gawk 3.1.5, patch for multibyte locales

[Aharon Robbins]

Greetings all.

The problem with gawk and multibyte locales has come up enough that it's
time to post a patch.  This is extracted from my current code base.
It is untested but should do the trick.

At least now it'll be in the mailing list archives.

Arnold
-------------------------------
Fri Mar 10 06:28:23 2006  Arnold D. Robbins  <address@hidden>

        * awk.h (free_wstr): New declaration if MBS_SUPPORT, empty macro
        otherwise.
        * node.c (free_wstr): New function, inside MBS_SUPPORT. Frees the wide
        string part of a node. Provided so that it can be used consistently
        everywhere.
        (format_val, r_dupnode, mk_number, make_str_node, unref): Use it.
        * builtin.c (sub_common): Call `free_wstr' instead of doing it manually.
        * eval.c (r_tree_eval): Same in Node_assign_concat case.
        * field.c (set_field, rebuild_record, set_record): Add calls to 
`free_wstr'.

Mon Feb 13 22:45:34 2006  Arnold D. Robbins  <address@hidden>

        * eval.c (r_tree_eval): Node_assign_concat. Release any
        wide string value and reset the WSTRCUR flag. Based on
        bug report by Karal Zak.

diff -u ../gawk-3.1.5/awk.h ./awk.h
--- ../gawk-3.1.5/awk.h 2005-07-26 21:07:43.000000000 +0300
+++ ./awk.h     2006-03-10 06:35:14.000000000 +0200
@@ -1166,6 +1166,9 @@
 #define force_wstring(n)       str2wstr(n, NULL)
 extern const wchar_t *wstrstr P((const wchar_t *haystack, size_t hs_len, const 
wchar_t *needle, size_t needle_len));
 extern const wchar_t *wcasestrstr P((const wchar_t *haystack, size_t hs_len, 
const wchar_t *needle, size_t needle_len));
+extern void free_wstr P((NODE *n));
+#else
+#define free_wstr(NODE)        /* empty */
 #endif
 /* re.c */
 extern Regexp *make_regexp P((const char *s, size_t len, int ignorecase, int 
dfa));
diff -u ../gawk-3.1.5/builtin.c ./builtin.c
--- ../gawk-3.1.5/builtin.c     2005-07-26 21:07:43.000000000 +0300
+++ ./builtin.c 2006-03-10 06:55:00.000000000 +0200
@@ -2462,6 +2462,7 @@
        free(t->stptr);
        t->stptr = buf;
        t->stlen = textlen;
+       free_wstr(t);
 
        free_temp(s);
        if (matches > 0 && lhs) {
diff -u ../gawk-3.1.5/eval.c ./eval.c
--- ../gawk-3.1.5/eval.c        2005-07-26 21:07:43.000000000 +0300
+++ ./eval.c    2006-07-04 22:41:25.000000000 +0300
@@ -1176,6 +1176,7 @@
                        memcpy(l->stptr + l->stlen, r->stptr, r->stlen);
                        l->stlen += r->stlen;
                        l->stptr[l->stlen] = '\0';
+                       free_wstr(l);
                } else {
                        char *nval;
                        size_t nlen = l->stlen + r->stlen + 2;
diff -u ../gawk-3.1.5/field.c ./field.c
--- ../gawk-3.1.5/field.c       2005-05-11 18:28:15.000000000 +0300
+++ ./field.c   2006-03-10 06:48:22.000000000 +0200
@@ -129,6 +129,7 @@
        n->stptr = str;
        n->stlen = len;
        n->flags = (STRCUR|STRING|MAYBE_NUM|FIELD);
+       free_wstr(n);
 }
 
 /* rebuild_record --- Someone assigned a value to $(something).
@@ -210,6 +211,7 @@
                        }
 
                        n->stptr = cops;
+                       free_wstr(n);
                        unref(fields_arr[i]);
                        fields_arr[i] = n;
                }
@@ -273,6 +275,7 @@
        n->type = Node_val;
        n->stfmt = -1;
        n->flags = (STRING|STRCUR|MAYBE_NUM|FIELD);
+       free_wstr(n);
        fields_arr[0] = n;
 
 #undef INITIAL_SIZE
diff -u ../gawk-3.1.5/node.c ./node.c
--- ../gawk-3.1.5/node.c        2005-07-26 21:07:43.000000000 +0300
+++ ./node.c    2006-03-10 07:04:25.000000000 +0200
@@ -216,15 +218,7 @@
 no_malloc:
        s->stref = 1;
        s->flags |= STRCUR;
-#if defined MBS_SUPPORT
-       if ((s->flags & WSTRCUR) != 0) {
-               assert(s->wstptr != NULL);
-               free(s->wstptr);
-               s->wstptr = NULL;
-               s->wstlen = 0;
-               s->flags &= ~WSTRCUR;
-       }
-#endif
+       free_wstr(s);
        return s;
 }
 
@@ -287,7 +281,12 @@
        *r = *n;
        r->flags &= ~(PERM|TEMP|FIELD);
        r->flags |= MALLOC;
-#if defined MBS_SUPPORT
+#ifdef MBS_SUPPORT
+       /*
+        * DON'T call free_wstr(r) here!
+        * r->wstptr still points at n->wstptr's value, and we
+        * don't want to free it!
+        */
        r->wstptr = NULL;
 #endif /* defined MBS_SUPPORT */
        if (n->type == Node_val && (n->flags & STRCUR) != 0) {
@@ -344,11 +343,7 @@
        r->stref = 1;
        r->stptr = NULL;
        r->stlen = 0;
-#if defined MBS_SUPPORT
-       r->wstptr = NULL;
-       r->wstlen = 0;
-       r->flags &= ~WSTRCUR;
-#endif /* MBS_SUPPORT */
+       free_wsptr(r);
 #endif /* GAWKDEBUG */
        return r;
 }
@@ -363,10 +358,7 @@
        getnode(r);
        r->type = Node_val;
        r->flags = (STRING|STRCUR|MALLOC);
-#if defined MBS_SUPPORT
-       r->wstptr = NULL;
-       r->wstlen = 0;
-#endif
+       free_wstr(r);
        if (flags & ALREADY_MALLOCED)
                r->stptr = s;
        else {
@@ -510,15 +502,7 @@
                                return;
                        }
                        free(tmp->stptr);
-#if defined MBS_SUPPORT
-                       if (tmp->wstptr != NULL) {
-                               assert((tmp->flags & WSTRCUR) != 0);
-                               free(tmp->wstptr);
-                       }
-                       tmp->flags &= ~WSTRCUR;
-                       tmp->wstptr = NULL;
-                       tmp->wstlen = 0;
-#endif
+                       free_wstr(tmp);
                }
                freenode(tmp);
                return;
@@ -775,6 +759,21 @@
        return n;
 }
 
+/* free_wstr --- release the wide string part of a node */
+
+void
+free_wstr(NODE *n)
+{
+
+       if ((n->flags & WSTRCUR) != 0) {
+               assert(n->wstptr != NULL);
+               free(n->wstptr);
+       }
+       n->wstptr = NULL;
+       n->wstlen = 0;
+       n->flags &= ~WSTRCUR;
+}
+
 #if 0
 static void
 dump_wstr(FILE *fp, const wchar_t *str, size_t len)