• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1# -*- coding: utf-8 -*-
2# This file should be kept compatible with both Python 2.6 and Python >= 3.0.
3
4import itertools
5import os
6import platform
7import re
8import sys
9import time
10from optparse import OptionParser
11
12out = sys.stdout
13
14TEXT_ENCODING = 'utf8'
15NEWLINES = 'lf'
16
17# Compatibility
18try:
19    xrange
20except NameError:
21    xrange = range
22
23def text_open(fn, mode, encoding=None):
24    try:
25        return open(fn, mode, encoding=encoding or TEXT_ENCODING)
26    except TypeError:
27        if 'r' in mode:
28            mode += 'U' # 'U' mode is needed only in Python 2.x
29        return open(fn, mode)
30
31def get_file_sizes():
32    for s in ['20 KiB', '400 KiB', '10 MiB']:
33        size, unit = s.split()
34        size = int(size) * {'KiB': 1024, 'MiB': 1024 ** 2}[unit]
35        yield s.replace(' ', ''), size
36
37def get_binary_files():
38    return ((name + ".bin", size) for name, size in get_file_sizes())
39
40def get_text_files():
41    return (("%s-%s-%s.txt" % (name, TEXT_ENCODING, NEWLINES), size)
42        for name, size in get_file_sizes())
43
44def with_open_mode(mode):
45    def decorate(f):
46        f.file_open_mode = mode
47        return f
48    return decorate
49
50def with_sizes(*sizes):
51    def decorate(f):
52        f.file_sizes = sizes
53        return f
54    return decorate
55
56
57# Here begin the tests
58
59@with_open_mode("r")
60@with_sizes("medium")
61def read_bytewise(f):
62    """ read one unit at a time """
63    f.seek(0)
64    while f.read(1):
65        pass
66
67@with_open_mode("r")
68@with_sizes("medium")
69def read_small_chunks(f):
70    """ read 20 units at a time """
71    f.seek(0)
72    while f.read(20):
73        pass
74
75@with_open_mode("r")
76@with_sizes("medium")
77def read_big_chunks(f):
78    """ read 4096 units at a time """
79    f.seek(0)
80    while f.read(4096):
81        pass
82
83@with_open_mode("r")
84@with_sizes("small", "medium", "large")
85def read_whole_file(f):
86    """ read whole contents at once """
87    f.seek(0)
88    while f.read():
89        pass
90
91@with_open_mode("rt")
92@with_sizes("medium")
93def read_lines(f):
94    """ read one line at a time """
95    f.seek(0)
96    for line in f:
97        pass
98
99@with_open_mode("r")
100@with_sizes("medium")
101def seek_forward_bytewise(f):
102    """ seek forward one unit at a time """
103    f.seek(0, 2)
104    size = f.tell()
105    f.seek(0, 0)
106    for i in xrange(0, size - 1):
107        f.seek(i, 0)
108
109@with_open_mode("r")
110@with_sizes("medium")
111def seek_forward_blockwise(f):
112    """ seek forward 1000 units at a time """
113    f.seek(0, 2)
114    size = f.tell()
115    f.seek(0, 0)
116    for i in xrange(0, size - 1, 1000):
117        f.seek(i, 0)
118
119@with_open_mode("rb")
120@with_sizes("medium")
121def read_seek_bytewise(f):
122    """ alternate read & seek one unit """
123    f.seek(0)
124    while f.read(1):
125        f.seek(1, 1)
126
127@with_open_mode("rb")
128@with_sizes("medium")
129def read_seek_blockwise(f):
130    """ alternate read & seek 1000 units """
131    f.seek(0)
132    while f.read(1000):
133        f.seek(1000, 1)
134
135
136@with_open_mode("w")
137@with_sizes("small")
138def write_bytewise(f, source):
139    """ write one unit at a time """
140    for i in xrange(0, len(source)):
141        f.write(source[i:i+1])
142
143@with_open_mode("w")
144@with_sizes("medium")
145def write_small_chunks(f, source):
146    """ write 20 units at a time """
147    for i in xrange(0, len(source), 20):
148        f.write(source[i:i+20])
149
150@with_open_mode("w")
151@with_sizes("medium")
152def write_medium_chunks(f, source):
153    """ write 4096 units at a time """
154    for i in xrange(0, len(source), 4096):
155        f.write(source[i:i+4096])
156
157@with_open_mode("w")
158@with_sizes("large")
159def write_large_chunks(f, source):
160    """ write 1e6 units at a time """
161    for i in xrange(0, len(source), 1000000):
162        f.write(source[i:i+1000000])
163
164
165@with_open_mode("w+")
166@with_sizes("small")
167def modify_bytewise(f, source):
168    """ modify one unit at a time """
169    f.seek(0)
170    for i in xrange(0, len(source)):
171        f.write(source[i:i+1])
172
173@with_open_mode("w+")
174@with_sizes("medium")
175def modify_small_chunks(f, source):
176    """ modify 20 units at a time """
177    f.seek(0)
178    for i in xrange(0, len(source), 20):
179        f.write(source[i:i+20])
180
181@with_open_mode("w+")
182@with_sizes("medium")
183def modify_medium_chunks(f, source):
184    """ modify 4096 units at a time """
185    f.seek(0)
186    for i in xrange(0, len(source), 4096):
187        f.write(source[i:i+4096])
188
189@with_open_mode("wb+")
190@with_sizes("medium")
191def modify_seek_forward_bytewise(f, source):
192    """ alternate write & seek one unit """
193    f.seek(0)
194    for i in xrange(0, len(source), 2):
195        f.write(source[i:i+1])
196        f.seek(i+2)
197
198@with_open_mode("wb+")
199@with_sizes("medium")
200def modify_seek_forward_blockwise(f, source):
201    """ alternate write & seek 1000 units """
202    f.seek(0)
203    for i in xrange(0, len(source), 2000):
204        f.write(source[i:i+1000])
205        f.seek(i+2000)
206
207# XXX the 2 following tests don't work with py3k's text IO
208@with_open_mode("wb+")
209@with_sizes("medium")
210def read_modify_bytewise(f, source):
211    """ alternate read & write one unit """
212    f.seek(0)
213    for i in xrange(0, len(source), 2):
214        f.read(1)
215        f.write(source[i+1:i+2])
216
217@with_open_mode("wb+")
218@with_sizes("medium")
219def read_modify_blockwise(f, source):
220    """ alternate read & write 1000 units """
221    f.seek(0)
222    for i in xrange(0, len(source), 2000):
223        f.read(1000)
224        f.write(source[i+1000:i+2000])
225
226
227read_tests = [
228    read_bytewise, read_small_chunks, read_lines, read_big_chunks,
229    None, read_whole_file, None,
230    seek_forward_bytewise, seek_forward_blockwise,
231    read_seek_bytewise, read_seek_blockwise,
232]
233
234write_tests = [
235    write_bytewise, write_small_chunks, write_medium_chunks, write_large_chunks,
236]
237
238modify_tests = [
239    modify_bytewise, modify_small_chunks, modify_medium_chunks,
240    None,
241    modify_seek_forward_bytewise, modify_seek_forward_blockwise,
242    read_modify_bytewise, read_modify_blockwise,
243]
244
245def run_during(duration, func):
246    _t = time.time
247    n = 0
248    start = os.times()
249    start_timestamp = _t()
250    real_start = start[4] or start_timestamp
251    while True:
252        func()
253        n += 1
254        if _t() - start_timestamp > duration:
255            break
256    end = os.times()
257    real = (end[4] if start[4] else time.time()) - real_start
258    return n, real, sum(end[0:2]) - sum(start[0:2])
259
260def warm_cache(filename):
261    with open(filename, "rb") as f:
262        f.read()
263
264
265def run_all_tests(options):
266    def print_label(filename, func):
267        name = re.split(r'[-.]', filename)[0]
268        out.write(
269            ("[%s] %s... "
270                % (name.center(7), func.__doc__.strip())
271            ).ljust(52))
272        out.flush()
273
274    def print_results(size, n, real, cpu):
275        bw = n * float(size) / 1024 ** 2 / real
276        bw = ("%4d MiB/s" if bw > 100 else "%.3g MiB/s") % bw
277        out.write(bw.rjust(12) + "\n")
278        if cpu < 0.90 * real:
279            out.write("   warning: test above used only %d%% CPU, "
280                "result may be flawed!\n" % (100.0 * cpu / real))
281
282    def run_one_test(name, size, open_func, test_func, *args):
283        mode = test_func.file_open_mode
284        print_label(name, test_func)
285        if "w" not in mode or "+" in mode:
286            warm_cache(name)
287        with open_func(name) as f:
288            n, real, cpu = run_during(1.5, lambda: test_func(f, *args))
289        print_results(size, n, real, cpu)
290
291    def run_test_family(tests, mode_filter, files, open_func, *make_args):
292        for test_func in tests:
293            if test_func is None:
294                out.write("\n")
295                continue
296            if mode_filter in test_func.file_open_mode:
297                continue
298            for s in test_func.file_sizes:
299                name, size = files[size_names[s]]
300                #name += file_ext
301                args = tuple(f(name, size) for f in make_args)
302                run_one_test(name, size,
303                    open_func, test_func, *args)
304
305    size_names = {
306        "small": 0,
307        "medium": 1,
308        "large": 2,
309    }
310
311    print("Python %s" % sys.version)
312    if sys.version_info < (3, 3):
313        if sys.maxunicode > 0xffff:
314            text = "UCS-4 (wide build)"
315        else:
316            text = "UTF-16 (narrow build)"
317    else:
318        text = "PEP 393"
319    print("Unicode: %s" % text)
320    print(platform.platform())
321    binary_files = list(get_binary_files())
322    text_files = list(get_text_files())
323    if "b" in options:
324        print("Binary unit = one byte")
325    if "t" in options:
326        print("Text unit = one character (%s-decoded)" % TEXT_ENCODING)
327
328    # Binary reads
329    if "b" in options and "r" in options:
330        print("\n** Binary input **\n")
331        run_test_family(read_tests, "t", binary_files, lambda fn: open(fn, "rb"))
332
333    # Text reads
334    if "t" in options and "r" in options:
335        print("\n** Text input **\n")
336        run_test_family(read_tests, "b", text_files, lambda fn: text_open(fn, "r"))
337
338    # Binary writes
339    if "b" in options and "w" in options:
340        print("\n** Binary append **\n")
341        def make_test_source(name, size):
342            with open(name, "rb") as f:
343                return f.read()
344        run_test_family(write_tests, "t", binary_files,
345            lambda fn: open(os.devnull, "wb"), make_test_source)
346
347    # Text writes
348    if "t" in options and "w" in options:
349        print("\n** Text append **\n")
350        def make_test_source(name, size):
351            with text_open(name, "r") as f:
352                return f.read()
353        run_test_family(write_tests, "b", text_files,
354            lambda fn: text_open(os.devnull, "w"), make_test_source)
355
356    # Binary overwrites
357    if "b" in options and "w" in options:
358        print("\n** Binary overwrite **\n")
359        def make_test_source(name, size):
360            with open(name, "rb") as f:
361                return f.read()
362        run_test_family(modify_tests, "t", binary_files,
363            lambda fn: open(fn, "r+b"), make_test_source)
364
365    # Text overwrites
366    if "t" in options and "w" in options:
367        print("\n** Text overwrite **\n")
368        def make_test_source(name, size):
369            with text_open(name, "r") as f:
370                return f.read()
371        run_test_family(modify_tests, "b", text_files,
372            lambda fn: text_open(fn, "r+"), make_test_source)
373
374
375def prepare_files():
376    print("Preparing files...")
377    # Binary files
378    for name, size in get_binary_files():
379        if os.path.isfile(name) and os.path.getsize(name) == size:
380            continue
381        with open(name, "wb") as f:
382            f.write(os.urandom(size))
383    # Text files
384    chunk = []
385    with text_open(__file__, "r", encoding='utf8') as f:
386        for line in f:
387            if line.startswith("# <iobench text chunk marker>"):
388                break
389        else:
390            raise RuntimeError(
391                "Couldn't find chunk marker in %s !" % __file__)
392        if NEWLINES == "all":
393            it = itertools.cycle(["\n", "\r", "\r\n"])
394        else:
395            it = itertools.repeat(
396                {"cr": "\r", "lf": "\n", "crlf": "\r\n"}[NEWLINES])
397        chunk = "".join(line.replace("\n", next(it)) for line in f)
398        if isinstance(chunk, bytes):
399            chunk = chunk.decode('utf8')
400        chunk = chunk.encode(TEXT_ENCODING)
401    for name, size in get_text_files():
402        if os.path.isfile(name) and os.path.getsize(name) == size:
403            continue
404        head = chunk * (size // len(chunk))
405        tail = chunk[:size % len(chunk)]
406        # Adjust tail to end on a character boundary
407        while True:
408            try:
409                tail.decode(TEXT_ENCODING)
410                break
411            except UnicodeDecodeError:
412                tail = tail[:-1]
413        with open(name, "wb") as f:
414            f.write(head)
415            f.write(tail)
416
417def main():
418    global TEXT_ENCODING, NEWLINES
419
420    usage = "usage: %prog [-h|--help] [options]"
421    parser = OptionParser(usage=usage)
422    parser.add_option("-b", "--binary",
423                      action="store_true", dest="binary", default=False,
424                      help="run binary I/O tests")
425    parser.add_option("-t", "--text",
426                      action="store_true", dest="text", default=False,
427                      help="run text I/O tests")
428    parser.add_option("-r", "--read",
429                      action="store_true", dest="read", default=False,
430                      help="run read tests")
431    parser.add_option("-w", "--write",
432                      action="store_true", dest="write", default=False,
433                      help="run write & modify tests")
434    parser.add_option("-E", "--encoding",
435                      action="store", dest="encoding", default=None,
436                      help="encoding for text tests (default: %s)" % TEXT_ENCODING)
437    parser.add_option("-N", "--newlines",
438                      action="store", dest="newlines", default='lf',
439                      help="line endings for text tests "
440                           "(one of: {lf (default), cr, crlf, all})")
441    parser.add_option("-m", "--io-module",
442                      action="store", dest="io_module", default=None,
443                      help="io module to test (default: builtin open())")
444    options, args = parser.parse_args()
445    if args:
446        parser.error("unexpected arguments")
447    NEWLINES = options.newlines.lower()
448    if NEWLINES not in ('lf', 'cr', 'crlf', 'all'):
449        parser.error("invalid 'newlines' option: %r" % NEWLINES)
450
451    test_options = ""
452    if options.read:
453        test_options += "r"
454    if options.write:
455        test_options += "w"
456    elif not options.read:
457        test_options += "rw"
458    if options.text:
459        test_options += "t"
460    if options.binary:
461        test_options += "b"
462    elif not options.text:
463        test_options += "tb"
464
465    if options.encoding:
466        TEXT_ENCODING = options.encoding
467
468    if options.io_module:
469        globals()['open'] = __import__(options.io_module, {}, {}, ['open']).open
470
471    prepare_files()
472    run_all_tests(test_options)
473
474if __name__ == "__main__":
475    main()
476
477
478# -- This part to exercise text reading. Don't change anything! --
479# <iobench text chunk marker>
480
481"""
4821.
483Gáttir allar,
484áðr gangi fram,
485um skoðask skyli,
486um skyggnast skyli,
487því at óvíst er at vita,
488hvar óvinir
489sitja á fleti fyrir.
490
4912.
492Gefendr heilir!
493Gestr er inn kominn,
494hvar skal sitja sjá?
495Mjök er bráðr,
496sá er á bröndum skal
497síns of freista frama.
498
4993.
500Elds er þörf,
501þeims inn er kominn
502ok á kné kalinn;
503matar ok váða
504er manni þörf,
505þeim er hefr um fjall farit.
506
5074.
508Vatns er þörf,
509þeim er til verðar kemr,
510þerru ok þjóðlaðar,
511góðs of æðis,
512ef sér geta mætti,
513orðs ok endrþögu.
514
5155.
516Vits er þörf,
517þeim er víða ratar;
518dælt er heima hvat;
519at augabragði verðr,
520sá er ekki kann
521ok með snotrum sitr.
522
5236.
524At hyggjandi sinni
525skyli-t maðr hræsinn vera,
526heldr gætinn at geði;
527þá er horskr ok þögull
528kemr heimisgarða til,
529sjaldan verðr víti vörum,
530því at óbrigðra vin
531fær maðr aldregi
532en mannvit mikit.
533
5347.
535Inn vari gestr,
536er til verðar kemr,
537þunnu hljóði þegir,
538eyrum hlýðir,
539en augum skoðar;
540svá nýsisk fróðra hverr fyrir.
541
5428.
543Hinn er sæll,
544er sér of getr
545lof ok líknstafi;
546ódælla er við þat,
547er maðr eiga skal
548annars brjóstum í.
549"""
550
551"""
552C'est revenir tard, je le sens, sur un sujet trop rebattu et déjà presque oublié. Mon état, qui ne me permet plus aucun travail suivi, mon aversion pour le genre polémique, ont causé ma lenteur à écrire et ma répugnance à publier. J'aurais même tout à fait supprimé ces Lettres, ou plutôt je lie les aurais point écrites, s'il n'eût été question que de moi : Mais ma patrie ne m'est pas tellement devenue étrangère que je puisse voir tranquillement opprimer ses citoyens, surtout lorsqu'ils n'ont compromis leurs droits qu'en défendant ma cause. Je serais le dernier des hommes si dans une telle occasion j'écoutais un sentiment qui n'est plus ni douceur ni patience, mais faiblesse et lâcheté, dans celui qu'il empêche de remplir son devoir.
553Rien de moins important pour le public, j'en conviens, que la matière de ces lettres. La constitution d'une petite République, le sort d'un petit particulier, l'exposé de quelques injustices, la réfutation de quelques sophismes ; tout cela n'a rien en soi d'assez considérable pour mériter beaucoup de lecteurs : mais si mes sujets sont petits mes objets sont grands, et dignes de l'attention de tout honnête homme. Laissons Genève à sa place, et Rousseau dans sa dépression ; mais la religion, mais la liberté, la justice ! voilà, qui que vous soyez, ce qui n'est pas au-dessous de vous.
554Qu'on ne cherche pas même ici dans le style le dédommagement de l'aridité de la matière. Ceux que quelques traits heureux de ma plume ont si fort irrités trouveront de quoi s'apaiser dans ces lettres, L'honneur de défendre un opprimé eût enflammé mon coeur si j'avais parlé pour un autre. Réduit au triste emploi de me défendre moi-même, j'ai dû me borner à raisonner ; m'échauffer eût été m'avilir. J'aurai donc trouvé grâce en ce point devant ceux qui s'imaginent qu'il est essentiel à la vérité d'être dite froidement ; opinion que pourtant j'ai peine à comprendre. Lorsqu'une vive persuasion nous anime, le moyen d'employer un langage glacé ? Quand Archimède tout transporté courait nu dans les rues de Syracuse, en avait-il moins trouvé la vérité parce qu'il se passionnait pour elle ? Tout au contraire, celui qui la sent ne peut s'abstenir de l'adorer ; celui qui demeure froid ne l'a pas vue.
555Quoi qu'il en soit, je prie les lecteurs de vouloir bien mettre à part mon beau style, et d'examiner seulement si je raisonne bien ou mal ; car enfin, de cela seul qu'un auteur s'exprime en bons termes, je ne vois pas comment il peut s'ensuivre que cet auteur ne sait ce qu'il dit.
556"""
557