benchmark.py

   1 #!/usr/bin/env python
   2 #
   3 # Copyright 2004-2006 Zuza Software Foundation
   4 #
   5 # This file is part of translate.
   6 #
   7 # translate is free software; you can redistribute it and/or modify
   8 # it under the terms of the GNU General Public License as published by
   9 # the Free Software Foundation; either version 2 of the License, or
  10 # (at your option) any later version.
  11 #
  12 # translate is distributed in the hope that it will be useful,
  13 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15 # GNU General Public License for more details.
  16 #
  17 # You should have received a copy of the GNU General Public License
  18 # along with translate; if not, write to the Free Software
  19 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  20
  21 from Pootle import pootlefile
  22 from Pootle import projects
  23 from Pootle import potree
  24 from Pootle import pootle
  25 from Pootle import users
  26 from translate.storage import po
  27 from translate.search import indexer
  28 import os
  29 import profile
  30 import pstats
  31
  32 class PootleBenchmarker:
  33     """class to aid in benchmarking pootle"""
  34     StoreClass = pootlefile.pootlefile
  35     UnitClass = pootlefile.pootleunit
  36     def __init__(self, test_dir):
  37         """sets up benchmarking on the test directory"""
  38         self.test_dir = os.path.abspath(test_dir)
  39         self.project_dir = os.path.join(self.test_dir, "benchmark")
  40         self.po_dir = os.path.join(self.project_dir, "zxx")
  41
  42     def clear_test_dir(self):
  43         """removes the given directory"""
  44         if os.path.exists(self.test_dir):
  45             for dirpath, subdirs, filenames in os.walk(self.test_dir, topdown=False):
  46                 for name in filenames:
  47                     os.remove(os.path.join(dirpath, name))
  48                 for name in subdirs:
  49                     os.rmdir(os.path.join(dirpath, name))
  50         if os.path.exists(self.test_dir): os.rmdir(self.test_dir)
  51         assert not os.path.exists(self.test_dir)
  52
  53     def setup_server(self):
  54         """gets a pootle server"""
  55         cwd = os.path.abspath(os.path.curdir)
  56         parser = pootle.PootleOptionParser()
  57         prefsfile = os.path.join(self.test_dir, "pootle.prefs")
  58         pootleprefsstr = """
  59 importmodules.pootleserver = 'Pootle.pootle'
  60 Pootle:
  61   serverclass = pootleserver.PootleServer
  62   sessionkey = 'dummy'
  63   baseurl = "/"
  64   userprefs = "users.prefs"
  65   podirectory = "%s"
  66   projects.benchmark:
  67     fullname = "Benchmark"
  68     description = "Benchmark auto-created files"
  69     checkstyle = "standard"
  70     localfiletype = "po"
  71   languages.zxx.fullname = "Test Language"
  72 """ % (self.test_dir)
  73         open(prefsfile, "w").write(pootleprefsstr)
  74         userprefsfile = os.path.join(self.test_dir, "users.prefs")
  75         open(userprefsfile, "w").write("testuser.activated=1\ntestuser.passwdhash = 'dd82c1882969461de74b46427961ea2c'\n")
  76         options, args = parser.parse_args(["--prefsfile=%s" % prefsfile])
  77         options.servertype = "dummy"
  78         self.server = parser.getserver(options)
  79         os.chdir(cwd)
  80         return self.server
  81
  82     def get_session(self):
  83         """gets a new session object"""
  84         return users.PootleSession(self.server.sessioncache, self.server)
  85
  86     def create_sample_files(self, num_dirs, files_per_dir, strings_per_file, source_words_per_string, target_words_per_string):
  87         """creates sample files for benchmarking"""
  88         if not os.path.exists(self.test_dir):
  89             os.mkdir(self.test_dir)
  90         if not os.path.exists(self.project_dir):
  91             os.mkdir(self.project_dir)
  92         if not os.path.exists(self.po_dir):
  93             os.mkdir(self.po_dir)
  94         for dirnum in range(num_dirs):
  95             if num_dirs > 1:
  96                 dirname = os.path.join(self.po_dir, "sample_%d" % dirnum)
  97                 if not os.path.exists(dirname):
  98                     os.mkdir(dirname)
  99             else:
 100                 dirname = self.po_dir
 101             for filenum in range(files_per_dir):
 102                 sample_file = self.StoreClass(pofilename=os.path.join(dirname, "file_%d.po" % filenum))
 103                 for stringnum in range(strings_per_file):
 104                     source_string = " ".join(["word%d" % i for i in range(source_words_per_string)])
 105                     sample_unit = sample_file.addsourceunit(source_string)
 106                     sample_unit.target = " ".join(["drow%d" % i for i in range(target_words_per_string)])
 107                 sample_file.savepofile()
 108
 109     def parse_po_files(self):
 110         """parses all the po files in the test directory into memory"""
 111         count = 0
 112         for dirpath, subdirs, filenames in os.walk(self.po_dir, topdown=False):
 113             for name in filenames:
 114                 pofilename = os.path.join(dirpath, name)
 115                 parsedfile = po.pofile(open(pofilename, 'r'))
 116                 count += len(parsedfile.units)
 117         print "counted %d units" % count
 118
 119     def parse_and_create_stats(self):
 120         """parses all the po files in the test directory into memory, using pootlefile, which creates Stats"""
 121         count = 0
 122         indexer.HAVE_INDEXER = False
 123         for dirpath, subdirs, filenames in os.walk(self.po_dir, topdown=False):
 124             for name in filenames:
 125                 pofilename = os.path.join(dirpath, name)
 126                 parsedfile = pootlefile.pootlefile(pofilename=pofilename)
 127                 count += len(parsedfile.units)
 128         print "stats on %d units" % count
 129
 130     def parse_and_create_index(self):
 131         """parses all the po files in the test directory into memory, using pootlefile, and allow index creation"""
 132         count = 0
 133         indexer.HAVE_INDEXER = True
 134         self.server.potree.projectcache.clear()
 135         project = self.server.potree.getproject("zxx", "benchmark")
 136         for name in project.browsefiles():
 137             count += len(project.getpofile(name).units)
 138         print "indexed %d units" % count
 139         assert os.path.exists(os.path.join(self.po_dir, ".poindex-%s-%s" % (project.projectcode, project.languagecode)))
 140
 141     def generate_main_page(self):
 142         """tests generating the main page"""
 143         session = self.get_session()
 144         page = self.server.getpage(["index.html"], session, {})
 145         print page.templatevars
 146
 147     def generate_projectindex_page(self):
 148         """tests generating the index page for the project"""
 149         session = self.get_session()
 150         assert self.server.potree.haslanguage("zxx")
 151         assert self.server.potree.hasproject("zxx", "benchmark")
 152         page = self.server.getpage(["zxx", "benchmark"], session, {})
 153         print page.templatevars
 154
 155     def generate_translation_page(self):
 156         """tests generating the translation page for the file"""
 157         session = self.get_session()
 158         page = self.server.getpage(["zxx", "benchmark", "translate.html"], session, {})
 159         print page.templatevars
 160
 161     def submit_translation_change(self):
 162         """tests generating the translation page for the file"""
 163         session = self.get_session()
 164         project = self.server.potree.getproject("zxx", "benchmark")
 165         project.setrights(None, ["view", "translate"])
 166         pofilename = project.browsefiles()[0]
 167         args = {"pofilename": pofilename, "submit0": "true", "trans0": "changed"}
 168         page = self.server.getpage(["zxx", "benchmark", "translate.html"], session, args)
 169         pofile = project.getpofile(pofilename)
 170         print str(pofile.getitem(0))
 171         # assert fails because of multistring
 172         # assert pofile.getitem(0).unquotedmsgstr == "changed"
 173         print page.templatevars
 174
 175 if __name__ == "__main__":
 176     for sample_file_sizes in [
 177       # num_dirs, files_per_dir, strings_per_file, source_words_per_string, target_words_per_string
 178       # (1, 1, 1, 1, 1),
 179       (1, 1, 30, 10, 10),
 180       # (1, 5, 10, 10, 10),
 181       (1, 10, 10, 10, 10),
 182       (5, 10, 10, 10, 10),
 183       # (5, 10, 100, 20, 20),
 184       # (10, 20, 100, 10, 10),
 185       ]:
 186         benchmarker = PootleBenchmarker("BenchmarkDir")
 187         benchmarker.clear_test_dir()
 188         benchmarker.create_sample_files(*sample_file_sizes)
 189         benchmarker.setup_server()
 190         methods = ["parse_po_files", "parse_and_create_stats", "parse_and_create_index",
 191                    "generate_main_page", "generate_projectindex_page", "generate_translation_page",
 192                    "submit_translation_change",
 193                   ]
 194         for methodname in methods:
 195             print methodname, "%d dirs, %d files, %d strings, %d/%d words" % sample_file_sizes
 196             print "_______________________________________________________"
 197             statsfile = methodname + '_%d_%d_%d_%d_%d.stats' % sample_file_sizes
 198             profile.run('benchmarker.%s()' % methodname, statsfile)
 199             stats = pstats.Stats(statsfile)
 200             stats.sort_stats('cumulative').print_stats(20)
 201             print "_______________________________________________________"
 202         benchmarker.clear_test_dir()
 203