Dear Maintainer,
html2text segfaults when converting the attached file twice:

$ html2text 1.html 1.html > 1.txt
Segmentation fault

control: retitle -1 html2text: segfault because of broken utf8 support
control: tags -1 +patch

Dear Maintainers,

I've found that the segfaults reported here are because of integer
overflows and other problems in utf8_* functions in file Area.C
I rewrite them, and it doesn't segfault anymore. Here is the patch:

--- a/Area.C
+++ b/Area.C
@@ -86,49 +86,29 @@

 /*           utf_length() and utf_width()
  *
- *     Very simplified algorithm of calculating length of UTF-8
- *   string. No check for errors. Counting only ASCII bytes and
- *   leading bytes of UTF-8 multibyte sequences. All bytes like
- *   10xxxxxx are dropped. If USE_UTF8 is false then returns
- *   usual length.               --YS
+ *   Easy and secure way: count "continuation"
+ *   (10xxxxxx) bytes and subtract that number.
+ *   If USE_UTF8 is false then returns usual length.
  */

--- a/Area.C
+++ b/Area.C
@@ -30,7 +30,7 @@

  /***************************************************************************/

-
+#include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <iostream>
@@ -45,10 +45,14 @@

 /* ------------------------------------------------------------------------- */

+static void* alloc_error() {
+  perror("html2text: error"); abort();
+}
+
 #define malloc_array(type, size)\
-((type *) malloc(sizeof(type) * (size)))
+((type *) (malloc(sizeof(type) * (size)) ? : alloc_error()))
 #define realloc_array(array, type, size) \
-((array) = (type *) realloc((array), sizeof(type) * (size)))
+((array) = (type *) (realloc((array), sizeof(type) * (size)) ? : alloc_error()))
 #define copy_array(from, to, type, count) \
 ((void) memcpy((to), (from), (count) * sizeof(type)))