Fixed when SS1 or SS2 followed by a set change by disallowing...

Fixed when SS1 or SS2 followed by a set change by disallowing controlcharacters after single shifts.

Made encoding_table_ptr and encoding_n_table_entries check for null tables.
moved 'Lm' type characters from marks to letters in mkunictype.

Version 0.08. Tagged as 'Unicode-0_08'

VersionNum

-/* (0.07)
+/* (0.08)
  *
  * This file is automatically maintained by srccommit, do not edit manually.
  *
  */
-#define Module_MajorVersion_CMHG     	0.07
+#define Module_MajorVersion_CMHG     	0.08
 #define Module_MinorVersion_CMHG	
-#define Module_Date_CMHG      		02 Dec 1997
+#define Module_Date_CMHG      		08 Dec 1997
 
-#define Module_MajorVersion     	"0.07"
+#define Module_MajorVersion     	"0.08"
 #define Module_MinorVersion		""
-#define Module_Date      		"02 Dec 1997"
+#define Module_Date      		"08 Dec 1997"
 

c/encoding

@@ -299,12 +299,12 @@ static encoding_table enc_table_list = NULL;
 
 UCS2 *encoding_table_ptr(encoding_table t)
 {
-    return t->table;
+    return t ? t->table : NULL;
 }
 
 int encoding_n_table_entries(encoding_table t)
 {
-    return t->n_entries;
+    return t ? t->n_entries : 0;
 }
 
 /* Table size is number of bytes including the length word

c/iso2022

@@ -240,6 +240,8 @@ static int iso2022_select_set(ISO2022_Encoding *i, int setno, int type, int id)
     encoding_table t;
     ISO2022_Set *set;
 
+    /* fdebugf(stderr, "iso2022_select_set: enc %p setno %d type %d\n", i, setno, type); */
+
     /* This can return NULL if the table is not present in which case
        we fall back on default mapping functions below */
     t = iso2022_find_table(type, id);
@@ -272,6 +274,8 @@ static int iso2022_select_set(ISO2022_Encoding *i, int setno, int type, int id)
 
     ((simple_set *)set)->table = t;
 
+    /* fdebugf(stderr, "iso2022_select_set: free %p new %p tempset %d oldset %p\n", i->Set[setno], set, i->tempset, i->oldset); */
+
     encoding__free(i->Set[setno]);
     i->Set[setno] = set;
 
@@ -484,8 +488,18 @@ static unsigned int iso2022_read(EncodingPriv *e,
         char c = *s++;
         UCS4 u;
 
-        if (i->esc_pending)
+	/* fdebugf(stderr, "iso2022: %02x\n", c); */
+
+	/* check for illegal single shifts */
+	if ((i->tempset == 1 && (c < 0x20 || c > 0x7F)) ||
+	    (i->tempset == 2 && (c < 0xA0)))
+	{
+	    u = 0xFFFD;
+	}
+	else if (i->esc_pending)
+	{
             u = iso2022_esc_cont(i, c);
+	}
         else if (c < 0x20)
         {
             i->sync[_GL] = i->sync[_GR] = 0;
@@ -504,13 +518,7 @@ static unsigned int iso2022_read(EncodingPriv *e,
         else
         {
             i->sync[_GL] = 0;
-#ifdef DEBUG
-/* 	    fprintf(stderr, "c: %02x GR %p next_code %p\n", c, i->GR, i->GR->next_code); */
-#endif
 	    u = i->GR->next_code(i->GR, c - 0xA0, _GR, i->sync + _GR);
-#ifdef DEBUG
-/* 	    fprintf(stderr, "u: %04x\n", u); */
-#endif
         }
 
         switch (u)

c/mkunictype

@@ -123,7 +123,6 @@ static int category_to_type(int category)
     {
     case cat('M', 'n'):
     case cat('M', 'c'):
-    case cat('L', 'm'):
 	return unictype_MARK;
 
     case cat('N', 'd'):
@@ -144,6 +143,7 @@ static int category_to_type(int category)
     case cat('L', 'l'):
     case cat('L', 't'):
     case cat('L', 'o'):
+    case cat('L', 'm'):		/* SJM: this is a spacing character not a mark */
 	return unictype_LETTER;
 
     case cat('P', 'd'):