Tianjie Xu | a5dcb7c | 2018-09-25 12:25:15 -0700 | [diff] [blame] | 1 | #include <cstddef> |
| 2 | #include <cstdio> |
| 3 | #include <cstring> |
| 4 | #include <fstream> |
| 5 | #include <vector> |
| 6 | |
| 7 | #include "./deorummolae.h" |
| 8 | #include "./durchschlag.h" |
| 9 | #include "./sieve.h" |
| 10 | |
| 11 | #define METHOD_DM 0 |
| 12 | #define METHOD_SIEVE 1 |
| 13 | #define METHOD_DURCHSCHLAG 2 |
| 14 | #define METHOD_DISTILL 3 |
| 15 | #define METHOD_PURIFY 4 |
| 16 | |
| 17 | static size_t readInt(const char* str) { |
| 18 | size_t result = 0; |
| 19 | if (str[0] == 0 || str[0] == '0') { |
| 20 | return 0; |
| 21 | } |
| 22 | for (size_t i = 0; i < 13; ++i) { |
| 23 | if (str[i] == 0) { |
| 24 | return result; |
| 25 | } |
| 26 | if (str[i] == 'k' || str[i] == 'K') { |
| 27 | if ((str[i + 1] == 0) && ((result << 10) > result)) { |
| 28 | return result << 10; |
| 29 | } |
| 30 | return 0; |
| 31 | } |
| 32 | if (str[i] == 'm' || str[i] == 'M') { |
| 33 | if ((str[i + 1] == 0) && ((result << 20) > result)) { |
| 34 | return result << 20; |
| 35 | } |
| 36 | return 0; |
| 37 | } |
| 38 | if (str[i] < '0' || str[i] > '9') { |
| 39 | return 0; |
| 40 | } |
| 41 | size_t next = (10 * result) + (str[i] - '0'); |
| 42 | if (next <= result) { |
| 43 | return 0; |
| 44 | } |
| 45 | result = next; |
| 46 | } |
| 47 | return 0; |
| 48 | } |
| 49 | |
| 50 | static std::string readFile(const std::string& path) { |
| 51 | std::ifstream file(path); |
| 52 | std::string content( |
| 53 | (std::istreambuf_iterator<char>(file)), std::istreambuf_iterator<char>()); |
| 54 | return content; |
| 55 | } |
| 56 | |
| 57 | static void writeFile(const char* file, const std::string& content) { |
| 58 | std::ofstream outfile(file, std::ofstream::binary); |
| 59 | outfile.write(content.c_str(), static_cast<std::streamsize>(content.size())); |
| 60 | outfile.close(); |
| 61 | } |
| 62 | |
| 63 | static void writeSamples(char const* argv[], const std::vector<int>& pathArgs, |
| 64 | const std::vector<size_t>& sizes, const uint8_t* data) { |
| 65 | size_t offset = 0; |
| 66 | for (size_t i = 0; i < pathArgs.size(); ++i) { |
| 67 | int j = pathArgs[i]; |
| 68 | const char* file = argv[j]; |
| 69 | size_t sampleSize = sizes[i]; |
| 70 | std::ofstream outfile(file, std::ofstream::binary); |
| 71 | outfile.write(reinterpret_cast<const char*>(data + offset), |
| 72 | static_cast<std::streamsize>(sampleSize)); |
| 73 | outfile.close(); |
| 74 | offset += sampleSize; |
| 75 | } |
| 76 | } |
| 77 | |
| 78 | /* Returns "base file name" or its tail, if it contains '/' or '\'. */ |
| 79 | static const char* fileName(const char* path) { |
| 80 | const char* separator_position = strrchr(path, '/'); |
| 81 | if (separator_position) path = separator_position + 1; |
| 82 | separator_position = strrchr(path, '\\'); |
| 83 | if (separator_position) path = separator_position + 1; |
| 84 | return path; |
| 85 | } |
| 86 | |
| 87 | static void printHelp(const char* name) { |
| 88 | fprintf(stderr, "Usage: %s [OPTION]... DICTIONARY [SAMPLE]...\n", name); |
| 89 | fprintf(stderr, |
| 90 | "Options:\n" |
| 91 | " --dm use 'deorummolae' engine\n" |
| 92 | " --distill rewrite samples; unique text parts are removed\n" |
| 93 | " --dsh use 'durchschlag' engine (default)\n" |
| 94 | " --purify rewrite samples; unique text parts are zeroed out\n" |
| 95 | " --sieve use 'sieve' engine\n" |
| 96 | " -b# set block length for 'durchschlag'; default: 1024\n" |
| 97 | " -s# set slice length for 'distill', 'durchschlag', 'purify'\n" |
| 98 | " and 'sieve'; default: 16\n" |
| 99 | " -t# set target dictionary size (limit); default: 16K\n" |
| 100 | " -u# set minimum slice population (for rewrites); default: 2\n" |
| 101 | "# is a decimal number with optional k/K/m/M suffix.\n" |
| 102 | "WARNING: 'distill' and 'purify' will overwrite original samples!\n" |
| 103 | " Completely unique samples might become empty files.\n\n"); |
| 104 | } |
| 105 | |
| 106 | int main(int argc, char const* argv[]) { |
| 107 | int dictionaryArg = -1; |
| 108 | int method = METHOD_DURCHSCHLAG; |
| 109 | size_t sliceLen = 16; |
| 110 | size_t targetSize = 16 << 10; |
| 111 | size_t blockSize = 1024; |
| 112 | size_t minimumPopulation = 2; |
| 113 | |
| 114 | std::vector<uint8_t> data; |
| 115 | std::vector<size_t> sizes; |
| 116 | std::vector<int> pathArgs; |
| 117 | size_t total = 0; |
| 118 | for (int i = 1; i < argc; ++i) { |
| 119 | if (argv[i] == nullptr) { |
| 120 | continue; |
| 121 | } |
| 122 | if (argv[i][0] == '-') { |
| 123 | if (argv[i][1] == '-') { |
| 124 | if (dictionaryArg != -1) { |
| 125 | fprintf(stderr, |
| 126 | "Method should be specified before dictionary / sample '%s'\n", |
| 127 | argv[i]); |
| 128 | exit(1); |
| 129 | } |
| 130 | if (std::strcmp("--sieve", argv[i]) == 0) { |
| 131 | method = METHOD_SIEVE; |
| 132 | continue; |
| 133 | } |
| 134 | if (std::strcmp("--dm", argv[i]) == 0) { |
| 135 | method = METHOD_DM; |
| 136 | continue; |
| 137 | } |
| 138 | if (std::strcmp("--dsh", argv[i]) == 0) { |
| 139 | method = METHOD_DURCHSCHLAG; |
| 140 | continue; |
| 141 | } |
| 142 | if (std::strcmp("--distill", argv[i]) == 0) { |
| 143 | method = METHOD_DISTILL; |
| 144 | continue; |
| 145 | } |
| 146 | if (std::strcmp("--purify", argv[i]) == 0) { |
| 147 | method = METHOD_PURIFY; |
| 148 | continue; |
| 149 | } |
| 150 | printHelp(fileName(argv[0])); |
| 151 | fprintf(stderr, "Invalid option '%s'\n", argv[i]); |
| 152 | exit(1); |
| 153 | } |
| 154 | if (argv[i][1] == 'b') { |
| 155 | blockSize = readInt(&argv[i][2]); |
| 156 | if (blockSize < 16 || blockSize > 65536) { |
| 157 | printHelp(fileName(argv[0])); |
| 158 | fprintf(stderr, "Invalid option '%s'\n", argv[i]); |
| 159 | exit(1); |
| 160 | } |
| 161 | } else if (argv[i][1] == 's') { |
| 162 | sliceLen = readInt(&argv[i][2]); |
| 163 | if (sliceLen < 4 || sliceLen > 256) { |
| 164 | printHelp(fileName(argv[0])); |
| 165 | fprintf(stderr, "Invalid option '%s'\n", argv[i]); |
| 166 | exit(1); |
| 167 | } |
| 168 | } else if (argv[i][1] == 't') { |
| 169 | targetSize = readInt(&argv[i][2]); |
| 170 | if (targetSize < 256 || targetSize > (1 << 25)) { |
| 171 | printHelp(fileName(argv[0])); |
| 172 | fprintf(stderr, "Invalid option '%s'\n", argv[i]); |
| 173 | exit(1); |
| 174 | } |
| 175 | } else if (argv[i][1] == 'u') { |
| 176 | minimumPopulation = readInt(&argv[i][2]); |
| 177 | if (minimumPopulation < 256 || minimumPopulation > 65536) { |
| 178 | printHelp(fileName(argv[0])); |
| 179 | fprintf(stderr, "Invalid option '%s'\n", argv[i]); |
| 180 | exit(1); |
| 181 | } |
| 182 | } else { |
| 183 | printHelp(fileName(argv[0])); |
| 184 | fprintf(stderr, "Unrecognized option '%s'\n", argv[i]); |
| 185 | exit(1); |
| 186 | } |
| 187 | continue; |
| 188 | } |
| 189 | if (dictionaryArg == -1) { |
| 190 | if (method != METHOD_DISTILL && method != METHOD_PURIFY) { |
| 191 | dictionaryArg = i; |
| 192 | continue; |
| 193 | } |
| 194 | } |
| 195 | std::string content = readFile(argv[i]); |
| 196 | data.insert(data.end(), content.begin(), content.end()); |
| 197 | total += content.size(); |
| 198 | pathArgs.push_back(i); |
| 199 | sizes.push_back(content.size()); |
| 200 | } |
| 201 | bool wantDictionary = (dictionaryArg == -1); |
| 202 | if (method == METHOD_DISTILL || method == METHOD_PURIFY) { |
| 203 | wantDictionary = false; |
| 204 | } |
| 205 | if (wantDictionary || total == 0) { |
| 206 | printHelp(fileName(argv[0])); |
| 207 | fprintf(stderr, "Not enough arguments\n"); |
| 208 | exit(1); |
| 209 | } |
| 210 | |
| 211 | if (method == METHOD_SIEVE) { |
| 212 | writeFile(argv[dictionaryArg], sieve_generate( |
| 213 | targetSize, sliceLen, sizes, data.data())); |
| 214 | } else if (method == METHOD_DM) { |
| 215 | writeFile(argv[dictionaryArg], DM_generate( |
| 216 | targetSize, sizes, data.data())); |
| 217 | } else if (method == METHOD_DURCHSCHLAG) { |
| 218 | writeFile(argv[dictionaryArg], durchschlag_generate( |
| 219 | targetSize, sliceLen, blockSize, sizes, data.data())); |
| 220 | } else if (method == METHOD_DISTILL) { |
| 221 | durchschlag_distill(sliceLen, minimumPopulation, &sizes, data.data()); |
| 222 | writeSamples(argv, pathArgs, sizes, data.data()); |
| 223 | } else if (method == METHOD_PURIFY) { |
| 224 | durchschlag_purify(sliceLen, minimumPopulation, sizes, data.data()); |
| 225 | writeSamples(argv, pathArgs, sizes, data.data()); |
| 226 | } else { |
| 227 | printHelp(fileName(argv[0])); |
| 228 | fprintf(stderr, "Unknown generator\n"); |
| 229 | exit(1); |
| 230 | } |
| 231 | return 0; |
| 232 | } |