[ad_1]
\n<\/p>\n
Its quite easy to implement a replacement for (Demonstration on ideone<\/a><\/strong>)<\/p>\n It doesn’t seem to make much sense to replace something which is already available. Thus, I want to mention some thouhgts about this.<\/p>\n The implementation can be adjusted to the precise personal requirements:<\/p>\n Extending this idea to other numeric types like e.g. As floating point numbers are definitely subject of localization this has to be considered. (Concerning decimal integer numbers I’m not sure what could be localized but even this might be the case.) If a text file reader with floating point number syntax (like in C) is implemented you may not forget to adjust the locale to Another fact is, that consideration of locale (even if adjusted to C) seems to be somehow expensive. Some years ago, we implemented an own floating point reader as replacement of Update:<\/strong><\/p>\n Encouraged by the feedback of Paul Floyd, I got curious how<\/em> faster I tried this on some platforms.<\/p>\n VS2013 on Windows 10 (64 bit), Release mode:<\/p>\n gcc 5.4.0 on cygwin, Windows 10 (64 bit), atoi()<\/code><\/a>:<\/p>\nint strToInt(const char *text)\n{\n int n = 0, sign = 1;\n switch (*text) {\n case '-': sign = -1;\n case '+': ++text;\n }\n for (; isdigit(*text); ++text) n *= 10, n += *text - '0';\n return n * sign;\n}\n<\/code><\/pre>\n\n
text<\/code> may be returned (as in strtol()<\/code>) to check how many characters have been processed or to do further parsing of other contents<\/li>\nunsigned<\/code> (which does not accept a sign).<\/li>\nfloat<\/code> or double<\/code>, it becomes even more interesting.<\/p>\nC<\/code> before using strtod()<\/code><\/a> (using setlocale()<\/code><\/a>). (Being a German I’m sensitive to this topic, as in the German locale, the meaning of ‘.’ and ‘,’ are just vice versa like in English.)<\/p>\n{ const char *localeOld = setlocale(LC_ALL, \"C\");\n value = strtod(text);\n setlocale(LC_ALL, localeOld);\n}\n<\/code><\/pre>\nstrtod()<\/code><\/a> which provided a speed-up of 60 … 100 in a COLLADA<\/a> reader (an XML file format where files often provide lots of floating point numbers).<\/p>\nstrToInt()<\/code> might be. Thus, I built a simple test suite and made some measurements:<\/p>\n#include <assert.h>\n#include <ctype.h>\n#include <stdio.h>\n#include <stdlib.h>\n#include <time.h>\n\nint strToInt(const char *text)\n{\n int n = 0, sign = 1;\n switch (*text) {\n case '-': sign = -1;\n case '+': ++text;\n }\n for (; isdigit(*text); ++text) n *= 10, n += *text - '0';\n return n * sign;\n}\n\nint main(int argc, char **argv)\n{\n int n = 10000000; \/* default number of measurements *\/\n \/* read command line options *\/\n if (argc > 1) n = atoi(argv[1]);\n if (n <= 0) return 1; \/* ERROR *\/\n \/* build samples *\/\n assert(sizeof(int) <= 8); \/* May be, I want to do it again 20 years ago. *\/\n \/* 24 characters should be capable to hold any decimal for int\n * (upto 64 bit)\n *\/\n char (*samples)[24] = malloc(n * 24 * sizeof(char));\n if (!samples) {\n printf(\"ERROR: Cannot allocate samples!\\n\"\n \"(Out of memory.)\\n\");\n return 1;\n }\n for (int i = 0; i < n; ++i) sprintf(samples[i], \"%d\", i - (i & 1) * n);\n \/* assert correct results, ensure fair caching, pre-heat CPU *\/\n int *retAToI = malloc(n * sizeof(int));\n if (!retAToI) {\n printf(\"ERROR: Cannot allocate result array for atoi()!\\n\"\n \"(Out of memory.)\\n\");\n return 1;\n }\n int *retStrToInt = malloc(n * sizeof(int));\n if (!retStrToInt) {\n printf(\"ERROR: Cannot allocate result array for strToInt()!\\n\"\n \"(Out of memory.)\\n\");\n return 1;\n }\n int nErrors = 0;\n for (int i = 0; i < n; ++i) {\n retAToI[i] = atoi(samples[i]); retStrToInt[i] = strToInt(samples[i]);\n if (retAToI[i] != retStrToInt[i]) {\n printf(\"ERROR: atoi(\\\"%s\\\"): %d, strToInt(\\\"%s\\\"): %d!\\n\",\n samples[i], retAToI[i], samples[i], retStrToInt[i]);\n ++nErrors;\n }\n }\n if (nErrors) {\n printf(\"%d ERRORs found!\", nErrors);\n return 2;\n }\n \/* do measurements *\/\n enum { nTries = 10 };\n time_t tTbl[nTries][2];\n for (int i = 0; i < nTries; ++i) {\n printf(\"Measurement %d:\\n\", i + 1);\n { time_t t0 = clock();\n for (int i = 0; i < n; ++i) retAToI[i] = atoi(samples[i]);\n tTbl[i][0] = clock() - t0;\n }\n { time_t t0 = clock();\n for (int i = 0; i < n; ++i) retStrToInt[i] = strToInt(samples[i]);\n tTbl[i][1] = clock() - t0;\n }\n \/* assert correct results (and prevent that measurement is optimized away) *\/\n for (int i = 0; i < n; ++i) if (retAToI[i] != retStrToInt[i]) return 3;\n }\n \/* report *\/\n printf(\"Report:\\n\");\n printf(\"%20s|%20s\\n\", \"atoi() \", \"strToInt() \");\n printf(\"--------------------+--------------------\\n\");\n double tAvg[2] = { 0.0, 0.0 }; const char *sep = \"|\\n\";\n for (int i = 0; i < nTries; ++i) {\n for (int j = 0; j < 2; ++j) {\n double t = (double)tTbl[i][j] \/ CLOCKS_PER_SEC;\n printf(\"%19.3f %c\", t, sep[j]);\n tAvg[j] += t;\n }\n }\n printf(\"--------------------+--------------------\\n\");\n for (int j = 0; j < 2; ++j) printf(\"%19.3f %c\", tAvg[j] \/ nTries, sep[j]);\n \/* done *\/\n return 0;\n}\n<\/code><\/pre>\nReport:\n atoi() | strToInt()\n--------------------+--------------------\n 0.232 | 0.200\n 0.310 | 0.240\n 0.253 | 0.199\n 0.231 | 0.201\n 0.232 | 0.253\n 0.247 | 0.201\n 0.238 | 0.201\n 0.247 | 0.223\n 0.248 | 0.200\n 0.249 | 0.200\n--------------------+--------------------\n 0.249 | 0.212\n<\/code><\/pre>\ngcc -std=c11 -O2<\/code>:<\/p>\nReport:\n atoi() | strToInt() \n--------------------+--------------------\n 0.360 | 0.312 \n 0.391 | 0.250 \n 0.360 | 0.328 \n 0.391 | 0.312 \n 0.375 | 0.281 \n 0.359 | 0.282 \n 0.375 | 0.297 \n 0.391 | 0.250 \n 0.359 | 0.297 \n 0.406 | 0.281 \n--------------------+--------------------\n 0.377 | 0.289\n<\/code><\/pre>\n