xref: /openbsd/gnu/llvm/llvm/lib/Support/YAMLParser.cpp (revision d415bd75)
1 //===- YAMLParser.cpp - Simple YAML parser --------------------------------===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 //  This file implements a YAML parser.
10 //
11 //===----------------------------------------------------------------------===//
12 
13 #include "llvm/Support/YAMLParser.h"
14 #include "llvm/ADT/AllocatorList.h"
15 #include "llvm/ADT/ArrayRef.h"
16 #include "llvm/ADT/STLExtras.h"
17 #include "llvm/ADT/SmallString.h"
18 #include "llvm/ADT/SmallVector.h"
19 #include "llvm/ADT/StringExtras.h"
20 #include "llvm/ADT/StringRef.h"
21 #include "llvm/ADT/Twine.h"
22 #include "llvm/Support/Compiler.h"
23 #include "llvm/Support/ErrorHandling.h"
24 #include "llvm/Support/MemoryBuffer.h"
25 #include "llvm/Support/SMLoc.h"
26 #include "llvm/Support/SourceMgr.h"
27 #include "llvm/Support/Unicode.h"
28 #include "llvm/Support/raw_ostream.h"
29 #include <cassert>
30 #include <cstddef>
31 #include <cstdint>
32 #include <map>
33 #include <memory>
34 #include <string>
35 #include <system_error>
36 #include <utility>
37 
38 using namespace llvm;
39 using namespace yaml;
40 
41 enum UnicodeEncodingForm {
42   UEF_UTF32_LE, ///< UTF-32 Little Endian
43   UEF_UTF32_BE, ///< UTF-32 Big Endian
44   UEF_UTF16_LE, ///< UTF-16 Little Endian
45   UEF_UTF16_BE, ///< UTF-16 Big Endian
46   UEF_UTF8,     ///< UTF-8 or ascii.
47   UEF_Unknown   ///< Not a valid Unicode encoding.
48 };
49 
50 /// EncodingInfo - Holds the encoding type and length of the byte order mark if
51 ///                it exists. Length is in {0, 2, 3, 4}.
52 using EncodingInfo = std::pair<UnicodeEncodingForm, unsigned>;
53 
54 /// getUnicodeEncoding - Reads up to the first 4 bytes to determine the Unicode
55 ///                      encoding form of \a Input.
56 ///
57 /// @param Input A string of length 0 or more.
58 /// @returns An EncodingInfo indicating the Unicode encoding form of the input
59 ///          and how long the byte order mark is if one exists.
getUnicodeEncoding(StringRef Input)60 static EncodingInfo getUnicodeEncoding(StringRef Input) {
61   if (Input.empty())
62     return std::make_pair(UEF_Unknown, 0);
63 
64   switch (uint8_t(Input[0])) {
65   case 0x00:
66     if (Input.size() >= 4) {
67       if (  Input[1] == 0
68          && uint8_t(Input[2]) == 0xFE
69          && uint8_t(Input[3]) == 0xFF)
70         return std::make_pair(UEF_UTF32_BE, 4);
71       if (Input[1] == 0 && Input[2] == 0 && Input[3] != 0)
72         return std::make_pair(UEF_UTF32_BE, 0);
73     }
74 
75     if (Input.size() >= 2 && Input[1] != 0)
76       return std::make_pair(UEF_UTF16_BE, 0);
77     return std::make_pair(UEF_Unknown, 0);
78   case 0xFF:
79     if (  Input.size() >= 4
80        && uint8_t(Input[1]) == 0xFE
81        && Input[2] == 0
82        && Input[3] == 0)
83       return std::make_pair(UEF_UTF32_LE, 4);
84 
85     if (Input.size() >= 2 && uint8_t(Input[1]) == 0xFE)
86       return std::make_pair(UEF_UTF16_LE, 2);
87     return std::make_pair(UEF_Unknown, 0);
88   case 0xFE:
89     if (Input.size() >= 2 && uint8_t(Input[1]) == 0xFF)
90       return std::make_pair(UEF_UTF16_BE, 2);
91     return std::make_pair(UEF_Unknown, 0);
92   case 0xEF:
93     if (  Input.size() >= 3
94        && uint8_t(Input[1]) == 0xBB
95        && uint8_t(Input[2]) == 0xBF)
96       return std::make_pair(UEF_UTF8, 3);
97     return std::make_pair(UEF_Unknown, 0);
98   }
99 
100   // It could still be utf-32 or utf-16.
101   if (Input.size() >= 4 && Input[1] == 0 && Input[2] == 0 && Input[3] == 0)
102     return std::make_pair(UEF_UTF32_LE, 0);
103 
104   if (Input.size() >= 2 && Input[1] == 0)
105     return std::make_pair(UEF_UTF16_LE, 0);
106 
107   return std::make_pair(UEF_UTF8, 0);
108 }
109 
110 /// Pin the vtables to this file.
anchor()111 void Node::anchor() {}
anchor()112 void NullNode::anchor() {}
anchor()113 void ScalarNode::anchor() {}
anchor()114 void BlockScalarNode::anchor() {}
anchor()115 void KeyValueNode::anchor() {}
anchor()116 void MappingNode::anchor() {}
anchor()117 void SequenceNode::anchor() {}
anchor()118 void AliasNode::anchor() {}
119 
120 namespace llvm {
121 namespace yaml {
122 
123 /// Token - A single YAML token.
124 struct Token {
125   enum TokenKind {
126     TK_Error, // Uninitialized token.
127     TK_StreamStart,
128     TK_StreamEnd,
129     TK_VersionDirective,
130     TK_TagDirective,
131     TK_DocumentStart,
132     TK_DocumentEnd,
133     TK_BlockEntry,
134     TK_BlockEnd,
135     TK_BlockSequenceStart,
136     TK_BlockMappingStart,
137     TK_FlowEntry,
138     TK_FlowSequenceStart,
139     TK_FlowSequenceEnd,
140     TK_FlowMappingStart,
141     TK_FlowMappingEnd,
142     TK_Key,
143     TK_Value,
144     TK_Scalar,
145     TK_BlockScalar,
146     TK_Alias,
147     TK_Anchor,
148     TK_Tag
149   } Kind = TK_Error;
150 
151   /// A string of length 0 or more whose begin() points to the logical location
152   /// of the token in the input.
153   StringRef Range;
154 
155   /// The value of a block scalar node.
156   std::string Value;
157 
158   Token() = default;
159 };
160 
161 } // end namespace yaml
162 } // end namespace llvm
163 
164 using TokenQueueT = BumpPtrList<Token>;
165 
166 namespace {
167 
168 /// This struct is used to track simple keys.
169 ///
170 /// Simple keys are handled by creating an entry in SimpleKeys for each Token
171 /// which could legally be the start of a simple key. When peekNext is called,
172 /// if the Token To be returned is referenced by a SimpleKey, we continue
173 /// tokenizing until that potential simple key has either been found to not be
174 /// a simple key (we moved on to the next line or went further than 1024 chars).
175 /// Or when we run into a Value, and then insert a Key token (and possibly
176 /// others) before the SimpleKey's Tok.
177 struct SimpleKey {
178   TokenQueueT::iterator Tok;
179   unsigned Column = 0;
180   unsigned Line = 0;
181   unsigned FlowLevel = 0;
182   bool IsRequired = false;
183 
operator ==__anon259698e90111::SimpleKey184   bool operator ==(const SimpleKey &Other) {
185     return Tok == Other.Tok;
186   }
187 };
188 
189 } // end anonymous namespace
190 
191 /// The Unicode scalar value of a UTF-8 minimal well-formed code unit
192 ///        subsequence and the subsequence's length in code units (uint8_t).
193 ///        A length of 0 represents an error.
194 using UTF8Decoded = std::pair<uint32_t, unsigned>;
195 
decodeUTF8(StringRef Range)196 static UTF8Decoded decodeUTF8(StringRef Range) {
197   StringRef::iterator Position= Range.begin();
198   StringRef::iterator End = Range.end();
199   // 1 byte: [0x00, 0x7f]
200   // Bit pattern: 0xxxxxxx
201   if (Position < End && (*Position & 0x80) == 0) {
202     return std::make_pair(*Position, 1);
203   }
204   // 2 bytes: [0x80, 0x7ff]
205   // Bit pattern: 110xxxxx 10xxxxxx
206   if (Position + 1 < End && ((*Position & 0xE0) == 0xC0) &&
207       ((*(Position + 1) & 0xC0) == 0x80)) {
208     uint32_t codepoint = ((*Position & 0x1F) << 6) |
209                           (*(Position + 1) & 0x3F);
210     if (codepoint >= 0x80)
211       return std::make_pair(codepoint, 2);
212   }
213   // 3 bytes: [0x8000, 0xffff]
214   // Bit pattern: 1110xxxx 10xxxxxx 10xxxxxx
215   if (Position + 2 < End && ((*Position & 0xF0) == 0xE0) &&
216       ((*(Position + 1) & 0xC0) == 0x80) &&
217       ((*(Position + 2) & 0xC0) == 0x80)) {
218     uint32_t codepoint = ((*Position & 0x0F) << 12) |
219                          ((*(Position + 1) & 0x3F) << 6) |
220                           (*(Position + 2) & 0x3F);
221     // Codepoints between 0xD800 and 0xDFFF are invalid, as
222     // they are high / low surrogate halves used by UTF-16.
223     if (codepoint >= 0x800 &&
224         (codepoint < 0xD800 || codepoint > 0xDFFF))
225       return std::make_pair(codepoint, 3);
226   }
227   // 4 bytes: [0x10000, 0x10FFFF]
228   // Bit pattern: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
229   if (Position + 3 < End && ((*Position & 0xF8) == 0xF0) &&
230       ((*(Position + 1) & 0xC0) == 0x80) &&
231       ((*(Position + 2) & 0xC0) == 0x80) &&
232       ((*(Position + 3) & 0xC0) == 0x80)) {
233     uint32_t codepoint = ((*Position & 0x07) << 18) |
234                          ((*(Position + 1) & 0x3F) << 12) |
235                          ((*(Position + 2) & 0x3F) << 6) |
236                           (*(Position + 3) & 0x3F);
237     if (codepoint >= 0x10000 && codepoint <= 0x10FFFF)
238       return std::make_pair(codepoint, 4);
239   }
240   return std::make_pair(0, 0);
241 }
242 
243 namespace llvm {
244 namespace yaml {
245 
246 /// Scans YAML tokens from a MemoryBuffer.
247 class Scanner {
248 public:
249   Scanner(StringRef Input, SourceMgr &SM, bool ShowColors = true,
250           std::error_code *EC = nullptr);
251   Scanner(MemoryBufferRef Buffer, SourceMgr &SM_, bool ShowColors = true,
252           std::error_code *EC = nullptr);
253 
254   /// Parse the next token and return it without popping it.
255   Token &peekNext();
256 
257   /// Parse the next token and pop it from the queue.
258   Token getNext();
259 
printError(SMLoc Loc,SourceMgr::DiagKind Kind,const Twine & Message,ArrayRef<SMRange> Ranges=std::nullopt)260   void printError(SMLoc Loc, SourceMgr::DiagKind Kind, const Twine &Message,
261                   ArrayRef<SMRange> Ranges = std::nullopt) {
262     SM.PrintMessage(Loc, Kind, Message, Ranges, /* FixIts= */ std::nullopt,
263                     ShowColors);
264   }
265 
setError(const Twine & Message,StringRef::iterator Position)266   void setError(const Twine &Message, StringRef::iterator Position) {
267     if (Position >= End)
268       Position = End - 1;
269 
270     // propagate the error if possible
271     if (EC)
272       *EC = make_error_code(std::errc::invalid_argument);
273 
274     // Don't print out more errors after the first one we encounter. The rest
275     // are just the result of the first, and have no meaning.
276     if (!Failed)
277       printError(SMLoc::getFromPointer(Position), SourceMgr::DK_Error, Message);
278     Failed = true;
279   }
280 
281   /// Returns true if an error occurred while parsing.
failed()282   bool failed() {
283     return Failed;
284   }
285 
286 private:
287   void init(MemoryBufferRef Buffer);
288 
currentInput()289   StringRef currentInput() {
290     return StringRef(Current, End - Current);
291   }
292 
293   /// Decode a UTF-8 minimal well-formed code unit subsequence starting
294   ///        at \a Position.
295   ///
296   /// If the UTF-8 code units starting at Position do not form a well-formed
297   /// code unit subsequence, then the Unicode scalar value is 0, and the length
298   /// is 0.
decodeUTF8(StringRef::iterator Position)299   UTF8Decoded decodeUTF8(StringRef::iterator Position) {
300     return ::decodeUTF8(StringRef(Position, End - Position));
301   }
302 
303   // The following functions are based on the gramar rules in the YAML spec. The
304   // style of the function names it meant to closely match how they are written
305   // in the spec. The number within the [] is the number of the grammar rule in
306   // the spec.
307   //
308   // See 4.2 [Production Naming Conventions] for the meaning of the prefixes.
309   //
310   // c-
311   //   A production starting and ending with a special character.
312   // b-
313   //   A production matching a single line break.
314   // nb-
315   //   A production starting and ending with a non-break character.
316   // s-
317   //   A production starting and ending with a white space character.
318   // ns-
319   //   A production starting and ending with a non-space character.
320   // l-
321   //   A production matching complete line(s).
322 
323   /// Skip a single nb-char[27] starting at Position.
324   ///
325   /// A nb-char is 0x9 | [0x20-0x7E] | 0x85 | [0xA0-0xD7FF] | [0xE000-0xFEFE]
326   ///                  | [0xFF00-0xFFFD] | [0x10000-0x10FFFF]
327   ///
328   /// @returns The code unit after the nb-char, or Position if it's not an
329   ///          nb-char.
330   StringRef::iterator skip_nb_char(StringRef::iterator Position);
331 
332   /// Skip a single b-break[28] starting at Position.
333   ///
334   /// A b-break is 0xD 0xA | 0xD | 0xA
335   ///
336   /// @returns The code unit after the b-break, or Position if it's not a
337   ///          b-break.
338   StringRef::iterator skip_b_break(StringRef::iterator Position);
339 
340   /// Skip a single s-space[31] starting at Position.
341   ///
342   /// An s-space is 0x20
343   ///
344   /// @returns The code unit after the s-space, or Position if it's not a
345   ///          s-space.
346   StringRef::iterator skip_s_space(StringRef::iterator Position);
347 
348   /// Skip a single s-white[33] starting at Position.
349   ///
350   /// A s-white is 0x20 | 0x9
351   ///
352   /// @returns The code unit after the s-white, or Position if it's not a
353   ///          s-white.
354   StringRef::iterator skip_s_white(StringRef::iterator Position);
355 
356   /// Skip a single ns-char[34] starting at Position.
357   ///
358   /// A ns-char is nb-char - s-white
359   ///
360   /// @returns The code unit after the ns-char, or Position if it's not a
361   ///          ns-char.
362   StringRef::iterator skip_ns_char(StringRef::iterator Position);
363 
364   using SkipWhileFunc = StringRef::iterator (Scanner::*)(StringRef::iterator);
365 
366   /// Skip minimal well-formed code unit subsequences until Func
367   ///        returns its input.
368   ///
369   /// @returns The code unit after the last minimal well-formed code unit
370   ///          subsequence that Func accepted.
371   StringRef::iterator skip_while( SkipWhileFunc Func
372                                 , StringRef::iterator Position);
373 
374   /// Skip minimal well-formed code unit subsequences until Func returns its
375   /// input.
376   void advanceWhile(SkipWhileFunc Func);
377 
378   /// Scan ns-uri-char[39]s starting at Cur.
379   ///
380   /// This updates Cur and Column while scanning.
381   void scan_ns_uri_char();
382 
383   /// Consume a minimal well-formed code unit subsequence starting at
384   ///        \a Cur. Return false if it is not the same Unicode scalar value as
385   ///        \a Expected. This updates \a Column.
386   bool consume(uint32_t Expected);
387 
388   /// Skip \a Distance UTF-8 code units. Updates \a Cur and \a Column.
389   void skip(uint32_t Distance);
390 
391   /// Return true if the minimal well-formed code unit subsequence at
392   ///        Pos is whitespace or a new line
393   bool isBlankOrBreak(StringRef::iterator Position);
394 
395   /// Return true if the line is a line break, false otherwise.
396   bool isLineEmpty(StringRef Line);
397 
398   /// Consume a single b-break[28] if it's present at the current position.
399   ///
400   /// Return false if the code unit at the current position isn't a line break.
401   bool consumeLineBreakIfPresent();
402 
403   /// If IsSimpleKeyAllowed, create and push_back a new SimpleKey.
404   void saveSimpleKeyCandidate( TokenQueueT::iterator Tok
405                              , unsigned AtColumn
406                              , bool IsRequired);
407 
408   /// Remove simple keys that can no longer be valid simple keys.
409   ///
410   /// Invalid simple keys are not on the current line or are further than 1024
411   /// columns back.
412   void removeStaleSimpleKeyCandidates();
413 
414   /// Remove all simple keys on FlowLevel \a Level.
415   void removeSimpleKeyCandidatesOnFlowLevel(unsigned Level);
416 
417   /// Unroll indentation in \a Indents back to \a Col. Creates BlockEnd
418   ///        tokens if needed.
419   bool unrollIndent(int ToColumn);
420 
421   /// Increase indent to \a Col. Creates \a Kind token at \a InsertPoint
422   ///        if needed.
423   bool rollIndent( int ToColumn
424                  , Token::TokenKind Kind
425                  , TokenQueueT::iterator InsertPoint);
426 
427   /// Skip a single-line comment when the comment starts at the current
428   /// position of the scanner.
429   void skipComment();
430 
431   /// Skip whitespace and comments until the start of the next token.
432   void scanToNextToken();
433 
434   /// Must be the first token generated.
435   bool scanStreamStart();
436 
437   /// Generate tokens needed to close out the stream.
438   bool scanStreamEnd();
439 
440   /// Scan a %BLAH directive.
441   bool scanDirective();
442 
443   /// Scan a ... or ---.
444   bool scanDocumentIndicator(bool IsStart);
445 
446   /// Scan a [ or { and generate the proper flow collection start token.
447   bool scanFlowCollectionStart(bool IsSequence);
448 
449   /// Scan a ] or } and generate the proper flow collection end token.
450   bool scanFlowCollectionEnd(bool IsSequence);
451 
452   /// Scan the , that separates entries in a flow collection.
453   bool scanFlowEntry();
454 
455   /// Scan the - that starts block sequence entries.
456   bool scanBlockEntry();
457 
458   /// Scan an explicit ? indicating a key.
459   bool scanKey();
460 
461   /// Scan an explicit : indicating a value.
462   bool scanValue();
463 
464   /// Scan a quoted scalar.
465   bool scanFlowScalar(bool IsDoubleQuoted);
466 
467   /// Scan an unquoted scalar.
468   bool scanPlainScalar();
469 
470   /// Scan an Alias or Anchor starting with * or &.
471   bool scanAliasOrAnchor(bool IsAlias);
472 
473   /// Scan a block scalar starting with | or >.
474   bool scanBlockScalar(bool IsLiteral);
475 
476   /// Scan a block scalar style indicator and header.
477   ///
478   /// Note: This is distinct from scanBlockScalarHeader to mirror the fact that
479   /// YAML does not consider the style indicator to be a part of the header.
480   ///
481   /// Return false if an error occurred.
482   bool scanBlockScalarIndicators(char &StyleIndicator, char &ChompingIndicator,
483                                  unsigned &IndentIndicator, bool &IsDone);
484 
485   /// Scan a style indicator in a block scalar header.
486   char scanBlockStyleIndicator();
487 
488   /// Scan a chomping indicator in a block scalar header.
489   char scanBlockChompingIndicator();
490 
491   /// Scan an indentation indicator in a block scalar header.
492   unsigned scanBlockIndentationIndicator();
493 
494   /// Scan a block scalar header.
495   ///
496   /// Return false if an error occurred.
497   bool scanBlockScalarHeader(char &ChompingIndicator, unsigned &IndentIndicator,
498                              bool &IsDone);
499 
500   /// Look for the indentation level of a block scalar.
501   ///
502   /// Return false if an error occurred.
503   bool findBlockScalarIndent(unsigned &BlockIndent, unsigned BlockExitIndent,
504                              unsigned &LineBreaks, bool &IsDone);
505 
506   /// Scan the indentation of a text line in a block scalar.
507   ///
508   /// Return false if an error occurred.
509   bool scanBlockScalarIndent(unsigned BlockIndent, unsigned BlockExitIndent,
510                              bool &IsDone);
511 
512   /// Scan a tag of the form !stuff.
513   bool scanTag();
514 
515   /// Dispatch to the next scanning function based on \a *Cur.
516   bool fetchMoreTokens();
517 
518   /// The SourceMgr used for diagnostics and buffer management.
519   SourceMgr &SM;
520 
521   /// The original input.
522   MemoryBufferRef InputBuffer;
523 
524   /// The current position of the scanner.
525   StringRef::iterator Current;
526 
527   /// The end of the input (one past the last character).
528   StringRef::iterator End;
529 
530   /// Current YAML indentation level in spaces.
531   int Indent;
532 
533   /// Current column number in Unicode code points.
534   unsigned Column;
535 
536   /// Current line number.
537   unsigned Line;
538 
539   /// How deep we are in flow style containers. 0 Means at block level.
540   unsigned FlowLevel;
541 
542   /// Are we at the start of the stream?
543   bool IsStartOfStream;
544 
545   /// Can the next token be the start of a simple key?
546   bool IsSimpleKeyAllowed;
547 
548   /// True if an error has occurred.
549   bool Failed;
550 
551   /// Should colors be used when printing out the diagnostic messages?
552   bool ShowColors;
553 
554   /// Queue of tokens. This is required to queue up tokens while looking
555   ///        for the end of a simple key. And for cases where a single character
556   ///        can produce multiple tokens (e.g. BlockEnd).
557   TokenQueueT TokenQueue;
558 
559   /// Indentation levels.
560   SmallVector<int, 4> Indents;
561 
562   /// Potential simple keys.
563   SmallVector<SimpleKey, 4> SimpleKeys;
564 
565   std::error_code *EC;
566 };
567 
568 } // end namespace yaml
569 } // end namespace llvm
570 
571 /// encodeUTF8 - Encode \a UnicodeScalarValue in UTF-8 and append it to result.
encodeUTF8(uint32_t UnicodeScalarValue,SmallVectorImpl<char> & Result)572 static void encodeUTF8( uint32_t UnicodeScalarValue
573                       , SmallVectorImpl<char> &Result) {
574   if (UnicodeScalarValue <= 0x7F) {
575     Result.push_back(UnicodeScalarValue & 0x7F);
576   } else if (UnicodeScalarValue <= 0x7FF) {
577     uint8_t FirstByte = 0xC0 | ((UnicodeScalarValue & 0x7C0) >> 6);
578     uint8_t SecondByte = 0x80 | (UnicodeScalarValue & 0x3F);
579     Result.push_back(FirstByte);
580     Result.push_back(SecondByte);
581   } else if (UnicodeScalarValue <= 0xFFFF) {
582     uint8_t FirstByte = 0xE0 | ((UnicodeScalarValue & 0xF000) >> 12);
583     uint8_t SecondByte = 0x80 | ((UnicodeScalarValue & 0xFC0) >> 6);
584     uint8_t ThirdByte = 0x80 | (UnicodeScalarValue & 0x3F);
585     Result.push_back(FirstByte);
586     Result.push_back(SecondByte);
587     Result.push_back(ThirdByte);
588   } else if (UnicodeScalarValue <= 0x10FFFF) {
589     uint8_t FirstByte = 0xF0 | ((UnicodeScalarValue & 0x1F0000) >> 18);
590     uint8_t SecondByte = 0x80 | ((UnicodeScalarValue & 0x3F000) >> 12);
591     uint8_t ThirdByte = 0x80 | ((UnicodeScalarValue & 0xFC0) >> 6);
592     uint8_t FourthByte = 0x80 | (UnicodeScalarValue & 0x3F);
593     Result.push_back(FirstByte);
594     Result.push_back(SecondByte);
595     Result.push_back(ThirdByte);
596     Result.push_back(FourthByte);
597   }
598 }
599 
dumpTokens(StringRef Input,raw_ostream & OS)600 bool yaml::dumpTokens(StringRef Input, raw_ostream &OS) {
601   SourceMgr SM;
602   Scanner scanner(Input, SM);
603   while (true) {
604     Token T = scanner.getNext();
605     switch (T.Kind) {
606     case Token::TK_StreamStart:
607       OS << "Stream-Start: ";
608       break;
609     case Token::TK_StreamEnd:
610       OS << "Stream-End: ";
611       break;
612     case Token::TK_VersionDirective:
613       OS << "Version-Directive: ";
614       break;
615     case Token::TK_TagDirective:
616       OS << "Tag-Directive: ";
617       break;
618     case Token::TK_DocumentStart:
619       OS << "Document-Start: ";
620       break;
621     case Token::TK_DocumentEnd:
622       OS << "Document-End: ";
623       break;
624     case Token::TK_BlockEntry:
625       OS << "Block-Entry: ";
626       break;
627     case Token::TK_BlockEnd:
628       OS << "Block-End: ";
629       break;
630     case Token::TK_BlockSequenceStart:
631       OS << "Block-Sequence-Start: ";
632       break;
633     case Token::TK_BlockMappingStart:
634       OS << "Block-Mapping-Start: ";
635       break;
636     case Token::TK_FlowEntry:
637       OS << "Flow-Entry: ";
638       break;
639     case Token::TK_FlowSequenceStart:
640       OS << "Flow-Sequence-Start: ";
641       break;
642     case Token::TK_FlowSequenceEnd:
643       OS << "Flow-Sequence-End: ";
644       break;
645     case Token::TK_FlowMappingStart:
646       OS << "Flow-Mapping-Start: ";
647       break;
648     case Token::TK_FlowMappingEnd:
649       OS << "Flow-Mapping-End: ";
650       break;
651     case Token::TK_Key:
652       OS << "Key: ";
653       break;
654     case Token::TK_Value:
655       OS << "Value: ";
656       break;
657     case Token::TK_Scalar:
658       OS << "Scalar: ";
659       break;
660     case Token::TK_BlockScalar:
661       OS << "Block Scalar: ";
662       break;
663     case Token::TK_Alias:
664       OS << "Alias: ";
665       break;
666     case Token::TK_Anchor:
667       OS << "Anchor: ";
668       break;
669     case Token::TK_Tag:
670       OS << "Tag: ";
671       break;
672     case Token::TK_Error:
673       break;
674     }
675     OS << T.Range << "\n";
676     if (T.Kind == Token::TK_StreamEnd)
677       break;
678     else if (T.Kind == Token::TK_Error)
679       return false;
680   }
681   return true;
682 }
683 
scanTokens(StringRef Input)684 bool yaml::scanTokens(StringRef Input) {
685   SourceMgr SM;
686   Scanner scanner(Input, SM);
687   while (true) {
688     Token T = scanner.getNext();
689     if (T.Kind == Token::TK_StreamEnd)
690       break;
691     else if (T.Kind == Token::TK_Error)
692       return false;
693   }
694   return true;
695 }
696 
escape(StringRef Input,bool EscapePrintable)697 std::string yaml::escape(StringRef Input, bool EscapePrintable) {
698   std::string EscapedInput;
699   for (StringRef::iterator i = Input.begin(), e = Input.end(); i != e; ++i) {
700     if (*i == '\\')
701       EscapedInput += "\\\\";
702     else if (*i == '"')
703       EscapedInput += "\\\"";
704     else if (*i == 0)
705       EscapedInput += "\\0";
706     else if (*i == 0x07)
707       EscapedInput += "\\a";
708     else if (*i == 0x08)
709       EscapedInput += "\\b";
710     else if (*i == 0x09)
711       EscapedInput += "\\t";
712     else if (*i == 0x0A)
713       EscapedInput += "\\n";
714     else if (*i == 0x0B)
715       EscapedInput += "\\v";
716     else if (*i == 0x0C)
717       EscapedInput += "\\f";
718     else if (*i == 0x0D)
719       EscapedInput += "\\r";
720     else if (*i == 0x1B)
721       EscapedInput += "\\e";
722     else if ((unsigned char)*i < 0x20) { // Control characters not handled above.
723       std::string HexStr = utohexstr(*i);
724       EscapedInput += "\\x" + std::string(2 - HexStr.size(), '0') + HexStr;
725     } else if (*i & 0x80) { // UTF-8 multiple code unit subsequence.
726       UTF8Decoded UnicodeScalarValue
727         = decodeUTF8(StringRef(i, Input.end() - i));
728       if (UnicodeScalarValue.second == 0) {
729         // Found invalid char.
730         SmallString<4> Val;
731         encodeUTF8(0xFFFD, Val);
732         llvm::append_range(EscapedInput, Val);
733         // FIXME: Error reporting.
734         return EscapedInput;
735       }
736       if (UnicodeScalarValue.first == 0x85)
737         EscapedInput += "\\N";
738       else if (UnicodeScalarValue.first == 0xA0)
739         EscapedInput += "\\_";
740       else if (UnicodeScalarValue.first == 0x2028)
741         EscapedInput += "\\L";
742       else if (UnicodeScalarValue.first == 0x2029)
743         EscapedInput += "\\P";
744       else if (!EscapePrintable &&
745                sys::unicode::isPrintable(UnicodeScalarValue.first))
746         EscapedInput += StringRef(i, UnicodeScalarValue.second);
747       else {
748         std::string HexStr = utohexstr(UnicodeScalarValue.first);
749         if (HexStr.size() <= 2)
750           EscapedInput += "\\x" + std::string(2 - HexStr.size(), '0') + HexStr;
751         else if (HexStr.size() <= 4)
752           EscapedInput += "\\u" + std::string(4 - HexStr.size(), '0') + HexStr;
753         else if (HexStr.size() <= 8)
754           EscapedInput += "\\U" + std::string(8 - HexStr.size(), '0') + HexStr;
755       }
756       i += UnicodeScalarValue.second - 1;
757     } else
758       EscapedInput.push_back(*i);
759   }
760   return EscapedInput;
761 }
762 
parseBool(StringRef S)763 std::optional<bool> yaml::parseBool(StringRef S) {
764   switch (S.size()) {
765   case 1:
766     switch (S.front()) {
767     case 'y':
768     case 'Y':
769       return true;
770     case 'n':
771     case 'N':
772       return false;
773     default:
774       return std::nullopt;
775     }
776   case 2:
777     switch (S.front()) {
778     case 'O':
779       if (S[1] == 'N') // ON
780         return true;
781       [[fallthrough]];
782     case 'o':
783       if (S[1] == 'n') //[Oo]n
784         return true;
785       return std::nullopt;
786     case 'N':
787       if (S[1] == 'O') // NO
788         return false;
789       [[fallthrough]];
790     case 'n':
791       if (S[1] == 'o') //[Nn]o
792         return false;
793       return std::nullopt;
794     default:
795       return std::nullopt;
796     }
797   case 3:
798     switch (S.front()) {
799     case 'O':
800       if (S.drop_front() == "FF") // OFF
801         return false;
802       [[fallthrough]];
803     case 'o':
804       if (S.drop_front() == "ff") //[Oo]ff
805         return false;
806       return std::nullopt;
807     case 'Y':
808       if (S.drop_front() == "ES") // YES
809         return true;
810       [[fallthrough]];
811     case 'y':
812       if (S.drop_front() == "es") //[Yy]es
813         return true;
814       return std::nullopt;
815     default:
816       return std::nullopt;
817     }
818   case 4:
819     switch (S.front()) {
820     case 'T':
821       if (S.drop_front() == "RUE") // TRUE
822         return true;
823       [[fallthrough]];
824     case 't':
825       if (S.drop_front() == "rue") //[Tt]rue
826         return true;
827       return std::nullopt;
828     default:
829       return std::nullopt;
830     }
831   case 5:
832     switch (S.front()) {
833     case 'F':
834       if (S.drop_front() == "ALSE") // FALSE
835         return false;
836       [[fallthrough]];
837     case 'f':
838       if (S.drop_front() == "alse") //[Ff]alse
839         return false;
840       return std::nullopt;
841     default:
842       return std::nullopt;
843     }
844   default:
845     return std::nullopt;
846   }
847 }
848 
Scanner(StringRef Input,SourceMgr & sm,bool ShowColors,std::error_code * EC)849 Scanner::Scanner(StringRef Input, SourceMgr &sm, bool ShowColors,
850                  std::error_code *EC)
851     : SM(sm), ShowColors(ShowColors), EC(EC) {
852   init(MemoryBufferRef(Input, "YAML"));
853 }
854 
Scanner(MemoryBufferRef Buffer,SourceMgr & SM_,bool ShowColors,std::error_code * EC)855 Scanner::Scanner(MemoryBufferRef Buffer, SourceMgr &SM_, bool ShowColors,
856                  std::error_code *EC)
857     : SM(SM_), ShowColors(ShowColors), EC(EC) {
858   init(Buffer);
859 }
860 
init(MemoryBufferRef Buffer)861 void Scanner::init(MemoryBufferRef Buffer) {
862   InputBuffer = Buffer;
863   Current = InputBuffer.getBufferStart();
864   End = InputBuffer.getBufferEnd();
865   Indent = -1;
866   Column = 0;
867   Line = 0;
868   FlowLevel = 0;
869   IsStartOfStream = true;
870   IsSimpleKeyAllowed = true;
871   Failed = false;
872   std::unique_ptr<MemoryBuffer> InputBufferOwner =
873       MemoryBuffer::getMemBuffer(Buffer, /*RequiresNullTerminator=*/false);
874   SM.AddNewSourceBuffer(std::move(InputBufferOwner), SMLoc());
875 }
876 
peekNext()877 Token &Scanner::peekNext() {
878   // If the current token is a possible simple key, keep parsing until we
879   // can confirm.
880   bool NeedMore = false;
881   while (true) {
882     if (TokenQueue.empty() || NeedMore) {
883       if (!fetchMoreTokens()) {
884         TokenQueue.clear();
885         SimpleKeys.clear();
886         TokenQueue.push_back(Token());
887         return TokenQueue.front();
888       }
889     }
890     assert(!TokenQueue.empty() &&
891             "fetchMoreTokens lied about getting tokens!");
892 
893     removeStaleSimpleKeyCandidates();
894     SimpleKey SK;
895     SK.Tok = TokenQueue.begin();
896     if (!is_contained(SimpleKeys, SK))
897       break;
898     else
899       NeedMore = true;
900   }
901   return TokenQueue.front();
902 }
903 
getNext()904 Token Scanner::getNext() {
905   Token Ret = peekNext();
906   // TokenQueue can be empty if there was an error getting the next token.
907   if (!TokenQueue.empty())
908     TokenQueue.pop_front();
909 
910   // There cannot be any referenced Token's if the TokenQueue is empty. So do a
911   // quick deallocation of them all.
912   if (TokenQueue.empty())
913     TokenQueue.resetAlloc();
914 
915   return Ret;
916 }
917 
skip_nb_char(StringRef::iterator Position)918 StringRef::iterator Scanner::skip_nb_char(StringRef::iterator Position) {
919   if (Position == End)
920     return Position;
921   // Check 7 bit c-printable - b-char.
922   if (   *Position == 0x09
923       || (*Position >= 0x20 && *Position <= 0x7E))
924     return Position + 1;
925 
926   // Check for valid UTF-8.
927   if (uint8_t(*Position) & 0x80) {
928     UTF8Decoded u8d = decodeUTF8(Position);
929     if (   u8d.second != 0
930         && u8d.first != 0xFEFF
931         && ( u8d.first == 0x85
932           || ( u8d.first >= 0xA0
933             && u8d.first <= 0xD7FF)
934           || ( u8d.first >= 0xE000
935             && u8d.first <= 0xFFFD)
936           || ( u8d.first >= 0x10000
937             && u8d.first <= 0x10FFFF)))
938       return Position + u8d.second;
939   }
940   return Position;
941 }
942 
skip_b_break(StringRef::iterator Position)943 StringRef::iterator Scanner::skip_b_break(StringRef::iterator Position) {
944   if (Position == End)
945     return Position;
946   if (*Position == 0x0D) {
947     if (Position + 1 != End && *(Position + 1) == 0x0A)
948       return Position + 2;
949     return Position + 1;
950   }
951 
952   if (*Position == 0x0A)
953     return Position + 1;
954   return Position;
955 }
956 
skip_s_space(StringRef::iterator Position)957 StringRef::iterator Scanner::skip_s_space(StringRef::iterator Position) {
958   if (Position == End)
959     return Position;
960   if (*Position == ' ')
961     return Position + 1;
962   return Position;
963 }
964 
skip_s_white(StringRef::iterator Position)965 StringRef::iterator Scanner::skip_s_white(StringRef::iterator Position) {
966   if (Position == End)
967     return Position;
968   if (*Position == ' ' || *Position == '\t')
969     return Position + 1;
970   return Position;
971 }
972 
skip_ns_char(StringRef::iterator Position)973 StringRef::iterator Scanner::skip_ns_char(StringRef::iterator Position) {
974   if (Position == End)
975     return Position;
976   if (*Position == ' ' || *Position == '\t')
977     return Position;
978   return skip_nb_char(Position);
979 }
980 
skip_while(SkipWhileFunc Func,StringRef::iterator Position)981 StringRef::iterator Scanner::skip_while( SkipWhileFunc Func
982                                        , StringRef::iterator Position) {
983   while (true) {
984     StringRef::iterator i = (this->*Func)(Position);
985     if (i == Position)
986       break;
987     Position = i;
988   }
989   return Position;
990 }
991 
advanceWhile(SkipWhileFunc Func)992 void Scanner::advanceWhile(SkipWhileFunc Func) {
993   auto Final = skip_while(Func, Current);
994   Column += Final - Current;
995   Current = Final;
996 }
997 
is_ns_hex_digit(const char C)998 static bool is_ns_hex_digit(const char C) { return isAlnum(C); }
999 
is_ns_word_char(const char C)1000 static bool is_ns_word_char(const char C) { return C == '-' || isAlpha(C); }
1001 
scan_ns_uri_char()1002 void Scanner::scan_ns_uri_char() {
1003   while (true) {
1004     if (Current == End)
1005       break;
1006     if ((   *Current == '%'
1007           && Current + 2 < End
1008           && is_ns_hex_digit(*(Current + 1))
1009           && is_ns_hex_digit(*(Current + 2)))
1010         || is_ns_word_char(*Current)
1011         || StringRef(Current, 1).find_first_of("#;/?:@&=+$,_.!~*'()[]")
1012           != StringRef::npos) {
1013       ++Current;
1014       ++Column;
1015     } else
1016       break;
1017   }
1018 }
1019 
consume(uint32_t Expected)1020 bool Scanner::consume(uint32_t Expected) {
1021   if (Expected >= 0x80) {
1022     setError("Cannot consume non-ascii characters", Current);
1023     return false;
1024   }
1025   if (Current == End)
1026     return false;
1027   if (uint8_t(*Current) >= 0x80) {
1028     setError("Cannot consume non-ascii characters", Current);
1029     return false;
1030   }
1031   if (uint8_t(*Current) == Expected) {
1032     ++Current;
1033     ++Column;
1034     return true;
1035   }
1036   return false;
1037 }
1038 
skip(uint32_t Distance)1039 void Scanner::skip(uint32_t Distance) {
1040   Current += Distance;
1041   Column += Distance;
1042   assert(Current <= End && "Skipped past the end");
1043 }
1044 
isBlankOrBreak(StringRef::iterator Position)1045 bool Scanner::isBlankOrBreak(StringRef::iterator Position) {
1046   if (Position == End)
1047     return false;
1048   return *Position == ' ' || *Position == '\t' || *Position == '\r' ||
1049          *Position == '\n';
1050 }
1051 
isLineEmpty(StringRef Line)1052 bool Scanner::isLineEmpty(StringRef Line) {
1053   for (const auto *Position = Line.begin(); Position != Line.end(); ++Position)
1054     if (!isBlankOrBreak(Position))
1055       return false;
1056   return true;
1057 }
1058 
consumeLineBreakIfPresent()1059 bool Scanner::consumeLineBreakIfPresent() {
1060   auto Next = skip_b_break(Current);
1061   if (Next == Current)
1062     return false;
1063   Column = 0;
1064   ++Line;
1065   Current = Next;
1066   return true;
1067 }
1068 
saveSimpleKeyCandidate(TokenQueueT::iterator Tok,unsigned AtColumn,bool IsRequired)1069 void Scanner::saveSimpleKeyCandidate( TokenQueueT::iterator Tok
1070                                     , unsigned AtColumn
1071                                     , bool IsRequired) {
1072   if (IsSimpleKeyAllowed) {
1073     SimpleKey SK;
1074     SK.Tok = Tok;
1075     SK.Line = Line;
1076     SK.Column = AtColumn;
1077     SK.IsRequired = IsRequired;
1078     SK.FlowLevel = FlowLevel;
1079     SimpleKeys.push_back(SK);
1080   }
1081 }
1082 
removeStaleSimpleKeyCandidates()1083 void Scanner::removeStaleSimpleKeyCandidates() {
1084   for (SmallVectorImpl<SimpleKey>::iterator i = SimpleKeys.begin();
1085                                             i != SimpleKeys.end();) {
1086     if (i->Line != Line || i->Column + 1024 < Column) {
1087       if (i->IsRequired)
1088         setError( "Could not find expected : for simple key"
1089                 , i->Tok->Range.begin());
1090       i = SimpleKeys.erase(i);
1091     } else
1092       ++i;
1093   }
1094 }
1095 
removeSimpleKeyCandidatesOnFlowLevel(unsigned Level)1096 void Scanner::removeSimpleKeyCandidatesOnFlowLevel(unsigned Level) {
1097   if (!SimpleKeys.empty() && (SimpleKeys.end() - 1)->FlowLevel == Level)
1098     SimpleKeys.pop_back();
1099 }
1100 
unrollIndent(int ToColumn)1101 bool Scanner::unrollIndent(int ToColumn) {
1102   Token T;
1103   // Indentation is ignored in flow.
1104   if (FlowLevel != 0)
1105     return true;
1106 
1107   while (Indent > ToColumn) {
1108     T.Kind = Token::TK_BlockEnd;
1109     T.Range = StringRef(Current, 1);
1110     TokenQueue.push_back(T);
1111     Indent = Indents.pop_back_val();
1112   }
1113 
1114   return true;
1115 }
1116 
rollIndent(int ToColumn,Token::TokenKind Kind,TokenQueueT::iterator InsertPoint)1117 bool Scanner::rollIndent( int ToColumn
1118                         , Token::TokenKind Kind
1119                         , TokenQueueT::iterator InsertPoint) {
1120   if (FlowLevel)
1121     return true;
1122   if (Indent < ToColumn) {
1123     Indents.push_back(Indent);
1124     Indent = ToColumn;
1125 
1126     Token T;
1127     T.Kind = Kind;
1128     T.Range = StringRef(Current, 0);
1129     TokenQueue.insert(InsertPoint, T);
1130   }
1131   return true;
1132 }
1133 
skipComment()1134 void Scanner::skipComment() {
1135   if (Current == End || *Current != '#')
1136     return;
1137   while (true) {
1138     // This may skip more than one byte, thus Column is only incremented
1139     // for code points.
1140     StringRef::iterator I = skip_nb_char(Current);
1141     if (I == Current)
1142       break;
1143     Current = I;
1144     ++Column;
1145   }
1146 }
1147 
scanToNextToken()1148 void Scanner::scanToNextToken() {
1149   while (true) {
1150     while (Current != End && (*Current == ' ' || *Current == '\t')) {
1151       skip(1);
1152     }
1153 
1154     skipComment();
1155 
1156     // Skip EOL.
1157     StringRef::iterator i = skip_b_break(Current);
1158     if (i == Current)
1159       break;
1160     Current = i;
1161     ++Line;
1162     Column = 0;
1163     // New lines may start a simple key.
1164     if (!FlowLevel)
1165       IsSimpleKeyAllowed = true;
1166   }
1167 }
1168 
scanStreamStart()1169 bool Scanner::scanStreamStart() {
1170   IsStartOfStream = false;
1171 
1172   EncodingInfo EI = getUnicodeEncoding(currentInput());
1173 
1174   Token T;
1175   T.Kind = Token::TK_StreamStart;
1176   T.Range = StringRef(Current, EI.second);
1177   TokenQueue.push_back(T);
1178   Current += EI.second;
1179   return true;
1180 }
1181 
scanStreamEnd()1182 bool Scanner::scanStreamEnd() {
1183   // Force an ending new line if one isn't present.
1184   if (Column != 0) {
1185     Column = 0;
1186     ++Line;
1187   }
1188 
1189   unrollIndent(-1);
1190   SimpleKeys.clear();
1191   IsSimpleKeyAllowed = false;
1192 
1193   Token T;
1194   T.Kind = Token::TK_StreamEnd;
1195   T.Range = StringRef(Current, 0);
1196   TokenQueue.push_back(T);
1197   return true;
1198 }
1199 
scanDirective()1200 bool Scanner::scanDirective() {
1201   // Reset the indentation level.
1202   unrollIndent(-1);
1203   SimpleKeys.clear();
1204   IsSimpleKeyAllowed = false;
1205 
1206   StringRef::iterator Start = Current;
1207   consume('%');
1208   StringRef::iterator NameStart = Current;
1209   Current = skip_while(&Scanner::skip_ns_char, Current);
1210   StringRef Name(NameStart, Current - NameStart);
1211   Current = skip_while(&Scanner::skip_s_white, Current);
1212 
1213   Token T;
1214   if (Name == "YAML") {
1215     Current = skip_while(&Scanner::skip_ns_char, Current);
1216     T.Kind = Token::TK_VersionDirective;
1217     T.Range = StringRef(Start, Current - Start);
1218     TokenQueue.push_back(T);
1219     return true;
1220   } else if(Name == "TAG") {
1221     Current = skip_while(&Scanner::skip_ns_char, Current);
1222     Current = skip_while(&Scanner::skip_s_white, Current);
1223     Current = skip_while(&Scanner::skip_ns_char, Current);
1224     T.Kind = Token::TK_TagDirective;
1225     T.Range = StringRef(Start, Current - Start);
1226     TokenQueue.push_back(T);
1227     return true;
1228   }
1229   return false;
1230 }
1231 
scanDocumentIndicator(bool IsStart)1232 bool Scanner::scanDocumentIndicator(bool IsStart) {
1233   unrollIndent(-1);
1234   SimpleKeys.clear();
1235   IsSimpleKeyAllowed = false;
1236 
1237   Token T;
1238   T.Kind = IsStart ? Token::TK_DocumentStart : Token::TK_DocumentEnd;
1239   T.Range = StringRef(Current, 3);
1240   skip(3);
1241   TokenQueue.push_back(T);
1242   return true;
1243 }
1244 
scanFlowCollectionStart(bool IsSequence)1245 bool Scanner::scanFlowCollectionStart(bool IsSequence) {
1246   Token T;
1247   T.Kind = IsSequence ? Token::TK_FlowSequenceStart
1248                       : Token::TK_FlowMappingStart;
1249   T.Range = StringRef(Current, 1);
1250   skip(1);
1251   TokenQueue.push_back(T);
1252 
1253   // [ and { may begin a simple key.
1254   saveSimpleKeyCandidate(--TokenQueue.end(), Column - 1, false);
1255 
1256   // And may also be followed by a simple key.
1257   IsSimpleKeyAllowed = true;
1258   ++FlowLevel;
1259   return true;
1260 }
1261 
scanFlowCollectionEnd(bool IsSequence)1262 bool Scanner::scanFlowCollectionEnd(bool IsSequence) {
1263   removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
1264   IsSimpleKeyAllowed = false;
1265   Token T;
1266   T.Kind = IsSequence ? Token::TK_FlowSequenceEnd
1267                       : Token::TK_FlowMappingEnd;
1268   T.Range = StringRef(Current, 1);
1269   skip(1);
1270   TokenQueue.push_back(T);
1271   if (FlowLevel)
1272     --FlowLevel;
1273   return true;
1274 }
1275 
scanFlowEntry()1276 bool Scanner::scanFlowEntry() {
1277   removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
1278   IsSimpleKeyAllowed = true;
1279   Token T;
1280   T.Kind = Token::TK_FlowEntry;
1281   T.Range = StringRef(Current, 1);
1282   skip(1);
1283   TokenQueue.push_back(T);
1284   return true;
1285 }
1286 
scanBlockEntry()1287 bool Scanner::scanBlockEntry() {
1288   rollIndent(Column, Token::TK_BlockSequenceStart, TokenQueue.end());
1289   removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
1290   IsSimpleKeyAllowed = true;
1291   Token T;
1292   T.Kind = Token::TK_BlockEntry;
1293   T.Range = StringRef(Current, 1);
1294   skip(1);
1295   TokenQueue.push_back(T);
1296   return true;
1297 }
1298 
scanKey()1299 bool Scanner::scanKey() {
1300   if (!FlowLevel)
1301     rollIndent(Column, Token::TK_BlockMappingStart, TokenQueue.end());
1302 
1303   removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
1304   IsSimpleKeyAllowed = !FlowLevel;
1305 
1306   Token T;
1307   T.Kind = Token::TK_Key;
1308   T.Range = StringRef(Current, 1);
1309   skip(1);
1310   TokenQueue.push_back(T);
1311   return true;
1312 }
1313 
scanValue()1314 bool Scanner::scanValue() {
1315   // If the previous token could have been a simple key, insert the key token
1316   // into the token queue.
1317   if (!SimpleKeys.empty()) {
1318     SimpleKey SK = SimpleKeys.pop_back_val();
1319     Token T;
1320     T.Kind = Token::TK_Key;
1321     T.Range = SK.Tok->Range;
1322     TokenQueueT::iterator i, e;
1323     for (i = TokenQueue.begin(), e = TokenQueue.end(); i != e; ++i) {
1324       if (i == SK.Tok)
1325         break;
1326     }
1327     if (i == e) {
1328       Failed = true;
1329       return false;
1330     }
1331     i = TokenQueue.insert(i, T);
1332 
1333     // We may also need to add a Block-Mapping-Start token.
1334     rollIndent(SK.Column, Token::TK_BlockMappingStart, i);
1335 
1336     IsSimpleKeyAllowed = false;
1337   } else {
1338     if (!FlowLevel)
1339       rollIndent(Column, Token::TK_BlockMappingStart, TokenQueue.end());
1340     IsSimpleKeyAllowed = !FlowLevel;
1341   }
1342 
1343   Token T;
1344   T.Kind = Token::TK_Value;
1345   T.Range = StringRef(Current, 1);
1346   skip(1);
1347   TokenQueue.push_back(T);
1348   return true;
1349 }
1350 
1351 // Forbidding inlining improves performance by roughly 20%.
1352 // FIXME: Remove once llvm optimizes this to the faster version without hints.
1353 LLVM_ATTRIBUTE_NOINLINE static bool
1354 wasEscaped(StringRef::iterator First, StringRef::iterator Position);
1355 
1356 // Returns whether a character at 'Position' was escaped with a leading '\'.
1357 // 'First' specifies the position of the first character in the string.
wasEscaped(StringRef::iterator First,StringRef::iterator Position)1358 static bool wasEscaped(StringRef::iterator First,
1359                        StringRef::iterator Position) {
1360   assert(Position - 1 >= First);
1361   StringRef::iterator I = Position - 1;
1362   // We calculate the number of consecutive '\'s before the current position
1363   // by iterating backwards through our string.
1364   while (I >= First && *I == '\\') --I;
1365   // (Position - 1 - I) now contains the number of '\'s before the current
1366   // position. If it is odd, the character at 'Position' was escaped.
1367   return (Position - 1 - I) % 2 == 1;
1368 }
1369 
scanFlowScalar(bool IsDoubleQuoted)1370 bool Scanner::scanFlowScalar(bool IsDoubleQuoted) {
1371   StringRef::iterator Start = Current;
1372   unsigned ColStart = Column;
1373   if (IsDoubleQuoted) {
1374     do {
1375       ++Current;
1376       while (Current != End && *Current != '"')
1377         ++Current;
1378       // Repeat until the previous character was not a '\' or was an escaped
1379       // backslash.
1380     } while (   Current != End
1381              && *(Current - 1) == '\\'
1382              && wasEscaped(Start + 1, Current));
1383   } else {
1384     skip(1);
1385     while (Current != End) {
1386       // Skip a ' followed by another '.
1387       if (Current + 1 < End && *Current == '\'' && *(Current + 1) == '\'') {
1388         skip(2);
1389         continue;
1390       } else if (*Current == '\'')
1391         break;
1392       StringRef::iterator i = skip_nb_char(Current);
1393       if (i == Current) {
1394         i = skip_b_break(Current);
1395         if (i == Current)
1396           break;
1397         Current = i;
1398         Column = 0;
1399         ++Line;
1400       } else {
1401         if (i == End)
1402           break;
1403         Current = i;
1404         ++Column;
1405       }
1406     }
1407   }
1408 
1409   if (Current == End) {
1410     setError("Expected quote at end of scalar", Current);
1411     return false;
1412   }
1413 
1414   skip(1); // Skip ending quote.
1415   Token T;
1416   T.Kind = Token::TK_Scalar;
1417   T.Range = StringRef(Start, Current - Start);
1418   TokenQueue.push_back(T);
1419 
1420   saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);
1421 
1422   IsSimpleKeyAllowed = false;
1423 
1424   return true;
1425 }
1426 
scanPlainScalar()1427 bool Scanner::scanPlainScalar() {
1428   StringRef::iterator Start = Current;
1429   unsigned ColStart = Column;
1430   unsigned LeadingBlanks = 0;
1431   assert(Indent >= -1 && "Indent must be >= -1 !");
1432   unsigned indent = static_cast<unsigned>(Indent + 1);
1433   while (Current != End) {
1434     if (*Current == '#')
1435       break;
1436 
1437     while (Current != End && !isBlankOrBreak(Current)) {
1438       if (FlowLevel && *Current == ':' &&
1439           (Current + 1 == End ||
1440            !(isBlankOrBreak(Current + 1) || *(Current + 1) == ','))) {
1441         setError("Found unexpected ':' while scanning a plain scalar", Current);
1442         return false;
1443       }
1444 
1445       // Check for the end of the plain scalar.
1446       if (  (*Current == ':' && isBlankOrBreak(Current + 1))
1447           || (  FlowLevel
1448           && (StringRef(Current, 1).find_first_of(",:?[]{}")
1449               != StringRef::npos)))
1450         break;
1451 
1452       StringRef::iterator i = skip_nb_char(Current);
1453       if (i == Current)
1454         break;
1455       Current = i;
1456       ++Column;
1457     }
1458 
1459     // Are we at the end?
1460     if (!isBlankOrBreak(Current))
1461       break;
1462 
1463     // Eat blanks.
1464     StringRef::iterator Tmp = Current;
1465     while (isBlankOrBreak(Tmp)) {
1466       StringRef::iterator i = skip_s_white(Tmp);
1467       if (i != Tmp) {
1468         if (LeadingBlanks && (Column < indent) && *Tmp == '\t') {
1469           setError("Found invalid tab character in indentation", Tmp);
1470           return false;
1471         }
1472         Tmp = i;
1473         ++Column;
1474       } else {
1475         i = skip_b_break(Tmp);
1476         if (!LeadingBlanks)
1477           LeadingBlanks = 1;
1478         Tmp = i;
1479         Column = 0;
1480         ++Line;
1481       }
1482     }
1483 
1484     if (!FlowLevel && Column < indent)
1485       break;
1486 
1487     Current = Tmp;
1488   }
1489   if (Start == Current) {
1490     setError("Got empty plain scalar", Start);
1491     return false;
1492   }
1493   Token T;
1494   T.Kind = Token::TK_Scalar;
1495   T.Range = StringRef(Start, Current - Start);
1496   TokenQueue.push_back(T);
1497 
1498   // Plain scalars can be simple keys.
1499   saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);
1500 
1501   IsSimpleKeyAllowed = false;
1502 
1503   return true;
1504 }
1505 
scanAliasOrAnchor(bool IsAlias)1506 bool Scanner::scanAliasOrAnchor(bool IsAlias) {
1507   StringRef::iterator Start = Current;
1508   unsigned ColStart = Column;
1509   skip(1);
1510   while (Current != End) {
1511     if (   *Current == '[' || *Current == ']'
1512         || *Current == '{' || *Current == '}'
1513         || *Current == ','
1514         || *Current == ':')
1515       break;
1516     StringRef::iterator i = skip_ns_char(Current);
1517     if (i == Current)
1518       break;
1519     Current = i;
1520     ++Column;
1521   }
1522 
1523   if (Start + 1 == Current) {
1524     setError("Got empty alias or anchor", Start);
1525     return false;
1526   }
1527 
1528   Token T;
1529   T.Kind = IsAlias ? Token::TK_Alias : Token::TK_Anchor;
1530   T.Range = StringRef(Start, Current - Start);
1531   TokenQueue.push_back(T);
1532 
1533   // Alias and anchors can be simple keys.
1534   saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);
1535 
1536   IsSimpleKeyAllowed = false;
1537 
1538   return true;
1539 }
1540 
scanBlockScalarIndicators(char & StyleIndicator,char & ChompingIndicator,unsigned & IndentIndicator,bool & IsDone)1541 bool Scanner::scanBlockScalarIndicators(char &StyleIndicator,
1542                                         char &ChompingIndicator,
1543                                         unsigned &IndentIndicator,
1544                                         bool &IsDone) {
1545   StyleIndicator = scanBlockStyleIndicator();
1546   if (!scanBlockScalarHeader(ChompingIndicator, IndentIndicator, IsDone))
1547     return false;
1548   return true;
1549 }
1550 
scanBlockStyleIndicator()1551 char Scanner::scanBlockStyleIndicator() {
1552   char Indicator = ' ';
1553   if (Current != End && (*Current == '>' || *Current == '|')) {
1554     Indicator = *Current;
1555     skip(1);
1556   }
1557   return Indicator;
1558 }
1559 
scanBlockChompingIndicator()1560 char Scanner::scanBlockChompingIndicator() {
1561   char Indicator = ' ';
1562   if (Current != End && (*Current == '+' || *Current == '-')) {
1563     Indicator = *Current;
1564     skip(1);
1565   }
1566   return Indicator;
1567 }
1568 
1569 /// Get the number of line breaks after chomping.
1570 ///
1571 /// Return the number of trailing line breaks to emit, depending on
1572 /// \p ChompingIndicator.
getChompedLineBreaks(char ChompingIndicator,unsigned LineBreaks,StringRef Str)1573 static unsigned getChompedLineBreaks(char ChompingIndicator,
1574                                      unsigned LineBreaks, StringRef Str) {
1575   if (ChompingIndicator == '-') // Strip all line breaks.
1576     return 0;
1577   if (ChompingIndicator == '+') // Keep all line breaks.
1578     return LineBreaks;
1579   // Clip trailing lines.
1580   return Str.empty() ? 0 : 1;
1581 }
1582 
scanBlockIndentationIndicator()1583 unsigned Scanner::scanBlockIndentationIndicator() {
1584   unsigned Indent = 0;
1585   if (Current != End && (*Current >= '1' && *Current <= '9')) {
1586     Indent = unsigned(*Current - '0');
1587     skip(1);
1588   }
1589   return Indent;
1590 }
1591 
scanBlockScalarHeader(char & ChompingIndicator,unsigned & IndentIndicator,bool & IsDone)1592 bool Scanner::scanBlockScalarHeader(char &ChompingIndicator,
1593                                     unsigned &IndentIndicator, bool &IsDone) {
1594   auto Start = Current;
1595 
1596   ChompingIndicator = scanBlockChompingIndicator();
1597   IndentIndicator = scanBlockIndentationIndicator();
1598   // Check for the chomping indicator once again.
1599   if (ChompingIndicator == ' ')
1600     ChompingIndicator = scanBlockChompingIndicator();
1601   Current = skip_while(&Scanner::skip_s_white, Current);
1602   skipComment();
1603 
1604   if (Current == End) { // EOF, we have an empty scalar.
1605     Token T;
1606     T.Kind = Token::TK_BlockScalar;
1607     T.Range = StringRef(Start, Current - Start);
1608     TokenQueue.push_back(T);
1609     IsDone = true;
1610     return true;
1611   }
1612 
1613   if (!consumeLineBreakIfPresent()) {
1614     setError("Expected a line break after block scalar header", Current);
1615     return false;
1616   }
1617   return true;
1618 }
1619 
findBlockScalarIndent(unsigned & BlockIndent,unsigned BlockExitIndent,unsigned & LineBreaks,bool & IsDone)1620 bool Scanner::findBlockScalarIndent(unsigned &BlockIndent,
1621                                     unsigned BlockExitIndent,
1622                                     unsigned &LineBreaks, bool &IsDone) {
1623   unsigned MaxAllSpaceLineCharacters = 0;
1624   StringRef::iterator LongestAllSpaceLine;
1625 
1626   while (true) {
1627     advanceWhile(&Scanner::skip_s_space);
1628     if (skip_nb_char(Current) != Current) {
1629       // This line isn't empty, so try and find the indentation.
1630       if (Column <= BlockExitIndent) { // End of the block literal.
1631         IsDone = true;
1632         return true;
1633       }
1634       // We found the block's indentation.
1635       BlockIndent = Column;
1636       if (MaxAllSpaceLineCharacters > BlockIndent) {
1637         setError(
1638             "Leading all-spaces line must be smaller than the block indent",
1639             LongestAllSpaceLine);
1640         return false;
1641       }
1642       return true;
1643     }
1644     if (skip_b_break(Current) != Current &&
1645         Column > MaxAllSpaceLineCharacters) {
1646       // Record the longest all-space line in case it's longer than the
1647       // discovered block indent.
1648       MaxAllSpaceLineCharacters = Column;
1649       LongestAllSpaceLine = Current;
1650     }
1651 
1652     // Check for EOF.
1653     if (Current == End) {
1654       IsDone = true;
1655       return true;
1656     }
1657 
1658     if (!consumeLineBreakIfPresent()) {
1659       IsDone = true;
1660       return true;
1661     }
1662     ++LineBreaks;
1663   }
1664   return true;
1665 }
1666 
scanBlockScalarIndent(unsigned BlockIndent,unsigned BlockExitIndent,bool & IsDone)1667 bool Scanner::scanBlockScalarIndent(unsigned BlockIndent,
1668                                     unsigned BlockExitIndent, bool &IsDone) {
1669   // Skip the indentation.
1670   while (Column < BlockIndent) {
1671     auto I = skip_s_space(Current);
1672     if (I == Current)
1673       break;
1674     Current = I;
1675     ++Column;
1676   }
1677 
1678   if (skip_nb_char(Current) == Current)
1679     return true;
1680 
1681   if (Column <= BlockExitIndent) { // End of the block literal.
1682     IsDone = true;
1683     return true;
1684   }
1685 
1686   if (Column < BlockIndent) {
1687     if (Current != End && *Current == '#') { // Trailing comment.
1688       IsDone = true;
1689       return true;
1690     }
1691     setError("A text line is less indented than the block scalar", Current);
1692     return false;
1693   }
1694   return true; // A normal text line.
1695 }
1696 
scanBlockScalar(bool IsLiteral)1697 bool Scanner::scanBlockScalar(bool IsLiteral) {
1698   assert(*Current == '|' || *Current == '>');
1699   char StyleIndicator;
1700   char ChompingIndicator;
1701   unsigned BlockIndent;
1702   bool IsDone = false;
1703   if (!scanBlockScalarIndicators(StyleIndicator, ChompingIndicator, BlockIndent,
1704                                  IsDone))
1705     return false;
1706   if (IsDone)
1707     return true;
1708   bool IsFolded = StyleIndicator == '>';
1709 
1710   const auto *Start = Current;
1711   unsigned BlockExitIndent = Indent < 0 ? 0 : (unsigned)Indent;
1712   unsigned LineBreaks = 0;
1713   if (BlockIndent == 0) {
1714     if (!findBlockScalarIndent(BlockIndent, BlockExitIndent, LineBreaks,
1715                                IsDone))
1716       return false;
1717   }
1718 
1719   // Scan the block's scalars body.
1720   SmallString<256> Str;
1721   while (!IsDone) {
1722     if (!scanBlockScalarIndent(BlockIndent, BlockExitIndent, IsDone))
1723       return false;
1724     if (IsDone)
1725       break;
1726 
1727     // Parse the current line.
1728     auto LineStart = Current;
1729     advanceWhile(&Scanner::skip_nb_char);
1730     if (LineStart != Current) {
1731       if (LineBreaks && IsFolded && !Scanner::isLineEmpty(Str)) {
1732         // The folded style "folds" any single line break between content into a
1733         // single space, except when that content is "empty" (only contains
1734         // whitespace) in which case the line break is left as-is.
1735         if (LineBreaks == 1) {
1736           Str.append(LineBreaks,
1737                      isLineEmpty(StringRef(LineStart, Current - LineStart))
1738                          ? '\n'
1739                          : ' ');
1740         }
1741         // If we saw a single line break, we are completely replacing it and so
1742         // want `LineBreaks == 0`. Otherwise this decrement accounts for the
1743         // fact that the first line break is "trimmed", only being used to
1744         // signal a sequence of line breaks which should not be folded.
1745         LineBreaks--;
1746       }
1747       Str.append(LineBreaks, '\n');
1748       Str.append(StringRef(LineStart, Current - LineStart));
1749       LineBreaks = 0;
1750     }
1751 
1752     // Check for EOF.
1753     if (Current == End)
1754       break;
1755 
1756     if (!consumeLineBreakIfPresent())
1757       break;
1758     ++LineBreaks;
1759   }
1760 
1761   if (Current == End && !LineBreaks)
1762     // Ensure that there is at least one line break before the end of file.
1763     LineBreaks = 1;
1764   Str.append(getChompedLineBreaks(ChompingIndicator, LineBreaks, Str), '\n');
1765 
1766   // New lines may start a simple key.
1767   if (!FlowLevel)
1768     IsSimpleKeyAllowed = true;
1769 
1770   Token T;
1771   T.Kind = Token::TK_BlockScalar;
1772   T.Range = StringRef(Start, Current - Start);
1773   T.Value = std::string(Str);
1774   TokenQueue.push_back(T);
1775   return true;
1776 }
1777 
scanTag()1778 bool Scanner::scanTag() {
1779   StringRef::iterator Start = Current;
1780   unsigned ColStart = Column;
1781   skip(1); // Eat !.
1782   if (Current == End || isBlankOrBreak(Current)); // An empty tag.
1783   else if (*Current == '<') {
1784     skip(1);
1785     scan_ns_uri_char();
1786     if (!consume('>'))
1787       return false;
1788   } else {
1789     // FIXME: Actually parse the c-ns-shorthand-tag rule.
1790     Current = skip_while(&Scanner::skip_ns_char, Current);
1791   }
1792 
1793   Token T;
1794   T.Kind = Token::TK_Tag;
1795   T.Range = StringRef(Start, Current - Start);
1796   TokenQueue.push_back(T);
1797 
1798   // Tags can be simple keys.
1799   saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);
1800 
1801   IsSimpleKeyAllowed = false;
1802 
1803   return true;
1804 }
1805 
fetchMoreTokens()1806 bool Scanner::fetchMoreTokens() {
1807   if (IsStartOfStream)
1808     return scanStreamStart();
1809 
1810   scanToNextToken();
1811 
1812   if (Current == End)
1813     return scanStreamEnd();
1814 
1815   removeStaleSimpleKeyCandidates();
1816 
1817   unrollIndent(Column);
1818 
1819   if (Column == 0 && *Current == '%')
1820     return scanDirective();
1821 
1822   if (Column == 0 && Current + 4 <= End
1823       && *Current == '-'
1824       && *(Current + 1) == '-'
1825       && *(Current + 2) == '-'
1826       && (Current + 3 == End || isBlankOrBreak(Current + 3)))
1827     return scanDocumentIndicator(true);
1828 
1829   if (Column == 0 && Current + 4 <= End
1830       && *Current == '.'
1831       && *(Current + 1) == '.'
1832       && *(Current + 2) == '.'
1833       && (Current + 3 == End || isBlankOrBreak(Current + 3)))
1834     return scanDocumentIndicator(false);
1835 
1836   if (*Current == '[')
1837     return scanFlowCollectionStart(true);
1838 
1839   if (*Current == '{')
1840     return scanFlowCollectionStart(false);
1841 
1842   if (*Current == ']')
1843     return scanFlowCollectionEnd(true);
1844 
1845   if (*Current == '}')
1846     return scanFlowCollectionEnd(false);
1847 
1848   if (*Current == ',')
1849     return scanFlowEntry();
1850 
1851   if (*Current == '-' && isBlankOrBreak(Current + 1))
1852     return scanBlockEntry();
1853 
1854   if (*Current == '?' && (FlowLevel || isBlankOrBreak(Current + 1)))
1855     return scanKey();
1856 
1857   if (*Current == ':' && (FlowLevel || isBlankOrBreak(Current + 1)))
1858     return scanValue();
1859 
1860   if (*Current == '*')
1861     return scanAliasOrAnchor(true);
1862 
1863   if (*Current == '&')
1864     return scanAliasOrAnchor(false);
1865 
1866   if (*Current == '!')
1867     return scanTag();
1868 
1869   if (*Current == '|' && !FlowLevel)
1870     return scanBlockScalar(true);
1871 
1872   if (*Current == '>' && !FlowLevel)
1873     return scanBlockScalar(false);
1874 
1875   if (*Current == '\'')
1876     return scanFlowScalar(false);
1877 
1878   if (*Current == '"')
1879     return scanFlowScalar(true);
1880 
1881   // Get a plain scalar.
1882   StringRef FirstChar(Current, 1);
1883   if (!(isBlankOrBreak(Current)
1884         || FirstChar.find_first_of("-?:,[]{}#&*!|>'\"%@`") != StringRef::npos)
1885       || (*Current == '-' && !isBlankOrBreak(Current + 1))
1886       || (!FlowLevel && (*Current == '?' || *Current == ':')
1887           && isBlankOrBreak(Current + 1))
1888       || (!FlowLevel && *Current == ':'
1889                       && Current + 2 < End
1890                       && *(Current + 1) == ':'
1891                       && !isBlankOrBreak(Current + 2)))
1892     return scanPlainScalar();
1893 
1894   setError("Unrecognized character while tokenizing.", Current);
1895   return false;
1896 }
1897 
Stream(StringRef Input,SourceMgr & SM,bool ShowColors,std::error_code * EC)1898 Stream::Stream(StringRef Input, SourceMgr &SM, bool ShowColors,
1899                std::error_code *EC)
1900     : scanner(new Scanner(Input, SM, ShowColors, EC)) {}
1901 
Stream(MemoryBufferRef InputBuffer,SourceMgr & SM,bool ShowColors,std::error_code * EC)1902 Stream::Stream(MemoryBufferRef InputBuffer, SourceMgr &SM, bool ShowColors,
1903                std::error_code *EC)
1904     : scanner(new Scanner(InputBuffer, SM, ShowColors, EC)) {}
1905 
1906 Stream::~Stream() = default;
1907 
failed()1908 bool Stream::failed() { return scanner->failed(); }
1909 
printError(Node * N,const Twine & Msg,SourceMgr::DiagKind Kind)1910 void Stream::printError(Node *N, const Twine &Msg, SourceMgr::DiagKind Kind) {
1911   printError(N ? N->getSourceRange() : SMRange(), Msg, Kind);
1912 }
1913 
printError(const SMRange & Range,const Twine & Msg,SourceMgr::DiagKind Kind)1914 void Stream::printError(const SMRange &Range, const Twine &Msg,
1915                         SourceMgr::DiagKind Kind) {
1916   scanner->printError(Range.Start, Kind, Msg, Range);
1917 }
1918 
begin()1919 document_iterator Stream::begin() {
1920   if (CurrentDoc)
1921     report_fatal_error("Can only iterate over the stream once");
1922 
1923   // Skip Stream-Start.
1924   scanner->getNext();
1925 
1926   CurrentDoc.reset(new Document(*this));
1927   return document_iterator(CurrentDoc);
1928 }
1929 
end()1930 document_iterator Stream::end() {
1931   return document_iterator();
1932 }
1933 
skip()1934 void Stream::skip() {
1935   for (Document &Doc : *this)
1936     Doc.skip();
1937 }
1938 
Node(unsigned int Type,std::unique_ptr<Document> & D,StringRef A,StringRef T)1939 Node::Node(unsigned int Type, std::unique_ptr<Document> &D, StringRef A,
1940            StringRef T)
1941     : Doc(D), TypeID(Type), Anchor(A), Tag(T) {
1942   SMLoc Start = SMLoc::getFromPointer(peekNext().Range.begin());
1943   SourceRange = SMRange(Start, Start);
1944 }
1945 
getVerbatimTag() const1946 std::string Node::getVerbatimTag() const {
1947   StringRef Raw = getRawTag();
1948   if (!Raw.empty() && Raw != "!") {
1949     std::string Ret;
1950     if (Raw.find_last_of('!') == 0) {
1951       Ret = std::string(Doc->getTagMap().find("!")->second);
1952       Ret += Raw.substr(1);
1953       return Ret;
1954     } else if (Raw.startswith("!!")) {
1955       Ret = std::string(Doc->getTagMap().find("!!")->second);
1956       Ret += Raw.substr(2);
1957       return Ret;
1958     } else {
1959       StringRef TagHandle = Raw.substr(0, Raw.find_last_of('!') + 1);
1960       std::map<StringRef, StringRef>::const_iterator It =
1961           Doc->getTagMap().find(TagHandle);
1962       if (It != Doc->getTagMap().end())
1963         Ret = std::string(It->second);
1964       else {
1965         Token T;
1966         T.Kind = Token::TK_Tag;
1967         T.Range = TagHandle;
1968         setError(Twine("Unknown tag handle ") + TagHandle, T);
1969       }
1970       Ret += Raw.substr(Raw.find_last_of('!') + 1);
1971       return Ret;
1972     }
1973   }
1974 
1975   switch (getType()) {
1976   case NK_Null:
1977     return "tag:yaml.org,2002:null";
1978   case NK_Scalar:
1979   case NK_BlockScalar:
1980     // TODO: Tag resolution.
1981     return "tag:yaml.org,2002:str";
1982   case NK_Mapping:
1983     return "tag:yaml.org,2002:map";
1984   case NK_Sequence:
1985     return "tag:yaml.org,2002:seq";
1986   }
1987 
1988   return "";
1989 }
1990 
peekNext()1991 Token &Node::peekNext() {
1992   return Doc->peekNext();
1993 }
1994 
getNext()1995 Token Node::getNext() {
1996   return Doc->getNext();
1997 }
1998 
parseBlockNode()1999 Node *Node::parseBlockNode() {
2000   return Doc->parseBlockNode();
2001 }
2002 
getAllocator()2003 BumpPtrAllocator &Node::getAllocator() {
2004   return Doc->NodeAllocator;
2005 }
2006 
setError(const Twine & Msg,Token & Tok) const2007 void Node::setError(const Twine &Msg, Token &Tok) const {
2008   Doc->setError(Msg, Tok);
2009 }
2010 
failed() const2011 bool Node::failed() const {
2012   return Doc->failed();
2013 }
2014 
getValue(SmallVectorImpl<char> & Storage) const2015 StringRef ScalarNode::getValue(SmallVectorImpl<char> &Storage) const {
2016   // TODO: Handle newlines properly. We need to remove leading whitespace.
2017   if (Value[0] == '"') { // Double quoted.
2018     // Pull off the leading and trailing "s.
2019     StringRef UnquotedValue = Value.substr(1, Value.size() - 2);
2020     // Search for characters that would require unescaping the value.
2021     StringRef::size_type i = UnquotedValue.find_first_of("\\\r\n");
2022     if (i != StringRef::npos)
2023       return unescapeDoubleQuoted(UnquotedValue, i, Storage);
2024     return UnquotedValue;
2025   } else if (Value[0] == '\'') { // Single quoted.
2026     // Pull off the leading and trailing 's.
2027     StringRef UnquotedValue = Value.substr(1, Value.size() - 2);
2028     StringRef::size_type i = UnquotedValue.find('\'');
2029     if (i != StringRef::npos) {
2030       // We're going to need Storage.
2031       Storage.clear();
2032       Storage.reserve(UnquotedValue.size());
2033       for (; i != StringRef::npos; i = UnquotedValue.find('\'')) {
2034         StringRef Valid(UnquotedValue.begin(), i);
2035         llvm::append_range(Storage, Valid);
2036         Storage.push_back('\'');
2037         UnquotedValue = UnquotedValue.substr(i + 2);
2038       }
2039       llvm::append_range(Storage, UnquotedValue);
2040       return StringRef(Storage.begin(), Storage.size());
2041     }
2042     return UnquotedValue;
2043   }
2044   // Plain or block.
2045   return Value.rtrim(' ');
2046 }
2047 
unescapeDoubleQuoted(StringRef UnquotedValue,StringRef::size_type i,SmallVectorImpl<char> & Storage) const2048 StringRef ScalarNode::unescapeDoubleQuoted( StringRef UnquotedValue
2049                                           , StringRef::size_type i
2050                                           , SmallVectorImpl<char> &Storage)
2051                                           const {
2052   // Use Storage to build proper value.
2053   Storage.clear();
2054   Storage.reserve(UnquotedValue.size());
2055   for (; i != StringRef::npos; i = UnquotedValue.find_first_of("\\\r\n")) {
2056     // Insert all previous chars into Storage.
2057     StringRef Valid(UnquotedValue.begin(), i);
2058     llvm::append_range(Storage, Valid);
2059     // Chop off inserted chars.
2060     UnquotedValue = UnquotedValue.substr(i);
2061 
2062     assert(!UnquotedValue.empty() && "Can't be empty!");
2063 
2064     // Parse escape or line break.
2065     switch (UnquotedValue[0]) {
2066     case '\r':
2067     case '\n':
2068       Storage.push_back('\n');
2069       if (   UnquotedValue.size() > 1
2070           && (UnquotedValue[1] == '\r' || UnquotedValue[1] == '\n'))
2071         UnquotedValue = UnquotedValue.substr(1);
2072       UnquotedValue = UnquotedValue.substr(1);
2073       break;
2074     default:
2075       if (UnquotedValue.size() == 1) {
2076         Token T;
2077         T.Range = StringRef(UnquotedValue.begin(), 1);
2078         setError("Unrecognized escape code", T);
2079         return "";
2080       }
2081       UnquotedValue = UnquotedValue.substr(1);
2082       switch (UnquotedValue[0]) {
2083       default: {
2084           Token T;
2085           T.Range = StringRef(UnquotedValue.begin(), 1);
2086           setError("Unrecognized escape code", T);
2087           return "";
2088         }
2089       case '\r':
2090       case '\n':
2091         // Remove the new line.
2092         if (   UnquotedValue.size() > 1
2093             && (UnquotedValue[1] == '\r' || UnquotedValue[1] == '\n'))
2094           UnquotedValue = UnquotedValue.substr(1);
2095         // If this was just a single byte newline, it will get skipped
2096         // below.
2097         break;
2098       case '0':
2099         Storage.push_back(0x00);
2100         break;
2101       case 'a':
2102         Storage.push_back(0x07);
2103         break;
2104       case 'b':
2105         Storage.push_back(0x08);
2106         break;
2107       case 't':
2108       case 0x09:
2109         Storage.push_back(0x09);
2110         break;
2111       case 'n':
2112         Storage.push_back(0x0A);
2113         break;
2114       case 'v':
2115         Storage.push_back(0x0B);
2116         break;
2117       case 'f':
2118         Storage.push_back(0x0C);
2119         break;
2120       case 'r':
2121         Storage.push_back(0x0D);
2122         break;
2123       case 'e':
2124         Storage.push_back(0x1B);
2125         break;
2126       case ' ':
2127         Storage.push_back(0x20);
2128         break;
2129       case '"':
2130         Storage.push_back(0x22);
2131         break;
2132       case '/':
2133         Storage.push_back(0x2F);
2134         break;
2135       case '\\':
2136         Storage.push_back(0x5C);
2137         break;
2138       case 'N':
2139         encodeUTF8(0x85, Storage);
2140         break;
2141       case '_':
2142         encodeUTF8(0xA0, Storage);
2143         break;
2144       case 'L':
2145         encodeUTF8(0x2028, Storage);
2146         break;
2147       case 'P':
2148         encodeUTF8(0x2029, Storage);
2149         break;
2150       case 'x': {
2151           if (UnquotedValue.size() < 3)
2152             // TODO: Report error.
2153             break;
2154           unsigned int UnicodeScalarValue;
2155           if (UnquotedValue.substr(1, 2).getAsInteger(16, UnicodeScalarValue))
2156             // TODO: Report error.
2157             UnicodeScalarValue = 0xFFFD;
2158           encodeUTF8(UnicodeScalarValue, Storage);
2159           UnquotedValue = UnquotedValue.substr(2);
2160           break;
2161         }
2162       case 'u': {
2163           if (UnquotedValue.size() < 5)
2164             // TODO: Report error.
2165             break;
2166           unsigned int UnicodeScalarValue;
2167           if (UnquotedValue.substr(1, 4).getAsInteger(16, UnicodeScalarValue))
2168             // TODO: Report error.
2169             UnicodeScalarValue = 0xFFFD;
2170           encodeUTF8(UnicodeScalarValue, Storage);
2171           UnquotedValue = UnquotedValue.substr(4);
2172           break;
2173         }
2174       case 'U': {
2175           if (UnquotedValue.size() < 9)
2176             // TODO: Report error.
2177             break;
2178           unsigned int UnicodeScalarValue;
2179           if (UnquotedValue.substr(1, 8).getAsInteger(16, UnicodeScalarValue))
2180             // TODO: Report error.
2181             UnicodeScalarValue = 0xFFFD;
2182           encodeUTF8(UnicodeScalarValue, Storage);
2183           UnquotedValue = UnquotedValue.substr(8);
2184           break;
2185         }
2186       }
2187       UnquotedValue = UnquotedValue.substr(1);
2188     }
2189   }
2190   llvm::append_range(Storage, UnquotedValue);
2191   return StringRef(Storage.begin(), Storage.size());
2192 }
2193 
getKey()2194 Node *KeyValueNode::getKey() {
2195   if (Key)
2196     return Key;
2197   // Handle implicit null keys.
2198   {
2199     Token &t = peekNext();
2200     if (   t.Kind == Token::TK_BlockEnd
2201         || t.Kind == Token::TK_Value
2202         || t.Kind == Token::TK_Error) {
2203       return Key = new (getAllocator()) NullNode(Doc);
2204     }
2205     if (t.Kind == Token::TK_Key)
2206       getNext(); // skip TK_Key.
2207   }
2208 
2209   // Handle explicit null keys.
2210   Token &t = peekNext();
2211   if (t.Kind == Token::TK_BlockEnd || t.Kind == Token::TK_Value) {
2212     return Key = new (getAllocator()) NullNode(Doc);
2213   }
2214 
2215   // We've got a normal key.
2216   return Key = parseBlockNode();
2217 }
2218 
getValue()2219 Node *KeyValueNode::getValue() {
2220   if (Value)
2221     return Value;
2222 
2223   if (Node* Key = getKey())
2224     Key->skip();
2225   else {
2226     setError("Null key in Key Value.", peekNext());
2227     return Value = new (getAllocator()) NullNode(Doc);
2228   }
2229 
2230   if (failed())
2231     return Value = new (getAllocator()) NullNode(Doc);
2232 
2233   // Handle implicit null values.
2234   {
2235     Token &t = peekNext();
2236     if (   t.Kind == Token::TK_BlockEnd
2237         || t.Kind == Token::TK_FlowMappingEnd
2238         || t.Kind == Token::TK_Key
2239         || t.Kind == Token::TK_FlowEntry
2240         || t.Kind == Token::TK_Error) {
2241       return Value = new (getAllocator()) NullNode(Doc);
2242     }
2243 
2244     if (t.Kind != Token::TK_Value) {
2245       setError("Unexpected token in Key Value.", t);
2246       return Value = new (getAllocator()) NullNode(Doc);
2247     }
2248     getNext(); // skip TK_Value.
2249   }
2250 
2251   // Handle explicit null values.
2252   Token &t = peekNext();
2253   if (t.Kind == Token::TK_BlockEnd || t.Kind == Token::TK_Key) {
2254     return Value = new (getAllocator()) NullNode(Doc);
2255   }
2256 
2257   // We got a normal value.
2258   return Value = parseBlockNode();
2259 }
2260 
increment()2261 void MappingNode::increment() {
2262   if (failed()) {
2263     IsAtEnd = true;
2264     CurrentEntry = nullptr;
2265     return;
2266   }
2267   if (CurrentEntry) {
2268     CurrentEntry->skip();
2269     if (Type == MT_Inline) {
2270       IsAtEnd = true;
2271       CurrentEntry = nullptr;
2272       return;
2273     }
2274   }
2275   Token T = peekNext();
2276   if (T.Kind == Token::TK_Key || T.Kind == Token::TK_Scalar) {
2277     // KeyValueNode eats the TK_Key. That way it can detect null keys.
2278     CurrentEntry = new (getAllocator()) KeyValueNode(Doc);
2279   } else if (Type == MT_Block) {
2280     switch (T.Kind) {
2281     case Token::TK_BlockEnd:
2282       getNext();
2283       IsAtEnd = true;
2284       CurrentEntry = nullptr;
2285       break;
2286     default:
2287       setError("Unexpected token. Expected Key or Block End", T);
2288       [[fallthrough]];
2289     case Token::TK_Error:
2290       IsAtEnd = true;
2291       CurrentEntry = nullptr;
2292     }
2293   } else {
2294     switch (T.Kind) {
2295     case Token::TK_FlowEntry:
2296       // Eat the flow entry and recurse.
2297       getNext();
2298       return increment();
2299     case Token::TK_FlowMappingEnd:
2300       getNext();
2301       [[fallthrough]];
2302     case Token::TK_Error:
2303       // Set this to end iterator.
2304       IsAtEnd = true;
2305       CurrentEntry = nullptr;
2306       break;
2307     default:
2308       setError( "Unexpected token. Expected Key, Flow Entry, or Flow "
2309                 "Mapping End."
2310               , T);
2311       IsAtEnd = true;
2312       CurrentEntry = nullptr;
2313     }
2314   }
2315 }
2316 
increment()2317 void SequenceNode::increment() {
2318   if (failed()) {
2319     IsAtEnd = true;
2320     CurrentEntry = nullptr;
2321     return;
2322   }
2323   if (CurrentEntry)
2324     CurrentEntry->skip();
2325   Token T = peekNext();
2326   if (SeqType == ST_Block) {
2327     switch (T.Kind) {
2328     case Token::TK_BlockEntry:
2329       getNext();
2330       CurrentEntry = parseBlockNode();
2331       if (!CurrentEntry) { // An error occurred.
2332         IsAtEnd = true;
2333         CurrentEntry = nullptr;
2334       }
2335       break;
2336     case Token::TK_BlockEnd:
2337       getNext();
2338       IsAtEnd = true;
2339       CurrentEntry = nullptr;
2340       break;
2341     default:
2342       setError( "Unexpected token. Expected Block Entry or Block End."
2343               , T);
2344       [[fallthrough]];
2345     case Token::TK_Error:
2346       IsAtEnd = true;
2347       CurrentEntry = nullptr;
2348     }
2349   } else if (SeqType == ST_Indentless) {
2350     switch (T.Kind) {
2351     case Token::TK_BlockEntry:
2352       getNext();
2353       CurrentEntry = parseBlockNode();
2354       if (!CurrentEntry) { // An error occurred.
2355         IsAtEnd = true;
2356         CurrentEntry = nullptr;
2357       }
2358       break;
2359     default:
2360     case Token::TK_Error:
2361       IsAtEnd = true;
2362       CurrentEntry = nullptr;
2363     }
2364   } else if (SeqType == ST_Flow) {
2365     switch (T.Kind) {
2366     case Token::TK_FlowEntry:
2367       // Eat the flow entry and recurse.
2368       getNext();
2369       WasPreviousTokenFlowEntry = true;
2370       return increment();
2371     case Token::TK_FlowSequenceEnd:
2372       getNext();
2373       [[fallthrough]];
2374     case Token::TK_Error:
2375       // Set this to end iterator.
2376       IsAtEnd = true;
2377       CurrentEntry = nullptr;
2378       break;
2379     case Token::TK_StreamEnd:
2380     case Token::TK_DocumentEnd:
2381     case Token::TK_DocumentStart:
2382       setError("Could not find closing ]!", T);
2383       // Set this to end iterator.
2384       IsAtEnd = true;
2385       CurrentEntry = nullptr;
2386       break;
2387     default:
2388       if (!WasPreviousTokenFlowEntry) {
2389         setError("Expected , between entries!", T);
2390         IsAtEnd = true;
2391         CurrentEntry = nullptr;
2392         break;
2393       }
2394       // Otherwise it must be a flow entry.
2395       CurrentEntry = parseBlockNode();
2396       if (!CurrentEntry) {
2397         IsAtEnd = true;
2398       }
2399       WasPreviousTokenFlowEntry = false;
2400       break;
2401     }
2402   }
2403 }
2404 
Document(Stream & S)2405 Document::Document(Stream &S) : stream(S), Root(nullptr) {
2406   // Tag maps starts with two default mappings.
2407   TagMap["!"] = "!";
2408   TagMap["!!"] = "tag:yaml.org,2002:";
2409 
2410   if (parseDirectives())
2411     expectToken(Token::TK_DocumentStart);
2412   Token &T = peekNext();
2413   if (T.Kind == Token::TK_DocumentStart)
2414     getNext();
2415 }
2416 
skip()2417 bool Document::skip()  {
2418   if (stream.scanner->failed())
2419     return false;
2420   if (!Root && !getRoot())
2421     return false;
2422   Root->skip();
2423   Token &T = peekNext();
2424   if (T.Kind == Token::TK_StreamEnd)
2425     return false;
2426   if (T.Kind == Token::TK_DocumentEnd) {
2427     getNext();
2428     return skip();
2429   }
2430   return true;
2431 }
2432 
peekNext()2433 Token &Document::peekNext() {
2434   return stream.scanner->peekNext();
2435 }
2436 
getNext()2437 Token Document::getNext() {
2438   return stream.scanner->getNext();
2439 }
2440 
setError(const Twine & Message,Token & Location) const2441 void Document::setError(const Twine &Message, Token &Location) const {
2442   stream.scanner->setError(Message, Location.Range.begin());
2443 }
2444 
failed() const2445 bool Document::failed() const {
2446   return stream.scanner->failed();
2447 }
2448 
parseBlockNode()2449 Node *Document::parseBlockNode() {
2450   Token T = peekNext();
2451   // Handle properties.
2452   Token AnchorInfo;
2453   Token TagInfo;
2454 parse_property:
2455   switch (T.Kind) {
2456   case Token::TK_Alias:
2457     getNext();
2458     return new (NodeAllocator) AliasNode(stream.CurrentDoc, T.Range.substr(1));
2459   case Token::TK_Anchor:
2460     if (AnchorInfo.Kind == Token::TK_Anchor) {
2461       setError("Already encountered an anchor for this node!", T);
2462       return nullptr;
2463     }
2464     AnchorInfo = getNext(); // Consume TK_Anchor.
2465     T = peekNext();
2466     goto parse_property;
2467   case Token::TK_Tag:
2468     if (TagInfo.Kind == Token::TK_Tag) {
2469       setError("Already encountered a tag for this node!", T);
2470       return nullptr;
2471     }
2472     TagInfo = getNext(); // Consume TK_Tag.
2473     T = peekNext();
2474     goto parse_property;
2475   default:
2476     break;
2477   }
2478 
2479   switch (T.Kind) {
2480   case Token::TK_BlockEntry:
2481     // We got an unindented BlockEntry sequence. This is not terminated with
2482     // a BlockEnd.
2483     // Don't eat the TK_BlockEntry, SequenceNode needs it.
2484     return new (NodeAllocator) SequenceNode( stream.CurrentDoc
2485                                            , AnchorInfo.Range.substr(1)
2486                                            , TagInfo.Range
2487                                            , SequenceNode::ST_Indentless);
2488   case Token::TK_BlockSequenceStart:
2489     getNext();
2490     return new (NodeAllocator)
2491       SequenceNode( stream.CurrentDoc
2492                   , AnchorInfo.Range.substr(1)
2493                   , TagInfo.Range
2494                   , SequenceNode::ST_Block);
2495   case Token::TK_BlockMappingStart:
2496     getNext();
2497     return new (NodeAllocator)
2498       MappingNode( stream.CurrentDoc
2499                  , AnchorInfo.Range.substr(1)
2500                  , TagInfo.Range
2501                  , MappingNode::MT_Block);
2502   case Token::TK_FlowSequenceStart:
2503     getNext();
2504     return new (NodeAllocator)
2505       SequenceNode( stream.CurrentDoc
2506                   , AnchorInfo.Range.substr(1)
2507                   , TagInfo.Range
2508                   , SequenceNode::ST_Flow);
2509   case Token::TK_FlowMappingStart:
2510     getNext();
2511     return new (NodeAllocator)
2512       MappingNode( stream.CurrentDoc
2513                  , AnchorInfo.Range.substr(1)
2514                  , TagInfo.Range
2515                  , MappingNode::MT_Flow);
2516   case Token::TK_Scalar:
2517     getNext();
2518     return new (NodeAllocator)
2519       ScalarNode( stream.CurrentDoc
2520                 , AnchorInfo.Range.substr(1)
2521                 , TagInfo.Range
2522                 , T.Range);
2523   case Token::TK_BlockScalar: {
2524     getNext();
2525     StringRef NullTerminatedStr(T.Value.c_str(), T.Value.length() + 1);
2526     StringRef StrCopy = NullTerminatedStr.copy(NodeAllocator).drop_back();
2527     return new (NodeAllocator)
2528         BlockScalarNode(stream.CurrentDoc, AnchorInfo.Range.substr(1),
2529                         TagInfo.Range, StrCopy, T.Range);
2530   }
2531   case Token::TK_Key:
2532     // Don't eat the TK_Key, KeyValueNode expects it.
2533     return new (NodeAllocator)
2534       MappingNode( stream.CurrentDoc
2535                  , AnchorInfo.Range.substr(1)
2536                  , TagInfo.Range
2537                  , MappingNode::MT_Inline);
2538   case Token::TK_DocumentStart:
2539   case Token::TK_DocumentEnd:
2540   case Token::TK_StreamEnd:
2541   default:
2542     // TODO: Properly handle tags. "[!!str ]" should resolve to !!str "", not
2543     //       !!null null.
2544     return new (NodeAllocator) NullNode(stream.CurrentDoc);
2545   case Token::TK_FlowMappingEnd:
2546   case Token::TK_FlowSequenceEnd:
2547   case Token::TK_FlowEntry: {
2548     if (Root && (isa<MappingNode>(Root) || isa<SequenceNode>(Root)))
2549       return new (NodeAllocator) NullNode(stream.CurrentDoc);
2550 
2551     setError("Unexpected token", T);
2552     return nullptr;
2553   }
2554   case Token::TK_Error:
2555     return nullptr;
2556   }
2557   llvm_unreachable("Control flow shouldn't reach here.");
2558   return nullptr;
2559 }
2560 
parseDirectives()2561 bool Document::parseDirectives() {
2562   bool isDirective = false;
2563   while (true) {
2564     Token T = peekNext();
2565     if (T.Kind == Token::TK_TagDirective) {
2566       parseTAGDirective();
2567       isDirective = true;
2568     } else if (T.Kind == Token::TK_VersionDirective) {
2569       parseYAMLDirective();
2570       isDirective = true;
2571     } else
2572       break;
2573   }
2574   return isDirective;
2575 }
2576 
parseYAMLDirective()2577 void Document::parseYAMLDirective() {
2578   getNext(); // Eat %YAML <version>
2579 }
2580 
parseTAGDirective()2581 void Document::parseTAGDirective() {
2582   Token Tag = getNext(); // %TAG <handle> <prefix>
2583   StringRef T = Tag.Range;
2584   // Strip %TAG
2585   T = T.substr(T.find_first_of(" \t")).ltrim(" \t");
2586   std::size_t HandleEnd = T.find_first_of(" \t");
2587   StringRef TagHandle = T.substr(0, HandleEnd);
2588   StringRef TagPrefix = T.substr(HandleEnd).ltrim(" \t");
2589   TagMap[TagHandle] = TagPrefix;
2590 }
2591 
expectToken(int TK)2592 bool Document::expectToken(int TK) {
2593   Token T = getNext();
2594   if (T.Kind != TK) {
2595     setError("Unexpected token", T);
2596     return false;
2597   }
2598   return true;
2599 }
2600