Good day, everybody !
These two patches replace broken getword function in file.c with the not
broken one (I hope). Also i made small warkaround about one bug, and i submit
another patch later, but i need this function, so it will be really cool, if
somebody merge these patches to the CVS.
Thank you in advance.
--
With best regards,
Vadik Mironov
<[EMAIL PROTECTED]>
--- ./e/etalon/e16/e/src/E.h 2004-08-31 21:24:39.000000000 +0400
+++ ./e/hacked/e16/e/src/E.h 2004-09-01 12:00:40.485221968 +0400
@@ -2083,7 +2083,7 @@
char *usershell(int uid);
const char *atword(const char *s, int num);
const char *atchar(const char *s, char c);
-char *getword(char *s, int num);
+unsigned char *getword(unsigned char *s,int num);
void word(const char *s, int num, char *wd);
int canread(const char *s);
int canwrite(const char *s);
--- ./e/etalon/e16/e/src/file.c 2004-07-13 03:33:15.000000000 +0400
+++ ./e/hacked/e16/e/src/file.c 2004-09-01 12:03:04.695298704 +0400
@@ -510,74 +510,153 @@
EDBUG_RETURN(NULL);
}
-char *
-getword(char *s, int num)
-{
+/**
+ * Search for a [num]-th word in a string s and return it in
+ * allocated buffer.
+ *
+ * Parameters :
+ * s - pointer to a null terminating string.
+ * Punctuation signs treats as whitespaces.
+ * num - number of a token to search.
+ *
+ * Return value :
+ * NULL - if (num < 1) or (num > number of tokens in s)
+ * or if s contained unquoted symbols other than
+ * spaces, digits and symbols.
+ * Otherwise, pointer to null terminated string returned.
+ * WARNING : Caller must free the buffer.
+ *
+ * Example : " {..\" Alice \" wants more \"cookies4 grandma \""
+ * Token 1 = Alice
+ * Token 2 = wants
+ * Token 3 = more
+ * Token 4 = cookies4 grandma
+ */
+unsigned char *
+getword(unsigned char *s, int num)
+{
+ int cnt = 1 ;
+
+ ptrdiff_t char_count = 0 ;
+
+ unsigned char *start = NULL ;
+ unsigned char *word = NULL ;
+ unsigned char *tmp_copy = NULL ;
- /* *********FIXME**************
- * This function is broken but it isn't in use so I'll fix it later
- * (DO NOT USE UNTIL FIXED
- */
- int cnt, i;
- char *start, *finish, *ss, *w;
- char *wd = NULL;
+ const unsigned char quota_sym = '"' ;
- EDBUG(9, "getword");
if (!s)
- EDBUG_RETURN(NULL);
- if (!wd)
- EDBUG_RETURN(NULL);
+ {
+ goto fail ;
+ }
+
+ //counting from 1, negative index is mistake
if (num <= 0)
- {
- *wd = 0;
- EDBUG_RETURN(NULL);
- }
- cnt = 0;
- i = 0;
- start = NULL;
- finish = NULL;
- ss = NULL;
- w = wd;
-
- while (s[i])
- {
- if ((cnt == num) && ((s[i] == ' ') || (s[i] == '\t')))
- {
- finish = &s[i];
- break;
- }
- if ((s[i] != ' ') && (s[i] != '\t'))
- {
- if (i == 0)
- {
- cnt++;
- if (cnt == num)
- start = &s[i];
- }
- else if ((s[i - 1] == ' ') || (s[i - 1] == '\t'))
- {
- cnt++;
- if (cnt == num)
- start = &s[i];
- }
- }
- i++;
- }
- if (cnt == num)
- {
- if ((start) && (finish))
- {
- for (ss = start; ss < finish; ss++)
- *wd++ = *ss;
- }
- else if (start)
- {
- for (ss = start; *ss != 0; ss++)
- *wd++ = *ss;
- }
- *wd = 0;
- }
- EDBUG_RETURN(wd);
+ {
+ goto fail ;
+ }
+
+ //skip all whitespaces at first
+ //except double quotas since they are meaningful
+ //also let us ignore other garbage
+ while(!isalnum(*s) && quota_sym != *s)
+ {
+ s++ ;
+ }
+
+ //main cycle till the end
+ while ('\0'!=*s)
+ {
+ //we are standing at the start of quotes
+ if (quota_sym == *s)
+ {
+ //step over quota
+ s++ ;
+ //save start in case of word
+ if(cnt == num)
+ {
+ start = s ;
+ }
+ //it is possible to write inside quotes any trash
+ //even bells and whistles :-)
+ while (quota_sym != *s)
+ {
+ //only one quote sign
+ if ('\0' == *s)
+ {
+ goto fail ;
+ }
+ s++ ;
+ }
+
+ if (cnt == num)
+ {
+ char_count = s - start ;
+ //so far we have start offset and char count
+ break ;
+ }
+ //quotation ended
+ cnt++ ;
+ } else if (isalnum(*s))
+ {
+ if(cnt == num)
+ {
+ start = s ;
+ }
+ //go through word
+ while (isalnum(*s))
+ {
+ s++ ;
+ }
+ //we found word
+ //note: it can be followed by any crap
+ if (cnt == num)
+ {
+ char_count = s - start ;
+ break ;
+ }
+
+ //word must be followed by space
+ if (!isspace(*s) && !ispunct(*s))
+ {
+ goto fail ;
+ }
+ //word ended
+ cnt++ ;
+ //return at the end of the word
+ s-- ;
+ } else if (isspace(*s) || ispunct(*s))
+ {
+ //just ignoring spaces
+ } else
+ {
+ goto fail ;
+ }
+ s++ ;
+ }
+
+ if (start==NULL)
+ {
+ //avoid allocation
+ goto fail ;
+ }
+ //at this point we found the WORD
+ tmp_copy = word = (unsigned char *)Emalloc((sizeof(char)*char_count) + 1);
+ if(NULL==tmp_copy)
+ {
+ goto fail ;
+ }
+ while (0<= --char_count)
+ {
+ *tmp_copy++ = *start++ ;
+ }
+ *(tmp_copy++) = '\0' ;
+ goto success ;
+
+success :
+ EDBUG_RETURN(word) ;
+fail :
+ EDBUG_RETURN(NULL) ;
}
void
-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
_______________________________________________
enlightenment-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
