Allocates and populate a JsonArray from a JSON string.

By design, the parser needs to alter the string to insert null-terminators and replace escaped chars. If the JSON string is read-only, it will have to duplicate the input string, this consume more space in the JsonBuffer. Therefore, it’s recommended to have a JSON input in a char[] or a char*.


// The first overload, which accepts a modifiable array of chars, is the most efficient
// since it allows the zero-copy feature.
JsonArray& parseArray(char* json, uint8_t nestingLimit=10);

// The following overloads, which accept read-only strings, require a bigger JsonBuffer
// because parts of the JSON input has to be copied.
JsonArray& parseArray(const char* json, uint8_t nestingLimit=10);
JsonArray& parseArray(const String& json, uint8_t nestingLimit=10);
JsonArray& parseArray(const std::string& json, uint8_t nestingLimit=10);
JsonArray& parseArray(const __FlashStringHelper* json, uint8_t nestingLimit=10);

// The two last overloads, which accept input streams, make copy of the input too.
JsonArray& parseArray(Stream& json, uint8_t nestingLimit=10);
JsonArray& parseArray(std::istream& json, uint8_t nestingLimit=10);


json is the input string to be parsed.

nestingLimit specifies the maximum level of nesting allowed in the JSON string. If set to 0, only a flat array can be parsed. If set to 1, the array can contain nested arrays or objects but only 1 level deep. And bigger values will allow more level of nesting. The purpose of this feature is to prevent stack overflow that could lead to a security risk.

Return value

Returns a reference to the new JsonArray or JsonArray::invalid() if the allocation fails.

How to view the JSON input?

When you pass a Stream to JsonBuffer::parseArray(), it consumes the input but doesn’t print anything to the serial port, which makes troubleshooting difficult.

If you want to see what JsonBuffer::parseArray() consumed, use ReadLoggingStream from the StreamUtils library.


When you pass a Stream to JsonBuffer::parseArray(), it consumes bytes one by one, which can be slow depending on the input you use. For example, if you read from a SPIFFS file, you can read twenty times faster by reading chunks of 64 bytes.

To read the stream in chunks, you can use ReadBufferingStream from the StreamUtils library.


char json[] = "[\"hello\",\"world\"]";
StaticJsonBuffer<200> jsonBuffer;
JsonArray& array = jsonBuffer.parseArray(json);
const char* hello = array[0];
const char* world = array[1];

See also