From 7aeb8651ba0f3b69ff53bca043fe339d595842ab Mon Sep 17 00:00:00 2001
From: Jan Vidar Krey <janvidar@extatic.org>
Date: Mon, 23 Mar 2009 21:47:05 +0100
Subject: [PATCH] Fix bug #12: asserts in adc_msg_parse -> enabled strict utf8
 parsing.

Signed-off-by: Jan Vidar Krey <janvidar@extatic.org>
---
 ChangeLog     |  1 +
 src/message.c |  9 +++++++--
 src/misc.c    | 35 +++++++++++++++++++++++++++++------
 src/misc.h    |  1 +
 version.h     |  2 +-
 5 files changed, 39 insertions(+), 9 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index de6de0a..6ceb293 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,6 +1,7 @@
 0.2.8:
 - Fix bug #13: getsockname() failure, use sockaddr from accept() instead.
 - Fix bug #10: Improve logging, ensure logs are machine readable.
+- Fix bug #12: asserts in adc_msg_parse -> enabled strict utf8 parsing.
 
 
 0.2.7:
diff --git a/src/message.c b/src/message.c
index 92d1a95..372c9ec 100644
--- a/src/message.c
+++ b/src/message.c
@@ -26,8 +26,6 @@
 	uhub_assert(X->capacity); \
 	uhub_assert(X->length); \
 	uhub_assert(X->length <= X->capacity); \
-	if (X->length != strlen(X->cache)) \
-		hub_log(log_error, "Assert: X->length (%d) != strlen(X->cache) (%d)", X->length, strlen(X->cache)); \
 	uhub_assert(X->length == strlen(X->cache)); \
 	uhub_assert(X->references >= 0);
 #else
@@ -265,6 +263,13 @@ struct adc_message* adc_msg_parse(const char* line, size_t length)
 	
 	if (command == NULL)
 		return NULL; /* OOM */
+		
+	if (!is_printable_utf8(line, length))
+	{
+		hub_log(log_debug, "Dropped message with non-printable UTF-8 characters.");
+		hub_free(command);
+		return NULL;
+	}
 
 	if (line[length-1] != '\n')
 	{
diff --git a/src/misc.c b/src/misc.c
index 20242f9..7434817 100644
--- a/src/misc.c
+++ b/src/misc.c
@@ -31,6 +31,16 @@ int is_white_space(char c)
 	return 0;
 }
 
+int is_printable(char c)
+{
+	if (c >= 32)
+		return 1;
+		
+	if (c == '\t' || c == '\r' || c == '\n')
+		return 1;
+	return 0;
+}
+
 
 char* strip_white_space(char* string)
 {
@@ -48,17 +58,15 @@ char* strip_white_space(char* string)
 	return string;
 }
 
-
-int is_valid_utf8(const char* string)
+static int is_valid_utf8_str(const char* string, size_t length)
 {
 	int expect = 0;
 	char div = 0;
-	int pos = 0;
-	int length = strlen(string);
+	size_t pos = 0;
 	
 	if (length == 0) return 1;
 	
-	for (pos = 0; pos < strlen(string); pos++)
+	for (pos = 0; pos < length; pos++)
 	{
 		if (expect)
 		{
@@ -74,13 +82,28 @@ int is_valid_utf8(const char* string)
 					if (string[pos] & div) expect++;
 					else break;
 				}
-				if ((string[pos] & div) || (pos+expect >= strlen(string))) return 0;
+				if ((string[pos] & div) || (pos+expect >= length)) return 0;
 			}
 		}
 	}
 	return 1;
 }
 
+int is_valid_utf8(const char* string)
+{
+	return is_valid_utf8_str(string, strlen(string));
+}
+
+int is_printable_utf8(const char* string, size_t length)
+{
+	size_t pos = 0;
+	for (pos = 0; pos < length; pos++)
+	{
+		if (!is_printable(string[pos]))
+			return 0;
+	}
+	return is_valid_utf8_str(string, length);
+}
 
 int is_valid_base32_char(char c)
 {
diff --git a/src/misc.h b/src/misc.h
index 789c616..bac0928 100644
--- a/src/misc.h
+++ b/src/misc.h
@@ -26,6 +26,7 @@ extern int is_num(char c);
 extern int is_space(char c);
 extern int is_white_space(char c);
 extern int is_valid_utf8(const char* string);
+extern int is_printable_utf8(const char* string, size_t length);
 extern int is_valid_base32_char(char c);
 extern void base32_encode(const unsigned char* buffer, size_t len, char* result);
 extern void base32_decode(const char* src, unsigned char* dst, size_t len);
diff --git a/version.h b/version.h
index 2bb34ed..335a101 100644
--- a/version.h
+++ b/version.h
@@ -7,7 +7,7 @@
 #endif
 
 #ifndef VERSION
-#define VERSION "0.2.7"
+#define VERSION "0.2.8-alpha"
 #endif
 
 #ifndef COPYRIGHT