Scott Baker | eee8dd8 | 2019-09-24 12:52:34 -0700 | [diff] [blame] | 1 | package yaml |
| 2 | |
| 3 | import ( |
| 4 | "bytes" |
| 5 | "fmt" |
| 6 | ) |
| 7 | |
| 8 | // Introduction |
| 9 | // ************ |
| 10 | // |
| 11 | // The following notes assume that you are familiar with the YAML specification |
| 12 | // (http://yaml.org/spec/1.2/spec.html). We mostly follow it, although in |
| 13 | // some cases we are less restrictive that it requires. |
| 14 | // |
| 15 | // The process of transforming a YAML stream into a sequence of events is |
| 16 | // divided on two steps: Scanning and Parsing. |
| 17 | // |
| 18 | // The Scanner transforms the input stream into a sequence of tokens, while the |
| 19 | // parser transform the sequence of tokens produced by the Scanner into a |
| 20 | // sequence of parsing events. |
| 21 | // |
| 22 | // The Scanner is rather clever and complicated. The Parser, on the contrary, |
| 23 | // is a straightforward implementation of a recursive-descendant parser (or, |
| 24 | // LL(1) parser, as it is usually called). |
| 25 | // |
| 26 | // Actually there are two issues of Scanning that might be called "clever", the |
| 27 | // rest is quite straightforward. The issues are "block collection start" and |
| 28 | // "simple keys". Both issues are explained below in details. |
| 29 | // |
| 30 | // Here the Scanning step is explained and implemented. We start with the list |
| 31 | // of all the tokens produced by the Scanner together with short descriptions. |
| 32 | // |
| 33 | // Now, tokens: |
| 34 | // |
| 35 | // STREAM-START(encoding) # The stream start. |
| 36 | // STREAM-END # The stream end. |
| 37 | // VERSION-DIRECTIVE(major,minor) # The '%YAML' directive. |
| 38 | // TAG-DIRECTIVE(handle,prefix) # The '%TAG' directive. |
| 39 | // DOCUMENT-START # '---' |
| 40 | // DOCUMENT-END # '...' |
| 41 | // BLOCK-SEQUENCE-START # Indentation increase denoting a block |
| 42 | // BLOCK-MAPPING-START # sequence or a block mapping. |
| 43 | // BLOCK-END # Indentation decrease. |
| 44 | // FLOW-SEQUENCE-START # '[' |
| 45 | // FLOW-SEQUENCE-END # ']' |
| 46 | // BLOCK-SEQUENCE-START # '{' |
| 47 | // BLOCK-SEQUENCE-END # '}' |
| 48 | // BLOCK-ENTRY # '-' |
| 49 | // FLOW-ENTRY # ',' |
| 50 | // KEY # '?' or nothing (simple keys). |
| 51 | // VALUE # ':' |
| 52 | // ALIAS(anchor) # '*anchor' |
| 53 | // ANCHOR(anchor) # '&anchor' |
| 54 | // TAG(handle,suffix) # '!handle!suffix' |
| 55 | // SCALAR(value,style) # A scalar. |
| 56 | // |
| 57 | // The following two tokens are "virtual" tokens denoting the beginning and the |
| 58 | // end of the stream: |
| 59 | // |
| 60 | // STREAM-START(encoding) |
| 61 | // STREAM-END |
| 62 | // |
| 63 | // We pass the information about the input stream encoding with the |
| 64 | // STREAM-START token. |
| 65 | // |
| 66 | // The next two tokens are responsible for tags: |
| 67 | // |
| 68 | // VERSION-DIRECTIVE(major,minor) |
| 69 | // TAG-DIRECTIVE(handle,prefix) |
| 70 | // |
| 71 | // Example: |
| 72 | // |
| 73 | // %YAML 1.1 |
| 74 | // %TAG ! !foo |
| 75 | // %TAG !yaml! tag:yaml.org,2002: |
| 76 | // --- |
| 77 | // |
| 78 | // The correspoding sequence of tokens: |
| 79 | // |
| 80 | // STREAM-START(utf-8) |
| 81 | // VERSION-DIRECTIVE(1,1) |
| 82 | // TAG-DIRECTIVE("!","!foo") |
| 83 | // TAG-DIRECTIVE("!yaml","tag:yaml.org,2002:") |
| 84 | // DOCUMENT-START |
| 85 | // STREAM-END |
| 86 | // |
| 87 | // Note that the VERSION-DIRECTIVE and TAG-DIRECTIVE tokens occupy a whole |
| 88 | // line. |
| 89 | // |
| 90 | // The document start and end indicators are represented by: |
| 91 | // |
| 92 | // DOCUMENT-START |
| 93 | // DOCUMENT-END |
| 94 | // |
| 95 | // Note that if a YAML stream contains an implicit document (without '---' |
| 96 | // and '...' indicators), no DOCUMENT-START and DOCUMENT-END tokens will be |
| 97 | // produced. |
| 98 | // |
| 99 | // In the following examples, we present whole documents together with the |
| 100 | // produced tokens. |
| 101 | // |
| 102 | // 1. An implicit document: |
| 103 | // |
| 104 | // 'a scalar' |
| 105 | // |
| 106 | // Tokens: |
| 107 | // |
| 108 | // STREAM-START(utf-8) |
| 109 | // SCALAR("a scalar",single-quoted) |
| 110 | // STREAM-END |
| 111 | // |
| 112 | // 2. An explicit document: |
| 113 | // |
| 114 | // --- |
| 115 | // 'a scalar' |
| 116 | // ... |
| 117 | // |
| 118 | // Tokens: |
| 119 | // |
| 120 | // STREAM-START(utf-8) |
| 121 | // DOCUMENT-START |
| 122 | // SCALAR("a scalar",single-quoted) |
| 123 | // DOCUMENT-END |
| 124 | // STREAM-END |
| 125 | // |
| 126 | // 3. Several documents in a stream: |
| 127 | // |
| 128 | // 'a scalar' |
| 129 | // --- |
| 130 | // 'another scalar' |
| 131 | // --- |
| 132 | // 'yet another scalar' |
| 133 | // |
| 134 | // Tokens: |
| 135 | // |
| 136 | // STREAM-START(utf-8) |
| 137 | // SCALAR("a scalar",single-quoted) |
| 138 | // DOCUMENT-START |
| 139 | // SCALAR("another scalar",single-quoted) |
| 140 | // DOCUMENT-START |
| 141 | // SCALAR("yet another scalar",single-quoted) |
| 142 | // STREAM-END |
| 143 | // |
| 144 | // We have already introduced the SCALAR token above. The following tokens are |
| 145 | // used to describe aliases, anchors, tag, and scalars: |
| 146 | // |
| 147 | // ALIAS(anchor) |
| 148 | // ANCHOR(anchor) |
| 149 | // TAG(handle,suffix) |
| 150 | // SCALAR(value,style) |
| 151 | // |
| 152 | // The following series of examples illustrate the usage of these tokens: |
| 153 | // |
| 154 | // 1. A recursive sequence: |
| 155 | // |
| 156 | // &A [ *A ] |
| 157 | // |
| 158 | // Tokens: |
| 159 | // |
| 160 | // STREAM-START(utf-8) |
| 161 | // ANCHOR("A") |
| 162 | // FLOW-SEQUENCE-START |
| 163 | // ALIAS("A") |
| 164 | // FLOW-SEQUENCE-END |
| 165 | // STREAM-END |
| 166 | // |
| 167 | // 2. A tagged scalar: |
| 168 | // |
| 169 | // !!float "3.14" # A good approximation. |
| 170 | // |
| 171 | // Tokens: |
| 172 | // |
| 173 | // STREAM-START(utf-8) |
| 174 | // TAG("!!","float") |
| 175 | // SCALAR("3.14",double-quoted) |
| 176 | // STREAM-END |
| 177 | // |
| 178 | // 3. Various scalar styles: |
| 179 | // |
| 180 | // --- # Implicit empty plain scalars do not produce tokens. |
| 181 | // --- a plain scalar |
| 182 | // --- 'a single-quoted scalar' |
| 183 | // --- "a double-quoted scalar" |
| 184 | // --- |- |
| 185 | // a literal scalar |
| 186 | // --- >- |
| 187 | // a folded |
| 188 | // scalar |
| 189 | // |
| 190 | // Tokens: |
| 191 | // |
| 192 | // STREAM-START(utf-8) |
| 193 | // DOCUMENT-START |
| 194 | // DOCUMENT-START |
| 195 | // SCALAR("a plain scalar",plain) |
| 196 | // DOCUMENT-START |
| 197 | // SCALAR("a single-quoted scalar",single-quoted) |
| 198 | // DOCUMENT-START |
| 199 | // SCALAR("a double-quoted scalar",double-quoted) |
| 200 | // DOCUMENT-START |
| 201 | // SCALAR("a literal scalar",literal) |
| 202 | // DOCUMENT-START |
| 203 | // SCALAR("a folded scalar",folded) |
| 204 | // STREAM-END |
| 205 | // |
| 206 | // Now it's time to review collection-related tokens. We will start with |
| 207 | // flow collections: |
| 208 | // |
| 209 | // FLOW-SEQUENCE-START |
| 210 | // FLOW-SEQUENCE-END |
| 211 | // FLOW-MAPPING-START |
| 212 | // FLOW-MAPPING-END |
| 213 | // FLOW-ENTRY |
| 214 | // KEY |
| 215 | // VALUE |
| 216 | // |
| 217 | // The tokens FLOW-SEQUENCE-START, FLOW-SEQUENCE-END, FLOW-MAPPING-START, and |
| 218 | // FLOW-MAPPING-END represent the indicators '[', ']', '{', and '}' |
| 219 | // correspondingly. FLOW-ENTRY represent the ',' indicator. Finally the |
| 220 | // indicators '?' and ':', which are used for denoting mapping keys and values, |
| 221 | // are represented by the KEY and VALUE tokens. |
| 222 | // |
| 223 | // The following examples show flow collections: |
| 224 | // |
| 225 | // 1. A flow sequence: |
| 226 | // |
| 227 | // [item 1, item 2, item 3] |
| 228 | // |
| 229 | // Tokens: |
| 230 | // |
| 231 | // STREAM-START(utf-8) |
| 232 | // FLOW-SEQUENCE-START |
| 233 | // SCALAR("item 1",plain) |
| 234 | // FLOW-ENTRY |
| 235 | // SCALAR("item 2",plain) |
| 236 | // FLOW-ENTRY |
| 237 | // SCALAR("item 3",plain) |
| 238 | // FLOW-SEQUENCE-END |
| 239 | // STREAM-END |
| 240 | // |
| 241 | // 2. A flow mapping: |
| 242 | // |
| 243 | // { |
| 244 | // a simple key: a value, # Note that the KEY token is produced. |
| 245 | // ? a complex key: another value, |
| 246 | // } |
| 247 | // |
| 248 | // Tokens: |
| 249 | // |
| 250 | // STREAM-START(utf-8) |
| 251 | // FLOW-MAPPING-START |
| 252 | // KEY |
| 253 | // SCALAR("a simple key",plain) |
| 254 | // VALUE |
| 255 | // SCALAR("a value",plain) |
| 256 | // FLOW-ENTRY |
| 257 | // KEY |
| 258 | // SCALAR("a complex key",plain) |
| 259 | // VALUE |
| 260 | // SCALAR("another value",plain) |
| 261 | // FLOW-ENTRY |
| 262 | // FLOW-MAPPING-END |
| 263 | // STREAM-END |
| 264 | // |
| 265 | // A simple key is a key which is not denoted by the '?' indicator. Note that |
| 266 | // the Scanner still produce the KEY token whenever it encounters a simple key. |
| 267 | // |
| 268 | // For scanning block collections, the following tokens are used (note that we |
| 269 | // repeat KEY and VALUE here): |
| 270 | // |
| 271 | // BLOCK-SEQUENCE-START |
| 272 | // BLOCK-MAPPING-START |
| 273 | // BLOCK-END |
| 274 | // BLOCK-ENTRY |
| 275 | // KEY |
| 276 | // VALUE |
| 277 | // |
| 278 | // The tokens BLOCK-SEQUENCE-START and BLOCK-MAPPING-START denote indentation |
| 279 | // increase that precedes a block collection (cf. the INDENT token in Python). |
| 280 | // The token BLOCK-END denote indentation decrease that ends a block collection |
| 281 | // (cf. the DEDENT token in Python). However YAML has some syntax pecularities |
| 282 | // that makes detections of these tokens more complex. |
| 283 | // |
| 284 | // The tokens BLOCK-ENTRY, KEY, and VALUE are used to represent the indicators |
| 285 | // '-', '?', and ':' correspondingly. |
| 286 | // |
| 287 | // The following examples show how the tokens BLOCK-SEQUENCE-START, |
| 288 | // BLOCK-MAPPING-START, and BLOCK-END are emitted by the Scanner: |
| 289 | // |
| 290 | // 1. Block sequences: |
| 291 | // |
| 292 | // - item 1 |
| 293 | // - item 2 |
| 294 | // - |
| 295 | // - item 3.1 |
| 296 | // - item 3.2 |
| 297 | // - |
| 298 | // key 1: value 1 |
| 299 | // key 2: value 2 |
| 300 | // |
| 301 | // Tokens: |
| 302 | // |
| 303 | // STREAM-START(utf-8) |
| 304 | // BLOCK-SEQUENCE-START |
| 305 | // BLOCK-ENTRY |
| 306 | // SCALAR("item 1",plain) |
| 307 | // BLOCK-ENTRY |
| 308 | // SCALAR("item 2",plain) |
| 309 | // BLOCK-ENTRY |
| 310 | // BLOCK-SEQUENCE-START |
| 311 | // BLOCK-ENTRY |
| 312 | // SCALAR("item 3.1",plain) |
| 313 | // BLOCK-ENTRY |
| 314 | // SCALAR("item 3.2",plain) |
| 315 | // BLOCK-END |
| 316 | // BLOCK-ENTRY |
| 317 | // BLOCK-MAPPING-START |
| 318 | // KEY |
| 319 | // SCALAR("key 1",plain) |
| 320 | // VALUE |
| 321 | // SCALAR("value 1",plain) |
| 322 | // KEY |
| 323 | // SCALAR("key 2",plain) |
| 324 | // VALUE |
| 325 | // SCALAR("value 2",plain) |
| 326 | // BLOCK-END |
| 327 | // BLOCK-END |
| 328 | // STREAM-END |
| 329 | // |
| 330 | // 2. Block mappings: |
| 331 | // |
| 332 | // a simple key: a value # The KEY token is produced here. |
| 333 | // ? a complex key |
| 334 | // : another value |
| 335 | // a mapping: |
| 336 | // key 1: value 1 |
| 337 | // key 2: value 2 |
| 338 | // a sequence: |
| 339 | // - item 1 |
| 340 | // - item 2 |
| 341 | // |
| 342 | // Tokens: |
| 343 | // |
| 344 | // STREAM-START(utf-8) |
| 345 | // BLOCK-MAPPING-START |
| 346 | // KEY |
| 347 | // SCALAR("a simple key",plain) |
| 348 | // VALUE |
| 349 | // SCALAR("a value",plain) |
| 350 | // KEY |
| 351 | // SCALAR("a complex key",plain) |
| 352 | // VALUE |
| 353 | // SCALAR("another value",plain) |
| 354 | // KEY |
| 355 | // SCALAR("a mapping",plain) |
| 356 | // BLOCK-MAPPING-START |
| 357 | // KEY |
| 358 | // SCALAR("key 1",plain) |
| 359 | // VALUE |
| 360 | // SCALAR("value 1",plain) |
| 361 | // KEY |
| 362 | // SCALAR("key 2",plain) |
| 363 | // VALUE |
| 364 | // SCALAR("value 2",plain) |
| 365 | // BLOCK-END |
| 366 | // KEY |
| 367 | // SCALAR("a sequence",plain) |
| 368 | // VALUE |
| 369 | // BLOCK-SEQUENCE-START |
| 370 | // BLOCK-ENTRY |
| 371 | // SCALAR("item 1",plain) |
| 372 | // BLOCK-ENTRY |
| 373 | // SCALAR("item 2",plain) |
| 374 | // BLOCK-END |
| 375 | // BLOCK-END |
| 376 | // STREAM-END |
| 377 | // |
| 378 | // YAML does not always require to start a new block collection from a new |
| 379 | // line. If the current line contains only '-', '?', and ':' indicators, a new |
| 380 | // block collection may start at the current line. The following examples |
| 381 | // illustrate this case: |
| 382 | // |
| 383 | // 1. Collections in a sequence: |
| 384 | // |
| 385 | // - - item 1 |
| 386 | // - item 2 |
| 387 | // - key 1: value 1 |
| 388 | // key 2: value 2 |
| 389 | // - ? complex key |
| 390 | // : complex value |
| 391 | // |
| 392 | // Tokens: |
| 393 | // |
| 394 | // STREAM-START(utf-8) |
| 395 | // BLOCK-SEQUENCE-START |
| 396 | // BLOCK-ENTRY |
| 397 | // BLOCK-SEQUENCE-START |
| 398 | // BLOCK-ENTRY |
| 399 | // SCALAR("item 1",plain) |
| 400 | // BLOCK-ENTRY |
| 401 | // SCALAR("item 2",plain) |
| 402 | // BLOCK-END |
| 403 | // BLOCK-ENTRY |
| 404 | // BLOCK-MAPPING-START |
| 405 | // KEY |
| 406 | // SCALAR("key 1",plain) |
| 407 | // VALUE |
| 408 | // SCALAR("value 1",plain) |
| 409 | // KEY |
| 410 | // SCALAR("key 2",plain) |
| 411 | // VALUE |
| 412 | // SCALAR("value 2",plain) |
| 413 | // BLOCK-END |
| 414 | // BLOCK-ENTRY |
| 415 | // BLOCK-MAPPING-START |
| 416 | // KEY |
| 417 | // SCALAR("complex key") |
| 418 | // VALUE |
| 419 | // SCALAR("complex value") |
| 420 | // BLOCK-END |
| 421 | // BLOCK-END |
| 422 | // STREAM-END |
| 423 | // |
| 424 | // 2. Collections in a mapping: |
| 425 | // |
| 426 | // ? a sequence |
| 427 | // : - item 1 |
| 428 | // - item 2 |
| 429 | // ? a mapping |
| 430 | // : key 1: value 1 |
| 431 | // key 2: value 2 |
| 432 | // |
| 433 | // Tokens: |
| 434 | // |
| 435 | // STREAM-START(utf-8) |
| 436 | // BLOCK-MAPPING-START |
| 437 | // KEY |
| 438 | // SCALAR("a sequence",plain) |
| 439 | // VALUE |
| 440 | // BLOCK-SEQUENCE-START |
| 441 | // BLOCK-ENTRY |
| 442 | // SCALAR("item 1",plain) |
| 443 | // BLOCK-ENTRY |
| 444 | // SCALAR("item 2",plain) |
| 445 | // BLOCK-END |
| 446 | // KEY |
| 447 | // SCALAR("a mapping",plain) |
| 448 | // VALUE |
| 449 | // BLOCK-MAPPING-START |
| 450 | // KEY |
| 451 | // SCALAR("key 1",plain) |
| 452 | // VALUE |
| 453 | // SCALAR("value 1",plain) |
| 454 | // KEY |
| 455 | // SCALAR("key 2",plain) |
| 456 | // VALUE |
| 457 | // SCALAR("value 2",plain) |
| 458 | // BLOCK-END |
| 459 | // BLOCK-END |
| 460 | // STREAM-END |
| 461 | // |
| 462 | // YAML also permits non-indented sequences if they are included into a block |
| 463 | // mapping. In this case, the token BLOCK-SEQUENCE-START is not produced: |
| 464 | // |
| 465 | // key: |
| 466 | // - item 1 # BLOCK-SEQUENCE-START is NOT produced here. |
| 467 | // - item 2 |
| 468 | // |
| 469 | // Tokens: |
| 470 | // |
| 471 | // STREAM-START(utf-8) |
| 472 | // BLOCK-MAPPING-START |
| 473 | // KEY |
| 474 | // SCALAR("key",plain) |
| 475 | // VALUE |
| 476 | // BLOCK-ENTRY |
| 477 | // SCALAR("item 1",plain) |
| 478 | // BLOCK-ENTRY |
| 479 | // SCALAR("item 2",plain) |
| 480 | // BLOCK-END |
| 481 | // |
| 482 | |
| 483 | // Ensure that the buffer contains the required number of characters. |
| 484 | // Return true on success, false on failure (reader error or memory error). |
| 485 | func cache(parser *yaml_parser_t, length int) bool { |
| 486 | // [Go] This was inlined: !cache(A, B) -> unread < B && !update(A, B) |
| 487 | return parser.unread >= length || yaml_parser_update_buffer(parser, length) |
| 488 | } |
| 489 | |
| 490 | // Advance the buffer pointer. |
| 491 | func skip(parser *yaml_parser_t) { |
| 492 | parser.mark.index++ |
| 493 | parser.mark.column++ |
| 494 | parser.unread-- |
| 495 | parser.buffer_pos += width(parser.buffer[parser.buffer_pos]) |
| 496 | } |
| 497 | |
| 498 | func skip_line(parser *yaml_parser_t) { |
| 499 | if is_crlf(parser.buffer, parser.buffer_pos) { |
| 500 | parser.mark.index += 2 |
| 501 | parser.mark.column = 0 |
| 502 | parser.mark.line++ |
| 503 | parser.unread -= 2 |
| 504 | parser.buffer_pos += 2 |
| 505 | } else if is_break(parser.buffer, parser.buffer_pos) { |
| 506 | parser.mark.index++ |
| 507 | parser.mark.column = 0 |
| 508 | parser.mark.line++ |
| 509 | parser.unread-- |
| 510 | parser.buffer_pos += width(parser.buffer[parser.buffer_pos]) |
| 511 | } |
| 512 | } |
| 513 | |
| 514 | // Copy a character to a string buffer and advance pointers. |
| 515 | func read(parser *yaml_parser_t, s []byte) []byte { |
| 516 | w := width(parser.buffer[parser.buffer_pos]) |
| 517 | if w == 0 { |
| 518 | panic("invalid character sequence") |
| 519 | } |
| 520 | if len(s) == 0 { |
| 521 | s = make([]byte, 0, 32) |
| 522 | } |
| 523 | if w == 1 && len(s)+w <= cap(s) { |
| 524 | s = s[:len(s)+1] |
| 525 | s[len(s)-1] = parser.buffer[parser.buffer_pos] |
| 526 | parser.buffer_pos++ |
| 527 | } else { |
| 528 | s = append(s, parser.buffer[parser.buffer_pos:parser.buffer_pos+w]...) |
| 529 | parser.buffer_pos += w |
| 530 | } |
| 531 | parser.mark.index++ |
| 532 | parser.mark.column++ |
| 533 | parser.unread-- |
| 534 | return s |
| 535 | } |
| 536 | |
| 537 | // Copy a line break character to a string buffer and advance pointers. |
| 538 | func read_line(parser *yaml_parser_t, s []byte) []byte { |
| 539 | buf := parser.buffer |
| 540 | pos := parser.buffer_pos |
| 541 | switch { |
| 542 | case buf[pos] == '\r' && buf[pos+1] == '\n': |
| 543 | // CR LF . LF |
| 544 | s = append(s, '\n') |
| 545 | parser.buffer_pos += 2 |
| 546 | parser.mark.index++ |
| 547 | parser.unread-- |
| 548 | case buf[pos] == '\r' || buf[pos] == '\n': |
| 549 | // CR|LF . LF |
| 550 | s = append(s, '\n') |
| 551 | parser.buffer_pos += 1 |
| 552 | case buf[pos] == '\xC2' && buf[pos+1] == '\x85': |
| 553 | // NEL . LF |
| 554 | s = append(s, '\n') |
| 555 | parser.buffer_pos += 2 |
| 556 | case buf[pos] == '\xE2' && buf[pos+1] == '\x80' && (buf[pos+2] == '\xA8' || buf[pos+2] == '\xA9'): |
| 557 | // LS|PS . LS|PS |
| 558 | s = append(s, buf[parser.buffer_pos:pos+3]...) |
| 559 | parser.buffer_pos += 3 |
| 560 | default: |
| 561 | return s |
| 562 | } |
| 563 | parser.mark.index++ |
| 564 | parser.mark.column = 0 |
| 565 | parser.mark.line++ |
| 566 | parser.unread-- |
| 567 | return s |
| 568 | } |
| 569 | |
| 570 | // Get the next token. |
| 571 | func yaml_parser_scan(parser *yaml_parser_t, token *yaml_token_t) bool { |
| 572 | // Erase the token object. |
| 573 | *token = yaml_token_t{} // [Go] Is this necessary? |
| 574 | |
| 575 | // No tokens after STREAM-END or error. |
| 576 | if parser.stream_end_produced || parser.error != yaml_NO_ERROR { |
| 577 | return true |
| 578 | } |
| 579 | |
| 580 | // Ensure that the tokens queue contains enough tokens. |
| 581 | if !parser.token_available { |
| 582 | if !yaml_parser_fetch_more_tokens(parser) { |
| 583 | return false |
| 584 | } |
| 585 | } |
| 586 | |
| 587 | // Fetch the next token from the queue. |
| 588 | *token = parser.tokens[parser.tokens_head] |
| 589 | parser.tokens_head++ |
| 590 | parser.tokens_parsed++ |
| 591 | parser.token_available = false |
| 592 | |
| 593 | if token.typ == yaml_STREAM_END_TOKEN { |
| 594 | parser.stream_end_produced = true |
| 595 | } |
| 596 | return true |
| 597 | } |
| 598 | |
| 599 | // Set the scanner error and return false. |
| 600 | func yaml_parser_set_scanner_error(parser *yaml_parser_t, context string, context_mark yaml_mark_t, problem string) bool { |
| 601 | parser.error = yaml_SCANNER_ERROR |
| 602 | parser.context = context |
| 603 | parser.context_mark = context_mark |
| 604 | parser.problem = problem |
| 605 | parser.problem_mark = parser.mark |
| 606 | return false |
| 607 | } |
| 608 | |
| 609 | func yaml_parser_set_scanner_tag_error(parser *yaml_parser_t, directive bool, context_mark yaml_mark_t, problem string) bool { |
| 610 | context := "while parsing a tag" |
| 611 | if directive { |
| 612 | context = "while parsing a %TAG directive" |
| 613 | } |
| 614 | return yaml_parser_set_scanner_error(parser, context, context_mark, problem) |
| 615 | } |
| 616 | |
| 617 | func trace(args ...interface{}) func() { |
| 618 | pargs := append([]interface{}{"+++"}, args...) |
| 619 | fmt.Println(pargs...) |
| 620 | pargs = append([]interface{}{"---"}, args...) |
| 621 | return func() { fmt.Println(pargs...) } |
| 622 | } |
| 623 | |
| 624 | // Ensure that the tokens queue contains at least one token which can be |
| 625 | // returned to the Parser. |
| 626 | func yaml_parser_fetch_more_tokens(parser *yaml_parser_t) bool { |
| 627 | // While we need more tokens to fetch, do it. |
| 628 | for { |
| 629 | // Check if we really need to fetch more tokens. |
| 630 | need_more_tokens := false |
| 631 | |
| 632 | if parser.tokens_head == len(parser.tokens) { |
| 633 | // Queue is empty. |
| 634 | need_more_tokens = true |
| 635 | } else { |
| 636 | // Check if any potential simple key may occupy the head position. |
| 637 | if !yaml_parser_stale_simple_keys(parser) { |
| 638 | return false |
| 639 | } |
| 640 | |
| 641 | for i := range parser.simple_keys { |
| 642 | simple_key := &parser.simple_keys[i] |
| 643 | if simple_key.possible && simple_key.token_number == parser.tokens_parsed { |
| 644 | need_more_tokens = true |
| 645 | break |
| 646 | } |
| 647 | } |
| 648 | } |
| 649 | |
| 650 | // We are finished. |
| 651 | if !need_more_tokens { |
| 652 | break |
| 653 | } |
| 654 | // Fetch the next token. |
| 655 | if !yaml_parser_fetch_next_token(parser) { |
| 656 | return false |
| 657 | } |
| 658 | } |
| 659 | |
| 660 | parser.token_available = true |
| 661 | return true |
| 662 | } |
| 663 | |
| 664 | // The dispatcher for token fetchers. |
| 665 | func yaml_parser_fetch_next_token(parser *yaml_parser_t) bool { |
| 666 | // Ensure that the buffer is initialized. |
| 667 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 668 | return false |
| 669 | } |
| 670 | |
| 671 | // Check if we just started scanning. Fetch STREAM-START then. |
| 672 | if !parser.stream_start_produced { |
| 673 | return yaml_parser_fetch_stream_start(parser) |
| 674 | } |
| 675 | |
| 676 | // Eat whitespaces and comments until we reach the next token. |
| 677 | if !yaml_parser_scan_to_next_token(parser) { |
| 678 | return false |
| 679 | } |
| 680 | |
| 681 | // Remove obsolete potential simple keys. |
| 682 | if !yaml_parser_stale_simple_keys(parser) { |
| 683 | return false |
| 684 | } |
| 685 | |
| 686 | // Check the indentation level against the current column. |
| 687 | if !yaml_parser_unroll_indent(parser, parser.mark.column) { |
| 688 | return false |
| 689 | } |
| 690 | |
| 691 | // Ensure that the buffer contains at least 4 characters. 4 is the length |
| 692 | // of the longest indicators ('--- ' and '... '). |
| 693 | if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) { |
| 694 | return false |
| 695 | } |
| 696 | |
| 697 | // Is it the end of the stream? |
| 698 | if is_z(parser.buffer, parser.buffer_pos) { |
| 699 | return yaml_parser_fetch_stream_end(parser) |
| 700 | } |
| 701 | |
| 702 | // Is it a directive? |
| 703 | if parser.mark.column == 0 && parser.buffer[parser.buffer_pos] == '%' { |
| 704 | return yaml_parser_fetch_directive(parser) |
| 705 | } |
| 706 | |
| 707 | buf := parser.buffer |
| 708 | pos := parser.buffer_pos |
| 709 | |
| 710 | // Is it the document start indicator? |
| 711 | if parser.mark.column == 0 && buf[pos] == '-' && buf[pos+1] == '-' && buf[pos+2] == '-' && is_blankz(buf, pos+3) { |
| 712 | return yaml_parser_fetch_document_indicator(parser, yaml_DOCUMENT_START_TOKEN) |
| 713 | } |
| 714 | |
| 715 | // Is it the document end indicator? |
| 716 | if parser.mark.column == 0 && buf[pos] == '.' && buf[pos+1] == '.' && buf[pos+2] == '.' && is_blankz(buf, pos+3) { |
| 717 | return yaml_parser_fetch_document_indicator(parser, yaml_DOCUMENT_END_TOKEN) |
| 718 | } |
| 719 | |
| 720 | // Is it the flow sequence start indicator? |
| 721 | if buf[pos] == '[' { |
| 722 | return yaml_parser_fetch_flow_collection_start(parser, yaml_FLOW_SEQUENCE_START_TOKEN) |
| 723 | } |
| 724 | |
| 725 | // Is it the flow mapping start indicator? |
| 726 | if parser.buffer[parser.buffer_pos] == '{' { |
| 727 | return yaml_parser_fetch_flow_collection_start(parser, yaml_FLOW_MAPPING_START_TOKEN) |
| 728 | } |
| 729 | |
| 730 | // Is it the flow sequence end indicator? |
| 731 | if parser.buffer[parser.buffer_pos] == ']' { |
| 732 | return yaml_parser_fetch_flow_collection_end(parser, |
| 733 | yaml_FLOW_SEQUENCE_END_TOKEN) |
| 734 | } |
| 735 | |
| 736 | // Is it the flow mapping end indicator? |
| 737 | if parser.buffer[parser.buffer_pos] == '}' { |
| 738 | return yaml_parser_fetch_flow_collection_end(parser, |
| 739 | yaml_FLOW_MAPPING_END_TOKEN) |
| 740 | } |
| 741 | |
| 742 | // Is it the flow entry indicator? |
| 743 | if parser.buffer[parser.buffer_pos] == ',' { |
| 744 | return yaml_parser_fetch_flow_entry(parser) |
| 745 | } |
| 746 | |
| 747 | // Is it the block entry indicator? |
| 748 | if parser.buffer[parser.buffer_pos] == '-' && is_blankz(parser.buffer, parser.buffer_pos+1) { |
| 749 | return yaml_parser_fetch_block_entry(parser) |
| 750 | } |
| 751 | |
| 752 | // Is it the key indicator? |
| 753 | if parser.buffer[parser.buffer_pos] == '?' && (parser.flow_level > 0 || is_blankz(parser.buffer, parser.buffer_pos+1)) { |
| 754 | return yaml_parser_fetch_key(parser) |
| 755 | } |
| 756 | |
| 757 | // Is it the value indicator? |
| 758 | if parser.buffer[parser.buffer_pos] == ':' && (parser.flow_level > 0 || is_blankz(parser.buffer, parser.buffer_pos+1)) { |
| 759 | return yaml_parser_fetch_value(parser) |
| 760 | } |
| 761 | |
| 762 | // Is it an alias? |
| 763 | if parser.buffer[parser.buffer_pos] == '*' { |
| 764 | return yaml_parser_fetch_anchor(parser, yaml_ALIAS_TOKEN) |
| 765 | } |
| 766 | |
| 767 | // Is it an anchor? |
| 768 | if parser.buffer[parser.buffer_pos] == '&' { |
| 769 | return yaml_parser_fetch_anchor(parser, yaml_ANCHOR_TOKEN) |
| 770 | } |
| 771 | |
| 772 | // Is it a tag? |
| 773 | if parser.buffer[parser.buffer_pos] == '!' { |
| 774 | return yaml_parser_fetch_tag(parser) |
| 775 | } |
| 776 | |
| 777 | // Is it a literal scalar? |
| 778 | if parser.buffer[parser.buffer_pos] == '|' && parser.flow_level == 0 { |
| 779 | return yaml_parser_fetch_block_scalar(parser, true) |
| 780 | } |
| 781 | |
| 782 | // Is it a folded scalar? |
| 783 | if parser.buffer[parser.buffer_pos] == '>' && parser.flow_level == 0 { |
| 784 | return yaml_parser_fetch_block_scalar(parser, false) |
| 785 | } |
| 786 | |
| 787 | // Is it a single-quoted scalar? |
| 788 | if parser.buffer[parser.buffer_pos] == '\'' { |
| 789 | return yaml_parser_fetch_flow_scalar(parser, true) |
| 790 | } |
| 791 | |
| 792 | // Is it a double-quoted scalar? |
| 793 | if parser.buffer[parser.buffer_pos] == '"' { |
| 794 | return yaml_parser_fetch_flow_scalar(parser, false) |
| 795 | } |
| 796 | |
| 797 | // Is it a plain scalar? |
| 798 | // |
| 799 | // A plain scalar may start with any non-blank characters except |
| 800 | // |
| 801 | // '-', '?', ':', ',', '[', ']', '{', '}', |
| 802 | // '#', '&', '*', '!', '|', '>', '\'', '\"', |
| 803 | // '%', '@', '`'. |
| 804 | // |
| 805 | // In the block context (and, for the '-' indicator, in the flow context |
| 806 | // too), it may also start with the characters |
| 807 | // |
| 808 | // '-', '?', ':' |
| 809 | // |
| 810 | // if it is followed by a non-space character. |
| 811 | // |
| 812 | // The last rule is more restrictive than the specification requires. |
| 813 | // [Go] Make this logic more reasonable. |
| 814 | //switch parser.buffer[parser.buffer_pos] { |
| 815 | //case '-', '?', ':', ',', '?', '-', ',', ':', ']', '[', '}', '{', '&', '#', '!', '*', '>', '|', '"', '\'', '@', '%', '-', '`': |
| 816 | //} |
| 817 | if !(is_blankz(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == '-' || |
| 818 | parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == ':' || |
| 819 | parser.buffer[parser.buffer_pos] == ',' || parser.buffer[parser.buffer_pos] == '[' || |
| 820 | parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '{' || |
| 821 | parser.buffer[parser.buffer_pos] == '}' || parser.buffer[parser.buffer_pos] == '#' || |
| 822 | parser.buffer[parser.buffer_pos] == '&' || parser.buffer[parser.buffer_pos] == '*' || |
| 823 | parser.buffer[parser.buffer_pos] == '!' || parser.buffer[parser.buffer_pos] == '|' || |
| 824 | parser.buffer[parser.buffer_pos] == '>' || parser.buffer[parser.buffer_pos] == '\'' || |
| 825 | parser.buffer[parser.buffer_pos] == '"' || parser.buffer[parser.buffer_pos] == '%' || |
| 826 | parser.buffer[parser.buffer_pos] == '@' || parser.buffer[parser.buffer_pos] == '`') || |
| 827 | (parser.buffer[parser.buffer_pos] == '-' && !is_blank(parser.buffer, parser.buffer_pos+1)) || |
| 828 | (parser.flow_level == 0 && |
| 829 | (parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == ':') && |
| 830 | !is_blankz(parser.buffer, parser.buffer_pos+1)) { |
| 831 | return yaml_parser_fetch_plain_scalar(parser) |
| 832 | } |
| 833 | |
| 834 | // If we don't determine the token type so far, it is an error. |
| 835 | return yaml_parser_set_scanner_error(parser, |
| 836 | "while scanning for the next token", parser.mark, |
| 837 | "found character that cannot start any token") |
| 838 | } |
| 839 | |
| 840 | // Check the list of potential simple keys and remove the positions that |
| 841 | // cannot contain simple keys anymore. |
| 842 | func yaml_parser_stale_simple_keys(parser *yaml_parser_t) bool { |
| 843 | // Check for a potential simple key for each flow level. |
| 844 | for i := range parser.simple_keys { |
| 845 | simple_key := &parser.simple_keys[i] |
| 846 | |
| 847 | // The specification requires that a simple key |
| 848 | // |
| 849 | // - is limited to a single line, |
| 850 | // - is shorter than 1024 characters. |
| 851 | if simple_key.possible && (simple_key.mark.line < parser.mark.line || simple_key.mark.index+1024 < parser.mark.index) { |
| 852 | |
| 853 | // Check if the potential simple key to be removed is required. |
| 854 | if simple_key.required { |
| 855 | return yaml_parser_set_scanner_error(parser, |
| 856 | "while scanning a simple key", simple_key.mark, |
| 857 | "could not find expected ':'") |
| 858 | } |
| 859 | simple_key.possible = false |
| 860 | } |
| 861 | } |
| 862 | return true |
| 863 | } |
| 864 | |
| 865 | // Check if a simple key may start at the current position and add it if |
| 866 | // needed. |
| 867 | func yaml_parser_save_simple_key(parser *yaml_parser_t) bool { |
| 868 | // A simple key is required at the current position if the scanner is in |
| 869 | // the block context and the current column coincides with the indentation |
| 870 | // level. |
| 871 | |
| 872 | required := parser.flow_level == 0 && parser.indent == parser.mark.column |
| 873 | |
| 874 | // |
| 875 | // If the current position may start a simple key, save it. |
| 876 | // |
| 877 | if parser.simple_key_allowed { |
| 878 | simple_key := yaml_simple_key_t{ |
| 879 | possible: true, |
| 880 | required: required, |
| 881 | token_number: parser.tokens_parsed + (len(parser.tokens) - parser.tokens_head), |
| 882 | } |
| 883 | simple_key.mark = parser.mark |
| 884 | |
| 885 | if !yaml_parser_remove_simple_key(parser) { |
| 886 | return false |
| 887 | } |
| 888 | parser.simple_keys[len(parser.simple_keys)-1] = simple_key |
| 889 | } |
| 890 | return true |
| 891 | } |
| 892 | |
| 893 | // Remove a potential simple key at the current flow level. |
| 894 | func yaml_parser_remove_simple_key(parser *yaml_parser_t) bool { |
| 895 | i := len(parser.simple_keys) - 1 |
| 896 | if parser.simple_keys[i].possible { |
| 897 | // If the key is required, it is an error. |
| 898 | if parser.simple_keys[i].required { |
| 899 | return yaml_parser_set_scanner_error(parser, |
| 900 | "while scanning a simple key", parser.simple_keys[i].mark, |
| 901 | "could not find expected ':'") |
| 902 | } |
| 903 | } |
| 904 | // Remove the key from the stack. |
| 905 | parser.simple_keys[i].possible = false |
| 906 | return true |
| 907 | } |
| 908 | |
| 909 | // Increase the flow level and resize the simple key list if needed. |
| 910 | func yaml_parser_increase_flow_level(parser *yaml_parser_t) bool { |
| 911 | // Reset the simple key on the next level. |
| 912 | parser.simple_keys = append(parser.simple_keys, yaml_simple_key_t{}) |
| 913 | |
| 914 | // Increase the flow level. |
| 915 | parser.flow_level++ |
| 916 | return true |
| 917 | } |
| 918 | |
| 919 | // Decrease the flow level. |
| 920 | func yaml_parser_decrease_flow_level(parser *yaml_parser_t) bool { |
| 921 | if parser.flow_level > 0 { |
| 922 | parser.flow_level-- |
| 923 | parser.simple_keys = parser.simple_keys[:len(parser.simple_keys)-1] |
| 924 | } |
| 925 | return true |
| 926 | } |
| 927 | |
| 928 | // Push the current indentation level to the stack and set the new level |
| 929 | // the current column is greater than the indentation level. In this case, |
| 930 | // append or insert the specified token into the token queue. |
| 931 | func yaml_parser_roll_indent(parser *yaml_parser_t, column, number int, typ yaml_token_type_t, mark yaml_mark_t) bool { |
| 932 | // In the flow context, do nothing. |
| 933 | if parser.flow_level > 0 { |
| 934 | return true |
| 935 | } |
| 936 | |
| 937 | if parser.indent < column { |
| 938 | // Push the current indentation level to the stack and set the new |
| 939 | // indentation level. |
| 940 | parser.indents = append(parser.indents, parser.indent) |
| 941 | parser.indent = column |
| 942 | |
| 943 | // Create a token and insert it into the queue. |
| 944 | token := yaml_token_t{ |
| 945 | typ: typ, |
| 946 | start_mark: mark, |
| 947 | end_mark: mark, |
| 948 | } |
| 949 | if number > -1 { |
| 950 | number -= parser.tokens_parsed |
| 951 | } |
| 952 | yaml_insert_token(parser, number, &token) |
| 953 | } |
| 954 | return true |
| 955 | } |
| 956 | |
| 957 | // Pop indentation levels from the indents stack until the current level |
| 958 | // becomes less or equal to the column. For each indentation level, append |
| 959 | // the BLOCK-END token. |
| 960 | func yaml_parser_unroll_indent(parser *yaml_parser_t, column int) bool { |
| 961 | // In the flow context, do nothing. |
| 962 | if parser.flow_level > 0 { |
| 963 | return true |
| 964 | } |
| 965 | |
| 966 | // Loop through the indentation levels in the stack. |
| 967 | for parser.indent > column { |
| 968 | // Create a token and append it to the queue. |
| 969 | token := yaml_token_t{ |
| 970 | typ: yaml_BLOCK_END_TOKEN, |
| 971 | start_mark: parser.mark, |
| 972 | end_mark: parser.mark, |
| 973 | } |
| 974 | yaml_insert_token(parser, -1, &token) |
| 975 | |
| 976 | // Pop the indentation level. |
| 977 | parser.indent = parser.indents[len(parser.indents)-1] |
| 978 | parser.indents = parser.indents[:len(parser.indents)-1] |
| 979 | } |
| 980 | return true |
| 981 | } |
| 982 | |
| 983 | // Initialize the scanner and produce the STREAM-START token. |
| 984 | func yaml_parser_fetch_stream_start(parser *yaml_parser_t) bool { |
| 985 | |
| 986 | // Set the initial indentation. |
| 987 | parser.indent = -1 |
| 988 | |
| 989 | // Initialize the simple key stack. |
| 990 | parser.simple_keys = append(parser.simple_keys, yaml_simple_key_t{}) |
| 991 | |
| 992 | // A simple key is allowed at the beginning of the stream. |
| 993 | parser.simple_key_allowed = true |
| 994 | |
| 995 | // We have started. |
| 996 | parser.stream_start_produced = true |
| 997 | |
| 998 | // Create the STREAM-START token and append it to the queue. |
| 999 | token := yaml_token_t{ |
| 1000 | typ: yaml_STREAM_START_TOKEN, |
| 1001 | start_mark: parser.mark, |
| 1002 | end_mark: parser.mark, |
| 1003 | encoding: parser.encoding, |
| 1004 | } |
| 1005 | yaml_insert_token(parser, -1, &token) |
| 1006 | return true |
| 1007 | } |
| 1008 | |
| 1009 | // Produce the STREAM-END token and shut down the scanner. |
| 1010 | func yaml_parser_fetch_stream_end(parser *yaml_parser_t) bool { |
| 1011 | |
| 1012 | // Force new line. |
| 1013 | if parser.mark.column != 0 { |
| 1014 | parser.mark.column = 0 |
| 1015 | parser.mark.line++ |
| 1016 | } |
| 1017 | |
| 1018 | // Reset the indentation level. |
| 1019 | if !yaml_parser_unroll_indent(parser, -1) { |
| 1020 | return false |
| 1021 | } |
| 1022 | |
| 1023 | // Reset simple keys. |
| 1024 | if !yaml_parser_remove_simple_key(parser) { |
| 1025 | return false |
| 1026 | } |
| 1027 | |
| 1028 | parser.simple_key_allowed = false |
| 1029 | |
| 1030 | // Create the STREAM-END token and append it to the queue. |
| 1031 | token := yaml_token_t{ |
| 1032 | typ: yaml_STREAM_END_TOKEN, |
| 1033 | start_mark: parser.mark, |
| 1034 | end_mark: parser.mark, |
| 1035 | } |
| 1036 | yaml_insert_token(parser, -1, &token) |
| 1037 | return true |
| 1038 | } |
| 1039 | |
| 1040 | // Produce a VERSION-DIRECTIVE or TAG-DIRECTIVE token. |
| 1041 | func yaml_parser_fetch_directive(parser *yaml_parser_t) bool { |
| 1042 | // Reset the indentation level. |
| 1043 | if !yaml_parser_unroll_indent(parser, -1) { |
| 1044 | return false |
| 1045 | } |
| 1046 | |
| 1047 | // Reset simple keys. |
| 1048 | if !yaml_parser_remove_simple_key(parser) { |
| 1049 | return false |
| 1050 | } |
| 1051 | |
| 1052 | parser.simple_key_allowed = false |
| 1053 | |
| 1054 | // Create the YAML-DIRECTIVE or TAG-DIRECTIVE token. |
| 1055 | token := yaml_token_t{} |
| 1056 | if !yaml_parser_scan_directive(parser, &token) { |
| 1057 | return false |
| 1058 | } |
| 1059 | // Append the token to the queue. |
| 1060 | yaml_insert_token(parser, -1, &token) |
| 1061 | return true |
| 1062 | } |
| 1063 | |
| 1064 | // Produce the DOCUMENT-START or DOCUMENT-END token. |
| 1065 | func yaml_parser_fetch_document_indicator(parser *yaml_parser_t, typ yaml_token_type_t) bool { |
| 1066 | // Reset the indentation level. |
| 1067 | if !yaml_parser_unroll_indent(parser, -1) { |
| 1068 | return false |
| 1069 | } |
| 1070 | |
| 1071 | // Reset simple keys. |
| 1072 | if !yaml_parser_remove_simple_key(parser) { |
| 1073 | return false |
| 1074 | } |
| 1075 | |
| 1076 | parser.simple_key_allowed = false |
| 1077 | |
| 1078 | // Consume the token. |
| 1079 | start_mark := parser.mark |
| 1080 | |
| 1081 | skip(parser) |
| 1082 | skip(parser) |
| 1083 | skip(parser) |
| 1084 | |
| 1085 | end_mark := parser.mark |
| 1086 | |
| 1087 | // Create the DOCUMENT-START or DOCUMENT-END token. |
| 1088 | token := yaml_token_t{ |
| 1089 | typ: typ, |
| 1090 | start_mark: start_mark, |
| 1091 | end_mark: end_mark, |
| 1092 | } |
| 1093 | // Append the token to the queue. |
| 1094 | yaml_insert_token(parser, -1, &token) |
| 1095 | return true |
| 1096 | } |
| 1097 | |
| 1098 | // Produce the FLOW-SEQUENCE-START or FLOW-MAPPING-START token. |
| 1099 | func yaml_parser_fetch_flow_collection_start(parser *yaml_parser_t, typ yaml_token_type_t) bool { |
| 1100 | // The indicators '[' and '{' may start a simple key. |
| 1101 | if !yaml_parser_save_simple_key(parser) { |
| 1102 | return false |
| 1103 | } |
| 1104 | |
| 1105 | // Increase the flow level. |
| 1106 | if !yaml_parser_increase_flow_level(parser) { |
| 1107 | return false |
| 1108 | } |
| 1109 | |
| 1110 | // A simple key may follow the indicators '[' and '{'. |
| 1111 | parser.simple_key_allowed = true |
| 1112 | |
| 1113 | // Consume the token. |
| 1114 | start_mark := parser.mark |
| 1115 | skip(parser) |
| 1116 | end_mark := parser.mark |
| 1117 | |
| 1118 | // Create the FLOW-SEQUENCE-START of FLOW-MAPPING-START token. |
| 1119 | token := yaml_token_t{ |
| 1120 | typ: typ, |
| 1121 | start_mark: start_mark, |
| 1122 | end_mark: end_mark, |
| 1123 | } |
| 1124 | // Append the token to the queue. |
| 1125 | yaml_insert_token(parser, -1, &token) |
| 1126 | return true |
| 1127 | } |
| 1128 | |
| 1129 | // Produce the FLOW-SEQUENCE-END or FLOW-MAPPING-END token. |
| 1130 | func yaml_parser_fetch_flow_collection_end(parser *yaml_parser_t, typ yaml_token_type_t) bool { |
| 1131 | // Reset any potential simple key on the current flow level. |
| 1132 | if !yaml_parser_remove_simple_key(parser) { |
| 1133 | return false |
| 1134 | } |
| 1135 | |
| 1136 | // Decrease the flow level. |
| 1137 | if !yaml_parser_decrease_flow_level(parser) { |
| 1138 | return false |
| 1139 | } |
| 1140 | |
| 1141 | // No simple keys after the indicators ']' and '}'. |
| 1142 | parser.simple_key_allowed = false |
| 1143 | |
| 1144 | // Consume the token. |
| 1145 | |
| 1146 | start_mark := parser.mark |
| 1147 | skip(parser) |
| 1148 | end_mark := parser.mark |
| 1149 | |
| 1150 | // Create the FLOW-SEQUENCE-END of FLOW-MAPPING-END token. |
| 1151 | token := yaml_token_t{ |
| 1152 | typ: typ, |
| 1153 | start_mark: start_mark, |
| 1154 | end_mark: end_mark, |
| 1155 | } |
| 1156 | // Append the token to the queue. |
| 1157 | yaml_insert_token(parser, -1, &token) |
| 1158 | return true |
| 1159 | } |
| 1160 | |
| 1161 | // Produce the FLOW-ENTRY token. |
| 1162 | func yaml_parser_fetch_flow_entry(parser *yaml_parser_t) bool { |
| 1163 | // Reset any potential simple keys on the current flow level. |
| 1164 | if !yaml_parser_remove_simple_key(parser) { |
| 1165 | return false |
| 1166 | } |
| 1167 | |
| 1168 | // Simple keys are allowed after ','. |
| 1169 | parser.simple_key_allowed = true |
| 1170 | |
| 1171 | // Consume the token. |
| 1172 | start_mark := parser.mark |
| 1173 | skip(parser) |
| 1174 | end_mark := parser.mark |
| 1175 | |
| 1176 | // Create the FLOW-ENTRY token and append it to the queue. |
| 1177 | token := yaml_token_t{ |
| 1178 | typ: yaml_FLOW_ENTRY_TOKEN, |
| 1179 | start_mark: start_mark, |
| 1180 | end_mark: end_mark, |
| 1181 | } |
| 1182 | yaml_insert_token(parser, -1, &token) |
| 1183 | return true |
| 1184 | } |
| 1185 | |
| 1186 | // Produce the BLOCK-ENTRY token. |
| 1187 | func yaml_parser_fetch_block_entry(parser *yaml_parser_t) bool { |
| 1188 | // Check if the scanner is in the block context. |
| 1189 | if parser.flow_level == 0 { |
| 1190 | // Check if we are allowed to start a new entry. |
| 1191 | if !parser.simple_key_allowed { |
| 1192 | return yaml_parser_set_scanner_error(parser, "", parser.mark, |
| 1193 | "block sequence entries are not allowed in this context") |
| 1194 | } |
| 1195 | // Add the BLOCK-SEQUENCE-START token if needed. |
| 1196 | if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_SEQUENCE_START_TOKEN, parser.mark) { |
| 1197 | return false |
| 1198 | } |
| 1199 | } else { |
| 1200 | // It is an error for the '-' indicator to occur in the flow context, |
| 1201 | // but we let the Parser detect and report about it because the Parser |
| 1202 | // is able to point to the context. |
| 1203 | } |
| 1204 | |
| 1205 | // Reset any potential simple keys on the current flow level. |
| 1206 | if !yaml_parser_remove_simple_key(parser) { |
| 1207 | return false |
| 1208 | } |
| 1209 | |
| 1210 | // Simple keys are allowed after '-'. |
| 1211 | parser.simple_key_allowed = true |
| 1212 | |
| 1213 | // Consume the token. |
| 1214 | start_mark := parser.mark |
| 1215 | skip(parser) |
| 1216 | end_mark := parser.mark |
| 1217 | |
| 1218 | // Create the BLOCK-ENTRY token and append it to the queue. |
| 1219 | token := yaml_token_t{ |
| 1220 | typ: yaml_BLOCK_ENTRY_TOKEN, |
| 1221 | start_mark: start_mark, |
| 1222 | end_mark: end_mark, |
| 1223 | } |
| 1224 | yaml_insert_token(parser, -1, &token) |
| 1225 | return true |
| 1226 | } |
| 1227 | |
| 1228 | // Produce the KEY token. |
| 1229 | func yaml_parser_fetch_key(parser *yaml_parser_t) bool { |
| 1230 | |
| 1231 | // In the block context, additional checks are required. |
| 1232 | if parser.flow_level == 0 { |
| 1233 | // Check if we are allowed to start a new key (not nessesary simple). |
| 1234 | if !parser.simple_key_allowed { |
| 1235 | return yaml_parser_set_scanner_error(parser, "", parser.mark, |
| 1236 | "mapping keys are not allowed in this context") |
| 1237 | } |
| 1238 | // Add the BLOCK-MAPPING-START token if needed. |
| 1239 | if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_MAPPING_START_TOKEN, parser.mark) { |
| 1240 | return false |
| 1241 | } |
| 1242 | } |
| 1243 | |
| 1244 | // Reset any potential simple keys on the current flow level. |
| 1245 | if !yaml_parser_remove_simple_key(parser) { |
| 1246 | return false |
| 1247 | } |
| 1248 | |
| 1249 | // Simple keys are allowed after '?' in the block context. |
| 1250 | parser.simple_key_allowed = parser.flow_level == 0 |
| 1251 | |
| 1252 | // Consume the token. |
| 1253 | start_mark := parser.mark |
| 1254 | skip(parser) |
| 1255 | end_mark := parser.mark |
| 1256 | |
| 1257 | // Create the KEY token and append it to the queue. |
| 1258 | token := yaml_token_t{ |
| 1259 | typ: yaml_KEY_TOKEN, |
| 1260 | start_mark: start_mark, |
| 1261 | end_mark: end_mark, |
| 1262 | } |
| 1263 | yaml_insert_token(parser, -1, &token) |
| 1264 | return true |
| 1265 | } |
| 1266 | |
| 1267 | // Produce the VALUE token. |
| 1268 | func yaml_parser_fetch_value(parser *yaml_parser_t) bool { |
| 1269 | |
| 1270 | simple_key := &parser.simple_keys[len(parser.simple_keys)-1] |
| 1271 | |
| 1272 | // Have we found a simple key? |
| 1273 | if simple_key.possible { |
| 1274 | // Create the KEY token and insert it into the queue. |
| 1275 | token := yaml_token_t{ |
| 1276 | typ: yaml_KEY_TOKEN, |
| 1277 | start_mark: simple_key.mark, |
| 1278 | end_mark: simple_key.mark, |
| 1279 | } |
| 1280 | yaml_insert_token(parser, simple_key.token_number-parser.tokens_parsed, &token) |
| 1281 | |
| 1282 | // In the block context, we may need to add the BLOCK-MAPPING-START token. |
| 1283 | if !yaml_parser_roll_indent(parser, simple_key.mark.column, |
| 1284 | simple_key.token_number, |
| 1285 | yaml_BLOCK_MAPPING_START_TOKEN, simple_key.mark) { |
| 1286 | return false |
| 1287 | } |
| 1288 | |
| 1289 | // Remove the simple key. |
| 1290 | simple_key.possible = false |
| 1291 | |
| 1292 | // A simple key cannot follow another simple key. |
| 1293 | parser.simple_key_allowed = false |
| 1294 | |
| 1295 | } else { |
| 1296 | // The ':' indicator follows a complex key. |
| 1297 | |
| 1298 | // In the block context, extra checks are required. |
| 1299 | if parser.flow_level == 0 { |
| 1300 | |
| 1301 | // Check if we are allowed to start a complex value. |
| 1302 | if !parser.simple_key_allowed { |
| 1303 | return yaml_parser_set_scanner_error(parser, "", parser.mark, |
| 1304 | "mapping values are not allowed in this context") |
| 1305 | } |
| 1306 | |
| 1307 | // Add the BLOCK-MAPPING-START token if needed. |
| 1308 | if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_MAPPING_START_TOKEN, parser.mark) { |
| 1309 | return false |
| 1310 | } |
| 1311 | } |
| 1312 | |
| 1313 | // Simple keys after ':' are allowed in the block context. |
| 1314 | parser.simple_key_allowed = parser.flow_level == 0 |
| 1315 | } |
| 1316 | |
| 1317 | // Consume the token. |
| 1318 | start_mark := parser.mark |
| 1319 | skip(parser) |
| 1320 | end_mark := parser.mark |
| 1321 | |
| 1322 | // Create the VALUE token and append it to the queue. |
| 1323 | token := yaml_token_t{ |
| 1324 | typ: yaml_VALUE_TOKEN, |
| 1325 | start_mark: start_mark, |
| 1326 | end_mark: end_mark, |
| 1327 | } |
| 1328 | yaml_insert_token(parser, -1, &token) |
| 1329 | return true |
| 1330 | } |
| 1331 | |
| 1332 | // Produce the ALIAS or ANCHOR token. |
| 1333 | func yaml_parser_fetch_anchor(parser *yaml_parser_t, typ yaml_token_type_t) bool { |
| 1334 | // An anchor or an alias could be a simple key. |
| 1335 | if !yaml_parser_save_simple_key(parser) { |
| 1336 | return false |
| 1337 | } |
| 1338 | |
| 1339 | // A simple key cannot follow an anchor or an alias. |
| 1340 | parser.simple_key_allowed = false |
| 1341 | |
| 1342 | // Create the ALIAS or ANCHOR token and append it to the queue. |
| 1343 | var token yaml_token_t |
| 1344 | if !yaml_parser_scan_anchor(parser, &token, typ) { |
| 1345 | return false |
| 1346 | } |
| 1347 | yaml_insert_token(parser, -1, &token) |
| 1348 | return true |
| 1349 | } |
| 1350 | |
| 1351 | // Produce the TAG token. |
| 1352 | func yaml_parser_fetch_tag(parser *yaml_parser_t) bool { |
| 1353 | // A tag could be a simple key. |
| 1354 | if !yaml_parser_save_simple_key(parser) { |
| 1355 | return false |
| 1356 | } |
| 1357 | |
| 1358 | // A simple key cannot follow a tag. |
| 1359 | parser.simple_key_allowed = false |
| 1360 | |
| 1361 | // Create the TAG token and append it to the queue. |
| 1362 | var token yaml_token_t |
| 1363 | if !yaml_parser_scan_tag(parser, &token) { |
| 1364 | return false |
| 1365 | } |
| 1366 | yaml_insert_token(parser, -1, &token) |
| 1367 | return true |
| 1368 | } |
| 1369 | |
| 1370 | // Produce the SCALAR(...,literal) or SCALAR(...,folded) tokens. |
| 1371 | func yaml_parser_fetch_block_scalar(parser *yaml_parser_t, literal bool) bool { |
| 1372 | // Remove any potential simple keys. |
| 1373 | if !yaml_parser_remove_simple_key(parser) { |
| 1374 | return false |
| 1375 | } |
| 1376 | |
| 1377 | // A simple key may follow a block scalar. |
| 1378 | parser.simple_key_allowed = true |
| 1379 | |
| 1380 | // Create the SCALAR token and append it to the queue. |
| 1381 | var token yaml_token_t |
| 1382 | if !yaml_parser_scan_block_scalar(parser, &token, literal) { |
| 1383 | return false |
| 1384 | } |
| 1385 | yaml_insert_token(parser, -1, &token) |
| 1386 | return true |
| 1387 | } |
| 1388 | |
| 1389 | // Produce the SCALAR(...,single-quoted) or SCALAR(...,double-quoted) tokens. |
| 1390 | func yaml_parser_fetch_flow_scalar(parser *yaml_parser_t, single bool) bool { |
| 1391 | // A plain scalar could be a simple key. |
| 1392 | if !yaml_parser_save_simple_key(parser) { |
| 1393 | return false |
| 1394 | } |
| 1395 | |
| 1396 | // A simple key cannot follow a flow scalar. |
| 1397 | parser.simple_key_allowed = false |
| 1398 | |
| 1399 | // Create the SCALAR token and append it to the queue. |
| 1400 | var token yaml_token_t |
| 1401 | if !yaml_parser_scan_flow_scalar(parser, &token, single) { |
| 1402 | return false |
| 1403 | } |
| 1404 | yaml_insert_token(parser, -1, &token) |
| 1405 | return true |
| 1406 | } |
| 1407 | |
| 1408 | // Produce the SCALAR(...,plain) token. |
| 1409 | func yaml_parser_fetch_plain_scalar(parser *yaml_parser_t) bool { |
| 1410 | // A plain scalar could be a simple key. |
| 1411 | if !yaml_parser_save_simple_key(parser) { |
| 1412 | return false |
| 1413 | } |
| 1414 | |
| 1415 | // A simple key cannot follow a flow scalar. |
| 1416 | parser.simple_key_allowed = false |
| 1417 | |
| 1418 | // Create the SCALAR token and append it to the queue. |
| 1419 | var token yaml_token_t |
| 1420 | if !yaml_parser_scan_plain_scalar(parser, &token) { |
| 1421 | return false |
| 1422 | } |
| 1423 | yaml_insert_token(parser, -1, &token) |
| 1424 | return true |
| 1425 | } |
| 1426 | |
| 1427 | // Eat whitespaces and comments until the next token is found. |
| 1428 | func yaml_parser_scan_to_next_token(parser *yaml_parser_t) bool { |
| 1429 | |
| 1430 | // Until the next token is not found. |
| 1431 | for { |
| 1432 | // Allow the BOM mark to start a line. |
| 1433 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1434 | return false |
| 1435 | } |
| 1436 | if parser.mark.column == 0 && is_bom(parser.buffer, parser.buffer_pos) { |
| 1437 | skip(parser) |
| 1438 | } |
| 1439 | |
| 1440 | // Eat whitespaces. |
| 1441 | // Tabs are allowed: |
| 1442 | // - in the flow context |
| 1443 | // - in the block context, but not at the beginning of the line or |
| 1444 | // after '-', '?', or ':' (complex value). |
| 1445 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1446 | return false |
| 1447 | } |
| 1448 | |
| 1449 | for parser.buffer[parser.buffer_pos] == ' ' || ((parser.flow_level > 0 || !parser.simple_key_allowed) && parser.buffer[parser.buffer_pos] == '\t') { |
| 1450 | skip(parser) |
| 1451 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1452 | return false |
| 1453 | } |
| 1454 | } |
| 1455 | |
| 1456 | // Eat a comment until a line break. |
| 1457 | if parser.buffer[parser.buffer_pos] == '#' { |
| 1458 | for !is_breakz(parser.buffer, parser.buffer_pos) { |
| 1459 | skip(parser) |
| 1460 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1461 | return false |
| 1462 | } |
| 1463 | } |
| 1464 | } |
| 1465 | |
| 1466 | // If it is a line break, eat it. |
| 1467 | if is_break(parser.buffer, parser.buffer_pos) { |
| 1468 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 1469 | return false |
| 1470 | } |
| 1471 | skip_line(parser) |
| 1472 | |
| 1473 | // In the block context, a new line may start a simple key. |
| 1474 | if parser.flow_level == 0 { |
| 1475 | parser.simple_key_allowed = true |
| 1476 | } |
| 1477 | } else { |
| 1478 | break // We have found a token. |
| 1479 | } |
| 1480 | } |
| 1481 | |
| 1482 | return true |
| 1483 | } |
| 1484 | |
| 1485 | // Scan a YAML-DIRECTIVE or TAG-DIRECTIVE token. |
| 1486 | // |
| 1487 | // Scope: |
| 1488 | // %YAML 1.1 # a comment \n |
| 1489 | // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 1490 | // %TAG !yaml! tag:yaml.org,2002: \n |
| 1491 | // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 1492 | // |
| 1493 | func yaml_parser_scan_directive(parser *yaml_parser_t, token *yaml_token_t) bool { |
| 1494 | // Eat '%'. |
| 1495 | start_mark := parser.mark |
| 1496 | skip(parser) |
| 1497 | |
| 1498 | // Scan the directive name. |
| 1499 | var name []byte |
| 1500 | if !yaml_parser_scan_directive_name(parser, start_mark, &name) { |
| 1501 | return false |
| 1502 | } |
| 1503 | |
| 1504 | // Is it a YAML directive? |
| 1505 | if bytes.Equal(name, []byte("YAML")) { |
| 1506 | // Scan the VERSION directive value. |
| 1507 | var major, minor int8 |
| 1508 | if !yaml_parser_scan_version_directive_value(parser, start_mark, &major, &minor) { |
| 1509 | return false |
| 1510 | } |
| 1511 | end_mark := parser.mark |
| 1512 | |
| 1513 | // Create a VERSION-DIRECTIVE token. |
| 1514 | *token = yaml_token_t{ |
| 1515 | typ: yaml_VERSION_DIRECTIVE_TOKEN, |
| 1516 | start_mark: start_mark, |
| 1517 | end_mark: end_mark, |
| 1518 | major: major, |
| 1519 | minor: minor, |
| 1520 | } |
| 1521 | |
| 1522 | // Is it a TAG directive? |
| 1523 | } else if bytes.Equal(name, []byte("TAG")) { |
| 1524 | // Scan the TAG directive value. |
| 1525 | var handle, prefix []byte |
| 1526 | if !yaml_parser_scan_tag_directive_value(parser, start_mark, &handle, &prefix) { |
| 1527 | return false |
| 1528 | } |
| 1529 | end_mark := parser.mark |
| 1530 | |
| 1531 | // Create a TAG-DIRECTIVE token. |
| 1532 | *token = yaml_token_t{ |
| 1533 | typ: yaml_TAG_DIRECTIVE_TOKEN, |
| 1534 | start_mark: start_mark, |
| 1535 | end_mark: end_mark, |
| 1536 | value: handle, |
| 1537 | prefix: prefix, |
| 1538 | } |
| 1539 | |
| 1540 | // Unknown directive. |
| 1541 | } else { |
| 1542 | yaml_parser_set_scanner_error(parser, "while scanning a directive", |
| 1543 | start_mark, "found unknown directive name") |
| 1544 | return false |
| 1545 | } |
| 1546 | |
| 1547 | // Eat the rest of the line including any comments. |
| 1548 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1549 | return false |
| 1550 | } |
| 1551 | |
| 1552 | for is_blank(parser.buffer, parser.buffer_pos) { |
| 1553 | skip(parser) |
| 1554 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1555 | return false |
| 1556 | } |
| 1557 | } |
| 1558 | |
| 1559 | if parser.buffer[parser.buffer_pos] == '#' { |
| 1560 | for !is_breakz(parser.buffer, parser.buffer_pos) { |
| 1561 | skip(parser) |
| 1562 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1563 | return false |
| 1564 | } |
| 1565 | } |
| 1566 | } |
| 1567 | |
| 1568 | // Check if we are at the end of the line. |
| 1569 | if !is_breakz(parser.buffer, parser.buffer_pos) { |
| 1570 | yaml_parser_set_scanner_error(parser, "while scanning a directive", |
| 1571 | start_mark, "did not find expected comment or line break") |
| 1572 | return false |
| 1573 | } |
| 1574 | |
| 1575 | // Eat a line break. |
| 1576 | if is_break(parser.buffer, parser.buffer_pos) { |
| 1577 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 1578 | return false |
| 1579 | } |
| 1580 | skip_line(parser) |
| 1581 | } |
| 1582 | |
| 1583 | return true |
| 1584 | } |
| 1585 | |
| 1586 | // Scan the directive name. |
| 1587 | // |
| 1588 | // Scope: |
| 1589 | // %YAML 1.1 # a comment \n |
| 1590 | // ^^^^ |
| 1591 | // %TAG !yaml! tag:yaml.org,2002: \n |
| 1592 | // ^^^ |
| 1593 | // |
| 1594 | func yaml_parser_scan_directive_name(parser *yaml_parser_t, start_mark yaml_mark_t, name *[]byte) bool { |
| 1595 | // Consume the directive name. |
| 1596 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1597 | return false |
| 1598 | } |
| 1599 | |
| 1600 | var s []byte |
| 1601 | for is_alpha(parser.buffer, parser.buffer_pos) { |
| 1602 | s = read(parser, s) |
| 1603 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1604 | return false |
| 1605 | } |
| 1606 | } |
| 1607 | |
| 1608 | // Check if the name is empty. |
| 1609 | if len(s) == 0 { |
| 1610 | yaml_parser_set_scanner_error(parser, "while scanning a directive", |
| 1611 | start_mark, "could not find expected directive name") |
| 1612 | return false |
| 1613 | } |
| 1614 | |
| 1615 | // Check for an blank character after the name. |
| 1616 | if !is_blankz(parser.buffer, parser.buffer_pos) { |
| 1617 | yaml_parser_set_scanner_error(parser, "while scanning a directive", |
| 1618 | start_mark, "found unexpected non-alphabetical character") |
| 1619 | return false |
| 1620 | } |
| 1621 | *name = s |
| 1622 | return true |
| 1623 | } |
| 1624 | |
| 1625 | // Scan the value of VERSION-DIRECTIVE. |
| 1626 | // |
| 1627 | // Scope: |
| 1628 | // %YAML 1.1 # a comment \n |
| 1629 | // ^^^^^^ |
| 1630 | func yaml_parser_scan_version_directive_value(parser *yaml_parser_t, start_mark yaml_mark_t, major, minor *int8) bool { |
| 1631 | // Eat whitespaces. |
| 1632 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1633 | return false |
| 1634 | } |
| 1635 | for is_blank(parser.buffer, parser.buffer_pos) { |
| 1636 | skip(parser) |
| 1637 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1638 | return false |
| 1639 | } |
| 1640 | } |
| 1641 | |
| 1642 | // Consume the major version number. |
| 1643 | if !yaml_parser_scan_version_directive_number(parser, start_mark, major) { |
| 1644 | return false |
| 1645 | } |
| 1646 | |
| 1647 | // Eat '.'. |
| 1648 | if parser.buffer[parser.buffer_pos] != '.' { |
| 1649 | return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive", |
| 1650 | start_mark, "did not find expected digit or '.' character") |
| 1651 | } |
| 1652 | |
| 1653 | skip(parser) |
| 1654 | |
| 1655 | // Consume the minor version number. |
| 1656 | if !yaml_parser_scan_version_directive_number(parser, start_mark, minor) { |
| 1657 | return false |
| 1658 | } |
| 1659 | return true |
| 1660 | } |
| 1661 | |
| 1662 | const max_number_length = 2 |
| 1663 | |
| 1664 | // Scan the version number of VERSION-DIRECTIVE. |
| 1665 | // |
| 1666 | // Scope: |
| 1667 | // %YAML 1.1 # a comment \n |
| 1668 | // ^ |
| 1669 | // %YAML 1.1 # a comment \n |
| 1670 | // ^ |
| 1671 | func yaml_parser_scan_version_directive_number(parser *yaml_parser_t, start_mark yaml_mark_t, number *int8) bool { |
| 1672 | |
| 1673 | // Repeat while the next character is digit. |
| 1674 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1675 | return false |
| 1676 | } |
| 1677 | var value, length int8 |
| 1678 | for is_digit(parser.buffer, parser.buffer_pos) { |
| 1679 | // Check if the number is too long. |
| 1680 | length++ |
| 1681 | if length > max_number_length { |
| 1682 | return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive", |
| 1683 | start_mark, "found extremely long version number") |
| 1684 | } |
| 1685 | value = value*10 + int8(as_digit(parser.buffer, parser.buffer_pos)) |
| 1686 | skip(parser) |
| 1687 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1688 | return false |
| 1689 | } |
| 1690 | } |
| 1691 | |
| 1692 | // Check if the number was present. |
| 1693 | if length == 0 { |
| 1694 | return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive", |
| 1695 | start_mark, "did not find expected version number") |
| 1696 | } |
| 1697 | *number = value |
| 1698 | return true |
| 1699 | } |
| 1700 | |
| 1701 | // Scan the value of a TAG-DIRECTIVE token. |
| 1702 | // |
| 1703 | // Scope: |
| 1704 | // %TAG !yaml! tag:yaml.org,2002: \n |
| 1705 | // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 1706 | // |
| 1707 | func yaml_parser_scan_tag_directive_value(parser *yaml_parser_t, start_mark yaml_mark_t, handle, prefix *[]byte) bool { |
| 1708 | var handle_value, prefix_value []byte |
| 1709 | |
| 1710 | // Eat whitespaces. |
| 1711 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1712 | return false |
| 1713 | } |
| 1714 | |
| 1715 | for is_blank(parser.buffer, parser.buffer_pos) { |
| 1716 | skip(parser) |
| 1717 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1718 | return false |
| 1719 | } |
| 1720 | } |
| 1721 | |
| 1722 | // Scan a handle. |
| 1723 | if !yaml_parser_scan_tag_handle(parser, true, start_mark, &handle_value) { |
| 1724 | return false |
| 1725 | } |
| 1726 | |
| 1727 | // Expect a whitespace. |
| 1728 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1729 | return false |
| 1730 | } |
| 1731 | if !is_blank(parser.buffer, parser.buffer_pos) { |
| 1732 | yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive", |
| 1733 | start_mark, "did not find expected whitespace") |
| 1734 | return false |
| 1735 | } |
| 1736 | |
| 1737 | // Eat whitespaces. |
| 1738 | for is_blank(parser.buffer, parser.buffer_pos) { |
| 1739 | skip(parser) |
| 1740 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1741 | return false |
| 1742 | } |
| 1743 | } |
| 1744 | |
| 1745 | // Scan a prefix. |
| 1746 | if !yaml_parser_scan_tag_uri(parser, true, nil, start_mark, &prefix_value) { |
| 1747 | return false |
| 1748 | } |
| 1749 | |
| 1750 | // Expect a whitespace or line break. |
| 1751 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1752 | return false |
| 1753 | } |
| 1754 | if !is_blankz(parser.buffer, parser.buffer_pos) { |
| 1755 | yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive", |
| 1756 | start_mark, "did not find expected whitespace or line break") |
| 1757 | return false |
| 1758 | } |
| 1759 | |
| 1760 | *handle = handle_value |
| 1761 | *prefix = prefix_value |
| 1762 | return true |
| 1763 | } |
| 1764 | |
| 1765 | func yaml_parser_scan_anchor(parser *yaml_parser_t, token *yaml_token_t, typ yaml_token_type_t) bool { |
| 1766 | var s []byte |
| 1767 | |
| 1768 | // Eat the indicator character. |
| 1769 | start_mark := parser.mark |
| 1770 | skip(parser) |
| 1771 | |
| 1772 | // Consume the value. |
| 1773 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1774 | return false |
| 1775 | } |
| 1776 | |
| 1777 | for is_alpha(parser.buffer, parser.buffer_pos) { |
| 1778 | s = read(parser, s) |
| 1779 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1780 | return false |
| 1781 | } |
| 1782 | } |
| 1783 | |
| 1784 | end_mark := parser.mark |
| 1785 | |
| 1786 | /* |
| 1787 | * Check if length of the anchor is greater than 0 and it is followed by |
| 1788 | * a whitespace character or one of the indicators: |
| 1789 | * |
| 1790 | * '?', ':', ',', ']', '}', '%', '@', '`'. |
| 1791 | */ |
| 1792 | |
| 1793 | if len(s) == 0 || |
| 1794 | !(is_blankz(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == '?' || |
| 1795 | parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == ',' || |
| 1796 | parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '}' || |
| 1797 | parser.buffer[parser.buffer_pos] == '%' || parser.buffer[parser.buffer_pos] == '@' || |
| 1798 | parser.buffer[parser.buffer_pos] == '`') { |
| 1799 | context := "while scanning an alias" |
| 1800 | if typ == yaml_ANCHOR_TOKEN { |
| 1801 | context = "while scanning an anchor" |
| 1802 | } |
| 1803 | yaml_parser_set_scanner_error(parser, context, start_mark, |
| 1804 | "did not find expected alphabetic or numeric character") |
| 1805 | return false |
| 1806 | } |
| 1807 | |
| 1808 | // Create a token. |
| 1809 | *token = yaml_token_t{ |
| 1810 | typ: typ, |
| 1811 | start_mark: start_mark, |
| 1812 | end_mark: end_mark, |
| 1813 | value: s, |
| 1814 | } |
| 1815 | |
| 1816 | return true |
| 1817 | } |
| 1818 | |
| 1819 | /* |
| 1820 | * Scan a TAG token. |
| 1821 | */ |
| 1822 | |
| 1823 | func yaml_parser_scan_tag(parser *yaml_parser_t, token *yaml_token_t) bool { |
| 1824 | var handle, suffix []byte |
| 1825 | |
| 1826 | start_mark := parser.mark |
| 1827 | |
| 1828 | // Check if the tag is in the canonical form. |
| 1829 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 1830 | return false |
| 1831 | } |
| 1832 | |
| 1833 | if parser.buffer[parser.buffer_pos+1] == '<' { |
| 1834 | // Keep the handle as '' |
| 1835 | |
| 1836 | // Eat '!<' |
| 1837 | skip(parser) |
| 1838 | skip(parser) |
| 1839 | |
| 1840 | // Consume the tag value. |
| 1841 | if !yaml_parser_scan_tag_uri(parser, false, nil, start_mark, &suffix) { |
| 1842 | return false |
| 1843 | } |
| 1844 | |
| 1845 | // Check for '>' and eat it. |
| 1846 | if parser.buffer[parser.buffer_pos] != '>' { |
| 1847 | yaml_parser_set_scanner_error(parser, "while scanning a tag", |
| 1848 | start_mark, "did not find the expected '>'") |
| 1849 | return false |
| 1850 | } |
| 1851 | |
| 1852 | skip(parser) |
| 1853 | } else { |
| 1854 | // The tag has either the '!suffix' or the '!handle!suffix' form. |
| 1855 | |
| 1856 | // First, try to scan a handle. |
| 1857 | if !yaml_parser_scan_tag_handle(parser, false, start_mark, &handle) { |
| 1858 | return false |
| 1859 | } |
| 1860 | |
| 1861 | // Check if it is, indeed, handle. |
| 1862 | if handle[0] == '!' && len(handle) > 1 && handle[len(handle)-1] == '!' { |
| 1863 | // Scan the suffix now. |
| 1864 | if !yaml_parser_scan_tag_uri(parser, false, nil, start_mark, &suffix) { |
| 1865 | return false |
| 1866 | } |
| 1867 | } else { |
| 1868 | // It wasn't a handle after all. Scan the rest of the tag. |
| 1869 | if !yaml_parser_scan_tag_uri(parser, false, handle, start_mark, &suffix) { |
| 1870 | return false |
| 1871 | } |
| 1872 | |
| 1873 | // Set the handle to '!'. |
| 1874 | handle = []byte{'!'} |
| 1875 | |
| 1876 | // A special case: the '!' tag. Set the handle to '' and the |
| 1877 | // suffix to '!'. |
| 1878 | if len(suffix) == 0 { |
| 1879 | handle, suffix = suffix, handle |
| 1880 | } |
| 1881 | } |
| 1882 | } |
| 1883 | |
| 1884 | // Check the character which ends the tag. |
| 1885 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1886 | return false |
| 1887 | } |
| 1888 | if !is_blankz(parser.buffer, parser.buffer_pos) { |
| 1889 | yaml_parser_set_scanner_error(parser, "while scanning a tag", |
| 1890 | start_mark, "did not find expected whitespace or line break") |
| 1891 | return false |
| 1892 | } |
| 1893 | |
| 1894 | end_mark := parser.mark |
| 1895 | |
| 1896 | // Create a token. |
| 1897 | *token = yaml_token_t{ |
| 1898 | typ: yaml_TAG_TOKEN, |
| 1899 | start_mark: start_mark, |
| 1900 | end_mark: end_mark, |
| 1901 | value: handle, |
| 1902 | suffix: suffix, |
| 1903 | } |
| 1904 | return true |
| 1905 | } |
| 1906 | |
| 1907 | // Scan a tag handle. |
| 1908 | func yaml_parser_scan_tag_handle(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, handle *[]byte) bool { |
| 1909 | // Check the initial '!' character. |
| 1910 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1911 | return false |
| 1912 | } |
| 1913 | if parser.buffer[parser.buffer_pos] != '!' { |
| 1914 | yaml_parser_set_scanner_tag_error(parser, directive, |
| 1915 | start_mark, "did not find expected '!'") |
| 1916 | return false |
| 1917 | } |
| 1918 | |
| 1919 | var s []byte |
| 1920 | |
| 1921 | // Copy the '!' character. |
| 1922 | s = read(parser, s) |
| 1923 | |
| 1924 | // Copy all subsequent alphabetical and numerical characters. |
| 1925 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1926 | return false |
| 1927 | } |
| 1928 | for is_alpha(parser.buffer, parser.buffer_pos) { |
| 1929 | s = read(parser, s) |
| 1930 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1931 | return false |
| 1932 | } |
| 1933 | } |
| 1934 | |
| 1935 | // Check if the trailing character is '!' and copy it. |
| 1936 | if parser.buffer[parser.buffer_pos] == '!' { |
| 1937 | s = read(parser, s) |
| 1938 | } else { |
| 1939 | // It's either the '!' tag or not really a tag handle. If it's a %TAG |
| 1940 | // directive, it's an error. If it's a tag token, it must be a part of URI. |
| 1941 | if directive && string(s) != "!" { |
| 1942 | yaml_parser_set_scanner_tag_error(parser, directive, |
| 1943 | start_mark, "did not find expected '!'") |
| 1944 | return false |
| 1945 | } |
| 1946 | } |
| 1947 | |
| 1948 | *handle = s |
| 1949 | return true |
| 1950 | } |
| 1951 | |
| 1952 | // Scan a tag. |
| 1953 | func yaml_parser_scan_tag_uri(parser *yaml_parser_t, directive bool, head []byte, start_mark yaml_mark_t, uri *[]byte) bool { |
| 1954 | //size_t length = head ? strlen((char *)head) : 0 |
| 1955 | var s []byte |
| 1956 | hasTag := len(head) > 0 |
| 1957 | |
| 1958 | // Copy the head if needed. |
| 1959 | // |
| 1960 | // Note that we don't copy the leading '!' character. |
| 1961 | if len(head) > 1 { |
| 1962 | s = append(s, head[1:]...) |
| 1963 | } |
| 1964 | |
| 1965 | // Scan the tag. |
| 1966 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1967 | return false |
| 1968 | } |
| 1969 | |
| 1970 | // The set of characters that may appear in URI is as follows: |
| 1971 | // |
| 1972 | // '0'-'9', 'A'-'Z', 'a'-'z', '_', '-', ';', '/', '?', ':', '@', '&', |
| 1973 | // '=', '+', '$', ',', '.', '!', '~', '*', '\'', '(', ')', '[', ']', |
| 1974 | // '%'. |
| 1975 | // [Go] Convert this into more reasonable logic. |
| 1976 | for is_alpha(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == ';' || |
| 1977 | parser.buffer[parser.buffer_pos] == '/' || parser.buffer[parser.buffer_pos] == '?' || |
| 1978 | parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == '@' || |
| 1979 | parser.buffer[parser.buffer_pos] == '&' || parser.buffer[parser.buffer_pos] == '=' || |
| 1980 | parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '$' || |
| 1981 | parser.buffer[parser.buffer_pos] == ',' || parser.buffer[parser.buffer_pos] == '.' || |
| 1982 | parser.buffer[parser.buffer_pos] == '!' || parser.buffer[parser.buffer_pos] == '~' || |
| 1983 | parser.buffer[parser.buffer_pos] == '*' || parser.buffer[parser.buffer_pos] == '\'' || |
| 1984 | parser.buffer[parser.buffer_pos] == '(' || parser.buffer[parser.buffer_pos] == ')' || |
| 1985 | parser.buffer[parser.buffer_pos] == '[' || parser.buffer[parser.buffer_pos] == ']' || |
| 1986 | parser.buffer[parser.buffer_pos] == '%' { |
| 1987 | // Check if it is a URI-escape sequence. |
| 1988 | if parser.buffer[parser.buffer_pos] == '%' { |
| 1989 | if !yaml_parser_scan_uri_escapes(parser, directive, start_mark, &s) { |
| 1990 | return false |
| 1991 | } |
| 1992 | } else { |
| 1993 | s = read(parser, s) |
| 1994 | } |
| 1995 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 1996 | return false |
| 1997 | } |
| 1998 | hasTag = true |
| 1999 | } |
| 2000 | |
| 2001 | if !hasTag { |
| 2002 | yaml_parser_set_scanner_tag_error(parser, directive, |
| 2003 | start_mark, "did not find expected tag URI") |
| 2004 | return false |
| 2005 | } |
| 2006 | *uri = s |
| 2007 | return true |
| 2008 | } |
| 2009 | |
| 2010 | // Decode an URI-escape sequence corresponding to a single UTF-8 character. |
| 2011 | func yaml_parser_scan_uri_escapes(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, s *[]byte) bool { |
| 2012 | |
| 2013 | // Decode the required number of characters. |
| 2014 | w := 1024 |
| 2015 | for w > 0 { |
| 2016 | // Check for a URI-escaped octet. |
| 2017 | if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) { |
| 2018 | return false |
| 2019 | } |
| 2020 | |
| 2021 | if !(parser.buffer[parser.buffer_pos] == '%' && |
| 2022 | is_hex(parser.buffer, parser.buffer_pos+1) && |
| 2023 | is_hex(parser.buffer, parser.buffer_pos+2)) { |
| 2024 | return yaml_parser_set_scanner_tag_error(parser, directive, |
| 2025 | start_mark, "did not find URI escaped octet") |
| 2026 | } |
| 2027 | |
| 2028 | // Get the octet. |
| 2029 | octet := byte((as_hex(parser.buffer, parser.buffer_pos+1) << 4) + as_hex(parser.buffer, parser.buffer_pos+2)) |
| 2030 | |
| 2031 | // If it is the leading octet, determine the length of the UTF-8 sequence. |
| 2032 | if w == 1024 { |
| 2033 | w = width(octet) |
| 2034 | if w == 0 { |
| 2035 | return yaml_parser_set_scanner_tag_error(parser, directive, |
| 2036 | start_mark, "found an incorrect leading UTF-8 octet") |
| 2037 | } |
| 2038 | } else { |
| 2039 | // Check if the trailing octet is correct. |
| 2040 | if octet&0xC0 != 0x80 { |
| 2041 | return yaml_parser_set_scanner_tag_error(parser, directive, |
| 2042 | start_mark, "found an incorrect trailing UTF-8 octet") |
| 2043 | } |
| 2044 | } |
| 2045 | |
| 2046 | // Copy the octet and move the pointers. |
| 2047 | *s = append(*s, octet) |
| 2048 | skip(parser) |
| 2049 | skip(parser) |
| 2050 | skip(parser) |
| 2051 | w-- |
| 2052 | } |
| 2053 | return true |
| 2054 | } |
| 2055 | |
| 2056 | // Scan a block scalar. |
| 2057 | func yaml_parser_scan_block_scalar(parser *yaml_parser_t, token *yaml_token_t, literal bool) bool { |
| 2058 | // Eat the indicator '|' or '>'. |
| 2059 | start_mark := parser.mark |
| 2060 | skip(parser) |
| 2061 | |
| 2062 | // Scan the additional block scalar indicators. |
| 2063 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2064 | return false |
| 2065 | } |
| 2066 | |
| 2067 | // Check for a chomping indicator. |
| 2068 | var chomping, increment int |
| 2069 | if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' { |
| 2070 | // Set the chomping method and eat the indicator. |
| 2071 | if parser.buffer[parser.buffer_pos] == '+' { |
| 2072 | chomping = +1 |
| 2073 | } else { |
| 2074 | chomping = -1 |
| 2075 | } |
| 2076 | skip(parser) |
| 2077 | |
| 2078 | // Check for an indentation indicator. |
| 2079 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2080 | return false |
| 2081 | } |
| 2082 | if is_digit(parser.buffer, parser.buffer_pos) { |
| 2083 | // Check that the indentation is greater than 0. |
| 2084 | if parser.buffer[parser.buffer_pos] == '0' { |
| 2085 | yaml_parser_set_scanner_error(parser, "while scanning a block scalar", |
| 2086 | start_mark, "found an indentation indicator equal to 0") |
| 2087 | return false |
| 2088 | } |
| 2089 | |
| 2090 | // Get the indentation level and eat the indicator. |
| 2091 | increment = as_digit(parser.buffer, parser.buffer_pos) |
| 2092 | skip(parser) |
| 2093 | } |
| 2094 | |
| 2095 | } else if is_digit(parser.buffer, parser.buffer_pos) { |
| 2096 | // Do the same as above, but in the opposite order. |
| 2097 | |
| 2098 | if parser.buffer[parser.buffer_pos] == '0' { |
| 2099 | yaml_parser_set_scanner_error(parser, "while scanning a block scalar", |
| 2100 | start_mark, "found an indentation indicator equal to 0") |
| 2101 | return false |
| 2102 | } |
| 2103 | increment = as_digit(parser.buffer, parser.buffer_pos) |
| 2104 | skip(parser) |
| 2105 | |
| 2106 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2107 | return false |
| 2108 | } |
| 2109 | if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' { |
| 2110 | if parser.buffer[parser.buffer_pos] == '+' { |
| 2111 | chomping = +1 |
| 2112 | } else { |
| 2113 | chomping = -1 |
| 2114 | } |
| 2115 | skip(parser) |
| 2116 | } |
| 2117 | } |
| 2118 | |
| 2119 | // Eat whitespaces and comments to the end of the line. |
| 2120 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2121 | return false |
| 2122 | } |
| 2123 | for is_blank(parser.buffer, parser.buffer_pos) { |
| 2124 | skip(parser) |
| 2125 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2126 | return false |
| 2127 | } |
| 2128 | } |
| 2129 | if parser.buffer[parser.buffer_pos] == '#' { |
| 2130 | for !is_breakz(parser.buffer, parser.buffer_pos) { |
| 2131 | skip(parser) |
| 2132 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2133 | return false |
| 2134 | } |
| 2135 | } |
| 2136 | } |
| 2137 | |
| 2138 | // Check if we are at the end of the line. |
| 2139 | if !is_breakz(parser.buffer, parser.buffer_pos) { |
| 2140 | yaml_parser_set_scanner_error(parser, "while scanning a block scalar", |
| 2141 | start_mark, "did not find expected comment or line break") |
| 2142 | return false |
| 2143 | } |
| 2144 | |
| 2145 | // Eat a line break. |
| 2146 | if is_break(parser.buffer, parser.buffer_pos) { |
| 2147 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2148 | return false |
| 2149 | } |
| 2150 | skip_line(parser) |
| 2151 | } |
| 2152 | |
| 2153 | end_mark := parser.mark |
| 2154 | |
| 2155 | // Set the indentation level if it was specified. |
| 2156 | var indent int |
| 2157 | if increment > 0 { |
| 2158 | if parser.indent >= 0 { |
| 2159 | indent = parser.indent + increment |
| 2160 | } else { |
| 2161 | indent = increment |
| 2162 | } |
| 2163 | } |
| 2164 | |
| 2165 | // Scan the leading line breaks and determine the indentation level if needed. |
| 2166 | var s, leading_break, trailing_breaks []byte |
| 2167 | if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) { |
| 2168 | return false |
| 2169 | } |
| 2170 | |
| 2171 | // Scan the block scalar content. |
| 2172 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2173 | return false |
| 2174 | } |
| 2175 | var leading_blank, trailing_blank bool |
| 2176 | for parser.mark.column == indent && !is_z(parser.buffer, parser.buffer_pos) { |
| 2177 | // We are at the beginning of a non-empty line. |
| 2178 | |
| 2179 | // Is it a trailing whitespace? |
| 2180 | trailing_blank = is_blank(parser.buffer, parser.buffer_pos) |
| 2181 | |
| 2182 | // Check if we need to fold the leading line break. |
| 2183 | if !literal && !leading_blank && !trailing_blank && len(leading_break) > 0 && leading_break[0] == '\n' { |
| 2184 | // Do we need to join the lines by space? |
| 2185 | if len(trailing_breaks) == 0 { |
| 2186 | s = append(s, ' ') |
| 2187 | } |
| 2188 | } else { |
| 2189 | s = append(s, leading_break...) |
| 2190 | } |
| 2191 | leading_break = leading_break[:0] |
| 2192 | |
| 2193 | // Append the remaining line breaks. |
| 2194 | s = append(s, trailing_breaks...) |
| 2195 | trailing_breaks = trailing_breaks[:0] |
| 2196 | |
| 2197 | // Is it a leading whitespace? |
| 2198 | leading_blank = is_blank(parser.buffer, parser.buffer_pos) |
| 2199 | |
| 2200 | // Consume the current line. |
| 2201 | for !is_breakz(parser.buffer, parser.buffer_pos) { |
| 2202 | s = read(parser, s) |
| 2203 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2204 | return false |
| 2205 | } |
| 2206 | } |
| 2207 | |
| 2208 | // Consume the line break. |
| 2209 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2210 | return false |
| 2211 | } |
| 2212 | |
| 2213 | leading_break = read_line(parser, leading_break) |
| 2214 | |
| 2215 | // Eat the following indentation spaces and line breaks. |
| 2216 | if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) { |
| 2217 | return false |
| 2218 | } |
| 2219 | } |
| 2220 | |
| 2221 | // Chomp the tail. |
| 2222 | if chomping != -1 { |
| 2223 | s = append(s, leading_break...) |
| 2224 | } |
| 2225 | if chomping == 1 { |
| 2226 | s = append(s, trailing_breaks...) |
| 2227 | } |
| 2228 | |
| 2229 | // Create a token. |
| 2230 | *token = yaml_token_t{ |
| 2231 | typ: yaml_SCALAR_TOKEN, |
| 2232 | start_mark: start_mark, |
| 2233 | end_mark: end_mark, |
| 2234 | value: s, |
| 2235 | style: yaml_LITERAL_SCALAR_STYLE, |
| 2236 | } |
| 2237 | if !literal { |
| 2238 | token.style = yaml_FOLDED_SCALAR_STYLE |
| 2239 | } |
| 2240 | return true |
| 2241 | } |
| 2242 | |
| 2243 | // Scan indentation spaces and line breaks for a block scalar. Determine the |
| 2244 | // indentation level if needed. |
| 2245 | func yaml_parser_scan_block_scalar_breaks(parser *yaml_parser_t, indent *int, breaks *[]byte, start_mark yaml_mark_t, end_mark *yaml_mark_t) bool { |
| 2246 | *end_mark = parser.mark |
| 2247 | |
| 2248 | // Eat the indentation spaces and line breaks. |
| 2249 | max_indent := 0 |
| 2250 | for { |
| 2251 | // Eat the indentation spaces. |
| 2252 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2253 | return false |
| 2254 | } |
| 2255 | for (*indent == 0 || parser.mark.column < *indent) && is_space(parser.buffer, parser.buffer_pos) { |
| 2256 | skip(parser) |
| 2257 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2258 | return false |
| 2259 | } |
| 2260 | } |
| 2261 | if parser.mark.column > max_indent { |
| 2262 | max_indent = parser.mark.column |
| 2263 | } |
| 2264 | |
| 2265 | // Check for a tab character messing the indentation. |
| 2266 | if (*indent == 0 || parser.mark.column < *indent) && is_tab(parser.buffer, parser.buffer_pos) { |
| 2267 | return yaml_parser_set_scanner_error(parser, "while scanning a block scalar", |
| 2268 | start_mark, "found a tab character where an indentation space is expected") |
| 2269 | } |
| 2270 | |
| 2271 | // Have we found a non-empty line? |
| 2272 | if !is_break(parser.buffer, parser.buffer_pos) { |
| 2273 | break |
| 2274 | } |
| 2275 | |
| 2276 | // Consume the line break. |
| 2277 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2278 | return false |
| 2279 | } |
| 2280 | // [Go] Should really be returning breaks instead. |
| 2281 | *breaks = read_line(parser, *breaks) |
| 2282 | *end_mark = parser.mark |
| 2283 | } |
| 2284 | |
| 2285 | // Determine the indentation level if needed. |
| 2286 | if *indent == 0 { |
| 2287 | *indent = max_indent |
| 2288 | if *indent < parser.indent+1 { |
| 2289 | *indent = parser.indent + 1 |
| 2290 | } |
| 2291 | if *indent < 1 { |
| 2292 | *indent = 1 |
| 2293 | } |
| 2294 | } |
| 2295 | return true |
| 2296 | } |
| 2297 | |
| 2298 | // Scan a quoted scalar. |
| 2299 | func yaml_parser_scan_flow_scalar(parser *yaml_parser_t, token *yaml_token_t, single bool) bool { |
| 2300 | // Eat the left quote. |
| 2301 | start_mark := parser.mark |
| 2302 | skip(parser) |
| 2303 | |
| 2304 | // Consume the content of the quoted scalar. |
| 2305 | var s, leading_break, trailing_breaks, whitespaces []byte |
| 2306 | for { |
| 2307 | // Check that there are no document indicators at the beginning of the line. |
| 2308 | if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) { |
| 2309 | return false |
| 2310 | } |
| 2311 | |
| 2312 | if parser.mark.column == 0 && |
| 2313 | ((parser.buffer[parser.buffer_pos+0] == '-' && |
| 2314 | parser.buffer[parser.buffer_pos+1] == '-' && |
| 2315 | parser.buffer[parser.buffer_pos+2] == '-') || |
| 2316 | (parser.buffer[parser.buffer_pos+0] == '.' && |
| 2317 | parser.buffer[parser.buffer_pos+1] == '.' && |
| 2318 | parser.buffer[parser.buffer_pos+2] == '.')) && |
| 2319 | is_blankz(parser.buffer, parser.buffer_pos+3) { |
| 2320 | yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar", |
| 2321 | start_mark, "found unexpected document indicator") |
| 2322 | return false |
| 2323 | } |
| 2324 | |
| 2325 | // Check for EOF. |
| 2326 | if is_z(parser.buffer, parser.buffer_pos) { |
| 2327 | yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar", |
| 2328 | start_mark, "found unexpected end of stream") |
| 2329 | return false |
| 2330 | } |
| 2331 | |
| 2332 | // Consume non-blank characters. |
| 2333 | leading_blanks := false |
| 2334 | for !is_blankz(parser.buffer, parser.buffer_pos) { |
| 2335 | if single && parser.buffer[parser.buffer_pos] == '\'' && parser.buffer[parser.buffer_pos+1] == '\'' { |
| 2336 | // Is is an escaped single quote. |
| 2337 | s = append(s, '\'') |
| 2338 | skip(parser) |
| 2339 | skip(parser) |
| 2340 | |
| 2341 | } else if single && parser.buffer[parser.buffer_pos] == '\'' { |
| 2342 | // It is a right single quote. |
| 2343 | break |
| 2344 | } else if !single && parser.buffer[parser.buffer_pos] == '"' { |
| 2345 | // It is a right double quote. |
| 2346 | break |
| 2347 | |
| 2348 | } else if !single && parser.buffer[parser.buffer_pos] == '\\' && is_break(parser.buffer, parser.buffer_pos+1) { |
| 2349 | // It is an escaped line break. |
| 2350 | if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) { |
| 2351 | return false |
| 2352 | } |
| 2353 | skip(parser) |
| 2354 | skip_line(parser) |
| 2355 | leading_blanks = true |
| 2356 | break |
| 2357 | |
| 2358 | } else if !single && parser.buffer[parser.buffer_pos] == '\\' { |
| 2359 | // It is an escape sequence. |
| 2360 | code_length := 0 |
| 2361 | |
| 2362 | // Check the escape character. |
| 2363 | switch parser.buffer[parser.buffer_pos+1] { |
| 2364 | case '0': |
| 2365 | s = append(s, 0) |
| 2366 | case 'a': |
| 2367 | s = append(s, '\x07') |
| 2368 | case 'b': |
| 2369 | s = append(s, '\x08') |
| 2370 | case 't', '\t': |
| 2371 | s = append(s, '\x09') |
| 2372 | case 'n': |
| 2373 | s = append(s, '\x0A') |
| 2374 | case 'v': |
| 2375 | s = append(s, '\x0B') |
| 2376 | case 'f': |
| 2377 | s = append(s, '\x0C') |
| 2378 | case 'r': |
| 2379 | s = append(s, '\x0D') |
| 2380 | case 'e': |
| 2381 | s = append(s, '\x1B') |
| 2382 | case ' ': |
| 2383 | s = append(s, '\x20') |
| 2384 | case '"': |
| 2385 | s = append(s, '"') |
| 2386 | case '\'': |
| 2387 | s = append(s, '\'') |
| 2388 | case '\\': |
| 2389 | s = append(s, '\\') |
| 2390 | case 'N': // NEL (#x85) |
| 2391 | s = append(s, '\xC2') |
| 2392 | s = append(s, '\x85') |
| 2393 | case '_': // #xA0 |
| 2394 | s = append(s, '\xC2') |
| 2395 | s = append(s, '\xA0') |
| 2396 | case 'L': // LS (#x2028) |
| 2397 | s = append(s, '\xE2') |
| 2398 | s = append(s, '\x80') |
| 2399 | s = append(s, '\xA8') |
| 2400 | case 'P': // PS (#x2029) |
| 2401 | s = append(s, '\xE2') |
| 2402 | s = append(s, '\x80') |
| 2403 | s = append(s, '\xA9') |
| 2404 | case 'x': |
| 2405 | code_length = 2 |
| 2406 | case 'u': |
| 2407 | code_length = 4 |
| 2408 | case 'U': |
| 2409 | code_length = 8 |
| 2410 | default: |
| 2411 | yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar", |
| 2412 | start_mark, "found unknown escape character") |
| 2413 | return false |
| 2414 | } |
| 2415 | |
| 2416 | skip(parser) |
| 2417 | skip(parser) |
| 2418 | |
| 2419 | // Consume an arbitrary escape code. |
| 2420 | if code_length > 0 { |
| 2421 | var value int |
| 2422 | |
| 2423 | // Scan the character value. |
| 2424 | if parser.unread < code_length && !yaml_parser_update_buffer(parser, code_length) { |
| 2425 | return false |
| 2426 | } |
| 2427 | for k := 0; k < code_length; k++ { |
| 2428 | if !is_hex(parser.buffer, parser.buffer_pos+k) { |
| 2429 | yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar", |
| 2430 | start_mark, "did not find expected hexdecimal number") |
| 2431 | return false |
| 2432 | } |
| 2433 | value = (value << 4) + as_hex(parser.buffer, parser.buffer_pos+k) |
| 2434 | } |
| 2435 | |
| 2436 | // Check the value and write the character. |
| 2437 | if (value >= 0xD800 && value <= 0xDFFF) || value > 0x10FFFF { |
| 2438 | yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar", |
| 2439 | start_mark, "found invalid Unicode character escape code") |
| 2440 | return false |
| 2441 | } |
| 2442 | if value <= 0x7F { |
| 2443 | s = append(s, byte(value)) |
| 2444 | } else if value <= 0x7FF { |
| 2445 | s = append(s, byte(0xC0+(value>>6))) |
| 2446 | s = append(s, byte(0x80+(value&0x3F))) |
| 2447 | } else if value <= 0xFFFF { |
| 2448 | s = append(s, byte(0xE0+(value>>12))) |
| 2449 | s = append(s, byte(0x80+((value>>6)&0x3F))) |
| 2450 | s = append(s, byte(0x80+(value&0x3F))) |
| 2451 | } else { |
| 2452 | s = append(s, byte(0xF0+(value>>18))) |
| 2453 | s = append(s, byte(0x80+((value>>12)&0x3F))) |
| 2454 | s = append(s, byte(0x80+((value>>6)&0x3F))) |
| 2455 | s = append(s, byte(0x80+(value&0x3F))) |
| 2456 | } |
| 2457 | |
| 2458 | // Advance the pointer. |
| 2459 | for k := 0; k < code_length; k++ { |
| 2460 | skip(parser) |
| 2461 | } |
| 2462 | } |
| 2463 | } else { |
| 2464 | // It is a non-escaped non-blank character. |
| 2465 | s = read(parser, s) |
| 2466 | } |
| 2467 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2468 | return false |
| 2469 | } |
| 2470 | } |
| 2471 | |
| 2472 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2473 | return false |
| 2474 | } |
| 2475 | |
| 2476 | // Check if we are at the end of the scalar. |
| 2477 | if single { |
| 2478 | if parser.buffer[parser.buffer_pos] == '\'' { |
| 2479 | break |
| 2480 | } |
| 2481 | } else { |
| 2482 | if parser.buffer[parser.buffer_pos] == '"' { |
| 2483 | break |
| 2484 | } |
| 2485 | } |
| 2486 | |
| 2487 | // Consume blank characters. |
| 2488 | for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) { |
| 2489 | if is_blank(parser.buffer, parser.buffer_pos) { |
| 2490 | // Consume a space or a tab character. |
| 2491 | if !leading_blanks { |
| 2492 | whitespaces = read(parser, whitespaces) |
| 2493 | } else { |
| 2494 | skip(parser) |
| 2495 | } |
| 2496 | } else { |
| 2497 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2498 | return false |
| 2499 | } |
| 2500 | |
| 2501 | // Check if it is a first line break. |
| 2502 | if !leading_blanks { |
| 2503 | whitespaces = whitespaces[:0] |
| 2504 | leading_break = read_line(parser, leading_break) |
| 2505 | leading_blanks = true |
| 2506 | } else { |
| 2507 | trailing_breaks = read_line(parser, trailing_breaks) |
| 2508 | } |
| 2509 | } |
| 2510 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2511 | return false |
| 2512 | } |
| 2513 | } |
| 2514 | |
| 2515 | // Join the whitespaces or fold line breaks. |
| 2516 | if leading_blanks { |
| 2517 | // Do we need to fold line breaks? |
| 2518 | if len(leading_break) > 0 && leading_break[0] == '\n' { |
| 2519 | if len(trailing_breaks) == 0 { |
| 2520 | s = append(s, ' ') |
| 2521 | } else { |
| 2522 | s = append(s, trailing_breaks...) |
| 2523 | } |
| 2524 | } else { |
| 2525 | s = append(s, leading_break...) |
| 2526 | s = append(s, trailing_breaks...) |
| 2527 | } |
| 2528 | trailing_breaks = trailing_breaks[:0] |
| 2529 | leading_break = leading_break[:0] |
| 2530 | } else { |
| 2531 | s = append(s, whitespaces...) |
| 2532 | whitespaces = whitespaces[:0] |
| 2533 | } |
| 2534 | } |
| 2535 | |
| 2536 | // Eat the right quote. |
| 2537 | skip(parser) |
| 2538 | end_mark := parser.mark |
| 2539 | |
| 2540 | // Create a token. |
| 2541 | *token = yaml_token_t{ |
| 2542 | typ: yaml_SCALAR_TOKEN, |
| 2543 | start_mark: start_mark, |
| 2544 | end_mark: end_mark, |
| 2545 | value: s, |
| 2546 | style: yaml_SINGLE_QUOTED_SCALAR_STYLE, |
| 2547 | } |
| 2548 | if !single { |
| 2549 | token.style = yaml_DOUBLE_QUOTED_SCALAR_STYLE |
| 2550 | } |
| 2551 | return true |
| 2552 | } |
| 2553 | |
| 2554 | // Scan a plain scalar. |
| 2555 | func yaml_parser_scan_plain_scalar(parser *yaml_parser_t, token *yaml_token_t) bool { |
| 2556 | |
| 2557 | var s, leading_break, trailing_breaks, whitespaces []byte |
| 2558 | var leading_blanks bool |
| 2559 | var indent = parser.indent + 1 |
| 2560 | |
| 2561 | start_mark := parser.mark |
| 2562 | end_mark := parser.mark |
| 2563 | |
| 2564 | // Consume the content of the plain scalar. |
| 2565 | for { |
| 2566 | // Check for a document indicator. |
| 2567 | if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) { |
| 2568 | return false |
| 2569 | } |
| 2570 | if parser.mark.column == 0 && |
| 2571 | ((parser.buffer[parser.buffer_pos+0] == '-' && |
| 2572 | parser.buffer[parser.buffer_pos+1] == '-' && |
| 2573 | parser.buffer[parser.buffer_pos+2] == '-') || |
| 2574 | (parser.buffer[parser.buffer_pos+0] == '.' && |
| 2575 | parser.buffer[parser.buffer_pos+1] == '.' && |
| 2576 | parser.buffer[parser.buffer_pos+2] == '.')) && |
| 2577 | is_blankz(parser.buffer, parser.buffer_pos+3) { |
| 2578 | break |
| 2579 | } |
| 2580 | |
| 2581 | // Check for a comment. |
| 2582 | if parser.buffer[parser.buffer_pos] == '#' { |
| 2583 | break |
| 2584 | } |
| 2585 | |
| 2586 | // Consume non-blank characters. |
| 2587 | for !is_blankz(parser.buffer, parser.buffer_pos) { |
| 2588 | |
| 2589 | // Check for indicators that may end a plain scalar. |
| 2590 | if (parser.buffer[parser.buffer_pos] == ':' && is_blankz(parser.buffer, parser.buffer_pos+1)) || |
| 2591 | (parser.flow_level > 0 && |
| 2592 | (parser.buffer[parser.buffer_pos] == ',' || |
| 2593 | parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == '[' || |
| 2594 | parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '{' || |
| 2595 | parser.buffer[parser.buffer_pos] == '}')) { |
| 2596 | break |
| 2597 | } |
| 2598 | |
| 2599 | // Check if we need to join whitespaces and breaks. |
| 2600 | if leading_blanks || len(whitespaces) > 0 { |
| 2601 | if leading_blanks { |
| 2602 | // Do we need to fold line breaks? |
| 2603 | if leading_break[0] == '\n' { |
| 2604 | if len(trailing_breaks) == 0 { |
| 2605 | s = append(s, ' ') |
| 2606 | } else { |
| 2607 | s = append(s, trailing_breaks...) |
| 2608 | } |
| 2609 | } else { |
| 2610 | s = append(s, leading_break...) |
| 2611 | s = append(s, trailing_breaks...) |
| 2612 | } |
| 2613 | trailing_breaks = trailing_breaks[:0] |
| 2614 | leading_break = leading_break[:0] |
| 2615 | leading_blanks = false |
| 2616 | } else { |
| 2617 | s = append(s, whitespaces...) |
| 2618 | whitespaces = whitespaces[:0] |
| 2619 | } |
| 2620 | } |
| 2621 | |
| 2622 | // Copy the character. |
| 2623 | s = read(parser, s) |
| 2624 | |
| 2625 | end_mark = parser.mark |
| 2626 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2627 | return false |
| 2628 | } |
| 2629 | } |
| 2630 | |
| 2631 | // Is it the end? |
| 2632 | if !(is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos)) { |
| 2633 | break |
| 2634 | } |
| 2635 | |
| 2636 | // Consume blank characters. |
| 2637 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2638 | return false |
| 2639 | } |
| 2640 | |
| 2641 | for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) { |
| 2642 | if is_blank(parser.buffer, parser.buffer_pos) { |
| 2643 | |
| 2644 | // Check for tab characters that abuse indentation. |
| 2645 | if leading_blanks && parser.mark.column < indent && is_tab(parser.buffer, parser.buffer_pos) { |
| 2646 | yaml_parser_set_scanner_error(parser, "while scanning a plain scalar", |
| 2647 | start_mark, "found a tab character that violates indentation") |
| 2648 | return false |
| 2649 | } |
| 2650 | |
| 2651 | // Consume a space or a tab character. |
| 2652 | if !leading_blanks { |
| 2653 | whitespaces = read(parser, whitespaces) |
| 2654 | } else { |
| 2655 | skip(parser) |
| 2656 | } |
| 2657 | } else { |
| 2658 | if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { |
| 2659 | return false |
| 2660 | } |
| 2661 | |
| 2662 | // Check if it is a first line break. |
| 2663 | if !leading_blanks { |
| 2664 | whitespaces = whitespaces[:0] |
| 2665 | leading_break = read_line(parser, leading_break) |
| 2666 | leading_blanks = true |
| 2667 | } else { |
| 2668 | trailing_breaks = read_line(parser, trailing_breaks) |
| 2669 | } |
| 2670 | } |
| 2671 | if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { |
| 2672 | return false |
| 2673 | } |
| 2674 | } |
| 2675 | |
| 2676 | // Check indentation level. |
| 2677 | if parser.flow_level == 0 && parser.mark.column < indent { |
| 2678 | break |
| 2679 | } |
| 2680 | } |
| 2681 | |
| 2682 | // Create a token. |
| 2683 | *token = yaml_token_t{ |
| 2684 | typ: yaml_SCALAR_TOKEN, |
| 2685 | start_mark: start_mark, |
| 2686 | end_mark: end_mark, |
| 2687 | value: s, |
| 2688 | style: yaml_PLAIN_SCALAR_STYLE, |
| 2689 | } |
| 2690 | |
| 2691 | // Note that we change the 'simple_key_allowed' flag. |
| 2692 | if leading_blanks { |
| 2693 | parser.simple_key_allowed = true |
| 2694 | } |
| 2695 | return true |
| 2696 | } |