Correctly parse json documents with a leading BOM

A leading byte order mark is valid in utf-8 and we should
parse documents starting with those correctly.

Change-Id: Id85398ff6e05b93ceefbaf4a6de5571d5e61ca13
Reviewed-by: Denis Dzyubenko <denis.dzyubenko@nokia.com>

2 files changed, 15 insertions, 2 deletions

diff --git a/src/corelib/json/qjsonparser.cpp b/src/corelib/json/qjsonparser.cpp
index 8c5693c9be..5fecb8d4e7 100644
--- a/src/corelib/json/qjsonparser.cpp
+++ b/src/corelib/json/qjsonparser.cpp
@@ -185,7 +185,16 @@ enum {
     Quote = 0x22
 };
 
-
+void Parser::eatBOM()
+{
+    // eat UTF-8 byte order mark
+    uchar utf8bom[3] = { 0xef, 0xbb, 0xbf };
+    if (end - json > 3 &&
+        (uchar)json[0] == utf8bom[0] &&
+        (uchar)json[1] == utf8bom[1] &&
+        (uchar)json[2] == utf8bom[2])
+        json += 3;
+}
 
 bool Parser::eatSpace()
 {
@@ -244,8 +253,10 @@ QJsonDocument Parser::parse(QJsonParseError *error)
 
     current = sizeof(QJsonPrivate::Header);
 
+    eatBOM();
     char token = nextToken();
-    DEBUG << token;
+
+    DEBUG << hex << (uint)token;
     if (token == BeginArray) {
         if (!parseArray())
             goto error;
@@ -253,6 +264,7 @@ QJsonDocument Parser::parse(QJsonParseError *error)
         if (!parseObject())
             goto error;
     } else {
+        lastError = QJsonParseError::IllegalValue;
         goto error;
     }
 
diff --git a/src/corelib/json/qjsonparser_p.h b/src/corelib/json/qjsonparser_p.h
index 20e57dcf26..8085edb2e0 100644
--- a/src/corelib/json/qjsonparser_p.h
+++ b/src/corelib/json/qjsonparser_p.h
@@ -84,6 +84,7 @@ public:
 
 
 private:
+    inline void eatBOM();
     inline bool eatSpace();
     inline char nextToken();