ignore bom at the beginning of the UTF-8 text

add a testcase about bom.

Author: dota17

src/lib_json/json_reader.cpp

@@ -18,6 +18,7 @@
 #include <memory>
 #include <set>
 #include <sstream>
+#include <string.h>
 #include <utility>
 
 #include <cstdio>
@@ -939,6 +940,7 @@ class OurReader {
 
   bool readToken(Token& token);
   void skipSpaces();
+  void skipBom(Location& beginPtr);
   bool match(const Char* pattern, int patternLength);
   bool readComment();
   bool readCStyleComment(bool* containsNewLineResult);
@@ -1009,7 +1011,8 @@ bool OurReader::parse(const char* beginDoc, const char* endDoc, Value& root,
   if (!features_.allowComments_) {
     collectComments = false;
   }
-
+  // skip byte order mark if it exists at the beginning of the UTF-8 text.
+  skipBom(beginDoc);
   begin_ = beginDoc;
   end_ = endDoc;
   collectComments_ = collectComments;
@@ -1268,6 +1271,12 @@ void OurReader::skipSpaces() {
   }
 }
 
+void OurReader::skipBom(Location& beginPtr) {
+  if (strncmp(beginPtr, "\xEF\xBB\xBF", 3) == 0) {
+    beginPtr += 3;
+  }
+}
+
 bool OurReader::match(const Char* pattern, int patternLength) {
   if (end_ - current_ < patternLength)
     return false;

src/test_lib_json/main.cpp

@@ -3578,6 +3578,20 @@ JSONTEST_FIXTURE_LOCAL(BuilderTest, settings) {
   }
 }
 
+struct BomTest : JsonTest::TestCase {};
+
+JSONTEST_FIXTURE_LOCAL(BomTest, withBom) {
+  const std::string with_bom = u8"\xEF\xBB\xBF{}";
+  Json::Value root;
+  JSONCPP_STRING errs;
+  std::istringstream iss(with_bom);
+  bool ok = parseFromStream(Json::CharReaderBuilder(), iss, &root, &errs);
+  JSONTEST_ASSERT(ok);
+  if (!ok) {
+    std::cout << "errs: " << errs << std::endl;
+  }
+}
+
 struct IteratorTest : JsonTest::TestCase {};
 
 JSONTEST_FIXTURE_LOCAL(IteratorTest, convert) {