Merge branch '3972_fix_illumos'

Signed-off-by: Yury V. Zaytsev <yury@shurup.com>

autogen.sh

@@ -11,16 +11,20 @@ cd "$srcdir"
 ${AUTORECONF:-autoreconf} --verbose --install --force -I m4 ${AUTORECONF_FLAGS}
 
 # Customize the INSTALL file
-rm -f INSTALL && ln -s doc/INSTALL
+rm -f INSTALL && ln -s doc/INSTALL .
 
 # Generate po/POTFILES.in
-${XGETTEXT:-xgettext} --keyword=_ --keyword=N_ --keyword=Q_ --output=- \
+if ! xgettext -h 2>&1 | grep -e '--keyword=' >/dev/null ; then
+  echo "gettext is unable to extract translations, set XGETTEXT to GNU gettext!" >&2
+else
+  ${XGETTEXT:-xgettext} --keyword=_ --keyword=N_ --keyword=Q_ --output=- \
 	`find . -name '*.[ch]'` | ${SED-sed} -ne '/^#:/{s/#://;s/:[0-9]*/\
 /g;s/ //g;p;}' | \
 	grep -v '^$' | sort | uniq >po/POTFILES.in
+fi
 
-$srcdir/version.sh "$srcdir"
+"$srcdir/version.sh" "$srcdir"
 
-if test -x $srcdir/configure.mc; then
-  $srcdir/configure.mc "$@"
+if test -x "$srcdir/configure.mc"; then
+  "$srcdir/configure.mc" "$@"
 fi

configure.ac

@@ -705,6 +705,7 @@ tests/lib/mcconfig/Makefile
 tests/lib/search/Makefile
 tests/lib/strutil/Makefile
 tests/lib/vfs/Makefile
+tests/lib/vfs/mc.charsets
 tests/lib/widget/Makefile
 tests/src/Makefile
 tests/src/filemanager/Makefile

lib/charsets.c

@@ -267,17 +267,16 @@ get_codepage_index (const char *id)
 gboolean
 is_supported_encoding (const char *encoding)
 {
-    gboolean result = FALSE;
-    guint t;
+    GIConv coder;
+    gboolean result;
 
-    for (t = 0; t < codepages->len; t++)
-    {
-        const char *id;
-
-        id = ((codepage_desc *) g_ptr_array_index (codepages, t))->id;
-        result |= (g_ascii_strncasecmp (encoding, id, strlen (id)) == 0);
-    }
+    if (encoding == NULL)
+        return FALSE;
 
+    coder = str_crt_conv_from (encoding);
+    result = coder != INVALID_CONV;
+    if (result)
+        str_close_conv (coder);
     return result;
 }
 
@@ -364,6 +363,8 @@ str_nconvert_to_display (const char *str, int len)
         return g_string_new (str);
 
     conv = str_crt_conv_from (cp_source);
+    if (conv == INVALID_CONV)
+        return g_string_new (str);
 
     buff = g_string_new ("");
     str_nconvert (conv, str, len, buff);
@@ -396,6 +397,8 @@ str_nconvert_to_input (const char *str, int len)
         return g_string_new (str);
 
     conv = str_crt_conv_to (cp_source);
+    if (conv == INVALID_CONV)
+        return g_string_new (str);
 
     buff = g_string_new ("");
     str_nconvert (conv, str, len, buff);

lib/strutil/strutil.c

@@ -57,18 +57,16 @@ static const char *const str_utf8_encodings[] = {
 
 /* standard 8bit encodings, no wide or multibytes characters */
 static const char *const str_8bit_encodings[] = {
-    /* Solaris has different names of Windows 1251 encoding */
-#ifdef __sun
-    "ansi-1251",
-    "ansi1251",
-#else
     "cp-1251",
     "cp1251",
-#endif
+    /* solaris */
+    "ansi-1251",
+    "ansi1251",
     "cp-1250",
     "cp1250",
     "cp-866",
     "cp866",
+    /* glibc */
     "ibm-866",
     "ibm866",
     "cp-850",
@@ -274,7 +272,7 @@ str_crt_conv_from (const char *from_enc)
 void
 str_close_conv (GIConv conv)
 {
-    if (conv != str_cnv_not_convert)
+    if (conv != INVALID_CONV && conv != str_cnv_not_convert)
         g_iconv_close (conv);
 }
 

lib/utilunix.c

@@ -920,10 +920,20 @@ canonicalize_pathname_custom (char *path, canon_path_flags_t flags)
                 {
                     /* "token/../foo" -> "foo" */
 #ifdef HAVE_CHARSET
-                    if ((strncmp (s, VFS_ENCODING_PREFIX, enc_prefix_len) == 0)
-                        && (is_supported_encoding (s + enc_prefix_len)))
-                        /* special case: remove encoding */
-                        str_move (s, p + 1);
+                    if (strncmp (s, VFS_ENCODING_PREFIX, enc_prefix_len) == 0)
+                    {
+                        char *enc;
+
+                        enc = vfs_get_encoding (s, -1);
+
+                        if (is_supported_encoding (enc))
+                            /* special case: remove encoding */
+                            str_move (s, p + 1);
+                        else
+                            str_move (s, p + 4);
+
+                        g_free (enc);
+                    }
                     else
 #endif /* HAVE_CHARSET */
                         str_move (s, p + 4);
@@ -947,9 +957,18 @@ canonicalize_pathname_custom (char *path, canon_path_flags_t flags)
                 if (s == lpath + 1)
                     s[0] = '\0';
 #ifdef HAVE_CHARSET
-                else if ((strncmp (s, VFS_ENCODING_PREFIX, enc_prefix_len) == 0)
-                         && (is_supported_encoding (s + enc_prefix_len)))
+                else if (strncmp (s, VFS_ENCODING_PREFIX, enc_prefix_len) == 0)
                 {
+                    char *enc;
+                    gboolean ok;
+
+                    enc = vfs_get_encoding (s, -1);
+                    ok = is_supported_encoding (enc);
+                    g_free (enc);
+
+                    if (!ok)
+                        goto last;
+
                     /* special case: remove encoding */
                     s[0] = '.';
                     s[1] = '.';
@@ -966,6 +985,7 @@ canonicalize_pathname_custom (char *path, canon_path_flags_t flags)
 #endif /* HAVE_CHARSET */
                 else
                 {
+                  last:
                     if (s >= lpath + url_delim_len
                         && strncmp (s - url_delim_len, VFS_PATH_URL_DELIMITER, url_delim_len) == 0)
                         *s = '\0';

lib/vfs/path.c

@@ -182,42 +182,6 @@ vfs_canon (const char *path)
     return result;
 }
 
-/* --------------------------------------------------------------------------------------------- */
-
-#ifdef HAVE_CHARSET
-/** get encoding after last #enc: or NULL, if part does not contain #enc:
- *
- * @param path null-terminated string
- * @param len the maximum length of path, where #enc: should be searched
- *
- * @return newly allocated string.
- */
-
-static char *
-vfs_get_encoding (const char *path, ssize_t len)
-{
-    char *semi;
-
-    /* try found #enc: */
-    semi = g_strrstr_len (path, len, VFS_ENCODING_PREFIX);
-    if (semi == NULL)
-        return NULL;
-
-    if (semi == path || IS_PATH_SEP (semi[-1]))
-    {
-        char *slash;
-
-        semi += strlen (VFS_ENCODING_PREFIX);   /* skip "#enc:" */
-        slash = strchr (semi, PATH_SEP);
-        if (slash != NULL)
-            return g_strndup (semi, slash - semi);
-        return g_strdup (semi);
-    }
-
-    return vfs_get_encoding (path, semi - path);
-}
-#endif
-
 /* --------------------------------------------------------------------------------------------- */
 /**  Extract the hostname and username from the path
  *
@@ -896,8 +860,8 @@ vfs_path_element_clone (const vfs_path_element_t *element)
     new_element->vfs_prefix = g_strdup (element->vfs_prefix);
 #ifdef HAVE_CHARSET
     new_element->encoding = g_strdup (element->encoding);
-    if (vfs_path_element_need_cleanup_converter (element) && new_element->encoding != NULL)
-        new_element->dir.converter = str_crt_conv_from (new_element->encoding);
+    if (vfs_path_element_need_cleanup_converter (element) && element->encoding != NULL)
+        new_element->dir.converter = str_crt_conv_from (element->encoding);
     else
         new_element->dir.converter = element->dir.converter;
 #endif
@@ -1071,6 +1035,39 @@ vfs_prefix_to_class (const char *prefix)
 
 #ifdef HAVE_CHARSET
 
+/** get encoding after last #enc: or NULL, if part does not contain #enc:
+ *
+ * @param path null-terminated string
+ * @param len the maximum length of path, where #enc: should be searched
+ *
+ * @return newly allocated string.
+ */
+
+char *
+vfs_get_encoding (const char *path, ssize_t len)
+{
+    char *semi;
+
+    /* try found #enc: */
+    semi = g_strrstr_len (path, len, VFS_ENCODING_PREFIX);
+    if (semi == NULL)
+        return NULL;
+
+    if (semi == path || IS_PATH_SEP (semi[-1]))
+    {
+        char *slash;
+
+        semi += strlen (VFS_ENCODING_PREFIX);   /* skip "#enc:" */
+        slash = strchr (semi, PATH_SEP);
+        if (slash != NULL)
+            return g_strndup (semi, slash - semi);
+        return g_strdup (semi);
+    }
+
+    return vfs_get_encoding (path, semi - path);
+}
+
+/* --------------------------------------------------------------------------------------------- */
 /**
  * Check if need cleanup charset converter for vfs_path_element_t
  *

lib/vfs/path.h

@@ -82,6 +82,7 @@ void vfs_path_element_free (vfs_path_element_t * element);
 struct vfs_class *vfs_prefix_to_class (const char *prefix);
 
 #ifdef HAVE_CHARSET
+char *vfs_get_encoding(const char *path, ssize_t len);
 gboolean vfs_path_element_need_cleanup_converter (const vfs_path_element_t * element);
 vfs_path_t *vfs_path_change_encoding (vfs_path_t * vpath, const char *encoding);
 #endif

m4.include/mc-i18n.m4

@@ -18,16 +18,27 @@ AC_DEFUN([mc_I18N],[
         have_charset=yes
         charset_msg="yes"
 
+        AC_CHECK_HEADERS([gnu/libc-version.h])
+
         dnl Solaris has different name of Windows 1251 encoding
         case $host_os in
             solaris*)
-                CP1251="ANSI-1251"
+                ENCODING_CP1251="ANSI-1251"
                 ;;
             *)
-                CP1251="CP1251"
+                ENCODING_CP1251="CP1251"
                 ;;
         esac
 
-        AC_SUBST(CP1251)
+        if test "x$ac_cv_header_gnu_libc_version_h" != "xno"; then
+            ENCODING_CP866="IBM866"
+            ENCODING_ISO8859="ISO-8859"
+        else
+            ENCODING_CP866="CP866"
+            ENCODING_ISO8859="ISO8859"
+        fi
+
+        AC_SUBST(ENCODING_CP1251)
+        AC_SUBST(ENCODING_CP866)
     fi
 ])

misc/mc.charsets.in

@@ -1,13 +1,13 @@
 ASCII		7-bit ASCII
-ISO-8859-1	ISO 8859-1
-ISO-8859-2	ISO 8859-2
-ISO-8859-5	ISO 8859-5
+@ENCODING_ISO8859@-1	ISO 8859-1
+@ENCODING_ISO8859@-2	ISO 8859-2
+@ENCODING_ISO8859@-5	ISO 8859-5
 CP1250		Windows 1250
-@CP1251@		Windows 1251
+@ENCODING_CP1251@		Windows 1251
 CP437		CP 437
 CP850		CP 850
 CP852		CP 852
-IBM866		CP 866
+@ENCODING_CP866@		CP 866
 KOI8-R		KOI8-R
 KOI8-U		KOI8-U
 UTF-8		UTF-8

src/diffviewer/ydiff.c

@@ -607,7 +607,7 @@ dview_get_utf (const char *str, int *ch, int *ch_length)
     }
     else
     {
-        char *next_ch;
+        const char *next_ch;
 
         /* Calculate UTF-8 char length */
         next_ch = g_utf8_next_char (str);

src/editor/editdraw.c

@@ -573,6 +573,7 @@ edit_draw_this_line (WEdit *edit, off_t b, long row, long start_col, long end_co
                 unsigned int c;
                 gboolean wide_width_char = FALSE;
                 gboolean control_char = FALSE;
+                gboolean printable;
 
                 p->ch = 0;
                 p->style = q == edit->buffer.curs1 ? MOD_CURSOR : 0;
@@ -759,34 +760,30 @@ edit_draw_this_line (WEdit *edit, off_t b, long row, long start_col, long end_co
                         control_char = TRUE;
                         break;
                     }
+
 #ifdef HAVE_CHARSET
                     if (edit->utf8)
                     {
-                        if (g_unichar_isprint (c))
-                            p->ch = c;
+                        if (mc_global.utf8_display)
+                            /* c is gunichar */
+                            printable = g_unichar_isprint (c);
                         else
-                        {
-                            p->ch = '.';
-                            p->style = abn_style;
-                        }
-                        p++;
+                            /* c was gunichar; now c is 8-bit char converted from gunichar */
+                            printable = is_printable (c);
                     }
                     else
 #endif
+                        /* c is 8-bit char */
+                        printable = is_printable (c);
+
+                    if (printable)
+                        p->ch = c;
+                    else
                     {
-                        if ((mc_global.utf8_display && g_unichar_isprint (c)) ||
-                            (!mc_global.utf8_display && is_printable (c)))
-                        {
-                            p->ch = c;
-                            p++;
-                        }
-                        else
-                        {
-                            p->ch = '.';
-                            p->style = abn_style;
-                            p++;
-                        }
+                        p->ch = '.';
+                        p->style = abn_style;
                     }
+                    p++;
                     col++;
                     break;
                 }               /* case */