github.com/keysonzzz/kmg@v0.0.0-20151121023212-05317bfd7d39/encoding/kmgYaml/scannerc.go (about) 1 package kmgYaml 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/cvs/current.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, "did not find URI escaped octet") 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 // A simple key is required only when it is the first token in the current 875 // line. Therefore it is always allowed. But we add a check anyway. 876 if required && !parser.simple_key_allowed { 877 panic("should not happen") 878 } 879 880 // 881 // If the current position may start a simple key, save it. 882 // 883 if parser.simple_key_allowed { 884 simple_key := yaml_simple_key_t{ 885 possible: true, 886 required: required, 887 token_number: parser.tokens_parsed + (len(parser.tokens) - parser.tokens_head), 888 } 889 simple_key.mark = parser.mark 890 891 if !yaml_parser_remove_simple_key(parser) { 892 return false 893 } 894 parser.simple_keys[len(parser.simple_keys)-1] = simple_key 895 } 896 return true 897 } 898 899 // Remove a potential simple key at the current flow level. 900 func yaml_parser_remove_simple_key(parser *yaml_parser_t) bool { 901 i := len(parser.simple_keys) - 1 902 if parser.simple_keys[i].possible { 903 // If the key is required, it is an error. 904 if parser.simple_keys[i].required { 905 return yaml_parser_set_scanner_error(parser, 906 "while scanning a simple key", parser.simple_keys[i].mark, 907 "could not find expected ':'") 908 } 909 } 910 // Remove the key from the stack. 911 parser.simple_keys[i].possible = false 912 return true 913 } 914 915 // Increase the flow level and resize the simple key list if needed. 916 func yaml_parser_increase_flow_level(parser *yaml_parser_t) bool { 917 // Reset the simple key on the next level. 918 parser.simple_keys = append(parser.simple_keys, yaml_simple_key_t{}) 919 920 // Increase the flow level. 921 parser.flow_level++ 922 return true 923 } 924 925 // Decrease the flow level. 926 func yaml_parser_decrease_flow_level(parser *yaml_parser_t) bool { 927 if parser.flow_level > 0 { 928 parser.flow_level-- 929 parser.simple_keys = parser.simple_keys[:len(parser.simple_keys)-1] 930 } 931 return true 932 } 933 934 // Push the current indentation level to the stack and set the new level 935 // the current column is greater than the indentation level. In this case, 936 // append or insert the specified token into the token queue. 937 func yaml_parser_roll_indent(parser *yaml_parser_t, column, number int, typ yaml_token_type_t, mark yaml_mark_t) bool { 938 // In the flow context, do nothing. 939 if parser.flow_level > 0 { 940 return true 941 } 942 943 if parser.indent < column { 944 // Push the current indentation level to the stack and set the new 945 // indentation level. 946 parser.indents = append(parser.indents, parser.indent) 947 parser.indent = column 948 949 // Create a token and insert it into the queue. 950 token := yaml_token_t{ 951 typ: typ, 952 start_mark: mark, 953 end_mark: mark, 954 } 955 if number > -1 { 956 number -= parser.tokens_parsed 957 } 958 yaml_insert_token(parser, number, &token) 959 } 960 return true 961 } 962 963 // Pop indentation levels from the indents stack until the current level 964 // becomes less or equal to the column. For each intendation level, append 965 // the BLOCK-END token. 966 func yaml_parser_unroll_indent(parser *yaml_parser_t, column int) bool { 967 // In the flow context, do nothing. 968 if parser.flow_level > 0 { 969 return true 970 } 971 972 // Loop through the intendation levels in the stack. 973 for parser.indent > column { 974 // Create a token and append it to the queue. 975 token := yaml_token_t{ 976 typ: yaml_BLOCK_END_TOKEN, 977 start_mark: parser.mark, 978 end_mark: parser.mark, 979 } 980 yaml_insert_token(parser, -1, &token) 981 982 // Pop the indentation level. 983 parser.indent = parser.indents[len(parser.indents)-1] 984 parser.indents = parser.indents[:len(parser.indents)-1] 985 } 986 return true 987 } 988 989 // Initialize the scanner and produce the STREAM-START token. 990 func yaml_parser_fetch_stream_start(parser *yaml_parser_t) bool { 991 992 // Set the initial indentation. 993 parser.indent = -1 994 995 // Initialize the simple key stack. 996 parser.simple_keys = append(parser.simple_keys, yaml_simple_key_t{}) 997 998 // A simple key is allowed at the beginning of the stream. 999 parser.simple_key_allowed = true 1000 1001 // We have started. 1002 parser.stream_start_produced = true 1003 1004 // Create the STREAM-START token and append it to the queue. 1005 token := yaml_token_t{ 1006 typ: yaml_STREAM_START_TOKEN, 1007 start_mark: parser.mark, 1008 end_mark: parser.mark, 1009 encoding: parser.encoding, 1010 } 1011 yaml_insert_token(parser, -1, &token) 1012 return true 1013 } 1014 1015 // Produce the STREAM-END token and shut down the scanner. 1016 func yaml_parser_fetch_stream_end(parser *yaml_parser_t) bool { 1017 1018 // Force new line. 1019 if parser.mark.column != 0 { 1020 parser.mark.column = 0 1021 parser.mark.line++ 1022 } 1023 1024 // Reset the indentation level. 1025 if !yaml_parser_unroll_indent(parser, -1) { 1026 return false 1027 } 1028 1029 // Reset simple keys. 1030 if !yaml_parser_remove_simple_key(parser) { 1031 return false 1032 } 1033 1034 parser.simple_key_allowed = false 1035 1036 // Create the STREAM-END token and append it to the queue. 1037 token := yaml_token_t{ 1038 typ: yaml_STREAM_END_TOKEN, 1039 start_mark: parser.mark, 1040 end_mark: parser.mark, 1041 } 1042 yaml_insert_token(parser, -1, &token) 1043 return true 1044 } 1045 1046 // Produce a VERSION-DIRECTIVE or TAG-DIRECTIVE token. 1047 func yaml_parser_fetch_directive(parser *yaml_parser_t) bool { 1048 // Reset the indentation level. 1049 if !yaml_parser_unroll_indent(parser, -1) { 1050 return false 1051 } 1052 1053 // Reset simple keys. 1054 if !yaml_parser_remove_simple_key(parser) { 1055 return false 1056 } 1057 1058 parser.simple_key_allowed = false 1059 1060 // Create the YAML-DIRECTIVE or TAG-DIRECTIVE token. 1061 token := yaml_token_t{} 1062 if !yaml_parser_scan_directive(parser, &token) { 1063 return false 1064 } 1065 // Append the token to the queue. 1066 yaml_insert_token(parser, -1, &token) 1067 return true 1068 } 1069 1070 // Produce the DOCUMENT-START or DOCUMENT-END token. 1071 func yaml_parser_fetch_document_indicator(parser *yaml_parser_t, typ yaml_token_type_t) bool { 1072 // Reset the indentation level. 1073 if !yaml_parser_unroll_indent(parser, -1) { 1074 return false 1075 } 1076 1077 // Reset simple keys. 1078 if !yaml_parser_remove_simple_key(parser) { 1079 return false 1080 } 1081 1082 parser.simple_key_allowed = false 1083 1084 // Consume the token. 1085 start_mark := parser.mark 1086 1087 skip(parser) 1088 skip(parser) 1089 skip(parser) 1090 1091 end_mark := parser.mark 1092 1093 // Create the DOCUMENT-START or DOCUMENT-END token. 1094 token := yaml_token_t{ 1095 typ: typ, 1096 start_mark: start_mark, 1097 end_mark: end_mark, 1098 } 1099 // Append the token to the queue. 1100 yaml_insert_token(parser, -1, &token) 1101 return true 1102 } 1103 1104 // Produce the FLOW-SEQUENCE-START or FLOW-MAPPING-START token. 1105 func yaml_parser_fetch_flow_collection_start(parser *yaml_parser_t, typ yaml_token_type_t) bool { 1106 // The indicators '[' and '{' may start a simple key. 1107 if !yaml_parser_save_simple_key(parser) { 1108 return false 1109 } 1110 1111 // Increase the flow level. 1112 if !yaml_parser_increase_flow_level(parser) { 1113 return false 1114 } 1115 1116 // A simple key may follow the indicators '[' and '{'. 1117 parser.simple_key_allowed = true 1118 1119 // Consume the token. 1120 start_mark := parser.mark 1121 skip(parser) 1122 end_mark := parser.mark 1123 1124 // Create the FLOW-SEQUENCE-START of FLOW-MAPPING-START token. 1125 token := yaml_token_t{ 1126 typ: typ, 1127 start_mark: start_mark, 1128 end_mark: end_mark, 1129 } 1130 // Append the token to the queue. 1131 yaml_insert_token(parser, -1, &token) 1132 return true 1133 } 1134 1135 // Produce the FLOW-SEQUENCE-END or FLOW-MAPPING-END token. 1136 func yaml_parser_fetch_flow_collection_end(parser *yaml_parser_t, typ yaml_token_type_t) bool { 1137 // Reset any potential simple key on the current flow level. 1138 if !yaml_parser_remove_simple_key(parser) { 1139 return false 1140 } 1141 1142 // Decrease the flow level. 1143 if !yaml_parser_decrease_flow_level(parser) { 1144 return false 1145 } 1146 1147 // No simple keys after the indicators ']' and '}'. 1148 parser.simple_key_allowed = false 1149 1150 // Consume the token. 1151 1152 start_mark := parser.mark 1153 skip(parser) 1154 end_mark := parser.mark 1155 1156 // Create the FLOW-SEQUENCE-END of FLOW-MAPPING-END token. 1157 token := yaml_token_t{ 1158 typ: typ, 1159 start_mark: start_mark, 1160 end_mark: end_mark, 1161 } 1162 // Append the token to the queue. 1163 yaml_insert_token(parser, -1, &token) 1164 return true 1165 } 1166 1167 // Produce the FLOW-ENTRY token. 1168 func yaml_parser_fetch_flow_entry(parser *yaml_parser_t) bool { 1169 // Reset any potential simple keys on the current flow level. 1170 if !yaml_parser_remove_simple_key(parser) { 1171 return false 1172 } 1173 1174 // Simple keys are allowed after ','. 1175 parser.simple_key_allowed = true 1176 1177 // Consume the token. 1178 start_mark := parser.mark 1179 skip(parser) 1180 end_mark := parser.mark 1181 1182 // Create the FLOW-ENTRY token and append it to the queue. 1183 token := yaml_token_t{ 1184 typ: yaml_FLOW_ENTRY_TOKEN, 1185 start_mark: start_mark, 1186 end_mark: end_mark, 1187 } 1188 yaml_insert_token(parser, -1, &token) 1189 return true 1190 } 1191 1192 // Produce the BLOCK-ENTRY token. 1193 func yaml_parser_fetch_block_entry(parser *yaml_parser_t) bool { 1194 // Check if the scanner is in the block context. 1195 if parser.flow_level == 0 { 1196 // Check if we are allowed to start a new entry. 1197 if !parser.simple_key_allowed { 1198 return yaml_parser_set_scanner_error(parser, "", parser.mark, 1199 "block sequence entries are not allowed in this context") 1200 } 1201 // Add the BLOCK-SEQUENCE-START token if needed. 1202 if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_SEQUENCE_START_TOKEN, parser.mark) { 1203 return false 1204 } 1205 } else { 1206 // It is an error for the '-' indicator to occur in the flow context, 1207 // but we let the Parser detect and report about it because the Parser 1208 // is able to point to the context. 1209 } 1210 1211 // Reset any potential simple keys on the current flow level. 1212 if !yaml_parser_remove_simple_key(parser) { 1213 return false 1214 } 1215 1216 // Simple keys are allowed after '-'. 1217 parser.simple_key_allowed = true 1218 1219 // Consume the token. 1220 start_mark := parser.mark 1221 skip(parser) 1222 end_mark := parser.mark 1223 1224 // Create the BLOCK-ENTRY token and append it to the queue. 1225 token := yaml_token_t{ 1226 typ: yaml_BLOCK_ENTRY_TOKEN, 1227 start_mark: start_mark, 1228 end_mark: end_mark, 1229 } 1230 yaml_insert_token(parser, -1, &token) 1231 return true 1232 } 1233 1234 // Produce the KEY token. 1235 func yaml_parser_fetch_key(parser *yaml_parser_t) bool { 1236 1237 // In the block context, additional checks are required. 1238 if parser.flow_level == 0 { 1239 // Check if we are allowed to start a new key (not nessesary simple). 1240 if !parser.simple_key_allowed { 1241 return yaml_parser_set_scanner_error(parser, "", parser.mark, 1242 "mapping keys are not allowed in this context") 1243 } 1244 // Add the BLOCK-MAPPING-START token if needed. 1245 if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_MAPPING_START_TOKEN, parser.mark) { 1246 return false 1247 } 1248 } 1249 1250 // Reset any potential simple keys on the current flow level. 1251 if !yaml_parser_remove_simple_key(parser) { 1252 return false 1253 } 1254 1255 // Simple keys are allowed after '?' in the block context. 1256 parser.simple_key_allowed = parser.flow_level == 0 1257 1258 // Consume the token. 1259 start_mark := parser.mark 1260 skip(parser) 1261 end_mark := parser.mark 1262 1263 // Create the KEY token and append it to the queue. 1264 token := yaml_token_t{ 1265 typ: yaml_KEY_TOKEN, 1266 start_mark: start_mark, 1267 end_mark: end_mark, 1268 } 1269 yaml_insert_token(parser, -1, &token) 1270 return true 1271 } 1272 1273 // Produce the VALUE token. 1274 func yaml_parser_fetch_value(parser *yaml_parser_t) bool { 1275 1276 simple_key := &parser.simple_keys[len(parser.simple_keys)-1] 1277 1278 // Have we found a simple key? 1279 if simple_key.possible { 1280 // Create the KEY token and insert it into the queue. 1281 token := yaml_token_t{ 1282 typ: yaml_KEY_TOKEN, 1283 start_mark: simple_key.mark, 1284 end_mark: simple_key.mark, 1285 } 1286 yaml_insert_token(parser, simple_key.token_number-parser.tokens_parsed, &token) 1287 1288 // In the block context, we may need to add the BLOCK-MAPPING-START token. 1289 if !yaml_parser_roll_indent(parser, simple_key.mark.column, 1290 simple_key.token_number, 1291 yaml_BLOCK_MAPPING_START_TOKEN, simple_key.mark) { 1292 return false 1293 } 1294 1295 // Remove the simple key. 1296 simple_key.possible = false 1297 1298 // A simple key cannot follow another simple key. 1299 parser.simple_key_allowed = false 1300 1301 } else { 1302 // The ':' indicator follows a complex key. 1303 1304 // In the block context, extra checks are required. 1305 if parser.flow_level == 0 { 1306 1307 // Check if we are allowed to start a complex value. 1308 if !parser.simple_key_allowed { 1309 panic("mapping values are not allowed in this context1") 1310 //return yaml_parser_set_scanner_error(parser, "", parser.mark, 1311 // "mapping values are not allowed in this context") 1312 } 1313 1314 // Add the BLOCK-MAPPING-START token if needed. 1315 if !yaml_parser_roll_indent(parser, parser.mark.column, -1, yaml_BLOCK_MAPPING_START_TOKEN, parser.mark) { 1316 return false 1317 } 1318 } 1319 1320 // Simple keys after ':' are allowed in the block context. 1321 parser.simple_key_allowed = parser.flow_level == 0 1322 } 1323 1324 // Consume the token. 1325 start_mark := parser.mark 1326 skip(parser) 1327 end_mark := parser.mark 1328 1329 // Create the VALUE token and append it to the queue. 1330 token := yaml_token_t{ 1331 typ: yaml_VALUE_TOKEN, 1332 start_mark: start_mark, 1333 end_mark: end_mark, 1334 } 1335 yaml_insert_token(parser, -1, &token) 1336 return true 1337 } 1338 1339 // Produce the ALIAS or ANCHOR token. 1340 func yaml_parser_fetch_anchor(parser *yaml_parser_t, typ yaml_token_type_t) bool { 1341 // An anchor or an alias could be a simple key. 1342 if !yaml_parser_save_simple_key(parser) { 1343 return false 1344 } 1345 1346 // A simple key cannot follow an anchor or an alias. 1347 parser.simple_key_allowed = false 1348 1349 // Create the ALIAS or ANCHOR token and append it to the queue. 1350 var token yaml_token_t 1351 if !yaml_parser_scan_anchor(parser, &token, typ) { 1352 return false 1353 } 1354 yaml_insert_token(parser, -1, &token) 1355 return true 1356 } 1357 1358 // Produce the TAG token. 1359 func yaml_parser_fetch_tag(parser *yaml_parser_t) bool { 1360 // A tag could be a simple key. 1361 if !yaml_parser_save_simple_key(parser) { 1362 return false 1363 } 1364 1365 // A simple key cannot follow a tag. 1366 parser.simple_key_allowed = false 1367 1368 // Create the TAG token and append it to the queue. 1369 var token yaml_token_t 1370 if !yaml_parser_scan_tag(parser, &token) { 1371 return false 1372 } 1373 yaml_insert_token(parser, -1, &token) 1374 return true 1375 } 1376 1377 // Produce the SCALAR(...,literal) or SCALAR(...,folded) tokens. 1378 func yaml_parser_fetch_block_scalar(parser *yaml_parser_t, literal bool) bool { 1379 // Remove any potential simple keys. 1380 if !yaml_parser_remove_simple_key(parser) { 1381 return false 1382 } 1383 1384 // A simple key may follow a block scalar. 1385 parser.simple_key_allowed = true 1386 1387 // Create the SCALAR token and append it to the queue. 1388 var token yaml_token_t 1389 if !yaml_parser_scan_block_scalar(parser, &token, literal) { 1390 return false 1391 } 1392 yaml_insert_token(parser, -1, &token) 1393 return true 1394 } 1395 1396 // Produce the SCALAR(...,single-quoted) or SCALAR(...,double-quoted) tokens. 1397 func yaml_parser_fetch_flow_scalar(parser *yaml_parser_t, single bool) bool { 1398 // A plain scalar could be a simple key. 1399 if !yaml_parser_save_simple_key(parser) { 1400 return false 1401 } 1402 1403 // A simple key cannot follow a flow scalar. 1404 parser.simple_key_allowed = false 1405 1406 // Create the SCALAR token and append it to the queue. 1407 var token yaml_token_t 1408 if !yaml_parser_scan_flow_scalar(parser, &token, single) { 1409 return false 1410 } 1411 yaml_insert_token(parser, -1, &token) 1412 return true 1413 } 1414 1415 // Produce the SCALAR(...,plain) token. 1416 func yaml_parser_fetch_plain_scalar(parser *yaml_parser_t) bool { 1417 // A plain scalar could be a simple key. 1418 if !yaml_parser_save_simple_key(parser) { 1419 return false 1420 } 1421 1422 // A simple key cannot follow a flow scalar. 1423 parser.simple_key_allowed = false 1424 1425 // Create the SCALAR token and append it to the queue. 1426 var token yaml_token_t 1427 if !yaml_parser_scan_plain_scalar(parser, &token) { 1428 return false 1429 } 1430 yaml_insert_token(parser, -1, &token) 1431 return true 1432 } 1433 1434 // Eat whitespaces and comments until the next token is found. 1435 func yaml_parser_scan_to_next_token(parser *yaml_parser_t) bool { 1436 1437 // Until the next token is not found. 1438 for { 1439 // Allow the BOM mark to start a line. 1440 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1441 return false 1442 } 1443 if parser.mark.column == 0 && is_bom(parser.buffer, parser.buffer_pos) { 1444 skip(parser) 1445 } 1446 1447 // Eat whitespaces. 1448 // Tabs are allowed: 1449 // - in the flow context 1450 // - in the block context, but not at the beginning of the line or 1451 // after '-', '?', or ':' (complex value). 1452 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1453 return false 1454 } 1455 1456 for parser.buffer[parser.buffer_pos] == ' ' || ((parser.flow_level > 0 || !parser.simple_key_allowed) && parser.buffer[parser.buffer_pos] == '\t') { 1457 skip(parser) 1458 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1459 return false 1460 } 1461 } 1462 1463 // Eat a comment until a line break. 1464 if parser.buffer[parser.buffer_pos] == '#' { 1465 for !is_breakz(parser.buffer, parser.buffer_pos) { 1466 skip(parser) 1467 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1468 return false 1469 } 1470 } 1471 } 1472 1473 // If it is a line break, eat it. 1474 if is_break(parser.buffer, parser.buffer_pos) { 1475 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 1476 return false 1477 } 1478 skip_line(parser) 1479 1480 // In the block context, a new line may start a simple key. 1481 if parser.flow_level == 0 { 1482 parser.simple_key_allowed = true 1483 } 1484 } else { 1485 break // We have found a token. 1486 } 1487 } 1488 1489 return true 1490 } 1491 1492 // Scan a YAML-DIRECTIVE or TAG-DIRECTIVE token. 1493 // 1494 // Scope: 1495 // %YAML 1.1 # a comment \n 1496 // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 1497 // %TAG !yaml! tag:yaml.org,2002: \n 1498 // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 1499 // 1500 func yaml_parser_scan_directive(parser *yaml_parser_t, token *yaml_token_t) bool { 1501 // Eat '%'. 1502 start_mark := parser.mark 1503 skip(parser) 1504 1505 // Scan the directive name. 1506 var name []byte 1507 if !yaml_parser_scan_directive_name(parser, start_mark, &name) { 1508 return false 1509 } 1510 1511 // Is it a YAML directive? 1512 if bytes.Equal(name, []byte("YAML")) { 1513 // Scan the VERSION directive value. 1514 var major, minor int8 1515 if !yaml_parser_scan_version_directive_value(parser, start_mark, &major, &minor) { 1516 return false 1517 } 1518 end_mark := parser.mark 1519 1520 // Create a VERSION-DIRECTIVE token. 1521 *token = yaml_token_t{ 1522 typ: yaml_VERSION_DIRECTIVE_TOKEN, 1523 start_mark: start_mark, 1524 end_mark: end_mark, 1525 major: major, 1526 minor: minor, 1527 } 1528 1529 // Is it a TAG directive? 1530 } else if bytes.Equal(name, []byte("TAG")) { 1531 // Scan the TAG directive value. 1532 var handle, prefix []byte 1533 if !yaml_parser_scan_tag_directive_value(parser, start_mark, &handle, &prefix) { 1534 return false 1535 } 1536 end_mark := parser.mark 1537 1538 // Create a TAG-DIRECTIVE token. 1539 *token = yaml_token_t{ 1540 typ: yaml_TAG_DIRECTIVE_TOKEN, 1541 start_mark: start_mark, 1542 end_mark: end_mark, 1543 value: handle, 1544 prefix: prefix, 1545 } 1546 1547 // Unknown directive. 1548 } else { 1549 yaml_parser_set_scanner_error(parser, "while scanning a directive", 1550 start_mark, "found uknown directive name") 1551 return false 1552 } 1553 1554 // Eat the rest of the line including any comments. 1555 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1556 return false 1557 } 1558 1559 for is_blank(parser.buffer, parser.buffer_pos) { 1560 skip(parser) 1561 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1562 return false 1563 } 1564 } 1565 1566 if parser.buffer[parser.buffer_pos] == '#' { 1567 for !is_breakz(parser.buffer, parser.buffer_pos) { 1568 skip(parser) 1569 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1570 return false 1571 } 1572 } 1573 } 1574 1575 // Check if we are at the end of the line. 1576 if !is_breakz(parser.buffer, parser.buffer_pos) { 1577 yaml_parser_set_scanner_error(parser, "while scanning a directive", 1578 start_mark, "did not find expected comment or line break") 1579 return false 1580 } 1581 1582 // Eat a line break. 1583 if is_break(parser.buffer, parser.buffer_pos) { 1584 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 1585 return false 1586 } 1587 skip_line(parser) 1588 } 1589 1590 return true 1591 } 1592 1593 // Scan the directive name. 1594 // 1595 // Scope: 1596 // %YAML 1.1 # a comment \n 1597 // ^^^^ 1598 // %TAG !yaml! tag:yaml.org,2002: \n 1599 // ^^^ 1600 // 1601 func yaml_parser_scan_directive_name(parser *yaml_parser_t, start_mark yaml_mark_t, name *[]byte) bool { 1602 // Consume the directive name. 1603 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1604 return false 1605 } 1606 1607 var s []byte 1608 for is_alpha(parser.buffer, parser.buffer_pos) { 1609 s = read(parser, s) 1610 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1611 return false 1612 } 1613 } 1614 1615 // Check if the name is empty. 1616 if len(s) == 0 { 1617 yaml_parser_set_scanner_error(parser, "while scanning a directive", 1618 start_mark, "could not find expected directive name") 1619 return false 1620 } 1621 1622 // Check for an blank character after the name. 1623 if !is_blankz(parser.buffer, parser.buffer_pos) { 1624 yaml_parser_set_scanner_error(parser, "while scanning a directive", 1625 start_mark, "found unexpected non-alphabetical character") 1626 return false 1627 } 1628 *name = s 1629 return true 1630 } 1631 1632 // Scan the value of VERSION-DIRECTIVE. 1633 // 1634 // Scope: 1635 // %YAML 1.1 # a comment \n 1636 // ^^^^^^ 1637 func yaml_parser_scan_version_directive_value(parser *yaml_parser_t, start_mark yaml_mark_t, major, minor *int8) bool { 1638 // Eat whitespaces. 1639 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1640 return false 1641 } 1642 for is_blank(parser.buffer, parser.buffer_pos) { 1643 skip(parser) 1644 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1645 return false 1646 } 1647 } 1648 1649 // Consume the major version number. 1650 if !yaml_parser_scan_version_directive_number(parser, start_mark, major) { 1651 return false 1652 } 1653 1654 // Eat '.'. 1655 if parser.buffer[parser.buffer_pos] != '.' { 1656 return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive", 1657 start_mark, "did not find expected digit or '.' character") 1658 } 1659 1660 skip(parser) 1661 1662 // Consume the minor version number. 1663 if !yaml_parser_scan_version_directive_number(parser, start_mark, minor) { 1664 return false 1665 } 1666 return true 1667 } 1668 1669 const max_number_length = 2 1670 1671 // Scan the version number of VERSION-DIRECTIVE. 1672 // 1673 // Scope: 1674 // %YAML 1.1 # a comment \n 1675 // ^ 1676 // %YAML 1.1 # a comment \n 1677 // ^ 1678 func yaml_parser_scan_version_directive_number(parser *yaml_parser_t, start_mark yaml_mark_t, number *int8) bool { 1679 1680 // Repeat while the next character is digit. 1681 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1682 return false 1683 } 1684 var value, length int8 1685 for is_digit(parser.buffer, parser.buffer_pos) { 1686 // Check if the number is too long. 1687 length++ 1688 if length > max_number_length { 1689 return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive", 1690 start_mark, "found extremely long version number") 1691 } 1692 value = value*10 + int8(as_digit(parser.buffer, parser.buffer_pos)) 1693 skip(parser) 1694 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1695 return false 1696 } 1697 } 1698 1699 // Check if the number was present. 1700 if length == 0 { 1701 return yaml_parser_set_scanner_error(parser, "while scanning a %YAML directive", 1702 start_mark, "did not find expected version number") 1703 } 1704 *number = value 1705 return true 1706 } 1707 1708 // Scan the value of a TAG-DIRECTIVE token. 1709 // 1710 // Scope: 1711 // %TAG !yaml! tag:yaml.org,2002: \n 1712 // ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 1713 // 1714 func yaml_parser_scan_tag_directive_value(parser *yaml_parser_t, start_mark yaml_mark_t, handle, prefix *[]byte) bool { 1715 var handle_value, prefix_value []byte 1716 1717 // Eat whitespaces. 1718 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1719 return false 1720 } 1721 1722 for is_blank(parser.buffer, parser.buffer_pos) { 1723 skip(parser) 1724 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1725 return false 1726 } 1727 } 1728 1729 // Scan a handle. 1730 if !yaml_parser_scan_tag_handle(parser, true, start_mark, &handle_value) { 1731 return false 1732 } 1733 1734 // Expect a whitespace. 1735 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1736 return false 1737 } 1738 if !is_blank(parser.buffer, parser.buffer_pos) { 1739 yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive", 1740 start_mark, "did not find expected whitespace") 1741 return false 1742 } 1743 1744 // Eat whitespaces. 1745 for is_blank(parser.buffer, parser.buffer_pos) { 1746 skip(parser) 1747 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1748 return false 1749 } 1750 } 1751 1752 // Scan a prefix. 1753 if !yaml_parser_scan_tag_uri(parser, true, nil, start_mark, &prefix_value) { 1754 return false 1755 } 1756 1757 // Expect a whitespace or line break. 1758 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1759 return false 1760 } 1761 if !is_blankz(parser.buffer, parser.buffer_pos) { 1762 yaml_parser_set_scanner_error(parser, "while scanning a %TAG directive", 1763 start_mark, "did not find expected whitespace or line break") 1764 return false 1765 } 1766 1767 *handle = handle_value 1768 *prefix = prefix_value 1769 return true 1770 } 1771 1772 func yaml_parser_scan_anchor(parser *yaml_parser_t, token *yaml_token_t, typ yaml_token_type_t) bool { 1773 var s []byte 1774 1775 // Eat the indicator character. 1776 start_mark := parser.mark 1777 skip(parser) 1778 1779 // Consume the value. 1780 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1781 return false 1782 } 1783 1784 for is_alpha(parser.buffer, parser.buffer_pos) { 1785 s = read(parser, s) 1786 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1787 return false 1788 } 1789 } 1790 1791 end_mark := parser.mark 1792 1793 /* 1794 * Check if length of the anchor is greater than 0 and it is followed by 1795 * a whitespace character or one of the indicators: 1796 * 1797 * '?', ':', ',', ']', '}', '%', '@', '`'. 1798 */ 1799 1800 if len(s) == 0 || 1801 !(is_blankz(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == '?' || 1802 parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == ',' || 1803 parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '}' || 1804 parser.buffer[parser.buffer_pos] == '%' || parser.buffer[parser.buffer_pos] == '@' || 1805 parser.buffer[parser.buffer_pos] == '`') { 1806 context := "while scanning an alias" 1807 if typ == yaml_ANCHOR_TOKEN { 1808 context = "while scanning an anchor" 1809 } 1810 yaml_parser_set_scanner_error(parser, context, start_mark, 1811 "did not find expected alphabetic or numeric character") 1812 return false 1813 } 1814 1815 // Create a token. 1816 *token = yaml_token_t{ 1817 typ: typ, 1818 start_mark: start_mark, 1819 end_mark: end_mark, 1820 value: s, 1821 } 1822 1823 return true 1824 } 1825 1826 /* 1827 * Scan a TAG token. 1828 */ 1829 1830 func yaml_parser_scan_tag(parser *yaml_parser_t, token *yaml_token_t) bool { 1831 var handle, suffix []byte 1832 1833 start_mark := parser.mark 1834 1835 // Check if the tag is in the canonical form. 1836 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 1837 return false 1838 } 1839 1840 if parser.buffer[parser.buffer_pos+1] == '<' { 1841 // Keep the handle as '' 1842 1843 // Eat '!<' 1844 skip(parser) 1845 skip(parser) 1846 1847 // Consume the tag value. 1848 if !yaml_parser_scan_tag_uri(parser, false, nil, start_mark, &suffix) { 1849 return false 1850 } 1851 1852 // Check for '>' and eat it. 1853 if parser.buffer[parser.buffer_pos] != '>' { 1854 yaml_parser_set_scanner_error(parser, "while scanning a tag", 1855 start_mark, "did not find the expected '>'") 1856 return false 1857 } 1858 1859 skip(parser) 1860 } else { 1861 // The tag has either the '!suffix' or the '!handle!suffix' form. 1862 1863 // First, try to scan a handle. 1864 if !yaml_parser_scan_tag_handle(parser, false, start_mark, &handle) { 1865 return false 1866 } 1867 1868 // Check if it is, indeed, handle. 1869 if handle[0] == '!' && len(handle) > 1 && handle[len(handle)-1] == '!' { 1870 // Scan the suffix now. 1871 if !yaml_parser_scan_tag_uri(parser, false, nil, start_mark, &suffix) { 1872 return false 1873 } 1874 } else { 1875 // It wasn't a handle after all. Scan the rest of the tag. 1876 if !yaml_parser_scan_tag_uri(parser, false, handle, start_mark, &suffix) { 1877 return false 1878 } 1879 1880 // Set the handle to '!'. 1881 handle = []byte{'!'} 1882 1883 // A special case: the '!' tag. Set the handle to '' and the 1884 // suffix to '!'. 1885 if len(suffix) == 0 { 1886 handle, suffix = suffix, handle 1887 } 1888 } 1889 } 1890 1891 // Check the character which ends the tag. 1892 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1893 return false 1894 } 1895 if !is_blankz(parser.buffer, parser.buffer_pos) { 1896 yaml_parser_set_scanner_error(parser, "while scanning a tag", 1897 start_mark, "did not find expected whitespace or line break") 1898 return false 1899 } 1900 1901 end_mark := parser.mark 1902 1903 // Create a token. 1904 *token = yaml_token_t{ 1905 typ: yaml_TAG_TOKEN, 1906 start_mark: start_mark, 1907 end_mark: end_mark, 1908 value: handle, 1909 suffix: suffix, 1910 } 1911 return true 1912 } 1913 1914 // Scan a tag handle. 1915 func yaml_parser_scan_tag_handle(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, handle *[]byte) bool { 1916 // Check the initial '!' character. 1917 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1918 return false 1919 } 1920 if parser.buffer[parser.buffer_pos] != '!' { 1921 yaml_parser_set_scanner_tag_error(parser, directive, 1922 start_mark, "did not find expected '!'") 1923 return false 1924 } 1925 1926 var s []byte 1927 1928 // Copy the '!' character. 1929 s = read(parser, s) 1930 1931 // Copy all subsequent alphabetical and numerical characters. 1932 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1933 return false 1934 } 1935 for is_alpha(parser.buffer, parser.buffer_pos) { 1936 s = read(parser, s) 1937 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1938 return false 1939 } 1940 } 1941 1942 // Check if the trailing character is '!' and copy it. 1943 if parser.buffer[parser.buffer_pos] == '!' { 1944 s = read(parser, s) 1945 } else { 1946 // It's either the '!' tag or not really a tag handle. If it's a %TAG 1947 // directive, it's an error. If it's a tag token, it must be a part of URI. 1948 if directive && !(s[0] == '!' && s[1] == 0) { 1949 yaml_parser_set_scanner_tag_error(parser, directive, 1950 start_mark, "did not find expected '!'") 1951 return false 1952 } 1953 } 1954 1955 *handle = s 1956 return true 1957 } 1958 1959 // Scan a tag. 1960 func yaml_parser_scan_tag_uri(parser *yaml_parser_t, directive bool, head []byte, start_mark yaml_mark_t, uri *[]byte) bool { 1961 //size_t length = head ? strlen((char *)head) : 0 1962 var s []byte 1963 1964 // Copy the head if needed. 1965 // 1966 // Note that we don't copy the leading '!' character. 1967 if len(head) > 1 { 1968 s = append(s, head[1:]...) 1969 } 1970 1971 // Scan the tag. 1972 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 1973 return false 1974 } 1975 1976 // The set of characters that may appear in URI is as follows: 1977 // 1978 // '0'-'9', 'A'-'Z', 'a'-'z', '_', '-', ';', '/', '?', ':', '@', '&', 1979 // '=', '+', '$', ',', '.', '!', '~', '*', '\'', '(', ')', '[', ']', 1980 // '%'. 1981 // [Go] Convert this into more reasonable logic. 1982 for is_alpha(parser.buffer, parser.buffer_pos) || parser.buffer[parser.buffer_pos] == ';' || 1983 parser.buffer[parser.buffer_pos] == '/' || parser.buffer[parser.buffer_pos] == '?' || 1984 parser.buffer[parser.buffer_pos] == ':' || parser.buffer[parser.buffer_pos] == '@' || 1985 parser.buffer[parser.buffer_pos] == '&' || parser.buffer[parser.buffer_pos] == '=' || 1986 parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '$' || 1987 parser.buffer[parser.buffer_pos] == ',' || parser.buffer[parser.buffer_pos] == '.' || 1988 parser.buffer[parser.buffer_pos] == '!' || parser.buffer[parser.buffer_pos] == '~' || 1989 parser.buffer[parser.buffer_pos] == '*' || parser.buffer[parser.buffer_pos] == '\'' || 1990 parser.buffer[parser.buffer_pos] == '(' || parser.buffer[parser.buffer_pos] == ')' || 1991 parser.buffer[parser.buffer_pos] == '[' || parser.buffer[parser.buffer_pos] == ']' || 1992 parser.buffer[parser.buffer_pos] == '%' { 1993 // Check if it is a URI-escape sequence. 1994 if parser.buffer[parser.buffer_pos] == '%' { 1995 if !yaml_parser_scan_uri_escapes(parser, directive, start_mark, &s) { 1996 return false 1997 } 1998 } else { 1999 s = read(parser, s) 2000 } 2001 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2002 return false 2003 } 2004 } 2005 2006 // Check if the tag is non-empty. 2007 if len(s) == 0 { 2008 yaml_parser_set_scanner_tag_error(parser, directive, 2009 start_mark, "did not find expected tag URI") 2010 return false 2011 } 2012 *uri = s 2013 return true 2014 } 2015 2016 // Decode an URI-escape sequence corresponding to a single UTF-8 character. 2017 func yaml_parser_scan_uri_escapes(parser *yaml_parser_t, directive bool, start_mark yaml_mark_t, s *[]byte) bool { 2018 2019 // Decode the required number of characters. 2020 w := 1024 2021 for w > 0 { 2022 // Check for a URI-escaped octet. 2023 if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) { 2024 return false 2025 } 2026 2027 if !(parser.buffer[parser.buffer_pos] == '%' && 2028 is_hex(parser.buffer, parser.buffer_pos+1) && 2029 is_hex(parser.buffer, parser.buffer_pos+2)) { 2030 return yaml_parser_set_scanner_tag_error(parser, directive, 2031 start_mark, "did not find URI escaped octet") 2032 } 2033 2034 // Get the octet. 2035 octet := byte((as_hex(parser.buffer, parser.buffer_pos+1) << 4) + as_hex(parser.buffer, parser.buffer_pos+2)) 2036 2037 // If it is the leading octet, determine the length of the UTF-8 sequence. 2038 if w == 1024 { 2039 w = width(octet) 2040 if w == 0 { 2041 return yaml_parser_set_scanner_tag_error(parser, directive, 2042 start_mark, "found an incorrect leading UTF-8 octet") 2043 } 2044 } else { 2045 // Check if the trailing octet is correct. 2046 if octet&0xC0 != 0x80 { 2047 return yaml_parser_set_scanner_tag_error(parser, directive, 2048 start_mark, "found an incorrect trailing UTF-8 octet") 2049 } 2050 } 2051 2052 // Copy the octet and move the pointers. 2053 *s = append(*s, octet) 2054 skip(parser) 2055 skip(parser) 2056 skip(parser) 2057 w-- 2058 } 2059 return true 2060 } 2061 2062 // Scan a block scalar. 2063 func yaml_parser_scan_block_scalar(parser *yaml_parser_t, token *yaml_token_t, literal bool) bool { 2064 // Eat the indicator '|' or '>'. 2065 start_mark := parser.mark 2066 skip(parser) 2067 2068 // Scan the additional block scalar indicators. 2069 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2070 return false 2071 } 2072 2073 // Check for a chomping indicator. 2074 var chomping, increment int 2075 if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' { 2076 // Set the chomping method and eat the indicator. 2077 if parser.buffer[parser.buffer_pos] == '+' { 2078 chomping = +1 2079 } else { 2080 chomping = -1 2081 } 2082 skip(parser) 2083 2084 // Check for an indentation indicator. 2085 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2086 return false 2087 } 2088 if is_digit(parser.buffer, parser.buffer_pos) { 2089 // Check that the intendation is greater than 0. 2090 if parser.buffer[parser.buffer_pos] == '0' { 2091 yaml_parser_set_scanner_error(parser, "while scanning a block scalar", 2092 start_mark, "found an intendation indicator equal to 0") 2093 return false 2094 } 2095 2096 // Get the intendation level and eat the indicator. 2097 increment = as_digit(parser.buffer, parser.buffer_pos) 2098 skip(parser) 2099 } 2100 2101 } else if is_digit(parser.buffer, parser.buffer_pos) { 2102 // Do the same as above, but in the opposite order. 2103 2104 if parser.buffer[parser.buffer_pos] == '0' { 2105 yaml_parser_set_scanner_error(parser, "while scanning a block scalar", 2106 start_mark, "found an intendation indicator equal to 0") 2107 return false 2108 } 2109 increment = as_digit(parser.buffer, parser.buffer_pos) 2110 skip(parser) 2111 2112 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2113 return false 2114 } 2115 if parser.buffer[parser.buffer_pos] == '+' || parser.buffer[parser.buffer_pos] == '-' { 2116 if parser.buffer[parser.buffer_pos] == '+' { 2117 chomping = +1 2118 } else { 2119 chomping = -1 2120 } 2121 skip(parser) 2122 } 2123 } 2124 2125 // Eat whitespaces and comments to the end of the line. 2126 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2127 return false 2128 } 2129 for is_blank(parser.buffer, parser.buffer_pos) { 2130 skip(parser) 2131 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2132 return false 2133 } 2134 } 2135 if parser.buffer[parser.buffer_pos] == '#' { 2136 for !is_breakz(parser.buffer, parser.buffer_pos) { 2137 skip(parser) 2138 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2139 return false 2140 } 2141 } 2142 } 2143 2144 // Check if we are at the end of the line. 2145 if !is_breakz(parser.buffer, parser.buffer_pos) { 2146 yaml_parser_set_scanner_error(parser, "while scanning a block scalar", 2147 start_mark, "did not find expected comment or line break") 2148 return false 2149 } 2150 2151 // Eat a line break. 2152 if is_break(parser.buffer, parser.buffer_pos) { 2153 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2154 return false 2155 } 2156 skip_line(parser) 2157 } 2158 2159 end_mark := parser.mark 2160 2161 // Set the intendation level if it was specified. 2162 var indent int 2163 if increment > 0 { 2164 if parser.indent >= 0 { 2165 indent = parser.indent + increment 2166 } else { 2167 indent = increment 2168 } 2169 } 2170 2171 // Scan the leading line breaks and determine the indentation level if needed. 2172 var s, leading_break, trailing_breaks []byte 2173 if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) { 2174 return false 2175 } 2176 2177 // Scan the block scalar content. 2178 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2179 return false 2180 } 2181 var leading_blank, trailing_blank bool 2182 for parser.mark.column == indent && !is_z(parser.buffer, parser.buffer_pos) { 2183 // We are at the beginning of a non-empty line. 2184 2185 // Is it a trailing whitespace? 2186 trailing_blank = is_blank(parser.buffer, parser.buffer_pos) 2187 2188 // Check if we need to fold the leading line break. 2189 if !literal && !leading_blank && !trailing_blank && len(leading_break) > 0 && leading_break[0] == '\n' { 2190 // Do we need to join the lines by space? 2191 if len(trailing_breaks) == 0 { 2192 s = append(s, ' ') 2193 } 2194 } else { 2195 s = append(s, leading_break...) 2196 } 2197 leading_break = leading_break[:0] 2198 2199 // Append the remaining line breaks. 2200 s = append(s, trailing_breaks...) 2201 trailing_breaks = trailing_breaks[:0] 2202 2203 // Is it a leading whitespace? 2204 leading_blank = is_blank(parser.buffer, parser.buffer_pos) 2205 2206 // Consume the current line. 2207 for !is_breakz(parser.buffer, parser.buffer_pos) { 2208 s = read(parser, s) 2209 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2210 return false 2211 } 2212 } 2213 2214 // Consume the line break. 2215 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2216 return false 2217 } 2218 2219 leading_break = read_line(parser, leading_break) 2220 2221 // Eat the following intendation spaces and line breaks. 2222 if !yaml_parser_scan_block_scalar_breaks(parser, &indent, &trailing_breaks, start_mark, &end_mark) { 2223 return false 2224 } 2225 } 2226 2227 // Chomp the tail. 2228 if chomping != -1 { 2229 s = append(s, leading_break...) 2230 } 2231 if chomping == 1 { 2232 s = append(s, trailing_breaks...) 2233 } 2234 2235 // Create a token. 2236 *token = yaml_token_t{ 2237 typ: yaml_SCALAR_TOKEN, 2238 start_mark: start_mark, 2239 end_mark: end_mark, 2240 value: s, 2241 style: yaml_LITERAL_SCALAR_STYLE, 2242 } 2243 if !literal { 2244 token.style = yaml_FOLDED_SCALAR_STYLE 2245 } 2246 return true 2247 } 2248 2249 // Scan intendation spaces and line breaks for a block scalar. Determine the 2250 // intendation level if needed. 2251 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 { 2252 *end_mark = parser.mark 2253 2254 // Eat the intendation spaces and line breaks. 2255 max_indent := 0 2256 for { 2257 // Eat the intendation spaces. 2258 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2259 return false 2260 } 2261 for (*indent == 0 || parser.mark.column < *indent) && is_space(parser.buffer, parser.buffer_pos) { 2262 skip(parser) 2263 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2264 return false 2265 } 2266 } 2267 if parser.mark.column > max_indent { 2268 max_indent = parser.mark.column 2269 } 2270 2271 // Check for a tab character messing the intendation. 2272 if (*indent == 0 || parser.mark.column < *indent) && is_tab(parser.buffer, parser.buffer_pos) { 2273 return yaml_parser_set_scanner_error(parser, "while scanning a block scalar", 2274 start_mark, "found a tab character where an intendation space is expected") 2275 } 2276 2277 // Have we found a non-empty line? 2278 if !is_break(parser.buffer, parser.buffer_pos) { 2279 break 2280 } 2281 2282 // Consume the line break. 2283 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2284 return false 2285 } 2286 // [Go] Should really be returning breaks instead. 2287 *breaks = read_line(parser, *breaks) 2288 *end_mark = parser.mark 2289 } 2290 2291 // Determine the indentation level if needed. 2292 if *indent == 0 { 2293 *indent = max_indent 2294 if *indent < parser.indent+1 { 2295 *indent = parser.indent + 1 2296 } 2297 if *indent < 1 { 2298 *indent = 1 2299 } 2300 } 2301 return true 2302 } 2303 2304 // Scan a quoted scalar. 2305 func yaml_parser_scan_flow_scalar(parser *yaml_parser_t, token *yaml_token_t, single bool) bool { 2306 // Eat the left quote. 2307 start_mark := parser.mark 2308 skip(parser) 2309 2310 // Consume the content of the quoted scalar. 2311 var s, leading_break, trailing_breaks, whitespaces []byte 2312 for { 2313 // Check that there are no document indicators at the beginning of the line. 2314 if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) { 2315 return false 2316 } 2317 2318 if parser.mark.column == 0 && 2319 ((parser.buffer[parser.buffer_pos+0] == '-' && 2320 parser.buffer[parser.buffer_pos+1] == '-' && 2321 parser.buffer[parser.buffer_pos+2] == '-') || 2322 (parser.buffer[parser.buffer_pos+0] == '.' && 2323 parser.buffer[parser.buffer_pos+1] == '.' && 2324 parser.buffer[parser.buffer_pos+2] == '.')) && 2325 is_blankz(parser.buffer, parser.buffer_pos+3) { 2326 yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar", 2327 start_mark, "found unexpected document indicator") 2328 return false 2329 } 2330 2331 // Check for EOF. 2332 if is_z(parser.buffer, parser.buffer_pos) { 2333 yaml_parser_set_scanner_error(parser, "while scanning a quoted scalar", 2334 start_mark, "found unexpected end of stream") 2335 return false 2336 } 2337 2338 // Consume non-blank characters. 2339 leading_blanks := false 2340 for !is_blankz(parser.buffer, parser.buffer_pos) { 2341 if single && parser.buffer[parser.buffer_pos] == '\'' && parser.buffer[parser.buffer_pos+1] == '\'' { 2342 // Is is an escaped single quote. 2343 s = append(s, '\'') 2344 skip(parser) 2345 skip(parser) 2346 2347 } else if single && parser.buffer[parser.buffer_pos] == '\'' { 2348 // It is a right single quote. 2349 break 2350 } else if !single && parser.buffer[parser.buffer_pos] == '"' { 2351 // It is a right double quote. 2352 break 2353 2354 } else if !single && parser.buffer[parser.buffer_pos] == '\\' && is_break(parser.buffer, parser.buffer_pos+1) { 2355 // It is an escaped line break. 2356 if parser.unread < 3 && !yaml_parser_update_buffer(parser, 3) { 2357 return false 2358 } 2359 skip(parser) 2360 skip_line(parser) 2361 leading_blanks = true 2362 break 2363 2364 } else if !single && parser.buffer[parser.buffer_pos] == '\\' { 2365 // It is an escape sequence. 2366 code_length := 0 2367 2368 // Check the escape character. 2369 switch parser.buffer[parser.buffer_pos+1] { 2370 case '0': 2371 s = append(s, 0) 2372 case 'a': 2373 s = append(s, '\x07') 2374 case 'b': 2375 s = append(s, '\x08') 2376 case 't', '\t': 2377 s = append(s, '\x09') 2378 case 'n': 2379 s = append(s, '\x0A') 2380 case 'v': 2381 s = append(s, '\x0B') 2382 case 'f': 2383 s = append(s, '\x0C') 2384 case 'r': 2385 s = append(s, '\x0D') 2386 case 'e': 2387 s = append(s, '\x1B') 2388 case ' ': 2389 s = append(s, '\x20') 2390 case '"': 2391 s = append(s, '"') 2392 case '\'': 2393 s = append(s, '\'') 2394 case '\\': 2395 s = append(s, '\\') 2396 case 'N': // NEL (#x85) 2397 s = append(s, '\xC2') 2398 s = append(s, '\x85') 2399 case '_': // #xA0 2400 s = append(s, '\xC2') 2401 s = append(s, '\xA0') 2402 case 'L': // LS (#x2028) 2403 s = append(s, '\xE2') 2404 s = append(s, '\x80') 2405 s = append(s, '\xA8') 2406 case 'P': // PS (#x2029) 2407 s = append(s, '\xE2') 2408 s = append(s, '\x80') 2409 s = append(s, '\xA9') 2410 case 'x': 2411 code_length = 2 2412 case 'u': 2413 code_length = 4 2414 case 'U': 2415 code_length = 8 2416 default: 2417 yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar", 2418 start_mark, "found unknown escape character") 2419 return false 2420 } 2421 2422 skip(parser) 2423 skip(parser) 2424 2425 // Consume an arbitrary escape code. 2426 if code_length > 0 { 2427 var value int 2428 2429 // Scan the character value. 2430 if parser.unread < code_length && !yaml_parser_update_buffer(parser, code_length) { 2431 return false 2432 } 2433 for k := 0; k < code_length; k++ { 2434 if !is_hex(parser.buffer, parser.buffer_pos+k) { 2435 yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar", 2436 start_mark, "did not find expected hexdecimal number") 2437 return false 2438 } 2439 value = (value << 4) + as_hex(parser.buffer, parser.buffer_pos+k) 2440 } 2441 2442 // Check the value and write the character. 2443 if (value >= 0xD800 && value <= 0xDFFF) || value > 0x10FFFF { 2444 yaml_parser_set_scanner_error(parser, "while parsing a quoted scalar", 2445 start_mark, "found invalid Unicode character escape code") 2446 return false 2447 } 2448 if value <= 0x7F { 2449 s = append(s, byte(value)) 2450 } else if value <= 0x7FF { 2451 s = append(s, byte(0xC0+(value>>6))) 2452 s = append(s, byte(0x80+(value&0x3F))) 2453 } else if value <= 0xFFFF { 2454 s = append(s, byte(0xE0+(value>>12))) 2455 s = append(s, byte(0x80+((value>>6)&0x3F))) 2456 s = append(s, byte(0x80+(value&0x3F))) 2457 } else { 2458 s = append(s, byte(0xF0+(value>>18))) 2459 s = append(s, byte(0x80+((value>>12)&0x3F))) 2460 s = append(s, byte(0x80+((value>>6)&0x3F))) 2461 s = append(s, byte(0x80+(value&0x3F))) 2462 } 2463 2464 // Advance the pointer. 2465 for k := 0; k < code_length; k++ { 2466 skip(parser) 2467 } 2468 } 2469 } else { 2470 // It is a non-escaped non-blank character. 2471 s = read(parser, s) 2472 } 2473 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2474 return false 2475 } 2476 } 2477 2478 // Check if we are at the end of the scalar. 2479 if single { 2480 if parser.buffer[parser.buffer_pos] == '\'' { 2481 break 2482 } 2483 } else { 2484 if parser.buffer[parser.buffer_pos] == '"' { 2485 break 2486 } 2487 } 2488 2489 // Consume blank characters. 2490 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2491 return false 2492 } 2493 2494 for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) { 2495 if is_blank(parser.buffer, parser.buffer_pos) { 2496 // Consume a space or a tab character. 2497 if !leading_blanks { 2498 whitespaces = read(parser, whitespaces) 2499 } else { 2500 skip(parser) 2501 } 2502 } else { 2503 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2504 return false 2505 } 2506 2507 // Check if it is a first line break. 2508 if !leading_blanks { 2509 whitespaces = whitespaces[:0] 2510 leading_break = read_line(parser, leading_break) 2511 leading_blanks = true 2512 } else { 2513 trailing_breaks = read_line(parser, trailing_breaks) 2514 } 2515 } 2516 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2517 return false 2518 } 2519 } 2520 2521 // Join the whitespaces or fold line breaks. 2522 if leading_blanks { 2523 // Do we need to fold line breaks? 2524 if len(leading_break) > 0 && leading_break[0] == '\n' { 2525 if len(trailing_breaks) == 0 { 2526 s = append(s, ' ') 2527 } else { 2528 s = append(s, trailing_breaks...) 2529 } 2530 } else { 2531 s = append(s, leading_break...) 2532 s = append(s, trailing_breaks...) 2533 } 2534 trailing_breaks = trailing_breaks[:0] 2535 leading_break = leading_break[:0] 2536 } else { 2537 s = append(s, whitespaces...) 2538 whitespaces = whitespaces[:0] 2539 } 2540 } 2541 2542 // Eat the right quote. 2543 skip(parser) 2544 end_mark := parser.mark 2545 2546 // Create a token. 2547 *token = yaml_token_t{ 2548 typ: yaml_SCALAR_TOKEN, 2549 start_mark: start_mark, 2550 end_mark: end_mark, 2551 value: s, 2552 style: yaml_SINGLE_QUOTED_SCALAR_STYLE, 2553 } 2554 if !single { 2555 token.style = yaml_DOUBLE_QUOTED_SCALAR_STYLE 2556 } 2557 return true 2558 } 2559 2560 // Scan a plain scalar. 2561 func yaml_parser_scan_plain_scalar(parser *yaml_parser_t, token *yaml_token_t) bool { 2562 2563 var s, leading_break, trailing_breaks, whitespaces []byte 2564 var leading_blanks bool 2565 var indent = parser.indent + 1 2566 2567 start_mark := parser.mark 2568 end_mark := parser.mark 2569 2570 // Consume the content of the plain scalar. 2571 for { 2572 // Check for a document indicator. 2573 if parser.unread < 4 && !yaml_parser_update_buffer(parser, 4) { 2574 return false 2575 } 2576 if parser.mark.column == 0 && 2577 ((parser.buffer[parser.buffer_pos+0] == '-' && 2578 parser.buffer[parser.buffer_pos+1] == '-' && 2579 parser.buffer[parser.buffer_pos+2] == '-') || 2580 (parser.buffer[parser.buffer_pos+0] == '.' && 2581 parser.buffer[parser.buffer_pos+1] == '.' && 2582 parser.buffer[parser.buffer_pos+2] == '.')) && 2583 is_blankz(parser.buffer, parser.buffer_pos+3) { 2584 break 2585 } 2586 2587 // Check for a comment. 2588 if parser.buffer[parser.buffer_pos] == '#' { 2589 break 2590 } 2591 2592 // Consume non-blank characters. 2593 for !is_blankz(parser.buffer, parser.buffer_pos) { 2594 2595 // Check for 'x:x' in the flow context. TODO: Fix the test "spec-08-13". 2596 if parser.flow_level > 0 && 2597 parser.buffer[parser.buffer_pos] == ':' && 2598 !is_blankz(parser.buffer, parser.buffer_pos+1) { 2599 yaml_parser_set_scanner_error(parser, "while scanning a plain scalar", 2600 start_mark, "found unexpected ':'") 2601 return false 2602 } 2603 2604 // Check for indicators that may end a plain scalar. 2605 if (parser.buffer[parser.buffer_pos] == ':' && is_blankz(parser.buffer, parser.buffer_pos+1)) || 2606 (parser.flow_level > 0 && 2607 (parser.buffer[parser.buffer_pos] == ',' || parser.buffer[parser.buffer_pos] == ':' || 2608 parser.buffer[parser.buffer_pos] == '?' || parser.buffer[parser.buffer_pos] == '[' || 2609 parser.buffer[parser.buffer_pos] == ']' || parser.buffer[parser.buffer_pos] == '{' || 2610 parser.buffer[parser.buffer_pos] == '}')) { 2611 break 2612 } 2613 2614 // Check if we need to join whitespaces and breaks. 2615 if leading_blanks || len(whitespaces) > 0 { 2616 if leading_blanks { 2617 // Do we need to fold line breaks? 2618 if leading_break[0] == '\n' { 2619 if len(trailing_breaks) == 0 { 2620 s = append(s, ' ') 2621 } else { 2622 s = append(s, trailing_breaks...) 2623 } 2624 } else { 2625 s = append(s, leading_break...) 2626 s = append(s, trailing_breaks...) 2627 } 2628 trailing_breaks = trailing_breaks[:0] 2629 leading_break = leading_break[:0] 2630 leading_blanks = false 2631 } else { 2632 s = append(s, whitespaces...) 2633 whitespaces = whitespaces[:0] 2634 } 2635 } 2636 2637 // Copy the character. 2638 s = read(parser, s) 2639 2640 end_mark = parser.mark 2641 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2642 return false 2643 } 2644 } 2645 2646 // Is it the end? 2647 if !(is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos)) { 2648 break 2649 } 2650 2651 // Consume blank characters. 2652 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2653 return false 2654 } 2655 2656 for is_blank(parser.buffer, parser.buffer_pos) || is_break(parser.buffer, parser.buffer_pos) { 2657 if is_blank(parser.buffer, parser.buffer_pos) { 2658 2659 // Check for tab character that abuse intendation. 2660 if leading_blanks && parser.mark.column < indent && is_tab(parser.buffer, parser.buffer_pos) { 2661 yaml_parser_set_scanner_error(parser, "while scanning a plain scalar", 2662 start_mark, "found a tab character that violate intendation") 2663 return false 2664 } 2665 2666 // Consume a space or a tab character. 2667 if !leading_blanks { 2668 whitespaces = read(parser, whitespaces) 2669 } else { 2670 skip(parser) 2671 } 2672 } else { 2673 if parser.unread < 2 && !yaml_parser_update_buffer(parser, 2) { 2674 return false 2675 } 2676 2677 // Check if it is a first line break. 2678 if !leading_blanks { 2679 whitespaces = whitespaces[:0] 2680 leading_break = read_line(parser, leading_break) 2681 leading_blanks = true 2682 } else { 2683 trailing_breaks = read_line(parser, trailing_breaks) 2684 } 2685 } 2686 if parser.unread < 1 && !yaml_parser_update_buffer(parser, 1) { 2687 return false 2688 } 2689 } 2690 2691 // Check intendation level. 2692 if parser.flow_level == 0 && parser.mark.column < indent { 2693 break 2694 } 2695 } 2696 2697 // Create a token. 2698 *token = yaml_token_t{ 2699 typ: yaml_SCALAR_TOKEN, 2700 start_mark: start_mark, 2701 end_mark: end_mark, 2702 value: s, 2703 style: yaml_PLAIN_SCALAR_STYLE, 2704 } 2705 2706 // Note that we change the 'simple_key_allowed' flag. 2707 if leading_blanks { 2708 parser.simple_key_allowed = true 2709 } 2710 return true 2711 }