[gmime] Fixed the text-to-html filtering logic to include non-ascii characters in URLs

[Jeffrey Stedfast <fejj src gnome org>]

commit c1135b0ab233c48eec3911a439a1d13b545c4008
Author: Jeffrey Stedfast <jeff xamarin com>
Date:   Sat Apr 11 18:44:29 2015 -0400

    Fixed the text-to-html filtering logic to include non-ascii characters in URLs
    
    2015-04-11  Jeffrey Stedfast  <fejj gnome org>
    
        * util/url-scanner.c (url_scanner_table_init): Treat all character
        values >= 128 as url-safe. Fixes bug #738583.

 ChangeLog          |    5 +++++
 util/url-scanner.c |   24 +++++++++++++-----------
 2 files changed, 18 insertions(+), 11 deletions(-)
---
diff --git a/ChangeLog b/ChangeLog
index b749a09..a194b53 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2015-04-11  Jeffrey Stedfast  <fejj gnome org>
+
+       * util/url-scanner.c (url_scanner_table_init): Treat all character
+       values >= 128 as url-safe. Fixes bug #738583.
+
 2014-07-18  Jeffrey Stedfast  <fejj gnome org>
 
        * gmime/gmime-header.c (g_mime_header_list_register_writer):
diff --git a/util/url-scanner.c b/util/url-scanner.c
index 20535ef..a942801 100644
--- a/util/url-scanner.c
+++ b/util/url-scanner.c
@@ -109,14 +109,14 @@ static unsigned char url_scanner_table[256] = {
         66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66,160,160,160,128,128,
        128, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66,
         66, 66, 66, 66, 66, 66, 66, 66, 66, 66, 66,128,128,128,128,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
-         1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,
+       128,128,128,128,128,128,128,128,128,128,128,128,128,128,128,128
 };
 
 enum {
@@ -525,12 +525,12 @@ url_web_end (const char *in, const char *pos, const char *inend, urlmatch_t *mat
 
 
 static void
-table_init_bits (unsigned int mask, const unsigned char *vals)
+table_init_bits (unsigned int mask, const char *vals)
 {
        int i;
        
        for (i = 0; vals[i] != '\0'; i++)
-               url_scanner_table[vals[i]] |= mask;
+               url_scanner_table[(unsigned char) vals[i]] |= mask;
 }
 
 static void
@@ -546,7 +546,9 @@ url_scanner_table_init (void)
                        url_scanner_table[i] |= IS_DIGIT | IS_DOMAIN;
                if ((i >= 'a' && i <= 'z') || (i >= 'A' && i <= 'Z'))
                        url_scanner_table[i] |= IS_ALPHA | IS_DOMAIN;
-               if (i >= 127)
+               if (i >= 128)
+                       url_scaller_table[i] |= IS_URLSAFE;
+               if (i == 127)
                        url_scanner_table[i] |= IS_CTRL;
        }