Class: Takagi::CBOR::Decoder
- Inherits:
-
Object
- Object
- Takagi::CBOR::Decoder
- Defined in:
- lib/takagi/cbor/decoder.rb,
sig/takagi/cbor/decoder.rbs
Overview
CBOR Decoder (RFC 8949)
Decodes CBOR binary format to Ruby objects. Optimized for IoT/CoAP workloads with minimal footprint.
Supported types:
- Integers → Integer
- Floats → Float
- Text strings → String (UTF-8)
- Byte strings → String (binary)
- Arrays → Array
- Maps → Hash (string keys)
- Booleans → true/false
- null → nil
- Timestamps (tag 1) → Time
Security features:
- Max nesting depth (prevents stack overflow)
- Max collection size (prevents memory exhaustion)
Constant Summary collapse
- MAJOR_TYPE_UNSIGNED_INT =
CBOR Major Types (RFC 8949 §3)
0- MAJOR_TYPE_NEGATIVE_INT =
1- MAJOR_TYPE_BYTE_STRING =
2- MAJOR_TYPE_TEXT_STRING =
3- MAJOR_TYPE_ARRAY =
4- MAJOR_TYPE_MAP =
5- MAJOR_TYPE_TAG =
6- MAJOR_TYPE_SIMPLE =
7- SIMPLE_FALSE =
Simple values (RFC 8949 §3.3)
20- SIMPLE_TRUE =
21- SIMPLE_NULL =
22- SIMPLE_FLOAT16 =
25- SIMPLE_FLOAT32 =
26- SIMPLE_FLOAT64 =
27- MAX_NESTING_DEPTH =
Security limits
100- MAX_COLLECTION_SIZE =
100_000- TAG_EPOCH_TIMESTAMP =
Tag values
1- MAJOR_TYPE_HANDLERS =
{ MAJOR_TYPE_UNSIGNED_INT => :handle_unsigned_int, MAJOR_TYPE_NEGATIVE_INT => :handle_negative_int, MAJOR_TYPE_BYTE_STRING => :read_bytes, MAJOR_TYPE_TEXT_STRING => :read_string, MAJOR_TYPE_ARRAY => :read_array, MAJOR_TYPE_MAP => :read_map, MAJOR_TYPE_TAG => :read_tagged, MAJOR_TYPE_SIMPLE => :read_simple }.freeze
Class Method Summary collapse
-
.decode(bytes) ⇒ Object
Decode CBOR bytes to Ruby object.
Instance Method Summary collapse
-
#check_available(needed) ⇒ nil, untyped
Check if enough bytes are available.
-
#check_collection_size(size) ⇒ nil, untyped
Check collection size to prevent memory exhaustion.
-
#check_depth! ⇒ nil, untyped
Check nesting depth to prevent stack overflow.
-
#decode ⇒ Object
Decode CBOR bytes to Ruby object.
-
#decode_additional_info(additional) ⇒ Object
Decode additional information (RFC 8949 §3).
- #handle_negative_int(value) ⇒ Object
- #handle_unsigned_int(value) ⇒ Object
-
#initialize(bytes) ⇒ Decoder
constructor
Initialize decoder with CBOR bytes.
-
#read_array(length) ⇒ Object
Read array.
-
#read_bytes(length) ⇒ Object
Read byte string (binary data).
-
#read_float16 ⇒ Object
Read 16-bit float (IEEE 754 half-precision).
-
#read_float32 ⇒ Object
Read 32-bit float (IEEE 754 single-precision).
-
#read_float64 ⇒ Object
Read 64-bit float (IEEE 754 double-precision).
-
#read_map(length) ⇒ Object
Read map (hash).
-
#read_simple(value) ⇒ Object
Read simple value (RFC 8949 §3.3).
-
#read_string(length) ⇒ Object
Read UTF-8 text string.
-
#read_tagged(tag) ⇒ Object
Read tagged value (RFC 8949 §3.4).
-
#read_type_and_value ⇒ ::Array[untyped]
Read major type and additional value RFC 8949 §3: Initial byte encoding.
-
#read_uint16 ⇒ Object
Read unsigned 16-bit integer (big-endian).
-
#read_uint32 ⇒ Object
Read unsigned 32-bit integer (big-endian).
-
#read_uint64 ⇒ Object
Read unsigned 64-bit integer (big-endian).
-
#read_uint8 ⇒ Object
Read unsigned 8-bit integer.
Constructor Details
#initialize(bytes) ⇒ Decoder
Initialize decoder with CBOR bytes
84 85 86 87 88 |
# File 'lib/takagi/cbor/decoder.rb', line 84 def initialize(bytes) @bytes = bytes.b @pos = 0 @depth = 0 end |
Class Method Details
.decode(bytes) ⇒ Object
Decode CBOR bytes to Ruby object
76 77 78 |
# File 'lib/takagi/cbor/decoder.rb', line 76 def decode(bytes) new(bytes).decode end |
Instance Method Details
#check_available(needed) ⇒ nil, untyped
Check if enough bytes are available
370 371 372 373 374 375 |
# File 'lib/takagi/cbor/decoder.rb', line 370 def check_available(needed) available = @bytes.bytesize - @pos return if available >= needed raise DecodeError, "Unexpected end of input (need #{needed} bytes, have #{available})" end |
#check_collection_size(size) ⇒ nil, untyped
Check collection size to prevent memory exhaustion
378 379 380 381 382 |
# File 'lib/takagi/cbor/decoder.rb', line 378 def check_collection_size(size) return if size <= MAX_COLLECTION_SIZE raise DecodeError, "Collection size #{size} exceeds maximum (#{MAX_COLLECTION_SIZE})" end |
#check_depth! ⇒ nil, untyped
Check nesting depth to prevent stack overflow
119 120 121 122 123 |
# File 'lib/takagi/cbor/decoder.rb', line 119 def check_depth! return if @depth < MAX_NESTING_DEPTH raise DecodeError, "Maximum nesting depth exceeded (#{MAX_NESTING_DEPTH})" end |
#decode ⇒ Object
Decode CBOR bytes to Ruby object
94 95 96 97 98 99 100 101 102 103 104 105 106 |
# File 'lib/takagi/cbor/decoder.rb', line 94 def decode check_depth! major_type, value = read_type_and_value handler = MAJOR_TYPE_HANDLERS[major_type] raise DecodeError, "Unknown major type: #{major_type}" unless handler send(handler, value) rescue DecodeError, UnsupportedError raise rescue StandardError => e raise DecodeError, "Decoding failed at position #{@pos}: #{e.}" end |
#decode_additional_info(additional) ⇒ Object
Decode additional information (RFC 8949 §3)
Additional info encoding: 0-23: Value is directly in additional info 24: 1-byte uint8 follows 25: 2-byte uint16 follows 26: 4-byte uint32 follows 27: 8-byte uint64 follows 28-30: Reserved (error) 31: Indefinite length (not supported in minimal impl)
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 |
# File 'lib/takagi/cbor/decoder.rb', line 159 def decode_additional_info(additional) case additional when 0..23 # Value directly encoded additional when 24 # 1-byte uint8 follows read_uint8 when 25 # 2-byte uint16 follows read_uint16 when 26 # 4-byte uint32 follows read_uint32 when 27 # 8-byte uint64 follows read_uint64 when 28, 29, 30 raise DecodeError, "Reserved additional info value: #{additional}" when 31 raise UnsupportedError, 'Indefinite-length items not supported in minimal implementation' end end |
#handle_negative_int(value) ⇒ Object
114 115 116 |
# File 'lib/takagi/cbor/decoder.rb', line 114 def handle_negative_int(value) -1 - value end |
#handle_unsigned_int(value) ⇒ Object
110 111 112 |
# File 'lib/takagi/cbor/decoder.rb', line 110 def handle_unsigned_int(value) value end |
#read_array(length) ⇒ Object
Read array
242 243 244 245 246 247 248 249 250 251 |
# File 'lib/takagi/cbor/decoder.rb', line 242 def read_array(length) check_collection_size(length) @depth += 1 arr = Array.new(length) { decode } @depth -= 1 arr end |
#read_bytes(length) ⇒ Object
Read byte string (binary data)
216 217 218 219 220 221 222 223 |
# File 'lib/takagi/cbor/decoder.rb', line 216 def read_bytes(length) check_collection_size(length) check_available(length) bytes = @bytes[@pos, length] @pos += length bytes end |
#read_float16 ⇒ Object
Read 16-bit float (IEEE 754 half-precision)
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 |
# File 'lib/takagi/cbor/decoder.rb', line 326 def read_float16 check_available(2) # Read 16-bit big-endian half = @bytes[@pos, 2].unpack1('n') @pos += 2 # Convert IEEE 754 half to Ruby float # Format: 1 sign bit, 5 exponent bits, 10 mantissa bits sign = (half >> 15) & 0x1 exponent = (half >> 10) & 0x1F mantissa = half & 0x3FF if exponent.zero? # Subnormal or zero result = mantissa.to_f / (2**24) elsif exponent == 0x1F # Infinity or NaN return mantissa.zero? ? Float::INFINITY : Float::NAN else # Normalized result = (1.0 + (mantissa.to_f / (2**10))) * (2**(exponent - 15)) end sign.zero? ? result : -result end |
#read_float32 ⇒ Object
Read 32-bit float (IEEE 754 single-precision)
354 355 356 357 358 359 |
# File 'lib/takagi/cbor/decoder.rb', line 354 def read_float32 check_available(4) float_bytes = @bytes[@pos, 4] @pos += 4 float_bytes.unpack1('g') # Big-endian single-precision float end |
#read_float64 ⇒ Object
Read 64-bit float (IEEE 754 double-precision)
362 363 364 365 366 367 |
# File 'lib/takagi/cbor/decoder.rb', line 362 def read_float64 check_available(8) float_bytes = @bytes[@pos, 8] @pos += 8 float_bytes.unpack1('G') # Big-endian double-precision float end |
#read_map(length) ⇒ Object
Read map (hash)
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 |
# File 'lib/takagi/cbor/decoder.rb', line 254 def read_map(length) check_collection_size(length) @depth += 1 hash = {} length.times do key = decode value = decode # Convert symbol keys to strings for consistency key = key.to_s if key.is_a?(Symbol) hash[key] = value end @depth -= 1 hash end |
#read_simple(value) ⇒ Object
Read simple value (RFC 8949 §3.3)
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 |
# File 'lib/takagi/cbor/decoder.rb', line 296 def read_simple(value) case value when SIMPLE_FALSE false when SIMPLE_TRUE true when SIMPLE_NULL nil when SIMPLE_FLOAT16 # 16-bit float (half-precision) read_float16 when SIMPLE_FLOAT32 # 32-bit float (single-precision) read_float32 when SIMPLE_FLOAT64 # 64-bit float (double-precision) read_float64 when 0..19 # Unassigned simple values (0-19) raise UnsupportedError, "Unassigned simple value: #{value}" when 24..31 # Should not reach here (handled in decode_additional_info) raise DecodeError, "Invalid simple value: #{value}" else # Simple values 32-255 (extended) raise UnsupportedError, "Extended simple values not supported: #{value}" end end |
#read_string(length) ⇒ Object
Read UTF-8 text string
226 227 228 229 230 231 232 233 234 235 236 237 238 239 |
# File 'lib/takagi/cbor/decoder.rb', line 226 def read_string(length) check_collection_size(length) check_available(length) str = @bytes[@pos, length] @pos += length # Force UTF-8 encoding and validate str.force_encoding('UTF-8') raise DecodeError, 'Invalid UTF-8 encoding in text string' unless str.valid_encoding? str end |
#read_tagged(tag) ⇒ Object
Read tagged value (RFC 8949 §3.4)
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 |
# File 'lib/takagi/cbor/decoder.rb', line 276 def read_tagged(tag) case tag when TAG_EPOCH_TIMESTAMP # Tag 1: Epoch-based timestamp = decode case when Integer, Float Time.at() else raise DecodeError, "Invalid timestamp value type: #{.class}" end else # Unknown tag: decode value but ignore tag # This allows forward compatibility decode end end |
#read_type_and_value ⇒ ::Array[untyped]
Read major type and additional value RFC 8949 §3: Initial byte encoding
Returns [major_type, value]
- major_type: 0-7 (3 bits)
- value: depends on additional info (5 bits)
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 |
# File 'lib/takagi/cbor/decoder.rb', line 131 def read_type_and_value raise DecodeError, 'Unexpected end of input' if @pos >= @bytes.bytesize initial_byte = @bytes[@pos].ord @pos += 1 major_type = initial_byte >> 5 additional = initial_byte & 0x1F value = if major_type == MAJOR_TYPE_SIMPLE additional else decode_additional_info(additional) end [major_type, value] end |
#read_uint16 ⇒ Object
Read unsigned 16-bit integer (big-endian)
192 193 194 195 196 197 |
# File 'lib/takagi/cbor/decoder.rb', line 192 def read_uint16 check_available(2) val = @bytes[@pos, 2].unpack1('n') @pos += 2 val end |
#read_uint32 ⇒ Object
Read unsigned 32-bit integer (big-endian)
200 201 202 203 204 205 |
# File 'lib/takagi/cbor/decoder.rb', line 200 def read_uint32 check_available(4) val = @bytes[@pos, 4].unpack1('N') @pos += 4 val end |
#read_uint64 ⇒ Object
Read unsigned 64-bit integer (big-endian)
208 209 210 211 212 213 |
# File 'lib/takagi/cbor/decoder.rb', line 208 def read_uint64 check_available(8) val = @bytes[@pos, 8].unpack1('Q>') @pos += 8 val end |
#read_uint8 ⇒ Object
Read unsigned 8-bit integer
184 185 186 187 188 189 |
# File 'lib/takagi/cbor/decoder.rb', line 184 def read_uint8 check_available(1) val = @bytes[@pos].ord @pos += 1 val end |