Fix [string tolower] buffer overflow for non-utf8

Reported-by: Andy <jimdevel@hummypkg.org.uk>
Signed-off-by: Steve Bennett <steveb@workware.net.au>

6 files changed, 15 insertions, 17 deletions

diff --git a/jim-format.c b/jim-format.c
index ed612cc..818eea4 100644
--- a/jim-format.c
+++ b/jim-format.c
@@ -302,7 +302,7 @@ Jim_Obj *Jim_FormatString(Jim_Interp *interp, Jim_Obj *fmtObjPtr, int objc, Jim_
 		goto error;
 	    }
 	    /* Just store the value in the 'spec' buffer */
-	    formatted_bytes = utf8_fromunicode(spec, code);
+	    formatted_bytes = utf8_getchars(spec, code);
 	    formatted_buf = spec;
 	    formatted_chars = 1;
 	    break;
diff --git a/jim.c b/jim.c
index 86b5479..1e26b38 100644
--- a/jim.c
+++ b/jim.c
@@ -2744,7 +2744,7 @@ static void JimStrCopyUpperLower(char *dest, const char *str, int uc)
     while (*str) {
         int c;
         str += utf8_tounicode(str, &c);
-        dest += utf8_fromunicode(dest, uc ? utf8_upper(c) : utf8_lower(c));
+        dest += utf8_getchars(dest, uc ? utf8_upper(c) : utf8_lower(c));
     }
     *dest = 0;
 }
@@ -2813,7 +2813,7 @@ static Jim_Obj *JimStringToTitle(Jim_Interp *interp, Jim_Obj *strObjPtr)
     buf = p = Jim_Alloc(len + 1);
 
     str += utf8_tounicode(str, &c);
-    p += utf8_fromunicode(p, utf8_title(c));
+    p += utf8_getchars(p, utf8_title(c));
 
     JimStrCopyUpperLower(p, str, 0);
 
diff --git a/jimregexp.c b/jimregexp.c
index efc56a1..fce33d9 100644
--- a/jimregexp.c
+++ b/jimregexp.c
@@ -1665,10 +1665,10 @@ static void regdump(regex_t *preg)
 			while (preg->program[s]) {
 				int len = preg->program[s++];
 				int first = preg->program[s++];
-				buf[utf8_fromunicode(buf, first)] = 0;
+				buf[utf8_setunicode(buf, first)] = 0;
 				printf("%s", buf);
 				if (len > 1) {
-					buf[utf8_fromunicode(buf, first + len - 1)] = 0;
+					buf[utf8_setunicode(buf, first + len - 1)] = 0;
 					printf("-%s", buf);
 				}
 			}
@@ -1678,7 +1678,7 @@ static void regdump(regex_t *preg)
 			/* Literal string, where present. */
 
 			while (preg->program[s]) {
-				buf[utf8_fromunicode(buf, preg->program[s])] = 0;
+				buf[utf8_setunicode(buf, preg->program[s])] = 0;
 				printf("%s", buf);
 				s++;
 			}
@@ -1690,7 +1690,7 @@ static void regdump(regex_t *preg)
 	if (op == END) {
 		/* Header fields of interest. */
 		if (preg->regstart) {
-			buf[utf8_fromunicode(buf, preg->regstart)] = 0;
+			buf[utf8_setunicode(buf, preg->regstart)] = 0;
 			printf("start '%s' ", buf);
 		}
 		if (preg->reganch)
diff --git a/linenoise.c b/linenoise.c
index 63d4899..e805356 100644
--- a/linenoise.c
+++ b/linenoise.c
@@ -717,16 +717,6 @@ static int getWindowSize(struct current *current)
 }
 #endif
 
-static int utf8_getchars(char *buf, int c)
-{
-#ifdef USE_UTF8
-    return utf8_fromunicode(buf, c);
-#else
-    *buf = c;
-    return 1;
-#endif
-}
-
 /**
  * Returns the unicode character at the given offset,
  * or -1 if none.
diff --git a/regtest.tcl b/regtest.tcl
index 2f09700..65f16af 100644
--- a/regtest.tcl
+++ b/regtest.tcl
@@ -210,6 +210,10 @@ set x [lindex {} 0]
 info source $x
 eval $x
 
+# REGTEST 30
+# non-UTF8 string tolower 
+string tolower "/mod/video/h\303\203\302\244xan_ witchcraft through the ages_20131101_0110.t"
+
 # TAKE THE FOLLOWING puts AS LAST LINE
 
 puts "--- ALL TESTS PASSED ---"
diff --git a/utf8.h b/utf8.h
index 71fd6ff..4dc5a02 100644
--- a/utf8.h
+++ b/utf8.h
@@ -26,6 +26,7 @@ int utf8_fromunicode(char *p, unsigned uc);
 /* No utf-8 support. 1 byte = 1 char */
 #define utf8_strlen(S, B) ((B) < 0 ? strlen(S) : (B))
 #define utf8_tounicode(S, CP) (*(CP) = (unsigned char)*(S), 1)
+#define utf8_getchars(CP, C) (*(CP) = (C), 1)
 #define utf8_upper(C) toupper(C)
 #define utf8_title(C) toupper(C)
 #define utf8_lower(C) tolower(C)
@@ -35,6 +36,9 @@ int utf8_fromunicode(char *p, unsigned uc);
 
 #else
 #if !defined(JIM_BOOTSTRAP)
+
+#define utf8_getchars utf8_fromunicode
+
 /**
  * Returns the length of the utf-8 sequence starting with 'c'.
  *