1 /* xbt_str.c - various helping functions to deal with strings */
3 /* Copyright (c) 2007, 2008, 2009, 2010. The SimGrid Team.
4 * All rights reserved. */
6 /* This program is free software; you can redistribute it and/or modify it
7 * under the terms of the license (GNU LGPL) which comes with this package. */
9 #define DJB2_HASH_FUNCTION
10 //#define FNV_HASH_FUNCTION
14 #include "xbt/sysdep.h"
15 #include "xbt/str.h" /* headers of these functions */
16 #include "xbt/strbuff.h"
17 #include "xbt/matrix.h" /* for the diff */
19 /** @brief Strip whitespace (or other characters) from the end of a string.
21 * Strips the whitespaces from the end of s.
22 * By default (when char_list=NULL), these characters get stripped:
24 * - " " (ASCII 32 (0x20)) space.
25 * - "\t" (ASCII 9 (0x09)) tab.
26 * - "\n" (ASCII 10 (0x0A)) line feed.
27 * - "\r" (ASCII 13 (0x0D)) carriage return.
28 * - "\0" (ASCII 0 (0x00)) NULL.
29 * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
31 * @param s The string to strip. Modified in place.
32 * @param char_list A string which contains the characters you want to strip.
35 void xbt_str_rtrim(char *s, const char *char_list)
38 const char *__char_list = " \t\n\r\x0B";
39 char white_char[256] = { 1, 0 };
45 while (*__char_list) {
46 white_char[(unsigned char) *__char_list++] = 1;
50 white_char[(unsigned char) *char_list++] = 1;
57 while ((cur >= s) && white_char[(unsigned char) *cur])
63 /** @brief Strip whitespace (or other characters) from the beginning of a string.
65 * Strips the whitespaces from the begining of s.
66 * By default (when char_list=NULL), these characters get stripped:
68 * - " " (ASCII 32 (0x20)) space.
69 * - "\t" (ASCII 9 (0x09)) tab.
70 * - "\n" (ASCII 10 (0x0A)) line feed.
71 * - "\r" (ASCII 13 (0x0D)) carriage return.
72 * - "\0" (ASCII 0 (0x00)) NULL.
73 * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
75 * @param s The string to strip. Modified in place.
76 * @param char_list A string which contains the characters you want to strip.
79 void xbt_str_ltrim(char *s, const char *char_list)
82 const char *__char_list = " \t\n\r\x0B";
83 char white_char[256] = { 1, 0 };
89 while (*__char_list) {
90 white_char[(unsigned char) *__char_list++] = 1;
94 white_char[(unsigned char) *char_list++] = 1;
98 while (*cur && white_char[(unsigned char) *cur])
101 memmove(s, cur, strlen(cur) + 1);
104 /** @brief Strip whitespace (or other characters) from the end and the begining of a string.
106 * Strips the whitespaces from both the beginning and the end of s.
107 * By default (when char_list=NULL), these characters get stripped:
109 * - " " (ASCII 32 (0x20)) space.
110 * - "\t" (ASCII 9 (0x09)) tab.
111 * - "\n" (ASCII 10 (0x0A)) line feed.
112 * - "\r" (ASCII 13 (0x0D)) carriage return.
113 * - "\0" (ASCII 0 (0x00)) NULL.
114 * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
116 * @param s The string to strip.
117 * @param char_list A string which contains the characters you want to strip.
120 void xbt_str_trim(char *s, const char *char_list)
126 xbt_str_rtrim(s, char_list);
127 xbt_str_ltrim(s, char_list);
130 /** @brief Replace double whitespaces (but no other characters) from the string.
132 * The function modifies the string so that each time that several spaces appear,
133 * they are replaced by a single space. It will only do so for spaces (ASCII 32, 0x20).
135 * @param s The string to strip. Modified in place.
138 void xbt_str_strip_spaces(char *s)
174 /** @brief Substitutes a char for another in a string
176 * @param str the string to modify
177 * @param from char to search
178 * @param to char to put instead
179 * @param occurence number of changes to do (=0 means all)
181 void xbt_str_subst(char *str, char from, char to, int occurence)
195 /** @brief Replaces a set of variables by their values
197 * @param str The input of the replacement process
198 * @param patterns The changes to apply
199 * @return The string modified
201 * Both '$toto' and '${toto}' are valid (and the two writing are equivalent).
203 * If the variable name contains spaces, use the brace version (ie, ${toto tutu})
205 * You can provide a default value to use if the variable is not set in the dict by using
206 * '${var:=default}' or '${var:-default}'. These two forms are equivalent, even if they
207 * shouldn't to respect the shell standard (:= form should set the value in the dict,
208 * but does not) (BUG).
211 char *xbt_str_varsubst(const char *str, xbt_dict_t patterns)
213 xbt_strbuff_t buff = xbt_strbuff_new_from(str);
215 xbt_strbuff_varsubst(buff, patterns);
217 xbt_strbuff_free_container(buff);
222 /** @brief Splits a string into a dynar of strings
224 * @param s: the string to split
225 * @param sep: a string of all chars to consider as separator.
227 * By default (with sep=NULL), these characters are used as separator:
229 * - " " (ASCII 32 (0x20)) space.
230 * - "\t" (ASCII 9 (0x09)) tab.
231 * - "\n" (ASCII 10 (0x0A)) line feed.
232 * - "\r" (ASCII 13 (0x0D)) carriage return.
233 * - "\0" (ASCII 0 (0x00)) NULL.
234 * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
237 xbt_dynar_t xbt_str_split(const char *s, const char *sep)
239 xbt_dynar_t res = xbt_dynar_new(sizeof(char *), &xbt_free_ref);
242 const char *sep_dflt = " \t\n\r\x0B";
243 char is_sep[256] = { 1, 0 };
245 /* check what are the separators */
246 memset(is_sep, 0, sizeof(is_sep));
249 is_sep[(unsigned char) *sep_dflt++] = 1;
252 is_sep[(unsigned char) *sep++] = 1;
254 is_sep[0] = 1; /* End of string is also separator */
265 while (!is_sep[(unsigned char) *q]) {
271 topush = xbt_malloc(q - p + 1);
272 memcpy(topush, p, q - p);
273 topush[q - p] = '\0';
274 xbt_dynar_push(res, &topush);
282 * \brief This functions splits a string after using another string as separator
283 * For example A!!B!!C splitted after !! will return the dynar {A,B,C}
284 * \return An array of dynars containing the string tokens
286 xbt_dynar_t xbt_str_split_str(const char *s, const char *sep)
288 xbt_dynar_t res = xbt_dynar_new(sizeof(char *), &xbt_free_ref);
297 if (sep[0] == '\0') {
299 xbt_dynar_push(res, &s);
306 //get the start of the first occurence of the substring
308 //if substring was not found add the entire string
311 to_push = malloc(v + 1);
312 memcpy(to_push, p, v);
314 xbt_dynar_push(res, &to_push);
318 to_push = malloc(q - p + 1);
319 memcpy(to_push, p, q - p);
320 //add string terminator
321 to_push[q - p] = '\0';
322 xbt_dynar_push(res, &to_push);
329 /** @brief Just like @ref xbt_str_split_quoted (Splits a string into a dynar of strings), but without memory allocation
331 * The string passed as argument must be writable (not const)
332 * The elements of the dynar are just parts of the string passed as argument.
334 * To free the structure constructed by this function, free the first element and free the dynar:
336 * free(xbt_dynar_get_ptr(dynar,0));
337 * xbt_dynar_free(&dynar);
339 xbt_dynar_t xbt_str_split_quoted_in_place(char *s) {
340 xbt_dynar_t res = xbt_dynar_new(sizeof(char *), NULL);
341 char *beg, *end; /* pointers around the parsed chunk */
342 int in_simple_quote = 0, in_double_quote = 0;
344 int ctn = 0; /* Got something in this block */
351 /* do not trim leading spaces: caller responsability to clean his cruft */
360 /* Protected char; move it closer */
361 memmove(end, end + 1, strlen(end));
363 THROWF(arg_error, 0, "String ends with \\");
364 end++; /* Pass the protected char */
369 if (!in_double_quote) {
370 in_simple_quote = !in_simple_quote;
371 memmove(end, end + 1, strlen(end));
373 /* simple quote protected by double ones */
379 if (!in_simple_quote) {
380 in_double_quote = !in_double_quote;
381 memmove(end, end + 1, strlen(end));
383 /* double quote protected by simple ones */
392 if (*end == '\0' && (in_simple_quote || in_double_quote)) {
394 "End of string found while searching for %c in %s",
395 (in_simple_quote ? '\'' : '"'), s);
397 if (in_simple_quote || in_double_quote) {
405 /* Found a separator. Push the string if contains something */
406 xbt_dynar_push(res, &beg);
414 /* trim within the string, manually to speed things up */
429 /** @brief Splits a string into a dynar of strings, taking quotes into account
431 * It basically does the same argument separation than the shell, where white
432 * spaces can be escaped and where arguments are never split within a
434 * Several subsequent spaces are ignored (unless within quotes, of course).
435 * You may want to trim the input string, if you want to avoid empty entries
439 xbt_dynar_t xbt_str_split_quoted(const char *s)
441 xbt_dynar_t res = xbt_dynar_new(sizeof(char *), &xbt_free_ref);
443 char *str_to_free; /* we have to copy the string before, to handle backslashes */
449 str_to_free = xbt_strdup(s);
451 parsed = xbt_str_split_quoted_in_place(str_to_free);
452 xbt_dynar_foreach(parsed,cursor,p) {
453 char *q=xbt_strdup(p);
454 xbt_dynar_push(res,&q);
457 xbt_dynar_shrink(res, 0);
458 xbt_dynar_free(&parsed);
462 /** @brief Join a set of strings as a single string */
463 char *xbt_str_join(xbt_dynar_t dyn, const char *sep)
465 int len = 1, dyn_len = xbt_dynar_length(dyn);
471 return xbt_strdup("");
473 /* compute the length */
474 xbt_dynar_foreach(dyn, cpt, cursor) {
475 len += strlen(cursor);
477 len += strlen(sep) * dyn_len;
479 res = xbt_malloc(len);
481 xbt_dynar_foreach(dyn, cpt, cursor) {
482 if ((int) cpt < dyn_len - 1)
483 p += sprintf(p, "%s%s", cursor, sep);
485 p += sprintf(p, "%s", cursor);
489 /** @brief Join a set of strings as a single string
491 * The parameter must be a NULL-terminated array of chars,
492 * just like xbt_dynar_to_array() produces
494 char *xbt_str_join_array(const char *const *strs, const char *sep)
497 int amount_strings=0;
501 if ((!strs) || (!strs[0]))
502 return xbt_strdup("");
504 /* compute the length before malloc */
505 for (i=0;strs[i];i++) {
506 len += strlen(strs[i]);
509 len += strlen(sep) * amount_strings;
512 q = res = xbt_malloc(len);
513 for (i=0;strs[i];i++) {
514 if (i!=0) { // not first loop
515 q += sprintf(q, "%s%s", sep, strs[i]);
517 q += sprintf(q,"%s",strs[i]);
523 #if defined(SIMGRID_NEED_GETLINE) || defined(DOXYGEN)
524 /** @brief Get a single line from the stream (reimplementation of the GNU getline)
526 * This is a redefinition of the GNU getline function, used on platforms where it does not exists.
528 * getline() reads an entire line from stream, storing the address of the buffer
529 * containing the text into *buf. The buffer is null-terminated and includes
530 * the newline character, if one was found.
532 * If *buf is NULL, then getline() will allocate a buffer for storing the line,
533 * which should be freed by the user program. Alternatively, before calling getline(),
534 * *buf can contain a pointer to a malloc()-allocated buffer *n bytes in size. If the buffer
535 * is not large enough to hold the line, getline() resizes it with realloc(), updating *buf and *n
536 * as necessary. In either case, on a successful call, *buf and *n will be updated to
537 * reflect the buffer address and allocated size respectively.
539 long getline(char **buf, size_t * n, FILE * stream)
546 *buf = xbt_malloc(512);
551 return (ssize_t) - 1;
553 for (i = 0; (ch = fgetc(stream)) != EOF; i++) {
556 *buf = xbt_realloc(*buf, *n += 512);
560 if ((*buf)[i] == '\n') {
568 *buf = xbt_realloc(*buf, *n += 1);
575 #endif /* HAVE_GETLINE */
578 * Diff related functions
580 * Implementation of the algorithm described in "An O(NP) Sequence Comparison
581 * Algorithm", by Sun Wu, Udi Manber, Gene Myers, and Webb Miller (Information
582 * Processing Letters 35(6):317-323, 1990), with the linear-space
583 * divide-and-conquer strategy described in "An O(ND) Difference Algorithm and
584 * Its Variations", by Eugene W. Myers (Algorithmica 1:251-266, 1986).
588 int x, y; /* starting coordinates */
589 int len; /* length */
593 void diff_snake(const char *vec_a[], int a0, int len_a,
594 const char *vec_b[], int b0, int len_b,
595 struct subsequence *seqs, int *fp, int k, int limit)
599 int fp_left = fp[k - 1] + 1;
600 int fp_right = fp[k + 1];
601 if (fp_left > fp_right) {
609 if (x + y <= limit) {
614 seqs[k] = seqs[record_seq];
616 while (x < len_a && y < len_b && !strcmp(vec_a[a0 + x], vec_b[b0 + y]))
620 seqs[k].len = x - seqs[k].x;
623 /* Returns the length of a shortest edit script, and a common
624 * subsequence from the middle.
626 static int diff_middle_subsequence(const char *vec_a[], int a0, int len_a,
627 const char *vec_b[], int b0, int len_b,
628 struct subsequence *subseq,
629 struct subsequence *seqs, int *fp)
631 const int delta = len_a - len_b;
632 const int limit = (len_a + len_b) / 2;
645 for (k = kmin; k <= kmax; k++)
649 fp[kmin - p - 1] = fp[kmax + p + 1] = -1;
650 for (k = kmax + p; k > delta; k--)
651 diff_snake(vec_a, a0, len_a, vec_b, b0, len_b, seqs, fp, k, limit);
652 for (k = kmin - p; k <= delta; k++)
653 diff_snake(vec_a, a0, len_a, vec_b, b0, len_b, seqs, fp, k, limit);
654 } while (fp[delta] != len_a);
656 subseq->x = a0 + seqs[delta].x;
657 subseq->y = b0 + seqs[delta].y;
658 subseq->len = seqs[delta].len;
659 return abs(delta) + 2 * p;;
662 /* Finds a longest common subsequence.
663 * Returns its length.
665 static int diff_compute_lcs(const char *vec_a[], int a0, int len_a,
666 const char *vec_b[], int b0, int len_b,
667 xbt_dynar_t common_sequence,
668 struct subsequence *seqs, int *fp)
670 if (len_a > 0 && len_b > 0) {
671 struct subsequence subseq;
672 int ses_len = diff_middle_subsequence(vec_a, a0, len_a, vec_b, b0, len_b,
674 int lcs_len = (len_a + len_b - ses_len) / 2;
677 } else if (ses_len > 1) {
678 int lcs_len1 = subseq.len;
679 if (lcs_len1 < lcs_len)
680 lcs_len1 += diff_compute_lcs(vec_a, a0, subseq.x - a0,
681 vec_b, b0, subseq.y - b0,
682 common_sequence, seqs, fp);
684 xbt_dynar_push(common_sequence, &subseq);
685 if (lcs_len1 < lcs_len) {
686 int u = subseq.x + subseq.len;
687 int v = subseq.y + subseq.len;
688 diff_compute_lcs(vec_a, u, a0 + len_a - u, vec_b, v, b0 + len_b - v,
689 common_sequence, seqs, fp);
692 int len = MIN(len_a, len_b) - subseq.len;
693 if (subseq.x == a0 && subseq.y == b0) {
695 xbt_dynar_push(common_sequence, &subseq);
697 struct subsequence subseq0 = {a0 + len_a - len,
698 b0 + len_b - len, len};
699 xbt_dynar_push(common_sequence, &subseq0);
703 struct subsequence subseq0 = {a0, b0, len};
704 xbt_dynar_push(common_sequence, &subseq0);
707 xbt_dynar_push(common_sequence, &subseq);
716 static int diff_member(const char *s, const char *vec[], int from, int to)
718 for ( ; from < to ; from++)
719 if (!strcmp(s, vec[from]))
724 /* Extract common prefix.
726 static void diff_easy_prefix(const char *vec_a[], int *a0_p, int *len_a_p,
727 const char *vec_b[], int *b0_p, int *len_b_p,
728 xbt_dynar_t common_sequence)
732 int len_a = *len_a_p;
733 int len_b = *len_b_p;
735 while (len_a > 0 && len_b > 0) {
736 struct subsequence subseq = {a0, b0, 0};
737 while (len_a > 0 && len_b > 0 && !strcmp(vec_a[a0], vec_b[b0])) {
743 xbt_dynar_push(common_sequence, &subseq);
744 if (len_a > 0 && len_b > 0 &&
745 !diff_member(vec_a[a0], vec_b, b0 + 1, b0 + len_b)) {
758 /* Extract common suffix.
760 static void diff_easy_suffix(const char *vec_a[], int *a0_p, int *len_a_p,
761 const char *vec_b[], int *b0_p, int *len_b_p,
762 xbt_dynar_t common_suffix)
766 int len_a = *len_a_p;
767 int len_b = *len_b_p;
769 while (len_a > 0 && len_b > 0){
770 struct subsequence subseq;
772 while (len_a > 0 && len_b > 0 &&
773 !strcmp(vec_a[a0 + len_a - 1], vec_b[b0 + len_b - 1])) {
778 if (subseq.len > 0) {
779 subseq.x = a0 + len_a;
780 subseq.y = b0 + len_b;
781 xbt_dynar_push(common_suffix, &subseq);
783 if (len_a > 0 && len_b > 0 &&
784 !diff_member(vec_b[b0 + len_b - 1], vec_a, a0, a0 + len_a - 1)) {
797 /** @brief Compute the unified diff of two strings */
798 char *xbt_str_diff(const char *a, const char *b)
800 xbt_dynar_t da = xbt_str_split(a, "\n");
801 xbt_dynar_t db = xbt_str_split(b, "\n");
802 xbt_dynar_t common_sequence, common_suffix;
804 const char **vec_a, **vec_b;
809 struct subsequence *seqs_base, *seqs;
810 struct subsequence subseq;
816 /* Clean empty lines at the end of da and db */
818 if (len > 0 && a[len - 1] == '\n')
819 xbt_dynar_pop(da, NULL);
821 if (len > 0 && b[len - 1] == '\n')
822 xbt_dynar_pop(db, NULL);
824 /* Various initializations */
825 /* Assume that dynar's content is contiguous */
827 len_a = xbt_dynar_length(da);
828 vec_a = len_a ? xbt_dynar_get_ptr(da, 0) : NULL;
830 len_b = xbt_dynar_length(db);
831 vec_b = len_b ? xbt_dynar_get_ptr(db, 0) : NULL;
832 max = MAX(len_a, len_b) + 1;
833 fp_base = xbt_new(int, 2 * max + 1);
834 fp = fp_base + max; /* indexes in [-max..max] */
835 seqs_base = xbt_new(struct subsequence, 2 * max + 1);
836 seqs = seqs_base + max; /* indexes in [-max..max] */
837 common_sequence = xbt_dynar_new(sizeof(struct subsequence), NULL);
838 common_suffix = xbt_dynar_new(sizeof(struct subsequence), NULL);
840 /* Add a sentinel a the end of the sequence */
844 xbt_dynar_push(common_suffix, &subseq);
846 /* Compute the Longest Common Subsequence */
847 diff_easy_prefix(vec_a, &a0, &len_a, vec_b, &b0, &len_b, common_sequence);
848 diff_easy_suffix(vec_a, &a0, &len_a, vec_b, &b0, &len_b, common_suffix);
849 diff_compute_lcs(vec_a, a0, len_a, vec_b, b0, len_b, common_sequence, seqs, fp);
850 while (!xbt_dynar_is_empty(common_suffix)) {
851 xbt_dynar_pop(common_suffix, &subseq);
852 xbt_dynar_push(common_sequence, &subseq);
855 /* Build a Shortest Edit Script, and the final result */
856 diff = xbt_dynar_new(sizeof(char *), &xbt_free_ref);
859 xbt_dynar_foreach(common_sequence, s, subseq) {
860 while (x < subseq.x) {
861 char *topush = bprintf("- %s", vec_a[x++]);
862 xbt_dynar_push_as(diff, char*, topush);
864 while (y < subseq.y) {
865 char *topush = bprintf("+ %s", vec_b[y++]);
866 xbt_dynar_push_as(diff, char*, topush);
868 while (x < subseq.x + subseq.len) {
869 char *topush = bprintf(" %s", vec_a[x++]);
870 xbt_dynar_push_as(diff, char*, topush);
874 res = xbt_str_join(diff, "\n");
880 xbt_dynar_free(&common_sequence);
881 xbt_dynar_free(&common_suffix);
882 xbt_dynar_free(&diff);
888 /** @brief creates a new string containing what can be read on a fd
891 char *xbt_str_from_file(FILE * file)
893 xbt_strbuff_t buff = xbt_strbuff_new();
896 memset(bread, 0, 1024);
898 while (!feof(file)) {
899 int got = fread(bread, 1, 1023, file);
901 xbt_strbuff_append(buff, bread);
905 xbt_strbuff_free_container(buff);
910 * @brief Returns the hash code of a string.
912 XBT_INLINE unsigned int xbt_dict_hash_ext(const char *str,
916 #ifdef DJB2_HASH_FUNCTION
917 /* fast implementation of djb2 algorithm */
919 register unsigned int hash = 5381;
923 hash = ((hash << 5) + hash) + c; /* hash * 33 + c */
925 # elif defined(FNV_HASH_FUNCTION)
926 register unsigned int hash = 0x811c9dc5;
927 unsigned char *bp = (unsigned char *) str; /* start of buffer */
928 unsigned char *be = bp + str_len; /* beyond end of buffer */
931 /* multiply by the 32 bit FNV magic prime mod 2^32 */
933 (hash << 1) + (hash << 4) + (hash << 7) + (hash << 8) +
936 /* xor the bottom with the current octet */
937 hash ^= (unsigned int) *bp++;
941 register unsigned int hash = 0;
944 hash += (*str) * (*str);
953 * @brief Returns the hash code of a string.
955 XBT_INLINE unsigned int xbt_dict_hash(const char *str)
957 #ifdef DJB2_HASH_FUNCTION
958 /* fast implementation of djb2 algorithm */
960 register unsigned int hash = 5381;
962 while ((c = *str++)) {
963 hash = ((hash << 5) + hash) + c; /* hash * 33 + c */
966 # elif defined(FNV_HASH_FUNCTION)
967 register unsigned int hash = 0x811c9dc5;
970 /* multiply by the 32 bit FNV magic prime mod 2^32 */
972 (hash << 1) + (hash << 4) + (hash << 7) + (hash << 8) +
975 /* xor the bottom with the current byte */
976 hash ^= (unsigned int) *str++;
980 register unsigned int hash = 0;
983 hash += (*str) * (*str);
993 #define mytest(name, input, expected) \
994 xbt_test_add(name); \
995 d=xbt_str_split_quoted(input); \
996 s=xbt_str_join(d,"XXX"); \
997 xbt_test_assert(!strcmp(s,expected),\
998 "Input (%s) leads to (%s) instead of (%s)", \
1003 XBT_TEST_SUITE("xbt_str", "String Handling");
1004 XBT_TEST_UNIT("xbt_str_split_quoted", test_split_quoted, "test the function xbt_str_split_quoted")
1009 mytest("Empty", "", "");
1010 mytest("Basic test", "toto tutu", "totoXXXtutu");
1011 mytest("Useless backslashes", "\\t\\o\\t\\o \\t\\u\\t\\u",
1013 mytest("Protected space", "toto\\ tutu", "toto tutu");
1014 mytest("Several spaces", "toto tutu", "totoXXXtutu");
1015 mytest("LTriming", " toto tatu", "totoXXXtatu");
1016 mytest("Triming", " toto tutu ", "totoXXXtutu");
1017 mytest("Single quotes", "'toto tutu' tata", "toto tutuXXXtata");
1018 mytest("Double quotes", "\"toto tutu\" tata", "toto tutuXXXtata");
1019 mytest("Mixed quotes", "\"toto' 'tutu\" tata", "toto' 'tutuXXXtata");
1020 mytest("Backslashed quotes", "\\'toto tutu\\' tata",
1021 "'totoXXXtutu'XXXtata");
1022 mytest("Backslashed quotes + quotes", "'toto \\'tutu' tata",
1023 "toto 'tutuXXXtata");
1027 #define mytest_str(name, input, separator, expected) \
1028 xbt_test_add(name); \
1029 d=xbt_str_split_str(input, separator); \
1030 s=xbt_str_join(d,"XXX"); \
1031 xbt_test_assert(!strcmp(s,expected),\
1032 "Input (%s) leads to (%s) instead of (%s)", \
1037 XBT_TEST_UNIT("xbt_str_split_str", test_split_str, "test the function xbt_str_split_str")
1042 mytest_str("Empty string and separator", "", "", "");
1043 mytest_str("Empty string", "", "##", "");
1044 mytest_str("Empty separator", "toto", "", "toto");
1045 mytest_str("String with no separator in it", "toto", "##", "toto");
1046 mytest_str("Basic test", "toto##tutu", "##", "totoXXXtutu");
1049 /* Last args are format string and parameters for xbt_test_add */
1050 #define mytest_diff(s1, s2, diff, ...) \
1052 char *mytest_diff_res; \
1053 xbt_test_add(__VA_ARGS__); \
1054 mytest_diff_res = xbt_str_diff(s1, s2); \
1055 xbt_test_assert(!strcmp(mytest_diff_res, diff), \
1056 "Wrong output:\n--- got:\n%s\n--- expected:\n%s\n---", \
1057 mytest_diff_res, diff); \
1058 free(mytest_diff_res); \
1061 XBT_TEST_UNIT("xbt_str_diff", test_diff, "test the function xbt_str_diff")
1066 mytest_diff("a", "a", " a", "1 word, no difference");
1067 mytest_diff("a", "A", "- a\n+ A", "1 word, different");
1068 mytest_diff("a\n", "a\n", " a", "1 line, no difference");
1069 mytest_diff("a\n", "A\n", "- a\n+ A", "1 line, different");
1072 mytest_diff("", "", "", "empty strings");
1073 mytest_diff("", "a", "+ a", "1 word, added");
1074 mytest_diff("a", "", "- a", "1 word, removed");
1075 mytest_diff("", "a\n", "+ a", "1 line, added");
1076 mytest_diff("a\n", "", "- a", "1 line, removed");
1077 mytest_diff("", "a\nb\nc\n", "+ a\n+ b\n+ c", "4 lines, all added");
1078 mytest_diff("a\nb\nc\n", "", "- a\n- b\n- c", "4 lines, all removed");
1081 mytest_diff("\n", "\n", " ", "empty lines");
1082 mytest_diff("", "\n", "+ ", "empty line, added");
1083 mytest_diff("\n", "", "- ", "empty line, removed");
1085 mytest_diff("a", "\na", "+ \n a", "empty line added before word");
1086 mytest_diff("a", "a\n\n", " a\n+ ", "empty line added after word");
1087 mytest_diff("\na", "a", "- \n a", "empty line removed before word");
1088 mytest_diff("a\n\n", "a", " a\n- ", "empty line removed after word");
1090 mytest_diff("a\n", "\na\n", "+ \n a", "empty line added before line");
1091 mytest_diff("a\n", "a\n\n", " a\n+ ", "empty line added after line");
1092 mytest_diff("\na\n", "a\n", "- \n a", "empty line removed before line");
1093 mytest_diff("a\n\n", "a\n", " a\n- ", "empty line removed after line");
1095 mytest_diff("a\nb\nc\nd\n", "\na\nb\nc\nd\n", "+ \n a\n b\n c\n d",
1096 "empty line added before 4 lines");
1097 mytest_diff("a\nb\nc\nd\n", "a\nb\nc\nd\n\n", " a\n b\n c\n d\n+ ",
1098 "empty line added after 4 lines");
1099 mytest_diff("\na\nb\nc\nd\n", "a\nb\nc\nd\n", "- \n a\n b\n c\n d",
1100 "empty line removed before 4 lines");
1101 mytest_diff("a\nb\nc\nd\n\n", "a\nb\nc\nd\n", " a\n b\n c\n d\n- ",
1102 "empty line removed after 4 lines");
1104 /* Missing newline at the end of one of the strings */
1105 mytest_diff("a\n", "a", " a", "1 line, 1 word, no difference");
1106 mytest_diff("a", "a\n", " a", "1 word, 1 line, no difference");
1107 mytest_diff("a\n", "A", "- a\n+ A", "1 line, 1 word, different");
1108 mytest_diff("a", "A\n", "- a\n+ A", "1 word, 1 line, different");
1110 mytest_diff("a\nb\nc\nd", "a\nb\nc\nd\n", " a\n b\n c\n d",
1111 "4 lines, no newline on first");
1112 mytest_diff("a\nb\nc\nd\n", "a\nb\nc\nd", " a\n b\n c\n d",
1113 "4 lines, no newline on second");
1115 /* Four lines, all combinations of differences */
1116 for (i = 0 ; i < (1U << 4) ; i++) {
1119 char res[4 * 8 + 1];
1126 for (/* j */ ; j < 4 && !(i & (1U << j)) ; j++) {
1128 ps += sprintf(ps, "%c\n", "abcd"[j]);
1129 pr += sprintf(pr, " %c\n", "abcd"[j]);
1131 for (k = j ; k < 4 && (i & (1U << k)) ; k++) {
1133 ps += sprintf(ps, "%c\n", "ABCD"[k]);
1134 pr += sprintf(pr, "- %c\n", "abcd"[k]);
1136 for (/* j */ ; j < k ; j++) {
1137 pr += sprintf(pr, "+ %c\n", "ABCD"[j]);
1141 *--pr = '\0'; /* strip last '\n' from expected result */
1142 mytest_diff("a\nb\nc\nd\n", s2, res,
1143 "compare (abcd) with changed (%s)", d2);
1146 /* Subsets of four lines, do not test for empty subset */
1147 for (i = 1 ; i < (1U << 4) ; i++) {
1150 char res[4 * 8 + 1];
1156 for (/* j */ ; j < 4 && (i & (1U << j)) ; j++) {
1158 ps += sprintf(ps, "%c\n", "abcd"[j]);
1159 pr += sprintf(pr, " %c\n", "abcd"[j]);
1161 for (/* j */; j < 4 && !(i & (1U << j)) ; j++) {
1162 pr += sprintf(pr, "- %c\n", "abcd"[j]);
1166 *--pr = '\0'; /* strip last '\n' from expected result */
1167 mytest_diff("a\nb\nc\nd\n", s2, res,
1168 "compare (abcd) with subset (%s)", d2);
1170 for (pr = res ; *pr != '\0' ; pr++)
1173 mytest_diff(s2, "a\nb\nc\nd\n", res,
1174 "compare subset (%s) with (abcd)", d2);
1178 #endif /* SIMGRID_TEST */