New function, a cross between g_strsplit() and strtok(). (#88329, Soeren

Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>

	* glib/gstrfuncs.h:
	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross
	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)

	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.

ChangeLog

@@ -1,3 +1,11 @@
+Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/gstrfuncs.h: 
+	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross 
+	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)
+
+	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.
+	
 Fri Jan 23 22:49:52 2004  Matthias Clasen  <maclas@gmx.de>
 
 	* glib/gutils.c (g_get_any_init): Sigh, protect against

ChangeLog.pre-2-10

@@ -1,3 +1,11 @@
+Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/gstrfuncs.h: 
+	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross 
+	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)
+
+	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.
+	
 Fri Jan 23 22:49:52 2004  Matthias Clasen  <maclas@gmx.de>
 
 	* glib/gutils.c (g_get_any_init): Sigh, protect against

ChangeLog.pre-2-12

@@ -1,3 +1,11 @@
+Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/gstrfuncs.h: 
+	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross 
+	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)
+
+	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.
+	
 Fri Jan 23 22:49:52 2004  Matthias Clasen  <maclas@gmx.de>
 
 	* glib/gutils.c (g_get_any_init): Sigh, protect against

ChangeLog.pre-2-4

@@ -1,3 +1,11 @@
+Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/gstrfuncs.h: 
+	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross 
+	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)
+
+	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.
+	
 Fri Jan 23 22:49:52 2004  Matthias Clasen  <maclas@gmx.de>
 
 	* glib/gutils.c (g_get_any_init): Sigh, protect against

ChangeLog.pre-2-6

@@ -1,3 +1,11 @@
+Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/gstrfuncs.h: 
+	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross 
+	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)
+
+	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.
+	
 Fri Jan 23 22:49:52 2004  Matthias Clasen  <maclas@gmx.de>
 
 	* glib/gutils.c (g_get_any_init): Sigh, protect against

ChangeLog.pre-2-8

@@ -1,3 +1,11 @@
+Wed Jan 28 01:39:21 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/gstrfuncs.h: 
+	* glib/gstrfuncs.c (g_strsplit_set): New function, a cross 
+	between g_strsplit() and strtok().  (#88329, Soeren Sandmann)
+
+	* tests/strfunc-test.c (main): Add g_strsplit_set() tests.
+	
 Fri Jan 23 22:49:52 2004  Matthias Clasen  <maclas@gmx.de>
 
 	* glib/gutils.c (g_get_any_init): Sigh, protect against

docs/reference/ChangeLog

@@ -1,3 +1,7 @@
+Wed Jan 28 01:39:59 2004  Matthias Clasen  <maclas@gmx.de>
+
+	* glib/glib-sections.txt: Add g_strsplit_set. (Soeren Sandmann)
+
 Thu Jan 22 14:51:19 2004  Owen Taylor  <otaylor@redhat.com>
 
 	* glib/glib-sections.txt glib/tmpl/timers.sgml: Document

docs/reference/glib/glib-sections.txt

@@ -1008,6 +1008,7 @@ g_strescape
 g_strcompress
 g_strcanon
 g_strsplit
+g_strsplit_set
 g_strfreev
 g_strconcat
 g_strjoin

glib/gstrfuncs.c

@@ -2171,6 +2171,106 @@ g_strsplit (const gchar *string,
   return str_array;
 }
 
+/**
+ * g_strsplit_set:
+ * @string: The string to be tokenized
+ * @delimiters: A nul-terminated string containing bytes that are used
+ *              to split the string.
+ * @max_tokens: The maximum number of tokens to split @string into. 
+ *              If this is less than 1, the string is split completely
+ * 
+ * Splits @string into a number of tokens not containing any of the characters
+ * in @delimiter. A token is the (possibly empty) longest string that does not
+ * contain any of the characters in @delimiters. If @max_tokens is reached, the
+ * remainder is appended to the last token.
+ *
+ * For example the result of g_strtokenize ("abc:def/ghi", ":/", -1) is a
+ * %NULL-terminated vector containing the three strings "abc", "def", 
+ * and "ghi".
+ *
+ * The result if g_strtokenize (":def/ghi:", ":/", -1) is a %NULL-terminated
+ * vector containing the four strings "", "def", "ghi", and "".
+ * 
+ * As a special case, the result of splitting the empty string "" is an empty
+ * vector, not a vector containing a single string. The reason for this
+ * special case is that being able to represent a empty vector is typically
+ * more useful than consistent handling of empty elements. If you do need
+ * to represent empty elements, you'll need to check for the empty string
+ * before calling g_strsplit().
+ *
+ * Note that this function works on bytes not characters, so it can't be used 
+ * to delimit UTF-8 strings for anything but ASCII characters.
+ * 
+ * Return value: a newly-allocated %NULL-terminated array of strings. Use 
+ *    g_strfreev() to free it.
+ * 
+ * Since: 2.4
+ **/
+gchar **
+g_strsplit_set (const gchar *string,
+	        const gchar *delimiters,
+	        gint         max_tokens)
+{
+  gboolean delim_table[256];
+  GSList *tokens, *list;
+  gint n_tokens;
+  const gchar *s;
+  const gchar *current;
+  gchar *token;
+  gchar **result;
+  
+  g_return_val_if_fail (string != NULL, NULL);
+  g_return_val_if_fail (delimiters != NULL, NULL);
+
+  if (max_tokens < 1)
+    max_tokens = G_MAXINT;
+
+  if (*string == '\0')
+    {
+      result = g_new (char *, 1);
+      result[0] = NULL;
+      return result;
+    }
+  
+  memset (delim_table, FALSE, sizeof (delim_table));
+  for (s = delimiters; *s != '\0'; ++s)
+    delim_table[*(guchar *)s] = TRUE;
+
+  tokens = NULL;
+  n_tokens = 0;
+
+  s = current = string;
+  while (*s != '\0')
+    {
+      if (delim_table[*(guchar *)s] && n_tokens + 1 < max_tokens)
+	{
+	  gchar *token;
+
+	  token = g_strndup (current, s - current);
+	  tokens = g_slist_prepend (tokens, token);
+	  ++n_tokens;
+
+	  current = s + 1;
+	}
+      
+      ++s;
+    }
+
+  token = g_strndup (current, s - current);
+  tokens = g_slist_prepend (tokens, token);
+  ++n_tokens;
+
+  result = g_new (gchar *, n_tokens + 1);
+
+  result[n_tokens] = NULL;
+  for (list = tokens; list != NULL; list = list->next)
+    result[--n_tokens] = list->data;
+
+  g_slist_free (tokens);
+  
+  return result;
+}
+
 /**
  * g_strfreev:
  * @str_array: a %NULL-terminated array of strings to free.

glib/gstrfuncs.h

@@ -213,8 +213,8 @@ gpointer              g_memdup	       (gconstpointer mem,
 					guint	       byte_size);
 
 /* NULL terminated string arrays.
- * g_strsplit() splits up string into max_tokens tokens at delim and
+ * g_strsplit(), g_strsplit_set() split up string into max_tokens tokens
- * returns a newly allocated string array.
+ * at delim and return a newly allocated string array.
  * g_strjoinv() concatenates all of str_array's strings, sliding in an
  * optional separator, the returned string is newly allocated.
  * g_strfreev() frees the array itself and all of its strings.
@@ -223,6 +223,9 @@ gpointer              g_memdup	       (gconstpointer mem,
 gchar**	              g_strsplit       (const gchar  *string,
 					const gchar  *delimiter,
 					gint          max_tokens);
+gchar **	      g_strsplit_set   (const gchar *string,
+					const gchar *delimiters,
+					gint         max_tokens);
 gchar*                g_strjoinv       (const gchar  *separator,
 					gchar       **str_array);
 void                  g_strfreev       (gchar       **str_array);

tests/strfunc-test.c

@@ -286,6 +286,7 @@ main (int   argc,
   TEST (NULL, strcmp (g_strrstr_len ("FooBarFooBarFoo", 14, "BarFoo"),
 		      "BarFooBarFoo") == 0);
 
+  /* Test g_strsplit() */
   TEST (NULL, strv_check (g_strsplit ("", ",", 0), NULL));
   TEST (NULL, strv_check (g_strsplit ("x", ",", 0), "x", NULL));
   TEST (NULL, strv_check (g_strsplit ("x,y", ",", 0), "x", "y", NULL));
@@ -325,6 +326,70 @@ main (int   argc,
   TEST (NULL, strv_check (g_strsplit (",,x,,y,,z,,", ",", 2), "", ",x,,y,,z,,", NULL));
   TEST (NULL, strv_check (g_strsplit (",,x,,y,,z,,", ",,", 2), "", "x,,y,,z,,", NULL));
 
+  /* Test g_strsplit_set() */
+  TEST (NULL, strv_check (g_strsplit_set ("", ",/", 0), NULL));
+  TEST (NULL, strv_check (g_strsplit_set (":def/ghi:", ":/", -1), "", "def", "ghi", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("abc:def/ghi", ":/", -1), "abc", "def", "ghi", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",;,;,;,;", ",;", -1), "", "", "", "", "", "", "", "", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,abc.def", ".,", -1), "", "", "abc", "def", NULL));
+
+  TEST (NULL, strv_check (g_strsplit_set (",x.y", ",.", 0), "", "x", "y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (".x,y,", ",.", 0), "", "x", "y", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y.z", ",.", 0), "x", "y", "z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x.y,z,", ",.", 0), "x", "y", "z", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x.y,z", ",.", 0), "", "x", "y", "z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z,", ",.", 0), "", "x", "y", "z", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",.x,,y,;z..", ".,;", 0), "", "", "x", "", "y", "", "z", "", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,x,,y,,z,,", ",,", 0), "", "", "x", "", "y", "", "z", "", "", NULL));
+
+  TEST (NULL, strv_check (g_strsplit_set ("x,y.z", ",.", 1), "x,y.z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x.y,z,", ",.", 1), "x.y,z,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z", ",.", 1), ",x,y,z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y.z,", ",.", 1), ",x,y.z,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,x,.y,,z,,", ",.", 1), ",,x,.y,,z,,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",.x,,y,,z,,", ",,..", 1), ",.x,,y,,z,,", NULL));
+   
+  TEST (NULL, strv_check (g_strsplit_set ("", ",", 0), NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x", ",", 0), "x", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y", ",", 0), "x", "y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,", ",", 0), "x", "y", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y", ",", 0), "", "x", "y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,", ",", 0), "", "x", "y", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,z", ",", 0), "x", "y", "z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,z,", ",", 0), "x", "y", "z", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z", ",", 0), "", "x", "y", "z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z,", ",", 0), "", "x", "y", "z", "", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,x,,y,,z,,", ",", 0), "", "", "x", "", "y", "", "z", "", "", NULL));
+
+  TEST (NULL, strv_check (g_strsplit_set ("", ",", 1), NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x", ",", 1), "x", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y", ",", 1), "x,y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,", ",", 1), "x,y,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y", ",", 1), ",x,y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,", ",", 1), ",x,y,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,z", ",", 1), "x,y,z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,z,", ",", 1), "x,y,z,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z", ",", 1), ",x,y,z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z,", ",", 1), ",x,y,z,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,x,,y,,z,,", ",", 1), ",,x,,y,,z,,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,x,,y,,z,,", ",,", 1), ",,x,,y,,z,,", NULL));
+
+  TEST (NULL, strv_check (g_strsplit_set ("", ",", 2), NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x", ",", 2), "x", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y", ",", 2), "x", "y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,", ",", 2), "x", "y,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y", ",", 2), "", "x,y", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,", ",", 2), "", "x,y,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,z", ",", 2), "x", "y,z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set ("x,y,z,", ",", 2), "x", "y,z,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z", ",", 2), "", "x,y,z", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",x,y,z,", ",", 2), "", "x,y,z,", NULL));
+  TEST (NULL, strv_check (g_strsplit_set (",,x,,y,,z,,", ",", 2), "", ",x,,y,,z,,", NULL));
+  
+  TEST (NULL, strv_check (g_strsplit_set (",,x,.y,..z,,", ",.", 3), "", "", "x,.y,..z,,", NULL));
+
+  
+  
   #define TEST_IS(name) test_is_function (#name, call_g_ascii_##name, call_##name, g_unichar_##name);
 
   FOR_ALL_CTYPE(TEST_IS)