benchtests/bench-strcoll.c

   1 /* Measure strcoll execution time in different locales.
   2    Copyright (C) 2015-2017 Free Software Foundation, Inc.
   3    This file is part of the GNU C Library.
   4
   5    The GNU C Library is free software; you can redistribute it and/or
   6    modify it under the terms of the GNU Lesser General Public
   7    License as published by the Free Software Foundation; either
   8    version 2.1 of the License, or (at your option) any later version.
   9
  10    The GNU C Library is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13    Lesser General Public License for more details.
  14
  15    You should have received a copy of the GNU Lesser General Public
  16    License along with the GNU C Library; if not, see
  17    <http://www.gnu.org/licenses/>.  */
  18
  19 #include <stdio.h>
  20 #include <fcntl.h>
  21 #include <assert.h>
  22 #include <stdlib.h>
  23 #include <locale.h>
  24 #include <unistd.h>
  25 #include "json-lib.h"
  26 #include "bench-timing.h"
  27 #include <string.h>
  28
  29 /* Many thanks to http://generator.lorem-ipsum.info/  */
  30 #define INPUT_PREFIX "strcoll-inputs/"
  31
  32 static const char *const input_files[] = {
  33   "filelist#C",
  34   "filelist#en_US.UTF-8",
  35   "lorem_ipsum#vi_VN.UTF-8",
  36   "lorem_ipsum#ar_SA.UTF-8",
  37   "lorem_ipsum#en_US.UTF-8",
  38   "lorem_ipsum#zh_CN.UTF-8",
  39   "lorem_ipsum#cs_CZ.UTF-8",
  40   "lorem_ipsum#en_GB.UTF-8",
  41   "lorem_ipsum#da_DK.UTF-8",
  42   "lorem_ipsum#pl_PL.UTF-8",
  43   "lorem_ipsum#fr_FR.UTF-8",
  44   "lorem_ipsum#pt_PT.UTF-8",
  45   "lorem_ipsum#el_GR.UTF-8",
  46   "lorem_ipsum#ru_RU.UTF-8",
  47   "lorem_ipsum#he_IL.UTF-8",
  48   "lorem_ipsum#es_ES.UTF-8",
  49   "lorem_ipsum#hi_IN.UTF-8",
  50   "lorem_ipsum#sv_SE.UTF-8",
  51   "lorem_ipsum#hu_HU.UTF-8",
  52   "lorem_ipsum#tr_TR.UTF-8",
  53   "lorem_ipsum#is_IS.UTF-8",
  54   "lorem_ipsum#it_IT.UTF-8",
  55   "lorem_ipsum#sr_RS.UTF-8",
  56   "lorem_ipsum#ja_JP.UTF-8"
  57 };
  58
  59 #define TEXTFILE_DELIMITER " \n\r\t.,?!"
  60
  61 static char *
  62 read_file (const char *filename)
  63 {
  64   struct stat stats;
  65   char *buffer = NULL;
  66   int fd = open (filename, O_CLOEXEC);
  67
  68   if (fd >= 0)
  69     {
  70       if (fstat (fd, &stats) == 0)
  71         {
  72           buffer = malloc (stats.st_size + 1);
  73           if (buffer)
  74             {
  75               if (read (fd, buffer, stats.st_size) == stats.st_size)
  76                 buffer[stats.st_size] = '\0';
  77               else
  78                 {
  79                   free (buffer);
  80                   buffer = NULL;
  81                 }
  82             }
  83         }
  84       close (fd);
  85     }
  86
  87   return buffer;
  88 }
  89
  90 static size_t
  91 count_words (const char *text, const char *delim)
  92 {
  93   size_t wordcount = 0;
  94   char *tmp = strdup (text);
  95
  96   char *token = strtok (tmp, delim);
  97   while (token != NULL)
  98     {
  99       if (*token != '\0')
 100         wordcount++;
 101       token = strtok (NULL, delim);
 102     }
 103
 104   free (tmp);
 105   return wordcount;
 106 }
 107
 108 typedef struct
 109 {
 110   size_t size;
 111   char **words;
 112 } word_list;
 113
 114 static word_list *
 115 new_word_list (size_t size)
 116 {
 117   word_list *list = malloc (sizeof (word_list));
 118   assert (list != NULL);
 119   list->size = size;
 120   list->words = malloc (size * sizeof (char *));
 121   assert (list->words != NULL);
 122   return list;
 123 }
 124
 125 static word_list *
 126 str_word_list (const char *str, const char *delim)
 127 {
 128   size_t n = 0;
 129   word_list *list = new_word_list (count_words (str, delim));
 130
 131   char *toks = strdup (str);
 132   char *word = strtok (toks, delim);
 133   while (word != NULL && n < list->size)
 134     {
 135       if (*word != '\0')
 136         list->words[n++] = strdup (word);
 137       word = strtok (NULL, delim);
 138     }
 139
 140   free (toks);
 141   return list;
 142 }
 143
 144 static word_list *
 145 copy_word_list (const word_list *list)
 146 {
 147   size_t i;
 148   word_list *copy = new_word_list (list->size);
 149
 150   for (i = 0; i < list->size; i++)
 151     copy->words[i] = strdup (list->words[i]);
 152
 153   return copy;
 154 }
 155
 156 static void
 157 free_word_list (word_list *list)
 158 {
 159   size_t i;
 160   for (i = 0; i < list->size; i++)
 161     free (list->words[i]);
 162
 163   free (list->words);
 164   free (list);
 165 }
 166
 167 static int
 168 compare_words (const void *a, const void *b)
 169 {
 170   const char *s1 = *(char **) a;
 171   const char *s2 = *(char **) b;
 172   return strcoll (s1, s2);
 173 }
 174
 175 #undef INNER_LOOP_ITERS
 176 #define INNER_LOOP_ITERS 16
 177
 178 static void
 179 bench_list (json_ctx_t *json_ctx, word_list *list)
 180 {
 181   size_t i;
 182   timing_t start, stop, cur;
 183
 184   word_list **tests = malloc (INNER_LOOP_ITERS * sizeof (word_list *));
 185   assert (tests != NULL);
 186   for (i = 0; i < INNER_LOOP_ITERS; i++)
 187     tests[i] = copy_word_list (list);
 188
 189   TIMING_NOW (start);
 190   for (i = 0; i < INNER_LOOP_ITERS; i++)
 191     qsort (tests[i]->words, tests[i]->size, sizeof (char *), compare_words);
 192   TIMING_NOW (stop);
 193
 194   TIMING_DIFF (cur, start, stop);
 195   setlocale (LC_ALL, "en_US.UTF-8");
 196   json_attr_double (json_ctx, "duration", cur);
 197   json_attr_double (json_ctx, "iterations", i);
 198   json_attr_double (json_ctx, "mean", (double) cur / i);
 199
 200   for (i = 0; i < INNER_LOOP_ITERS; i++)
 201     free_word_list (tests[i]);
 202   free (tests);
 203 }
 204
 205 typedef enum
 206 {
 207   OK,
 208   ERROR_FILENAME,
 209   ERROR_LOCALE,
 210   ERROR_IO
 211 } result_t;
 212
 213 static result_t
 214 bench_file (json_ctx_t *json_ctx, const char *testname, const char *filename,
 215             const char *locale)
 216 {
 217   if (setlocale (LC_ALL, locale) == NULL)
 218     return ERROR_LOCALE;
 219
 220   char *text = read_file (filename);
 221   if (text == NULL)
 222     return ERROR_IO;
 223
 224   word_list *list = str_word_list (text, TEXTFILE_DELIMITER);
 225
 226   json_attr_object_begin (json_ctx, testname);
 227   bench_list (json_ctx, list);
 228   json_attr_object_end (json_ctx);
 229
 230   free_word_list (list);
 231   free (text);
 232   return OK;
 233 }
 234
 235 int
 236 main (void)
 237 {
 238   json_ctx_t *json_ctx = malloc (sizeof (json_ctx_t));
 239   assert (json_ctx != NULL);
 240   json_init (json_ctx, 2, stdout);
 241   json_attr_object_begin (json_ctx, "strcoll");
 242
 243   size_t i;
 244   result_t result = OK;
 245   for (i = 0; i < (sizeof (input_files) / sizeof (input_files[0])); i++)
 246     {
 247       char *locale = strchr (input_files[i], '#');
 248       if (locale == NULL)
 249         {
 250           printf ("Failed to get locale from filename %s, aborting!\n",
 251                   input_files[i]);
 252           return ERROR_FILENAME;
 253         }
 254
 255       char *filename;
 256       asprintf (&filename, INPUT_PREFIX "%s", input_files[i]);
 257       result = bench_file (json_ctx, input_files[i], filename, locale + 1);
 258
 259       if (result != OK)
 260         {
 261           if (result == ERROR_LOCALE)
 262             printf ("Failed to set locale %s, aborting!\n", locale);
 263           else if (result == ERROR_IO)
 264             printf ("Failed to read file %s, aborting!\n", filename);
 265           free (filename);
 266           goto out;
 267         }
 268       free (filename);
 269     }
 270
 271 out:
 272   json_attr_object_end (json_ctx);
 273   free (json_ctx);
 274   return result;
 275 }