Can't use symlinks on Windows -- replace with real files (#159)

The regex_compile benchmark uses two symlinks, but those don't work reliably on Windows. Replace them with real files.

See faster-cpython/ideas#333

Author: gvanrossum

pyperformance/data-files/benchmarks/bm_regex_compile/bm_regex_effbot.py

@@ -0,0 +1,175 @@
+
+"""Benchmarks for Python's regex engine.
+
+These are some of the original benchmarks used to tune Python's regex engine
+in 2000 written by Fredrik Lundh. Retreived from
+http://mail.python.org/pipermail/python-dev/2000-August/007797.html and
+integrated into Unladen Swallow's pyperf.py in 2009 by David Laing.
+
+These benchmarks are of interest since they helped to guide the original
+optimization of the sre engine, and we shouldn't necessarily ignore them just
+because they're "old".
+"""
+
+# Python imports
+import re
+
+# Local imports
+import pyperf
+
+USE_BYTES = False
+
+
+def re_compile(s):
+    if USE_BYTES:
+        return re.compile(s.encode('latin1'))
+    else:
+        return re.compile(s)
+
+# These are the regular expressions to be tested. These sync up,
+# index-for-index with the list of strings generated by gen_string_table()
+# below.
+
+
+def gen_regex_table():
+    return [
+        re_compile('Python|Perl'),
+        re_compile('Python|Perl'),
+        re_compile('(Python|Perl)'),
+        re_compile('(?:Python|Perl)'),
+        re_compile('Python'),
+        re_compile('Python'),
+        re_compile('.*Python'),
+        re_compile('.*Python.*'),
+        re_compile('.*(Python)'),
+        re_compile('.*(?:Python)'),
+        re_compile('Python|Perl|Tcl'),
+        re_compile('Python|Perl|Tcl'),
+        re_compile('(Python|Perl|Tcl)'),
+        re_compile('(?:Python|Perl|Tcl)'),
+        re_compile('(Python)\\1'),
+        re_compile('(Python)\\1'),
+        re_compile('([0a-z][a-z0-9]*,)+'),
+        re_compile('(?:[0a-z][a-z0-9]*,)+'),
+        re_compile('([a-z][a-z0-9]*,)+'),
+        re_compile('(?:[a-z][a-z0-9]*,)+'),
+        re_compile('.*P.*y.*t.*h.*o.*n.*')]
+
+
+def gen_string_table(n):
+    """Generates the list of strings that will be used in the benchmarks.
+
+    All strings have repeated prefixes and suffices, and n specifies the
+    number of repetitions.
+    """
+    strings = []
+
+    def append(s):
+        if USE_BYTES:
+            strings.append(s.encode('latin1'))
+        else:
+            strings.append(s)
+    append('-' * n + 'Perl' + '-' * n)
+    append('P' * n + 'Perl' + 'P' * n)
+    append('-' * n + 'Perl' + '-' * n)
+    append('-' * n + 'Perl' + '-' * n)
+    append('-' * n + 'Python' + '-' * n)
+    append('P' * n + 'Python' + 'P' * n)
+    append('-' * n + 'Python' + '-' * n)
+    append('-' * n + 'Python' + '-' * n)
+    append('-' * n + 'Python' + '-' * n)
+    append('-' * n + 'Python' + '-' * n)
+    append('-' * n + 'Perl' + '-' * n)
+    append('P' * n + 'Perl' + 'P' * n)
+    append('-' * n + 'Perl' + '-' * n)
+    append('-' * n + 'Perl' + '-' * n)
+    append('-' * n + 'PythonPython' + '-' * n)
+    append('P' * n + 'PythonPython' + 'P' * n)
+    append('-' * n + 'a5,b7,c9,' + '-' * n)
+    append('-' * n + 'a5,b7,c9,' + '-' * n)
+    append('-' * n + 'a5,b7,c9,' + '-' * n)
+    append('-' * n + 'a5,b7,c9,' + '-' * n)
+    append('-' * n + 'Python' + '-' * n)
+    return strings
+
+
+def init_benchmarks(n_values=None):
+    """Initialize the strings we'll run the regexes against.
+
+    The strings used in the benchmark are prefixed and suffixed by
+    strings that are repeated n times.
+
+    The sequence n_values contains the values for n.
+    If n_values is None the values of n from the original benchmark
+    are used.
+
+    The generated list of strings is cached in the string_tables
+    variable, which is indexed by n.
+
+    Returns:
+    A list of string prefix/suffix lengths.
+    """
+
+    if n_values is None:
+        n_values = (0, 5, 50, 250, 1000, 5000, 10000)
+
+    string_tables = {n: gen_string_table(n) for n in n_values}
+    regexs = gen_regex_table()
+
+    data = []
+    for n in n_values:
+        for id in range(len(regexs)):
+            regex = regexs[id]
+            string = string_tables[n][id]
+            data.append((regex, string))
+    return data
+
+
+def bench_regex_effbot(loops):
+    if bench_regex_effbot.data is None:
+        bench_regex_effbot.data = init_benchmarks()
+    data = bench_regex_effbot.data
+
+    range_it = range(loops)
+    search = re.search
+    t0 = pyperf.perf_counter()
+
+    for _ in range_it:
+        # Runs all of the benchmarks for a given value of n.
+        for regex, string in data:
+            # search 10 times
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+            search(regex, string)
+
+    return pyperf.perf_counter() - t0
+
+
+# cached data, generated at the first call
+bench_regex_effbot.data = None
+
+
+def add_cmdline_args(cmd, args):
+    if args.force_bytes:
+        cmd.append("--force_bytes")
+
+
+if __name__ == '__main__':
+    runner = pyperf.Runner(add_cmdline_args=add_cmdline_args)
+    runner.metadata['description'] = ("Test the performance of regexps "
+                                      "using Fredik Lundh's benchmarks.")
+    runner.argparser.add_argument("-B", "--force_bytes", action="store_true",
+                                  help="test bytes regexps")
+    options = runner.parse_args()
+    if options.force_bytes:
+        USE_BYTES = True
+
+    runner.bench_time_func('regex_effbot', bench_regex_effbot,
+                           inner_loops=10)