/* xbt_str.c - various helping functions to deal with strings */
-/* Copyright (c) 2007, 2008, 2009, 2010. The SimGrid Team.
+/* Copyright (c) 2007-2014. The SimGrid Team.
* All rights reserved. */
/* This program is free software; you can redistribute it and/or modify it
* under the terms of the license (GNU LGPL) which comes with this package. */
-#define DJB2_HASH_FUNCTION
-//#define FNV_HASH_FUNCTION
-
-#include "portable.h"
+#include "src/portable.h"
#include "xbt/misc.h"
#include "xbt/sysdep.h"
#include "xbt/str.h" /* headers of these functions */
* Strips the whitespaces from the end of s.
* By default (when char_list=NULL), these characters get stripped:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*
* @param s The string to strip. Modified in place.
* @param char_list A string which contains the characters you want to strip.
* Strips the whitespaces from the begining of s.
* By default (when char_list=NULL), these characters get stripped:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*
* @param s The string to strip. Modified in place.
* @param char_list A string which contains the characters you want to strip.
* Strips the whitespaces from both the beginning and the end of s.
* By default (when char_list=NULL), these characters get stripped:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*
* @param s The string to strip.
* @param char_list A string which contains the characters you want to strip.
xbt_str_ltrim(s, char_list);
}
-/** @brief Replace double whitespaces (but no other characters) from the string.
- *
- * The function modifies the string so that each time that several spaces appear,
- * they are replaced by a single space. It will only do so for spaces (ASCII 32, 0x20).
- *
- * @param s The string to strip. Modified in place.
- *
- */
-void xbt_str_strip_spaces(char *s)
-{
- char *p = s;
- int e = 0;
-
- if (!s)
- return;
-
- while (1) {
- if (!*p)
- goto end;
-
- if (*p != ' ')
- break;
-
- p++;
- }
-
- e = 1;
-
- do {
- if (e)
- *s++ = *p;
-
- if (!*++p)
- goto end;
-
- if (e ^ (*p != ' '))
- if ((e = !e))
- *s++ = ' ';
- } while (1);
-
-end:
- *s = '\0';
-}
-
/** @brief Substitutes a char for another in a string
*
* @param str the string to modify
*
* By default (with sep=NULL), these characters are used as separator:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*/
xbt_dynar_t xbt_str_split(const char *s, const char *sep)
//if substring was not found add the entire string
if (NULL == q) {
v = strlen(p);
- to_push = malloc(v + 1);
+ to_push = xbt_malloc(v + 1);
memcpy(to_push, p, v);
to_push[v] = '\0';
xbt_dynar_push(res, &to_push);
done = 1;
} else {
//get the appearance
- to_push = malloc(q - p + 1);
+ to_push = xbt_malloc(q - p + 1);
memcpy(to_push, p, q - p);
//add string terminator
to_push[q - p] = '\0';
*
* The string passed as argument must be writable (not const)
* The elements of the dynar are just parts of the string passed as argument.
+ * So if you don't store that argument elsewhere, you should free it in addition
+ * to freeing the dynar. This can be done by simply freeing the first argument
+ * of the dynar:
+ * free(xbt_dynar_get_ptr(dynar,0));
*
- * To free the structure constructed by this function, free the first element and free the dynar:
- *
- * free(xbt_dynar_get_ptr(dynar,0));
- * xbt_dynar_free(&dynar);
+ * Actually this function puts a bunch of \0 in the memory area you passed as
+ * argument to separate the elements, and pushes the address of each chunk
+ * in the resulting dynar. Yes, that's uneven. Yes, that's gory. But that's efficient.
*/
xbt_dynar_t xbt_str_split_quoted_in_place(char *s) {
xbt_dynar_t res = xbt_dynar_new(sizeof(char *), NULL);
beg = s;
- /* do not trim leading spaces: caller responsability to clean his cruft */
+ /* do not trim leading spaces: caller responsibility to clean his cruft */
end = beg;
while (!done) {
return res;
}
-#if defined(SIMGRID_NEED_GETLINE) || defined(DOXYGEN)
-/** @brief Get a single line from the stream (reimplementation of the GNU getline)
- *
- * This is a redefinition of the GNU getline function, used on platforms where it does not exists.
- *
- * getline() reads an entire line from stream, storing the address of the buffer
- * containing the text into *buf. The buffer is null-terminated and includes
- * the newline character, if one was found.
- *
- * If *buf is NULL, then getline() will allocate a buffer for storing the line,
- * which should be freed by the user program. Alternatively, before calling getline(),
- * *buf can contain a pointer to a malloc()-allocated buffer *n bytes in size. If the buffer
- * is not large enough to hold the line, getline() resizes it with realloc(), updating *buf and *n
- * as necessary. In either case, on a successful call, *buf and *n will be updated to
- * reflect the buffer address and allocated size respectively.
- */
-long getline(char **buf, size_t * n, FILE * stream)
-{
-
- size_t i;
- int ch;
-
- if (!*buf) {
- *buf = xbt_malloc(512);
- *n = 512;
- }
-
- if (feof(stream))
- return (ssize_t) - 1;
-
- for (i = 0; (ch = fgetc(stream)) != EOF; i++) {
-
- if (i >= (*n) + 1)
- *buf = xbt_realloc(*buf, *n += 512);
-
- (*buf)[i] = ch;
-
- if ((*buf)[i] == '\n') {
- i++;
- (*buf)[i] = '\0';
- break;
- }
- }
-
- if (i == *n)
- *buf = xbt_realloc(*buf, *n += 1);
-
- (*buf)[i] = '\0';
-
- return (ssize_t) i;
-}
-
-#endif /* HAVE_GETLINE */
-
/*
* Diff related functions
*
return res;
}
-/**
- * @brief Returns the hash code of a string.
+/** @brief Parse an integer out of a string, or raise an error
+ *
+ * The #str is passed as argument to your #error_msg, as follows:
+ * THROWF(arg_error, 0, error_msg, str);
*/
-XBT_INLINE unsigned int xbt_dict_hash_ext(const char *str,
- int str_len)
+long int xbt_str_parse_int(const char* str, const char* error_msg)
{
+ char *endptr;
+ if (str == NULL || str[0] == '\0')
+ THROWF(arg_error, 0, error_msg, str);
-#ifdef DJB2_HASH_FUNCTION
- /* fast implementation of djb2 algorithm */
- int c;
- register unsigned int hash = 5381;
-
- while (str_len--) {
- c = *str++;
- hash = ((hash << 5) + hash) + c; /* hash * 33 + c */
- }
-# elif defined(FNV_HASH_FUNCTION)
- register unsigned int hash = 0x811c9dc5;
- unsigned char *bp = (unsigned char *) str; /* start of buffer */
- unsigned char *be = bp + str_len; /* beyond end of buffer */
-
- while (bp < be) {
- /* multiply by the 32 bit FNV magic prime mod 2^32 */
- hash +=
- (hash << 1) + (hash << 4) + (hash << 7) + (hash << 8) +
- (hash << 24);
-
- /* xor the bottom with the current octet */
- hash ^= (unsigned int) *bp++;
- }
-
-# else
- register unsigned int hash = 0;
-
- while (str_len--) {
- hash += (*str) * (*str);
- str++;
- }
-#endif
+ long int res = strtol(str, &endptr, 10);
+ if (endptr[0] != '\0')
+ THROWF(arg_error, 0, error_msg, str);
- return hash;
+ return res;
}
-
-/**
- * @brief Returns the hash code of a string.
+/** @brief Parse a double out of a string, or raise an error
+ *
+ * The #str is passed as argument to your #error_msg, as follows:
+ * THROWF(arg_error, 0, error_msg, str);
*/
-XBT_INLINE unsigned int xbt_dict_hash(const char *str)
+double xbt_str_parse_double(const char* str, const char* error_msg)
{
-#ifdef DJB2_HASH_FUNCTION
- /* fast implementation of djb2 algorithm */
- int c;
- register unsigned int hash = 5381;
+ char *endptr;
+ if (str == NULL || str[0] == '\0')
+ THROWF(arg_error, 0, error_msg, str);
- while ((c = *str++)) {
- hash = ((hash << 5) + hash) + c; /* hash * 33 + c */
- }
+ double res = strtod(str, &endptr);
+ if (endptr[0] != '\0')
+ THROWF(arg_error, 0, error_msg, str);
-# elif defined(FNV_HASH_FUNCTION)
- register unsigned int hash = 0x811c9dc5;
-
- while (*str) {
- /* multiply by the 32 bit FNV magic prime mod 2^32 */
- hash +=
- (hash << 1) + (hash << 4) + (hash << 7) + (hash << 8) +
- (hash << 24);
-
- /* xor the bottom with the current byte */
- hash ^= (unsigned int) *str++;
- }
-
-# else
- register unsigned int hash = 0;
-
- while (*str) {
- hash += (*str) * (*str);
- str++;
- }
-#endif
- return hash;
+ return res;
}
#ifdef SIMGRID_TEST
#include "xbt/str.h"
+XBT_TEST_SUITE("xbt_str", "String Handling");
+
#define mytest(name, input, expected) \
xbt_test_add(name); \
d=xbt_str_split_quoted(input); \
input,s,expected);\
free(s); \
xbt_dynar_free(&d);
-
-XBT_TEST_SUITE("xbt_str", "String Handling");
XBT_TEST_UNIT("xbt_str_split_quoted", test_split_quoted, "test the function xbt_str_split_quoted")
{
xbt_dynar_t d;
}
}
+#define test_parse_error(function, name, variable, str) \
+ do { \
+ xbt_test_add(name); \
+ TRY { \
+ variable = function(str, "Parse error"); \
+ xbt_test_fail("The test '%s' did not detect the problem",name ); \
+ } CATCH(e) { \
+ if (e.category != arg_error) { \
+ xbt_test_exception(e); \
+ } else { \
+ xbt_ex_free(e); \
+ } \
+ } \
+ } while (0)
+#define test_parse_ok(function, name, variable, str, value) \
+ do { \
+ xbt_test_add(name); \
+ TRY { \
+ variable = function(str, "Parse error"); \
+ } CATCH(e) { \
+ xbt_test_exception(e); \
+ } \
+ xbt_test_assert(variable == value, "Fail to parse '%s'", str); \
+ } while (0)
+
+XBT_TEST_UNIT("xbt_str_parse", test_parse, "Test the parsing functions")
+{
+ xbt_ex_t e;
+ int rint = -9999;
+ test_parse_ok(xbt_str_parse_int, "Parse int", rint, "42", 42);
+ test_parse_ok(xbt_str_parse_int, "Parse 0 as an int", rint, "0", 0);
+ test_parse_ok(xbt_str_parse_int, "Parse -1 as an int", rint, "-1", -1);
+
+ test_parse_error(xbt_str_parse_int, "Parse int + noise", rint, "342 cruft");
+ test_parse_error(xbt_str_parse_int, "Parse NULL as an int", rint, NULL);
+ test_parse_error(xbt_str_parse_int, "Parse '' as an int", rint, "");
+ test_parse_error(xbt_str_parse_int, "Parse cruft as an int", rint, "cruft");
+
+ double rdouble = -9999;
+ test_parse_ok(xbt_str_parse_double, "Parse 42 as a double", rdouble, "42", 42);
+ test_parse_ok(xbt_str_parse_double, "Parse 42.5 as a double", rdouble, "42.5", 42.5);
+ test_parse_ok(xbt_str_parse_double, "Parse 0 as a double", rdouble, "0", 0);
+ test_parse_ok(xbt_str_parse_double, "Parse -1 as a double", rdouble, "-1", -1);
+
+ test_parse_error(xbt_str_parse_double, "Parse double + noise", rdouble, "342 cruft");
+ test_parse_error(xbt_str_parse_double, "Parse NULL as a double", rdouble, NULL);
+ test_parse_error(xbt_str_parse_double, "Parse '' as a double", rdouble, "");
+ test_parse_error(xbt_str_parse_double, "Parse cruft as a double", rdouble, "cruft");
+}
+
#endif /* SIMGRID_TEST */
