1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
|
#!/usr/bin/env python3
import argparse
import ast
import sys
import os
from time import time
import _peg_parser
try:
import memory_profiler
except ModuleNotFoundError:
print("Please run `make venv` to create a virtual environment and install"
" all the dependencies, before running this script.")
sys.exit(1)
sys.path.insert(0, os.getcwd())
from scripts.test_parse_directory import parse_directory
argparser = argparse.ArgumentParser(
prog="benchmark", description="Reproduce the various pegen benchmarks"
)
argparser.add_argument(
"--parser",
action="store",
choices=["new", "old"],
default="pegen",
help="Which parser to benchmark (default is pegen)",
)
argparser.add_argument(
"--target",
action="store",
choices=["xxl", "stdlib"],
default="xxl",
help="Which target to use for the benchmark (default is xxl.py)",
)
subcommands = argparser.add_subparsers(title="Benchmarks", dest="subcommand")
command_compile = subcommands.add_parser(
"compile", help="Benchmark parsing and compiling to bytecode"
)
command_parse = subcommands.add_parser(
"parse", help="Benchmark parsing and generating an ast.AST"
)
command_notree = subcommands.add_parser(
"notree", help="Benchmark parsing and dumping the tree"
)
def benchmark(func):
def wrapper(*args):
times = list()
for _ in range(3):
start = time()
result = func(*args)
end = time()
times.append(end - start)
memory = memory_profiler.memory_usage((func, args))
print(f"{func.__name__}")
print(f"\tTime: {sum(times)/3:.3f} seconds on an average of 3 runs")
print(f"\tMemory: {max(memory)} MiB on an average of 3 runs")
return result
return wrapper
@benchmark
def time_compile(source, parser):
if parser == "old":
return _peg_parser.compile_string(
source,
oldparser=True,
)
else:
return _peg_parser.compile_string(source)
@benchmark
def time_parse(source, parser):
if parser == "old":
return _peg_parser.parse_string(source, oldparser=True)
else:
return _peg_parser.parse_string(source)
@benchmark
def time_notree(source, parser):
if parser == "old":
return _peg_parser.parse_string(source, oldparser=True, ast=False)
else:
return _peg_parser.parse_string(source, ast=False)
def run_benchmark_xxl(subcommand, parser, source):
if subcommand == "compile":
time_compile(source, parser)
elif subcommand == "parse":
time_parse(source, parser)
elif subcommand == "notree":
time_notree(source, parser)
def run_benchmark_stdlib(subcommand, parser):
modes = {"compile": 2, "parse": 1, "notree": 0}
for _ in range(3):
parse_directory(
"../../Lib",
verbose=False,
excluded_files=["*/bad*", "*/lib2to3/tests/data/*",],
tree_arg=0,
short=True,
mode=modes[subcommand],
oldparser=(parser == "old"),
)
def main():
args = argparser.parse_args()
subcommand = args.subcommand
parser = args.parser
target = args.target
if subcommand is None:
argparser.error("A benchmark to run is required")
if target == "xxl":
with open(os.path.join("data", "xxl.py"), "r") as f:
source = f.read()
run_benchmark_xxl(subcommand, parser, source)
elif target == "stdlib":
run_benchmark_stdlib(subcommand, parser)
if __name__ == "__main__":
main()
|
