, ct, newcallers)
self.top_level = new_top = set()
new_top.add(func_strip_path(func))
self.max_name_len = max_name_len
self.all_callees = all_callees = {}
for func, (cc, nc, tt, ct, callers) in self.stats.items():
if not func in all_callees:
for func2, caller in callers.items():
if not func2 in all_callees:
all_callees[func2][func] = caller
#******************************************************************
# The following functions support actual printing of reports
#******************************************************************
# Optional "amount" is either a line count, or a percentage of lines.
def eval_print_amount(self, sel, list, msg):
msg += " <Invalid regular expression %r>\n" % sel
if rex.search(func_std_string(func)):
if isinstance(sel, float) and 0.0 <= sel < 1.0:
count = int(count * sel + .5)
elif isinstance(sel, int) and 0 <= sel < count:
if len(list) != len(new_list):
msg += " List reduced from %r to %r due to restriction <%r>\n" % (
len(list), len(new_list), sel)
def get_stats_profile(self):
"""This method returns an instance of StatsProfile, which contains a mapping
of function names to instances of FunctionProfile. Each FunctionProfile
instance holds information related to the function's profile such as how
long the function took to run, how many times it was called, etc...
func_list = self.fcn_list[:] if self.fcn_list else list(self.stats.keys())
return StatsProfile(0, {})
total_tt = float(f8(self.total_tt))
stats_profile = StatsProfile(total_tt, func_profiles)
cc, nc, tt, ct, callers = self.stats[func]
file_name, line_number, func_name = func
ncalls = str(nc) if nc == cc else (str(nc) + '/' + str(cc))
percall_tottime = -1 if nc == 0 else float(f8(tt/nc))
percall_cumtime = -1 if cc == 0 else float(f8(ct/cc))
func_profile = FunctionProfile(
tottime, # time spent in this function alone
cumtime, # time spent in the function plus all functions that this function called,
func_profiles[func_name] = func_profile
def get_print_list(self, sel_list):
width = self.max_name_len
stat_list = self.fcn_list[:]
msg = " Ordered by: " + self.sort_type + '\n'
stat_list = list(self.stats.keys())
msg = " Random listing order was used\n"
for selection in sel_list:
stat_list, msg = self.eval_print_amount(selection, stat_list, msg)
print(msg, file=self.stream)
if count < len(self.stats):
if len(func_std_string(func)) > width:
width = len(func_std_string(func))
return width+2, stat_list
def print_stats(self, *amount):
for filename in self.files:
print(filename, file=self.stream)
for func in self.top_level:
print(indent, func_get_function_name(func), file=self.stream)
print(indent, self.total_calls, "function calls", end=' ', file=self.stream)
if self.total_calls != self.prim_calls:
print("(%d primitive calls)" % self.prim_calls, end=' ', file=self.stream)
print("in %.3f seconds" % self.total_tt, file=self.stream)
width, list = self.get_print_list(amount)
def print_callees(self, *amount):
width, list = self.get_print_list(amount)
self.print_call_heading(width, "called...")
if func in self.all_callees:
self.print_call_line(width, func, self.all_callees[func])
self.print_call_line(width, func, {})
def print_callers(self, *amount):
width, list = self.get_print_list(amount)
self.print_call_heading(width, "was called by...")
cc, nc, tt, ct, callers = self.stats[func]
self.print_call_line(width, func, callers, "<-")
def print_call_heading(self, name_size, column_title):
print("Function ".ljust(name_size) + column_title, file=self.stream)
# print sub-header only if we have new-style callers
for cc, nc, tt, ct, callers in self.stats.values():
value = next(iter(callers.values()))
subheader = isinstance(value, tuple)
print(" "*name_size + " ncalls tottime cumtime", file=self.stream)
def print_call_line(self, name_size, source, call_dict, arrow="->"):
print(func_std_string(source).ljust(name_size) + arrow, end=' ', file=self.stream)
clist = sorted(call_dict.keys())
name = func_std_string(func)
if isinstance(value, tuple):
substats = '%d/%d' % (nc, cc)
substats = '%s %s %s %s' % (substats.rjust(7+2*len(indent)),
left_width = name_size + 1
substats = '%s(%r) %s' % (name, value, f8(self.stats[func][3]))
left_width = name_size + 3
print(indent*left_width + substats, file=self.stream)
print(' ncalls tottime percall cumtime percall', end=' ', file=self.stream)
It is recommended that you Edit text format, this type of Fix handles quite a lot in one request
