| 1 | // =============================================================== // |
|---|
| 2 | // // |
|---|
| 3 | // File : adfile.cxx // |
|---|
| 4 | // Purpose : various IO related functions // |
|---|
| 5 | // // |
|---|
| 6 | // Institute of Microbiology (Technical University Munich) // |
|---|
| 7 | // http://www.arb-home.de/ // |
|---|
| 8 | // // |
|---|
| 9 | // =============================================================== // |
|---|
| 10 | |
|---|
| 11 | #include <climits> |
|---|
| 12 | #include <cctype> |
|---|
| 13 | |
|---|
| 14 | #include <dirent.h> |
|---|
| 15 | #include <unistd.h> |
|---|
| 16 | #include <sys/stat.h> |
|---|
| 17 | |
|---|
| 18 | #include <list> |
|---|
| 19 | #include <string> |
|---|
| 20 | |
|---|
| 21 | #include <arb_sort.h> |
|---|
| 22 | #include <arb_str.h> |
|---|
| 23 | #include <arb_strarray.h> |
|---|
| 24 | #include <arb_file.h> |
|---|
| 25 | |
|---|
| 26 | #include "gb_local.h" |
|---|
| 27 | #include "gb_load.h" |
|---|
| 28 | |
|---|
| 29 | #define FILE_PATH_MAX PATH_MAX |
|---|
| 30 | |
|---|
| 31 | GB_CSTR GB_getcwd() { |
|---|
| 32 | // get the current working directory |
|---|
| 33 | // (directory from which application has been started) |
|---|
| 34 | RETURN_ONETIME_ALLOC(getcwd(0, FILE_PATH_MAX)); |
|---|
| 35 | } |
|---|
| 36 | |
|---|
| 37 | GB_ERROR gb_scan_directory(char *basename, gb_scandir *sd) { |
|---|
| 38 | // goes to header: __ATTR__USERESULT_TODO |
|---|
| 39 | // look for quick saves (basename = yyy/xxx no arb ending !!!!) |
|---|
| 40 | char *path = strdup(basename); |
|---|
| 41 | const char *fulldir = "."; |
|---|
| 42 | char *file = strrchr(path, '/'); |
|---|
| 43 | DIR *dirp; |
|---|
| 44 | int curindex; |
|---|
| 45 | char *suffix; |
|---|
| 46 | dirent *dp; |
|---|
| 47 | struct stat st; |
|---|
| 48 | const char *oldstyle = ".arb.quick"; |
|---|
| 49 | char buffer[FILE_PATH_MAX]; |
|---|
| 50 | int oldstylelen = strlen(oldstyle); |
|---|
| 51 | int filelen; |
|---|
| 52 | |
|---|
| 53 | if (file) { |
|---|
| 54 | *(file++) = 0; |
|---|
| 55 | fulldir = path; |
|---|
| 56 | } |
|---|
| 57 | else { |
|---|
| 58 | file = path; |
|---|
| 59 | } |
|---|
| 60 | |
|---|
| 61 | memset((char*)sd, 0, sizeof(*sd)); |
|---|
| 62 | sd->type = GB_SCAN_NO_QUICK; |
|---|
| 63 | sd->highest_quick_index = -1; |
|---|
| 64 | sd->newest_quick_index = -1; |
|---|
| 65 | sd->date_of_quick_file = 0; |
|---|
| 66 | |
|---|
| 67 | dirp = opendir(fulldir); |
|---|
| 68 | if (!dirp) { |
|---|
| 69 | GB_ERROR error = GBS_global_string("Directory %s of file %s.arb not readable", fulldir, file); |
|---|
| 70 | free(path); |
|---|
| 71 | return error; |
|---|
| 72 | } |
|---|
| 73 | filelen = strlen(file); |
|---|
| 74 | for (dp = readdir(dirp); dp != NULL; dp = readdir(dirp)) { |
|---|
| 75 | if (strncmp(dp->d_name, file, filelen)) continue; |
|---|
| 76 | suffix = dp->d_name + filelen; |
|---|
| 77 | if (suffix[0] != '.') continue; |
|---|
| 78 | if (!strncmp(suffix, oldstyle, oldstylelen)) { |
|---|
| 79 | if (sd->type == GB_SCAN_NEW_QUICK) { |
|---|
| 80 | printf("Warning: Found new and old changes files, using new\n"); |
|---|
| 81 | continue; |
|---|
| 82 | } |
|---|
| 83 | sd->type = GB_SCAN_OLD_QUICK; |
|---|
| 84 | curindex = atoi(suffix+oldstylelen); |
|---|
| 85 | goto check_time_and_date; |
|---|
| 86 | } |
|---|
| 87 | if (strlen(suffix) == 4 && |
|---|
| 88 | suffix[0] == '.' && |
|---|
| 89 | suffix[1] == 'a' && |
|---|
| 90 | isdigit(suffix[2]) && |
|---|
| 91 | isdigit(suffix[3])) { |
|---|
| 92 | if (sd->type == GB_SCAN_OLD_QUICK) { |
|---|
| 93 | printf("Warning: Found new and old changes files, using new\n"); |
|---|
| 94 | } |
|---|
| 95 | sd->type = GB_SCAN_NEW_QUICK; |
|---|
| 96 | curindex = atoi(suffix+2); |
|---|
| 97 | goto check_time_and_date; |
|---|
| 98 | } |
|---|
| 99 | continue; |
|---|
| 100 | check_time_and_date : |
|---|
| 101 | if (curindex > sd->highest_quick_index) sd->highest_quick_index = curindex; |
|---|
| 102 | sprintf(buffer, "%s/%s", fulldir, dp->d_name); |
|---|
| 103 | stat(buffer, &st); |
|---|
| 104 | if ((unsigned long)st.st_mtime > sd->date_of_quick_file) { |
|---|
| 105 | sd->date_of_quick_file = st.st_mtime; |
|---|
| 106 | sd->newest_quick_index = curindex; |
|---|
| 107 | } |
|---|
| 108 | continue; |
|---|
| 109 | } |
|---|
| 110 | closedir(dirp); |
|---|
| 111 | free(path); |
|---|
| 112 | return 0; |
|---|
| 113 | } |
|---|
| 114 | |
|---|
| 115 | |
|---|
| 116 | char *GB_find_all_files(const char *dir, const char *mask, bool filename_only) { |
|---|
| 117 | /* Returns a string containing the filenames of all files matching mask. |
|---|
| 118 | The single filenames are separated by '*'. |
|---|
| 119 | if 'filename_only' is true -> string contains only filenames w/o path |
|---|
| 120 | |
|---|
| 121 | returns 0 if no files found (or directory not found). |
|---|
| 122 | in this case an error may be exported |
|---|
| 123 | |
|---|
| 124 | 'mask' may contain wildcards (*?) or |
|---|
| 125 | it may be a regular expression ('/regexp/') |
|---|
| 126 | */ |
|---|
| 127 | |
|---|
| 128 | DIR *dirp; |
|---|
| 129 | dirent *dp; |
|---|
| 130 | struct stat st; |
|---|
| 131 | char *result = 0; |
|---|
| 132 | char buffer[FILE_PATH_MAX]; |
|---|
| 133 | |
|---|
| 134 | dirp = opendir(dir); |
|---|
| 135 | if (dirp) { |
|---|
| 136 | GBS_string_matcher *matcher = GBS_compile_matcher(mask, GB_IGNORE_CASE); |
|---|
| 137 | if (matcher) { |
|---|
| 138 | for (dp = readdir(dirp); dp != NULL; dp = readdir(dirp)) { |
|---|
| 139 | if (GBS_string_matches_regexp(dp->d_name, matcher)) { |
|---|
| 140 | sprintf(buffer, "%s/%s", dir, dp->d_name); |
|---|
| 141 | if (stat(buffer, &st) == 0 && S_ISREG(st.st_mode)) { // regular file ? |
|---|
| 142 | if (filename_only) strcpy(buffer, dp->d_name); |
|---|
| 143 | if (result) { |
|---|
| 144 | freeset(result, GBS_global_string_copy("%s*%s", result, buffer)); |
|---|
| 145 | } |
|---|
| 146 | else { |
|---|
| 147 | result = strdup(buffer); |
|---|
| 148 | } |
|---|
| 149 | } |
|---|
| 150 | } |
|---|
| 151 | } |
|---|
| 152 | GBS_free_matcher(matcher); |
|---|
| 153 | } |
|---|
| 154 | closedir(dirp); |
|---|
| 155 | } |
|---|
| 156 | |
|---|
| 157 | return result; |
|---|
| 158 | } |
|---|
| 159 | |
|---|
| 160 | char *GB_find_latest_file(const char *dir, const char *mask) { |
|---|
| 161 | /* returns the name of the newest file in dir 'dir' matching 'mask' |
|---|
| 162 | * or NULL (in this case an error may be exported) |
|---|
| 163 | * |
|---|
| 164 | * 'mask' may contain wildcards (*?) or |
|---|
| 165 | * it may be a regular expression ('/regexp/') |
|---|
| 166 | */ |
|---|
| 167 | |
|---|
| 168 | DIR *dirp; |
|---|
| 169 | dirent *dp; |
|---|
| 170 | char buffer[FILE_PATH_MAX]; |
|---|
| 171 | struct stat st; |
|---|
| 172 | GB_ULONG newest = 0; |
|---|
| 173 | char *result = 0; |
|---|
| 174 | |
|---|
| 175 | dirp = opendir(dir); |
|---|
| 176 | if (dirp) { |
|---|
| 177 | GBS_string_matcher *matcher = GBS_compile_matcher(mask, GB_IGNORE_CASE); |
|---|
| 178 | if (matcher) { |
|---|
| 179 | for (dp = readdir(dirp); dp != NULL; dp = readdir(dirp)) { |
|---|
| 180 | if (GBS_string_matches_regexp(dp->d_name, matcher)) { |
|---|
| 181 | sprintf(buffer, "%s/%s", dir, dp->d_name); |
|---|
| 182 | if (stat(buffer, &st) == 0) { |
|---|
| 183 | if ((GB_ULONG)st.st_mtime > newest) { |
|---|
| 184 | newest = st.st_mtime; |
|---|
| 185 | freedup(result, dp->d_name); |
|---|
| 186 | } |
|---|
| 187 | } |
|---|
| 188 | } |
|---|
| 189 | } |
|---|
| 190 | GBS_free_matcher(matcher); |
|---|
| 191 | } |
|---|
| 192 | closedir(dirp); |
|---|
| 193 | } |
|---|
| 194 | return result; |
|---|
| 195 | } |
|---|
| 196 | |
|---|
| 197 | static const char *GB_existing_file(const char *file, bool warn_when_not_found) { |
|---|
| 198 | // return 'file' if it's an existing readable file |
|---|
| 199 | // return NULL otherwise |
|---|
| 200 | |
|---|
| 201 | gb_assert(file); |
|---|
| 202 | if (GB_is_readablefile(file)) return file; |
|---|
| 203 | if (warn_when_not_found) GB_warningf("Could not find '%s'", file); |
|---|
| 204 | return NULL; |
|---|
| 205 | } |
|---|
| 206 | |
|---|
| 207 | char *GB_lib_file(bool warn_when_not_found, const char *libprefix, const char *filename) { |
|---|
| 208 | // Search a file in '$ARBHOME/lib/libprefix' |
|---|
| 209 | // Return NULL if not found |
|---|
| 210 | return nulldup(GB_existing_file(GB_path_in_ARBLIB(libprefix, filename), warn_when_not_found)); |
|---|
| 211 | } |
|---|
| 212 | |
|---|
| 213 | char *GB_property_file(bool warn_when_not_found, const char *filename) { |
|---|
| 214 | // Search a file in '~/.arb_prop' or its default in '$ARBHOME/lib/arb_default' |
|---|
| 215 | // Return NULL if neighter found |
|---|
| 216 | |
|---|
| 217 | char *result = nulldup(GB_existing_file(GB_path_in_arbprop(filename), warn_when_not_found)); |
|---|
| 218 | if (!result) result = GB_lib_file(warn_when_not_found, "arb_default", filename); |
|---|
| 219 | return result; |
|---|
| 220 | } |
|---|
| 221 | |
|---|
| 222 | void GBS_read_dir(StrArray& names, const char *dir, const char *mask) { |
|---|
| 223 | /* Return names of files in directory 'dir'. |
|---|
| 224 | * Filter through 'mask': |
|---|
| 225 | * - mask == NULL -> return all files |
|---|
| 226 | * - in format '/expr/' -> use regular expression (case sensitive) |
|---|
| 227 | * - else it does a simple string match with wildcards ("?*") |
|---|
| 228 | * |
|---|
| 229 | * Result are inserted into 'names' and 'names' is sorted alphanumerically. |
|---|
| 230 | * Note: 'names' are not cleared, so several calls with the same StrArray get collected. |
|---|
| 231 | * |
|---|
| 232 | * In case of error, result is empty and error is exported. |
|---|
| 233 | * |
|---|
| 234 | * Special case: If 'dir' is the name of a file, return an array with file as only element |
|---|
| 235 | */ |
|---|
| 236 | |
|---|
| 237 | gb_assert(dir); // dir == NULL was allowed before 12/2008, forbidden now! |
|---|
| 238 | |
|---|
| 239 | char *fulldir = strdup(GB_canonical_path(dir)); |
|---|
| 240 | DIR *dirstream = opendir(fulldir); |
|---|
| 241 | |
|---|
| 242 | if (!dirstream) { |
|---|
| 243 | if (GB_is_readablefile(fulldir)) { // fixed: returned true for directories before (4/2012) |
|---|
| 244 | names.put(strdup(fulldir)); |
|---|
| 245 | } |
|---|
| 246 | else { |
|---|
| 247 | // @@@ does too much voodoo here - fix |
|---|
| 248 | char *lslash = strrchr(fulldir, '/'); |
|---|
| 249 | |
|---|
| 250 | if (lslash) { |
|---|
| 251 | lslash[0] = 0; |
|---|
| 252 | char *name = lslash+1; |
|---|
| 253 | if (GB_is_directory(fulldir)) { |
|---|
| 254 | GBS_read_dir(names, fulldir, name); // @@@ concat mask to name? |
|---|
| 255 | } |
|---|
| 256 | lslash[0] = '/'; |
|---|
| 257 | } |
|---|
| 258 | |
|---|
| 259 | if (names.empty()) GB_export_errorf("can't read directory '%s'", fulldir); // @@@ wrong place for error; maybe return error as result |
|---|
| 260 | } |
|---|
| 261 | } |
|---|
| 262 | else { |
|---|
| 263 | if (mask == NULL) mask = "*"; |
|---|
| 264 | |
|---|
| 265 | GBS_string_matcher *matcher = GBS_compile_matcher(mask, GB_MIND_CASE); |
|---|
| 266 | if (matcher) { |
|---|
| 267 | dirent *entry; |
|---|
| 268 | while ((entry = readdir(dirstream)) != 0) { |
|---|
| 269 | const char *name = entry->d_name; |
|---|
| 270 | |
|---|
| 271 | if (name[0] == '.' && (name[1] == 0 || (name[1] == '.' && name[2] == 0))) { |
|---|
| 272 | ; // skip '.' and '..' |
|---|
| 273 | } |
|---|
| 274 | else { |
|---|
| 275 | if (GBS_string_matches_regexp(name, matcher)) { |
|---|
| 276 | const char *full = GB_concat_path(fulldir, name); |
|---|
| 277 | if (!GB_is_directory(full)) { // skip directories |
|---|
| 278 | names.put(strdup(full)); |
|---|
| 279 | } |
|---|
| 280 | } |
|---|
| 281 | } |
|---|
| 282 | } |
|---|
| 283 | |
|---|
| 284 | names.sort(GB_string_comparator, 0); |
|---|
| 285 | |
|---|
| 286 | GBS_free_matcher(matcher); |
|---|
| 287 | } |
|---|
| 288 | |
|---|
| 289 | closedir(dirstream); |
|---|
| 290 | } |
|---|
| 291 | |
|---|
| 292 | free(fulldir); |
|---|
| 293 | } |
|---|
| 294 | |
|---|
| 295 | // -------------------------------------------------------------------------------- |
|---|
| 296 | |
|---|
| 297 | #ifdef UNIT_TESTS |
|---|
| 298 | #include <test_unit.h> |
|---|
| 299 | |
|---|
| 300 | // GB_test_textfile_difflines + GB_test_files_equal are helper functions used |
|---|
| 301 | // by unit tests. |
|---|
| 302 | // |
|---|
| 303 | // @@@ GB_test_textfile_difflines + GB_test_files_equal -> ARBCORE |
|---|
| 304 | |
|---|
| 305 | #define MAX_REGS 13 |
|---|
| 306 | |
|---|
| 307 | class difflineMode : virtual Noncopyable { |
|---|
| 308 | int mode; |
|---|
| 309 | GBS_regex *reg[MAX_REGS]; |
|---|
| 310 | const char *replace[MAX_REGS]; |
|---|
| 311 | int count; |
|---|
| 312 | mutable GB_ERROR error; |
|---|
| 313 | |
|---|
| 314 | static bool is_may; |
|---|
| 315 | |
|---|
| 316 | void add(const char *regEx, GB_CASE case_flag, const char *replacement) { |
|---|
| 317 | if (!error) { |
|---|
| 318 | gb_assert(count<MAX_REGS); |
|---|
| 319 | reg[count] = GBS_compile_regexpr(regEx, case_flag, &error); |
|---|
| 320 | if (!error) { |
|---|
| 321 | replace[count] = replacement; |
|---|
| 322 | count++; |
|---|
| 323 | } |
|---|
| 324 | } |
|---|
| 325 | } |
|---|
| 326 | |
|---|
| 327 | public: |
|---|
| 328 | difflineMode(int mode_) |
|---|
| 329 | : mode(mode_), |
|---|
| 330 | count(0), |
|---|
| 331 | error(NULL) |
|---|
| 332 | { |
|---|
| 333 | memset(reg, 0, sizeof(reg)); |
|---|
| 334 | switch (mode) { |
|---|
| 335 | case 0: break; |
|---|
| 336 | case 1: { |
|---|
| 337 | add("[0-9]{2}:[0-9]{2}:[0-9]{2}", GB_MIND_CASE, "<TIME>"); |
|---|
| 338 | add("(Mon|Tue|Wed|Thu|Fri|Sat|Sun)", GB_IGNORE_CASE, "<DOW>"); |
|---|
| 339 | |
|---|
| 340 | add("(January|February|March|April|May|June|July|August|September|October|November|December)", GB_IGNORE_CASE, "<Month>"); |
|---|
| 341 | add("(Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec)", GB_IGNORE_CASE, "<MON>"); |
|---|
| 342 | |
|---|
| 343 | add("<MON>[ -][0-9]{4}", GB_IGNORE_CASE, "<MON> <YEAR>"); |
|---|
| 344 | add("<Month>[ -][0-9]{4}", GB_IGNORE_CASE, "<Month> <YEAR>"); |
|---|
| 345 | |
|---|
| 346 | add("<TIME>[ -][0-9]{4}", GB_IGNORE_CASE, "<TIME> <YEAR>"); |
|---|
| 347 | |
|---|
| 348 | add("<MON>[ -][0-9 ]?[0-9]", GB_IGNORE_CASE, "<MON> <DAY>"); |
|---|
| 349 | add("<Month>[ -][0-9 ]?[0-9]", GB_IGNORE_CASE, "<Month> <DAY>"); |
|---|
| 350 | |
|---|
| 351 | add("[0-9]{2}[ -]<MON>", GB_IGNORE_CASE, "<DAY> <MON>"); |
|---|
| 352 | add("[0-9]{2}[ -]<Month>", GB_IGNORE_CASE, "<DAY> <Month>"); |
|---|
| 353 | |
|---|
| 354 | add("<DAY>, [0-9]{4}", GB_IGNORE_CASE, "<DAY> <YEAR>"); |
|---|
| 355 | |
|---|
| 356 | if (is_may) { |
|---|
| 357 | // 'May' does not differ between short/long monthname |
|---|
| 358 | // -> use less accurate tests in May |
|---|
| 359 | add("<Month>", GB_MIND_CASE, "<MON>"); |
|---|
| 360 | } |
|---|
| 361 | |
|---|
| 362 | break; |
|---|
| 363 | } |
|---|
| 364 | default: gb_assert(0); break; |
|---|
| 365 | } |
|---|
| 366 | } |
|---|
| 367 | ~difflineMode() { |
|---|
| 368 | for (int i = 0; i<count; ++i) { |
|---|
| 369 | GBS_free_regexpr(reg[i]); |
|---|
| 370 | reg[i] = NULL; |
|---|
| 371 | } |
|---|
| 372 | } |
|---|
| 373 | |
|---|
| 374 | const char *get_error() const { return error; } |
|---|
| 375 | int get_count() const { return count; } |
|---|
| 376 | |
|---|
| 377 | void replaceAll(char*& str) const { |
|---|
| 378 | for (int i = 0; i<count; ++i) { |
|---|
| 379 | size_t matchlen; |
|---|
| 380 | const char *matched = GBS_regmatch_compiled(str, reg[i], &matchlen); |
|---|
| 381 | |
|---|
| 382 | if (matched) { |
|---|
| 383 | char *prefix = GB_strpartdup(str, matched-1); |
|---|
| 384 | const char *suffix = matched+matchlen; |
|---|
| 385 | |
|---|
| 386 | freeset(str, GBS_global_string_copy("%s%s%s", prefix, replace[i], suffix)); |
|---|
| 387 | free(prefix); |
|---|
| 388 | } |
|---|
| 389 | } |
|---|
| 390 | } |
|---|
| 391 | void replaceAll(char*& str1, char*& str2) const { replaceAll(str1); replaceAll(str2); } |
|---|
| 392 | }; |
|---|
| 393 | |
|---|
| 394 | bool difflineMode::is_may = strstr(GB_date_string(), "May") != 0; |
|---|
| 395 | |
|---|
| 396 | static void cutEOL(char *s) { |
|---|
| 397 | char *lf = strchr(s, '\n'); |
|---|
| 398 | if (lf) lf[0] = 0; |
|---|
| 399 | } |
|---|
| 400 | |
|---|
| 401 | static bool test_accept_diff_lines(const char *line1, const char *line2, const difflineMode& mode) { |
|---|
| 402 | if (*line1++ != '-') return false; |
|---|
| 403 | if (*line2++ != '+') return false; |
|---|
| 404 | |
|---|
| 405 | char *dup1 = strdup(line1); |
|---|
| 406 | char *dup2 = strdup(line2); |
|---|
| 407 | |
|---|
| 408 | cutEOL(dup1); // side-effect: accepts missing trailing newline |
|---|
| 409 | cutEOL(dup2); |
|---|
| 410 | |
|---|
| 411 | mode.replaceAll(dup1, dup2); |
|---|
| 412 | |
|---|
| 413 | bool equalNow = strcmp(dup1, dup2) == 0; |
|---|
| 414 | #if defined(DEBUG) |
|---|
| 415 | // printf("dup1='%s'\ndup2='%s'\n", dup1, dup2); // uncomment this line to trace replaces |
|---|
| 416 | #endif // DEBUG |
|---|
| 417 | |
|---|
| 418 | free(dup2); |
|---|
| 419 | free(dup1); |
|---|
| 420 | |
|---|
| 421 | return equalNow; |
|---|
| 422 | } |
|---|
| 423 | |
|---|
| 424 | class DiffLines { |
|---|
| 425 | typedef std::list<std::string> Lines; |
|---|
| 426 | typedef Lines::iterator LinesIter; |
|---|
| 427 | typedef Lines::const_iterator LinesCIter; |
|---|
| 428 | |
|---|
| 429 | Lines added_lines; |
|---|
| 430 | Lines deleted_lines; |
|---|
| 431 | |
|---|
| 432 | LinesIter added_last_checked; |
|---|
| 433 | LinesIter deleted_last_checked; |
|---|
| 434 | |
|---|
| 435 | static LinesIter next(LinesIter iter) { advance(iter, 1); return iter; } |
|---|
| 436 | static LinesIter last(Lines& lines) { return (++lines.rbegin()).base(); } |
|---|
| 437 | |
|---|
| 438 | void set_checked() { |
|---|
| 439 | added_last_checked = last(added_lines); |
|---|
| 440 | deleted_last_checked = last(deleted_lines); |
|---|
| 441 | } |
|---|
| 442 | |
|---|
| 443 | public: |
|---|
| 444 | DiffLines() { set_checked(); } |
|---|
| 445 | |
|---|
| 446 | bool add(const char *diffline) { |
|---|
| 447 | bool did_add = true; |
|---|
| 448 | switch (diffline[0]) { |
|---|
| 449 | case '-': deleted_lines.push_back(diffline); break; |
|---|
| 450 | case '+': added_lines.push_back(diffline); break; |
|---|
| 451 | default : did_add = false; break; |
|---|
| 452 | } |
|---|
| 453 | // fputs(diffline, stdout); // uncomment to show all difflines |
|---|
| 454 | return did_add; |
|---|
| 455 | } |
|---|
| 456 | |
|---|
| 457 | int added() const { return added_lines.size(); } |
|---|
| 458 | int deleted() const { return deleted_lines.size(); } |
|---|
| 459 | |
|---|
| 460 | void remove_accepted_lines(const difflineMode& mode) { |
|---|
| 461 | LinesIter d = next(deleted_last_checked); |
|---|
| 462 | LinesIter dEnd = deleted_lines.end(); |
|---|
| 463 | LinesIter a = next(added_last_checked); |
|---|
| 464 | LinesIter aEnd = added_lines.end(); |
|---|
| 465 | |
|---|
| 466 | |
|---|
| 467 | while (d != dEnd && a != aEnd) { |
|---|
| 468 | if (test_accept_diff_lines(d->c_str(), a->c_str(), mode)) { |
|---|
| 469 | d = deleted_lines.erase(d); |
|---|
| 470 | a = added_lines.erase(a); |
|---|
| 471 | } |
|---|
| 472 | else { |
|---|
| 473 | ++d; |
|---|
| 474 | ++a; |
|---|
| 475 | } |
|---|
| 476 | } |
|---|
| 477 | set_checked(); |
|---|
| 478 | } |
|---|
| 479 | |
|---|
| 480 | void print_from(FILE *out, LinesCIter a, LinesCIter d) const { |
|---|
| 481 | LinesCIter dEnd = deleted_lines.end(); |
|---|
| 482 | LinesCIter aEnd = added_lines.end(); |
|---|
| 483 | |
|---|
| 484 | while (d != dEnd && a != aEnd) { |
|---|
| 485 | fputs(d->c_str(), out); ++d; |
|---|
| 486 | fputs(a->c_str(), out); ++a; |
|---|
| 487 | } |
|---|
| 488 | while (d != dEnd) { fputs(d->c_str(), out); ++d; } |
|---|
| 489 | while (a != aEnd) { fputs(a->c_str(), out); ++a; } |
|---|
| 490 | } |
|---|
| 491 | |
|---|
| 492 | void print(FILE *out) const { |
|---|
| 493 | LinesCIter d = deleted_lines.begin(); |
|---|
| 494 | LinesCIter a = added_lines.begin(); |
|---|
| 495 | print_from(out, a, d); |
|---|
| 496 | } |
|---|
| 497 | }; |
|---|
| 498 | |
|---|
| 499 | |
|---|
| 500 | bool GB_test_textfile_difflines(const char *file1, const char *file2, int expected_difflines, int special_mode) { |
|---|
| 501 | // special_mode: 0 = none, 1 = accept date and time changes as equal |
|---|
| 502 | const char *error = NULL; |
|---|
| 503 | |
|---|
| 504 | if (!GB_is_regularfile(file1)) error = GBS_global_string("No such file '%s'", file1); |
|---|
| 505 | else if (!GB_is_regularfile(file2)) error = GBS_global_string("No such file '%s'", file2); |
|---|
| 506 | else { |
|---|
| 507 | char *cmd = GBS_global_string_copy("/usr/bin/diff --unified %s %s", file1, file2); |
|---|
| 508 | FILE *diffout = popen(cmd, "r"); |
|---|
| 509 | |
|---|
| 510 | if (diffout) { |
|---|
| 511 | #define BUFSIZE 5000 |
|---|
| 512 | char *buffer = (char*)malloc(BUFSIZE); |
|---|
| 513 | bool inHunk = false; |
|---|
| 514 | DiffLines diff_lines; |
|---|
| 515 | difflineMode mode(special_mode); |
|---|
| 516 | |
|---|
| 517 | TEST_ASSERT_NO_ERROR(mode.get_error()); |
|---|
| 518 | |
|---|
| 519 | while (!feof(diffout)) { |
|---|
| 520 | char *line = fgets(buffer, BUFSIZE, diffout); |
|---|
| 521 | if (!line) break; |
|---|
| 522 | |
|---|
| 523 | #if defined(ASSERTION_USED) |
|---|
| 524 | size_t len = strlen(line); |
|---|
| 525 | gb_assert(line && len<(BUFSIZE-1)); // increase BUFSIZE |
|---|
| 526 | #endif |
|---|
| 527 | |
|---|
| 528 | bool remove_now = true; |
|---|
| 529 | if (strncmp(line, "@@", 2) == 0) inHunk = true; |
|---|
| 530 | else if (!inHunk && strncmp(line, "Index: ", 7) == 0) inHunk = false; |
|---|
| 531 | else if (inHunk) { |
|---|
| 532 | remove_now = !diff_lines.add(line); |
|---|
| 533 | } |
|---|
| 534 | |
|---|
| 535 | if (remove_now) diff_lines.remove_accepted_lines(mode); |
|---|
| 536 | } |
|---|
| 537 | |
|---|
| 538 | diff_lines.remove_accepted_lines(mode); |
|---|
| 539 | |
|---|
| 540 | int added = diff_lines.added(); |
|---|
| 541 | int deleted = diff_lines.deleted(); |
|---|
| 542 | |
|---|
| 543 | if (added != deleted) { |
|---|
| 544 | error = GBS_global_string("added lines (=%i) differ from deleted lines(=%i)", added, deleted); |
|---|
| 545 | } |
|---|
| 546 | else if (added != expected_difflines) { |
|---|
| 547 | error = GBS_global_string("files differ in %i lines (expected=%i)", added, expected_difflines); |
|---|
| 548 | } |
|---|
| 549 | if (error) { |
|---|
| 550 | fputs("Different lines:\n", stdout); |
|---|
| 551 | diff_lines.print(stdout); |
|---|
| 552 | fputc('\n', stdout); |
|---|
| 553 | } |
|---|
| 554 | |
|---|
| 555 | free(buffer); |
|---|
| 556 | IF_ASSERTION_USED(int err =) pclose(diffout); |
|---|
| 557 | gb_assert(err != -1); |
|---|
| 558 | #undef BUFSIZE |
|---|
| 559 | } |
|---|
| 560 | else { |
|---|
| 561 | error = GBS_global_string("failed to run diff (%s)", cmd); |
|---|
| 562 | } |
|---|
| 563 | free(cmd); |
|---|
| 564 | } |
|---|
| 565 | // return result; |
|---|
| 566 | if (error) printf("GB_test_textfile_difflines(%s, %s) fails: %s\n", file1, file2, error); |
|---|
| 567 | return !error; |
|---|
| 568 | } |
|---|
| 569 | |
|---|
| 570 | size_t GB_test_mem_equal(const unsigned char *buf1, const unsigned char *buf2, size_t common) { // used by unit tests |
|---|
| 571 | size_t equal_bytes; |
|---|
| 572 | if (memcmp(buf1, buf2, common) == 0) { |
|---|
| 573 | equal_bytes = common; |
|---|
| 574 | } |
|---|
| 575 | else { |
|---|
| 576 | equal_bytes = 0; |
|---|
| 577 | size_t x = 0; // position inside memory |
|---|
| 578 | while (buf1[x] == buf2[x]) { |
|---|
| 579 | x++; |
|---|
| 580 | equal_bytes++; |
|---|
| 581 | } |
|---|
| 582 | |
|---|
| 583 | const size_t DUMP = 7; |
|---|
| 584 | size_t y1 = x >= DUMP ? x-DUMP : 0; |
|---|
| 585 | size_t y2 = (x+DUMP)>common ? common : (x+DUMP); |
|---|
| 586 | size_t blockstart = equal_bytes-x; |
|---|
| 587 | |
|---|
| 588 | for (size_t y = y1; y <= y2; y++) { |
|---|
| 589 | fprintf(stderr, "[0x%04zx]", blockstart+y); |
|---|
| 590 | arb_test::print_pair(buf1[y], buf2[y]); |
|---|
| 591 | fputc(' ', stderr); |
|---|
| 592 | arb_test::print_hex_pair(buf1[y], buf2[y]); |
|---|
| 593 | if (x == y) fputs(" <- diff", stderr); |
|---|
| 594 | fputc('\n', stderr); |
|---|
| 595 | } |
|---|
| 596 | if (y2 == common) { |
|---|
| 597 | fputs("[end of block - truncated]\n", stderr); |
|---|
| 598 | } |
|---|
| 599 | } |
|---|
| 600 | return equal_bytes; |
|---|
| 601 | } |
|---|
| 602 | |
|---|
| 603 | bool GB_test_files_equal(const char *file1, const char *file2) { |
|---|
| 604 | const char *error = NULL; |
|---|
| 605 | FILE *fp1 = fopen(file1, "rb"); |
|---|
| 606 | |
|---|
| 607 | if (!fp1) { |
|---|
| 608 | printf("can't open '%s'\n", file1); |
|---|
| 609 | error = "i/o error"; |
|---|
| 610 | } |
|---|
| 611 | else { |
|---|
| 612 | FILE *fp2 = fopen(file2, "rb"); |
|---|
| 613 | if (!fp2) { |
|---|
| 614 | printf("can't open '%s'\n", file2); |
|---|
| 615 | error = "i/o error"; |
|---|
| 616 | } |
|---|
| 617 | else { |
|---|
| 618 | const int BLOCKSIZE = 4096; |
|---|
| 619 | unsigned char *buf1 = (unsigned char*)malloc(BLOCKSIZE); |
|---|
| 620 | unsigned char *buf2 = (unsigned char*)malloc(BLOCKSIZE); |
|---|
| 621 | int equal_bytes = 0; |
|---|
| 622 | |
|---|
| 623 | while (!error) { |
|---|
| 624 | int read1 = fread(buf1, 1, BLOCKSIZE, fp1); |
|---|
| 625 | int read2 = fread(buf2, 1, BLOCKSIZE, fp2); |
|---|
| 626 | size_t common = read1<read2 ? read1 : read2; |
|---|
| 627 | |
|---|
| 628 | if (!common) { |
|---|
| 629 | if (read1 != read2) error = "filesize differs"; |
|---|
| 630 | break; |
|---|
| 631 | } |
|---|
| 632 | |
|---|
| 633 | size_t thiseq = GB_test_mem_equal(buf1, buf2, common); |
|---|
| 634 | if (thiseq != common) { |
|---|
| 635 | error = "content differs"; |
|---|
| 636 | } |
|---|
| 637 | equal_bytes += thiseq; |
|---|
| 638 | } |
|---|
| 639 | |
|---|
| 640 | if (error) printf("test_files_equal: equal_bytes=%i\n", equal_bytes); |
|---|
| 641 | gb_assert(error || equal_bytes); // comparing empty files is nonsense |
|---|
| 642 | |
|---|
| 643 | free(buf2); |
|---|
| 644 | free(buf1); |
|---|
| 645 | fclose(fp2); |
|---|
| 646 | } |
|---|
| 647 | fclose(fp1); |
|---|
| 648 | } |
|---|
| 649 | |
|---|
| 650 | if (error) printf("test_files_equal(%s, %s) fails: %s\n", file1, file2, error); |
|---|
| 651 | return !error; |
|---|
| 652 | } |
|---|
| 653 | |
|---|
| 654 | void TEST_diff_files() { |
|---|
| 655 | const char *file = "diff/base.input"; |
|---|
| 656 | const char *file_swapped = "diff/swapped.input"; |
|---|
| 657 | const char *file_date_swapped = "diff/date_swapped.input"; |
|---|
| 658 | const char *file_date_changed = "diff/date_changed.input"; |
|---|
| 659 | |
|---|
| 660 | TEST_ASSERT(GB_test_textfile_difflines(file, file, 0, 0)); // check identity |
|---|
| 661 | |
|---|
| 662 | // check if swapped lines are detected properly |
|---|
| 663 | TEST_ASSERT(GB_test_textfile_difflines(file, file_swapped, 1, 0)); |
|---|
| 664 | TEST_ASSERT(GB_test_textfile_difflines(file, file_swapped, 1, 1)); |
|---|
| 665 | TEST_ASSERT(GB_test_textfile_difflines(file, file_date_swapped, 3, 0)); |
|---|
| 666 | TEST_ASSERT(GB_test_textfile_difflines(file, file_date_swapped, 3, 1)); |
|---|
| 667 | |
|---|
| 668 | TEST_ASSERT(GB_test_textfile_difflines(file, file_date_changed, 0, 1)); |
|---|
| 669 | TEST_ASSERT(GB_test_textfile_difflines(file, file_date_changed, 6, 0)); |
|---|
| 670 | |
|---|
| 671 | TEST_ASSERT(GB_test_textfile_difflines(file_date_swapped, file_date_changed, 6, 0)); |
|---|
| 672 | TEST_ASSERT(GB_test_textfile_difflines(file_date_swapped, file_date_changed, 0, 1)); |
|---|
| 673 | } |
|---|
| 674 | |
|---|
| 675 | static char *remove_path(const char *fullname, void *cl_path) { |
|---|
| 676 | const char *path = (const char *)cl_path; |
|---|
| 677 | return strdup(fullname+(ARB_strscmp(fullname, path) == 0 ? strlen(path) : 0)); |
|---|
| 678 | } |
|---|
| 679 | |
|---|
| 680 | static void GBT_transform_names(StrArray& dest, const StrArray& source, char *transform(const char *, void *), void *client_data) { |
|---|
| 681 | for (int i = 0; source[i]; ++i) dest.put(transform(source[i], client_data)); |
|---|
| 682 | } |
|---|
| 683 | |
|---|
| 684 | #define TEST_JOINED_DIR_CONTENT_EQUALS(subdir,mask,expected) do { \ |
|---|
| 685 | char *fulldir = strdup(GB_path_in_ARBHOME(subdir)); \ |
|---|
| 686 | StrArray contents; \ |
|---|
| 687 | GBS_read_dir(contents, fulldir, mask); \ |
|---|
| 688 | StrArray contents_no_path; \ |
|---|
| 689 | GBT_transform_names(contents_no_path, contents, remove_path, (void*)fulldir); \ |
|---|
| 690 | char *joined = GBT_join_names(contents_no_path, '!'); \ |
|---|
| 691 | TEST_ASSERT_EQUAL(joined, expected); \ |
|---|
| 692 | free(joined); \ |
|---|
| 693 | free(fulldir); \ |
|---|
| 694 | } while(0) |
|---|
| 695 | |
|---|
| 696 | void TEST_GBS_read_dir() { |
|---|
| 697 | TEST_JOINED_DIR_CONTENT_EQUALS("GDE/CLUSTAL", "*.c", "/amenu.c!/clustalv.c!/gcgcheck.c!/myers.c!/sequence.c!/showpair.c!/trees.c!/upgma.c!/util.c"); |
|---|
| 698 | TEST_JOINED_DIR_CONTENT_EQUALS("GDE/CLUSTAL", "/s.*\\.c/", "/clustalv.c!/myers.c!/sequence.c!/showpair.c!/trees.c"); |
|---|
| 699 | TEST_JOINED_DIR_CONTENT_EQUALS("GDE", NULL, "/Makefile!/README"); |
|---|
| 700 | TEST_JOINED_DIR_CONTENT_EQUALS("GDE", "*", "/Makefile!/README"); |
|---|
| 701 | } |
|---|
| 702 | |
|---|
| 703 | void TEST_find_file() { |
|---|
| 704 | TEST_ASSERT_EQUAL(GB_existing_file("min_ascii.arb", false), "min_ascii.arb"); |
|---|
| 705 | TEST_ASSERT_NULL(GB_existing_file("nosuchfile", false)); |
|---|
| 706 | |
|---|
| 707 | char *tcporg = GB_lib_file(false, "", "arb_tcp_org.dat"); |
|---|
| 708 | TEST_ASSERT_EQUAL(tcporg, GB_path_in_ARBHOME("lib/arb_tcp_org.dat")); |
|---|
| 709 | TEST_ASSERT_NULL(GB_lib_file(true, "bla", "blub")); |
|---|
| 710 | free(tcporg); |
|---|
| 711 | |
|---|
| 712 | char *status = GB_property_file(false, "status.arb"); |
|---|
| 713 | TEST_ASSERT_EQUAL(status, GB_path_in_ARBHOME("lib/arb_default/status.arb")); |
|---|
| 714 | TEST_ASSERT_NULL(GB_property_file(true, "undhepp")); |
|---|
| 715 | free(status); |
|---|
| 716 | } |
|---|
| 717 | |
|---|
| 718 | // -------------------------------------------------------------------------------- |
|---|
| 719 | // tests for global code included from central ARB headers (located in $ARBHOME/TEMPLATES) |
|---|
| 720 | #if defined(WARN_TODO) |
|---|
| 721 | #warning move to ARB_CORE library |
|---|
| 722 | #endif |
|---|
| 723 | |
|---|
| 724 | // gcc reports: "warning: logical 'or' of collectively exhaustive tests is always true" |
|---|
| 725 | // for 'implicated(any, any)'. True, obviously. Nevertheless annoying. |
|---|
| 726 | #pragma GCC diagnostic ignored "-Wlogical-op" |
|---|
| 727 | |
|---|
| 728 | void TEST_logic() { |
|---|
| 729 | #define FOR_ANY_BOOL(name) for (int name = 0; name<2; ++name) |
|---|
| 730 | |
|---|
| 731 | TEST_ASSERT(implicated(true, true)); |
|---|
| 732 | // for (int any = 0; any<2; ++any) { |
|---|
| 733 | FOR_ANY_BOOL(any) { |
|---|
| 734 | TEST_ASSERT(implicated(false, any)); // "..aus Falschem folgt Beliebiges.." |
|---|
| 735 | TEST_ASSERT(implicated(any, any)); |
|---|
| 736 | |
|---|
| 737 | TEST_ASSERT( correlated(any, any)); |
|---|
| 738 | TEST_ASSERT(!correlated(any, !any)); |
|---|
| 739 | TEST_ASSERT(!contradicted(any, any)); |
|---|
| 740 | TEST_ASSERT( contradicted(any, !any)); |
|---|
| 741 | } |
|---|
| 742 | |
|---|
| 743 | TEST_ASSERT(correlated(false, false)); |
|---|
| 744 | |
|---|
| 745 | TEST_ASSERT(contradicted(true, false)); |
|---|
| 746 | TEST_ASSERT(contradicted(false, true)); |
|---|
| 747 | |
|---|
| 748 | FOR_ANY_BOOL(kermitIsFrog) { |
|---|
| 749 | FOR_ANY_BOOL(kermitIsGreen) { |
|---|
| 750 | bool allFrogsAreGreen = implicated(kermitIsFrog, kermitIsGreen); |
|---|
| 751 | bool onlyFrogsAreGreen = implicated(kermitIsGreen, kermitIsFrog); |
|---|
| 752 | |
|---|
| 753 | TEST_ASSERT(implicated( kermitIsFrog && allFrogsAreGreen, kermitIsGreen)); |
|---|
| 754 | TEST_ASSERT(implicated(!kermitIsGreen && allFrogsAreGreen, !kermitIsFrog)); |
|---|
| 755 | |
|---|
| 756 | TEST_ASSERT(implicated( kermitIsGreen && onlyFrogsAreGreen, kermitIsFrog)); |
|---|
| 757 | TEST_ASSERT(implicated(!kermitIsFrog && onlyFrogsAreGreen, !kermitIsGreen)); |
|---|
| 758 | |
|---|
| 759 | TEST_ASSERT(implicated(kermitIsFrog && !kermitIsGreen, !allFrogsAreGreen)); |
|---|
| 760 | TEST_ASSERT(implicated(kermitIsGreen && !kermitIsFrog, !onlyFrogsAreGreen)); |
|---|
| 761 | |
|---|
| 762 | TEST_ASSERT(correlated( |
|---|
| 763 | correlated(kermitIsGreen, kermitIsFrog), |
|---|
| 764 | allFrogsAreGreen && onlyFrogsAreGreen |
|---|
| 765 | )); |
|---|
| 766 | } |
|---|
| 767 | } |
|---|
| 768 | } |
|---|
| 769 | |
|---|
| 770 | |
|---|
| 771 | #endif // UNIT_TESTS |
|---|