Added structured error reporting to Reader.

This allows applications for interactively viewing or editing JSON to do a better job of highlighting errors. Also added offset accessors to Value, offering the same sort of functionality even for non-errors. Thanks to Zach Clifford ([email protected]) for the patch.
raythorn · Apr 23, 2014 · 68db655 · 68db655
1 parent 642befc
commit 68db655
Show file tree

Hide file tree

Showing 5 changed files with 269 additions and 3 deletions.
diff --git a/include/json/reader.h b/include/json/reader.h
@@ -33,6 +33,19 @@ namespace Json {
       typedef char Char;
       typedef const Char *Location;
 
+      /** \brief An error tagged with where in the JSON text it was encountered.
+       *
+       * The offsets give the [start, limit) range of bytes within the text. Note
+       * that this is bytes, not codepoints.
+       *
+       */
+      struct StructuredError
+      {
+         size_t offset_start;
+         size_t offset_limit;
+         std::string message;
+      };
+
       /** \brief Constructs a Reader allowing all features
        * for parsing.
        */
@@ -95,6 +108,14 @@ namespace Json {
        */
       std::string getFormattedErrorMessages() const;
 
+     /** \brief Returns a vector of structured erros encounted while parsing.
+      * \return A (possibly empty) vector of StructuredError objects. Currently
+      *         only one error can be returned, but the caller should tolerate multiple
+      *         errors.  This can occur if the parser recovers from a non-fatal
+      *         parse error and then encounters additional errors.
+      */
+     std::vector<StructuredError> getStructuredErrors() const;
+
    private:
       enum TokenType
       {

diff --git a/include/json/value.h b/include/json/value.h
@@ -442,6 +442,13 @@ namespace Json {
       iterator begin();
       iterator end();
 
+      // Accessors for the [start, limit) range of bytes within the JSON text from
+      // which this value was parsed, if any.
+      void setOffsetStart( size_t start );
+      void setOffsetLimit( size_t limit );
+      size_t getOffsetStart() const;
+      size_t getOffsetLimit() const;
+
    private:
       Value &resolveReference( const char *key, 
                                bool isStatic );
@@ -509,6 +516,11 @@ namespace Json {
       int memberNameIsStatic_ : 1;       // used by the ValueInternalMap container.
 # endif
       CommentInfo *comments_;
+
+      // [start, limit) byte offsets in the source JSON text from which this Value
+      // was extracted.
+      size_t start_;
+      size_t limit_;
    };
 
 

diff --git a/src/lib_json/json_reader.cpp b/src/lib_json/json_reader.cpp
@@ -215,9 +215,11 @@ Reader::readValue()
    {
    case tokenObjectBegin:
       successful = readObject( token );
+      currentValue().setOffsetLimit(current_ - begin_);
       break;
    case tokenArrayBegin:
       successful = readArray( token );
+      currentValue().setOffsetLimit(current_ - begin_);
       break;
    case tokenNumber:
       successful = decodeNumber( token );
@@ -227,12 +229,18 @@ Reader::readValue()
       break;
    case tokenTrue:
       currentValue() = true;
+      currentValue().setOffsetStart(token.start_ - begin_);
+      currentValue().setOffsetLimit(token.end_ - begin_);
       break;
    case tokenFalse:
       currentValue() = false;
+      currentValue().setOffsetStart(token.start_ - begin_);
+      currentValue().setOffsetLimit(token.end_ - begin_);
       break;
    case tokenNull:
       currentValue() = Value();
+      currentValue().setOffsetStart(token.start_ - begin_);
+      currentValue().setOffsetLimit(token.end_ - begin_);
       break;
    case tokenArraySeparator:
       if ( features_.allowDroppedNullPlaceholders_ )
@@ -241,10 +249,14 @@ Reader::readValue()
          // token.
          current_--;
          currentValue() = Value();
+         currentValue().setOffsetStart(current_ - begin_ - 1);
+         currentValue().setOffsetLimit(current_ - begin_);
          break;
       }
       // Else, fall through...
    default:
+      currentValue().setOffsetStart(token.start_ - begin_);
+      currentValue().setOffsetLimit(token.end_ - begin_);
       return addError( "Syntax error: value, object or array expected.", token );
    }
 
@@ -493,11 +505,12 @@ Reader::readString()
 
 
 bool 
-Reader::readObject( Token &/*tokenStart*/ )
+Reader::readObject( Token &tokenStart )
 {
    Token tokenName;
    std::string name;
    currentValue() = Value( objectValue );
+   currentValue().setOffsetStart(tokenStart.start_ - begin_);
    while ( readToken( tokenName ) )
    {
       bool initialTokenOk = true;
@@ -564,9 +577,10 @@ Reader::readObject( Token &/*tokenStart*/ )
 
 
 bool 
-Reader::readArray( Token &/*tokenStart*/ )
+Reader::readArray( Token &tokenStart )
 {
    currentValue() = Value( arrayValue );
+   currentValue().setOffsetStart(tokenStart.start_ - begin_);
    skipSpaces();
    if ( *current_ == ']' ) // empty array
    {
@@ -613,6 +627,8 @@ Reader::decodeNumber( Token &token )
   if ( !decodeNumber( token, decoded ) )
      return false;
   currentValue() = decoded;
+  currentValue().setOffsetStart(token.start_ - begin_);
+  currentValue().setOffsetLimit(token.end_ - begin_);
   return true;
 }
 
@@ -678,6 +694,8 @@ Reader::decodeDouble( Token &token )
   if ( !decodeDouble( token, decoded ) )
      return false;
   currentValue() = decoded;
+  currentValue().setOffsetStart(token.start_ - begin_);
+  currentValue().setOffsetLimit(token.end_ - begin_);
   return true;
 }
 
@@ -729,6 +747,8 @@ Reader::decodeString( Token &token )
    if ( !decodeString( token, decoded ) )
       return false;
    currentValue() = decoded;
+   currentValue().setOffsetStart(token.start_ - begin_);
+   currentValue().setOffsetLimit(token.end_ - begin_);
    return true;
 }
 
@@ -963,6 +983,25 @@ Reader::getFormattedErrorMessages() const
 }
 
 
+std::vector<Reader::StructuredError>
+Reader::getStructuredErrors() const
+{
+   std::vector<Reader::StructuredError> allErrors;
+   for ( Errors::const_iterator itError = errors_.begin();
+         itError != errors_.end();
+         ++itError )
+   {
+      const ErrorInfo &error = *itError;
+      Reader::StructuredError structured;
+      structured.offset_start = error.token_.start_ - begin_;
+      structured.offset_limit = error.token_.end_ - begin_;
+      structured.message = error.message_;
+      allErrors.push_back(structured);
+   }
+   return allErrors;
+}
+
+
 std::istream& operator>>( std::istream &sin, Value &root )
 {
     Json::Reader reader;

diff --git a/src/lib_json/json_value.cpp b/src/lib_json/json_value.cpp
@@ -274,6 +274,8 @@ Value::Value( ValueType type )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    switch ( type )
    {
@@ -318,6 +320,8 @@ Value::Value( UInt value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.uint_ = value;
 }
@@ -329,6 +333,8 @@ Value::Value( Int value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.int_ = value;
 }
@@ -342,6 +348,8 @@ Value::Value( Int64 value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.int_ = value;
 }
@@ -354,6 +362,8 @@ Value::Value( UInt64 value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.uint_ = value;
 }
@@ -366,6 +376,8 @@ Value::Value( double value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.real_ = value;
 }
@@ -377,6 +389,8 @@ Value::Value( const char *value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.string_ = duplicateStringValue( value );
 }
@@ -390,6 +404,8 @@ Value::Value( const char *beginValue,
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.string_ = duplicateStringValue( beginValue, 
                                           (unsigned int)(endValue - beginValue) );
@@ -403,6 +419,8 @@ Value::Value( const std::string &value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.string_ = duplicateStringValue( value.c_str(), 
                                           (unsigned int)value.length() );
@@ -416,6 +434,8 @@ Value::Value( const StaticString &value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.string_ = const_cast<char *>( value.c_str() );
 }
@@ -429,6 +449,8 @@ Value::Value( const CppTL::ConstString &value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.string_ = duplicateStringValue( value, value.length() );
 }
@@ -441,6 +463,8 @@ Value::Value( bool value )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( 0 )
+   , limit_( 0 )
 {
    value_.bool_ = value;
 }
@@ -453,6 +477,8 @@ Value::Value( const Value &other )
    , itemIsUsed_( 0 )
 #endif
    , comments_( 0 )
+   , start_( other.start_ )
+   , limit_( other.limit_ )
 {
    switch ( type_ )
    {
@@ -557,6 +583,8 @@ Value::swap( Value &other )
    int temp2 = allocated_;
    allocated_ = other.allocated_;
    other.allocated_ = temp2;
+   std::swap( start_, other.start_ );
+   std::swap( limit_, other.limit_ );
 }
 
 ValueType 
@@ -1027,7 +1055,8 @@ void
 Value::clear()
 {
    JSON_ASSERT_MESSAGE( type_ == nullValue  ||  type_ == arrayValue  || type_ == objectValue, "in Json::Value::clear(): requires complex value" );
-
+   start_ = 0;
+   limit_ = 0;
    switch ( type_ )
    {
 #ifndef JSON_VALUE_USE_INTERNAL_MAP
@@ -1556,6 +1585,34 @@ Value::getComment( CommentPlacement placement ) const
 }
 
 
+void
+Value::setOffsetStart( size_t start )
+{
+   start_ = start;
+}
+
+
+void
+Value::setOffsetLimit( size_t limit )
+{
+   limit_ = limit;
+}
+
+
+size_t
+Value::getOffsetStart() const
+{
+   return start_;
+}
+
+
+size_t
+Value::getOffsetLimit() const
+{
+   return limit_;
+}
+
+
 std::string 
 Value::toStyledString() const
 {