This fixes (a rare) segfault. The actual reason for this segfault is
that lynx ignores non-breakable space in most situations; the table
logic was not expecting this, thus a branch was entered which should
not have been entered at all.
Until the question with non-breakable spaces is resolved, this fix
hardens the branch in question so that it does not break even if
entered this way. ;-)
Enjoy,
Ilya
--- ./src/TRSTable.c~ Fri Jun 20 18:57:27 2003
+++ ./src/TRSTable.c Thu Jul 17 16:39:32 2003
@@ -585,12 +585,17 @@ PRIVATE int Stbl_addCellToRow ARGS9(
while (me->ncells + colspan + 1 > me->allocated + growby)
growby += CELLS_GROWBY;
if (growby) {
- if (me->allocated == 0 && !me->cells) {
+ if (!me->cells) {
cells = typecallocn(STable_cellinfo, growby);
} else {
- cells = realloc(me->cells,
- (me->allocated + growby)
- * sizeof(STable_cellinfo));
+ if (me->allocated == 0) { /* Put into pool */
+ cells = typecallocn(STable_cellinfo, growby);
+ memcpy(cells, me->cells,
+ me->ncells * sizeof(STable_cellinfo));
+ } else
+ cells = realloc(me->cells,
+ (me->allocated + growby)
+ * sizeof(STable_cellinfo));
for (i = 0; cells && i < growby; i++) {
cells[me->allocated + i].alignment = HT_ALIGN_NONE;
}
; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to [EMAIL PROTECTED]
