emo/busybox
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

awk: fix dodgy multi-char separators splitting logic

Browse Source 
function                                             old     new   delta
awk_split                                            521     484     -37

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
This commit is contained in:
Denys Vlasenko 2020-11-16 10:40:32 +01:00
parent 89f063b900
commit 5323af7f51
1 changed files with 16 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
editors/awk.c
View File

@ -1765,10 +1765,9 @@ static void fsrealloc(int size)
static int awk_split(const char *s, node *spl, char **slist) static int awk_split(const char *s, node *spl, char **slist)
{ {
int l, n; int n;
char c[4]; char c[4];
char *s1; char *s1;
regmatch_t pmatch[2]; // TODO: why [2]? [1] is enough...
/* in worst case, each char would be a separate field */ /* in worst case, each char would be a separate field */
*slist = s1 = xzalloc(strlen(s) * 2 + 3); *slist = s1 = xzalloc(strlen(s) * 2 + 3);
@ -1785,12 +1784,18 @@ static int awk_split(const char *s, node *spl, char **slist)
return n; /* "": zero fields */ return n; /* "": zero fields */
n++; /* at least one field will be there */ n++; /* at least one field will be there */
do { do {
int l;
regmatch_t pmatch[2]; // TODO: why [2]? [1] is enough...
l = strcspn(s, c+2); /* len till next NUL or \n */ l = strcspn(s, c+2); /* len till next NUL or \n */
if (regexec(icase ? spl->r.ire : spl->l.re, s, 1, pmatch, 0) == 0 if (regexec(icase ? spl->r.ire : spl->l.re, s, 1, pmatch, 0) == 0
&& pmatch[0].rm_so <= l && pmatch[0].rm_so <= l
) { ) {
l = pmatch[0].rm_so; l = pmatch[0].rm_so;
if (pmatch[0].rm_eo == 0) { if (pmatch[0].rm_eo == 0) {
/* For example, happens when FS can match
* an empthy string (awk -F ' *')
*/
l++; l++;
pmatch[0].rm_eo++; pmatch[0].rm_eo++;
} }
@ -1800,14 +1805,16 @@ static int awk_split(const char *s, node *spl, char **slist)
if (s[l]) if (s[l])
pmatch[0].rm_eo++; pmatch[0].rm_eo++;
} }
memcpy(s1, s, l); s1 = mempcpy(s1, s, l);
/* make sure we remove *all* of the separator chars */ *s1++ = '\0';
do {
s1[l] = '\0';
} while (++l < pmatch[0].rm_eo);
nextword(&s1);
s += pmatch[0].rm_eo; s += pmatch[0].rm_eo;
} while (*s); } while (*s);
/* echo a-- | awk -F-- '{ print NF, length($NF), $NF }'
* should print "2 0 ":
*/
*s1 = '\0';
return n; return n;
} }
if (c[0] == '\0') { /* null split */ if (c[0] == '\0') { /* null split */
@ -2011,7 +2018,7 @@ static int ptest(node *pattern)
static int awk_getline(rstream *rsm, var *v) static int awk_getline(rstream *rsm, var *v)
{ {
char *b; char *b;
regmatch_t pmatch[2]; regmatch_t pmatch[2]; // TODO: why [2]? [1] is enough...
int size, a, p, pp = 0; int size, a, p, pp = 0;
int fd, so, eo, r, rp; int fd, so, eo, r, rp;
char c, *m, *s; char c, *m, *s;