#ifndef JSON_BINARY_INCLUDED #define JSON_BINARY_INCLUDED /* Copyright (c) 2015, 2019, Oracle and/or its affiliates. All rights reserved. This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License, version 2.0, as published by the Free Software Foundation. This program is also distributed with certain software (including but not limited to OpenSSL) that is licensed under separate terms, as designated in a particular file or component or in included license documentation. The authors of MySQL hereby grant you an additional permission to link the program and your derivative works with the separately licensed software that they have included with MySQL. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License, version 2.0, for more details. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ /** @file This file specifies the interface for serializing JSON values into binary representation, and for reading values back from the binary representation. The binary format is as follows: Each JSON value (scalar, object or array) has a one byte type identifier followed by the actual value. If the value is a JSON object, its binary representation will have a header that contains: - the member count - the size of the binary value in bytes - a list of pointers to each key - a list of pointers to each value The actual keys and values will come after the header, in the same order as in the header. Similarly, if the value is a JSON array, the binary representation will have a header with - the element count - the size of the binary value in bytes - a list of pointers to each value followed by the actual values, in the same order as in the header. @verbatim doc ::= type value type ::= 0x00 | // small JSON object 0x01 | // large JSON object 0x02 | // small JSON array 0x03 | // large JSON array 0x04 | // literal (true/false/null) 0x05 | // int16 0x06 | // uint16 0x07 | // int32 0x08 | // uint32 0x09 | // int64 0x0a | // uint64 0x0b | // double 0x0c | // utf8mb4 string 0x0f // custom data (any MySQL data type) value ::= object | array | literal | number | string | custom-data object ::= element-count size key-entry* value-entry* key* value* array ::= element-count size value-entry* value* // number of members in object or number of elements in array element-count ::= uint16 | // if used in small JSON object/array uint32 // if used in large JSON object/array // number of bytes in the binary representation of the object or array size ::= uint16 | // if used in small JSON object/array uint32 // if used in large JSON object/array key-entry ::= key-offset key-length key-offset ::= uint16 | // if used in small JSON object uint32 // if used in large JSON object key-length ::= uint16 // key length must be less than 64KB value-entry ::= type offset-or-inlined-value // This field holds either the offset to where the value is stored, // or the value itself if it is small enough to be inlined (that is, // if it is a JSON literal or a small enough [u]int). offset-or-inlined-value ::= uint16 | // if used in small JSON object/array uint32 // if used in large JSON object/array key ::= utf8mb4-data literal ::= 0x00 | // JSON null literal 0x01 | // JSON true literal 0x02 | // JSON false literal number ::= .... // little-endian format for [u]int(16|32|64), whereas // double is stored in a platform-independent, eight-byte // format using float8store() string ::= data-length utf8mb4-data custom-data ::= custom-type data-length binary-data custom-type ::= uint8 // type identifier that matches the // internal enum_field_types enum data-length ::= uint8* // If the high bit of a byte is 1, the length // field is continued in the next byte, // otherwise it is the last byte of the length // field. So we need 1 byte to represent // lengths up to 127, 2 bytes to represent // lengths up to 16383, and so on... @endverbatim */ #include #include #include "field_types.h" // enum_field_types #include "my_dbug.h" // DBUG_ASSERT #include "my_inttypes.h" class Field_json; class Json_dom; class Json_wrapper; class String; class THD; namespace json_binary { /** Serialize the JSON document represented by dom to binary format in the destination string, replacing any content already in the destination string. @param[in] thd THD handle @param[in] dom the input DOM tree @param[in,out] dest the destination string @retval false on success @retval true if an error occurred */ #ifdef MYSQL_SERVER bool serialize(const THD *thd, const Json_dom *dom, String *dest); #endif /** Class used for reading JSON values that are stored in the binary format. Values are parsed lazily, so that only the parts of the value that are interesting to the caller, are read. Array elements can be looked up in constant time using the element() function. Object members can be looked up in O(log n) time using the lookup() function. */ class Value { public: enum enum_type : uint8 { OBJECT, ARRAY, STRING, INT, UINT, DOUBLE, LITERAL_NULL, LITERAL_TRUE, LITERAL_FALSE, OPAQUE, ERROR /* Not really a type. Used to signal that an error was detected. */ }; /** Does this value, and all of its members, represent a valid JSON value? */ bool is_valid() const; enum_type type() const { return m_type; } /// Does this value use the large storage format? bool large_format() const { return m_large; } /** Get a pointer to the beginning of the STRING or OPAQUE data represented by this instance. */ const char *get_data() const { DBUG_ASSERT(m_type == STRING || m_type == OPAQUE); return m_data; } /** Get the length in bytes of the STRING or OPAQUE value represented by this instance. */ uint32 get_data_length() const { DBUG_ASSERT(m_type == STRING || m_type == OPAQUE); return m_length; } /** Get the value of an INT. */ int64 get_int64() const { DBUG_ASSERT(m_type == INT); return m_int_value; } /** Get the value of a UINT. */ uint64 get_uint64() const { DBUG_ASSERT(m_type == UINT); return static_cast(m_int_value); } /** Get the value of a DOUBLE. */ double get_double() const { DBUG_ASSERT(m_type == DOUBLE); return m_double_value; } /** Get the number of elements in an array, or the number of members in an object. */ uint32 element_count() const { DBUG_ASSERT(m_type == ARRAY || m_type == OBJECT); return m_element_count; } /** Get the MySQL field type of an opaque value. Identifies the type of the value stored in the data portion of an opaque value. */ enum_field_types field_type() const { DBUG_ASSERT(m_type == OPAQUE); return m_field_type; } Value element(size_t pos) const; Value key(size_t pos) const; Value lookup(const char *key, size_t length) const; Value lookup(const std::string &key) const { return lookup(key.c_str(), key.length()); } size_t lookup_index(const char *key, size_t length) const; size_t lookup_index(const std::string &key) const { return lookup_index(key.c_str(), key.length()); } bool is_backed_by(const String *str) const; bool raw_binary(const THD *thd, String *buf) const; bool get_free_space(const THD *thd, size_t *space) const; bool has_space(size_t pos, size_t needed, size_t *offset) const; bool update_in_shadow(const Field_json *field, size_t pos, Json_wrapper *new_value, size_t data_offset, size_t data_length, const char *original, char *destination, bool *changed) const; bool remove_in_shadow(const Field_json *field, size_t pos, const char *original, char *destination) const; /** Constructor for values that represent literals or errors. */ explicit Value(enum_type t) : m_data(nullptr), m_type(t) { DBUG_ASSERT(t == LITERAL_NULL || t == LITERAL_TRUE || t == LITERAL_FALSE || t == ERROR); } /** Constructor for values that represent ints or uints. */ explicit Value(enum_type t, int64 val) : m_int_value(val), m_type(t) { DBUG_ASSERT(t == INT || t == UINT); } /** Constructor for values that represent doubles. */ explicit Value(double val) : m_double_value(val), m_type(DOUBLE) {} /** Constructor for values that represent strings. */ Value(const char *data, uint32 len) : m_data(data), m_length(len), m_type(STRING) {} /** Constructor for values that represent arrays or objects. @param t type @param data pointer to the start of the binary representation @param bytes the number of bytes in the binary representation of the value @param element_count the number of elements or members in the value @param large true if the value should be stored in the large storage format with 4 byte offsets instead of 2 byte offsets */ Value(enum_type t, const char *data, uint32 bytes, uint32 element_count, bool large) : m_data(data), m_element_count(element_count), m_length(bytes), m_type(t), m_large(large) { DBUG_ASSERT(t == ARRAY || t == OBJECT); } /** Constructor for values that represent opaque data. */ Value(enum_field_types ft, const char *data, uint32 len) : m_data(data), m_length(len), m_field_type(ft), m_type(OPAQUE) {} /** Empty constructor. Produces a value that represents an error condition. */ Value() : Value(ERROR) {} /** Is this value an array? */ bool is_array() const { return m_type == ARRAY; } /** Is this value an object? */ bool is_object() const { return m_type == OBJECT; } /** Compare two Values @note This function is limited to scalars only, for objects/arrays it asserts. The main purpose is to separate old/new scalar values for updates on multi-valued indexes. @returns -1 this < val 0 this == val 1 this > val */ int eq(const Value &val) const; private: /* Instances use only one of m_data, m_int_value and m_double_value, so keep them in a union to save space in memory. */ union { /** Pointer to the start of the binary representation of the value. Only used by STRING, OPAQUE, OBJECT and ARRAY. The memory pointed to by this member is not owned by this Value object. Callers that create Value objects must make sure that the memory is not freed as long as the Value object is alive. */ const char *m_data; /** The value if the type is INT or UINT. */ int64 m_int_value; /** The value if the type is DOUBLE. */ double m_double_value; }; /** Element count for arrays and objects. Unused for other types. */ uint32 m_element_count; /** The full length (in bytes) of the binary representation of an array or object, or the length of a string or opaque value. Unused for other types. */ uint32 m_length; /** The MySQL field type of the value, in case the type of the value is OPAQUE. Otherwise, it is unused. */ enum_field_types m_field_type; /** The JSON type of the value. */ enum_type m_type; /** True if an array or an object uses the large storage format with 4 byte offsets instead of 2 byte offsets. */ bool m_large; size_t key_entry_offset(size_t pos) const; size_t value_entry_offset(size_t pos) const; bool first_value_offset(size_t *offset) const; bool element_offsets(size_t pos, size_t *start, size_t *end, bool *inlined) const; }; /** Parse a JSON binary document. @param[in] data a pointer to the binary data @param[in] len the size of the binary document in bytes @return an object that allows access to the contents of the document */ Value parse_binary(const char *data, size_t len); /** How much space is needed for a JSON value when it is stored in the binary format. @param[in] thd THD handle @param[in] value the JSON value to add to a document @param[in] large true if the large storage format is used @param[out] needed gets set to the amount of bytes needed to store the value @retval false if successful @retval true if an error occurred while calculating the needed space */ #ifdef MYSQL_SERVER bool space_needed(const THD *thd, const Json_wrapper *value, bool large, size_t *needed); #endif /** Apply a function to every value in a JSON document. That is, apply the function to the root node of the JSON document, to all its children, grandchildren and so on. @param value the root of the JSON document @param func the function to apply @retval true if the processing was stopped @retval false if the processing was completed @tparam Func a functor type that takes a #json_binary::Value parameter and returns a `bool` which is `true` if the processing should stop or `false` if the processing should continue with the next node */ template bool for_each_node(const Value &value, const Func &func) { if (func(value)) return true; if (value.is_array() || value.is_object()) for (size_t i = 0, size = value.element_count(); i < size; ++i) if (for_each_node(value.element(i), func)) return true; return false; } } // namespace json_binary #endif /* JSON_BINARY_INCLUDED */