/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp

Bug Summary

File:	llvm/lib/Support/YAMLParser.cpp
Warning:	line 1597, column 9 2nd function call argument is an uninitialized value
Annotated Source Code

Press '?' to see keyboard shortcuts
Show analyzer invocation
clang -cc1 -cc1 -triple x86_64-pc-linux-gnu -analyze -disable-free -disable-llvm-verifier -discard-value-names -main-file-name YAMLParser.cpp -analyzer-store=region -analyzer-opt-analyze-nested-blocks -analyzer-checker=core -analyzer-checker=apiModeling -analyzer-checker=unix -analyzer-checker=deadcode -analyzer-checker=cplusplus -analyzer-checker=security.insecureAPI.UncheckedReturn -analyzer-checker=security.insecureAPI.getpw -analyzer-checker=security.insecureAPI.gets -analyzer-checker=security.insecureAPI.mktemp -analyzer-checker=security.insecureAPI.mkstemp -analyzer-checker=security.insecureAPI.vfork -analyzer-checker=nullability.NullPassedToNonnull -analyzer-checker=nullability.NullReturnedFromNonnull -analyzer-output plist -w -setup-static-analyzer -analyzer-config-compatibility-mode=true -mrelocation-model pic -pic-level 2 -mframe-pointer=none -fmath-errno -fno-rounding-math -mconstructor-aliases -munwind-tables -target-cpu x86-64 -tune-cpu generic -debugger-tuning=gdb -ffunction-sections -fdata-sections -fcoverage-compilation-dir=/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/build-llvm/lib/Support -resource-dir /usr/lib/llvm-14/lib/clang/14.0.0 -D _DEBUG -D _GNU_SOURCE -D __STDC_CONSTANT_MACROS -D __STDC_FORMAT_MACROS -D __STDC_LIMIT_MACROS -I /build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/build-llvm/lib/Support -I /build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support -I /build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/build-llvm/include -I /build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/include -D NDEBUG -U NDEBUG -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../include/c++/10 -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../include/x86_64-linux-gnu/c++/10 -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../include/c++/10/backward -internal-isystem /usr/lib/llvm-14/lib/clang/14.0.0/include -internal-isystem /usr/local/include -internal-isystem /usr/lib/gcc/x86_64-linux-gnu/10/../../../../x86_64-linux-gnu/include -internal-externc-isystem /usr/include/x86_64-linux-gnu -internal-externc-isystem /include -internal-externc-isystem /usr/include -O2 -Wno-unused-parameter -Wwrite-strings -Wno-missing-field-initializers -Wno-long-long -Wno-maybe-uninitialized -Wno-class-memaccess -Wno-redundant-move -Wno-pessimizing-move -Wno-noexcept-type -Wno-comment -std=c++14 -fdeprecated-macro -fdebug-compilation-dir=/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/build-llvm/lib/Support -fdebug-prefix-map=/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0=. -ferror-limit 19 -fvisibility-inlines-hidden -stack-protector 2 -fgnuc-version=4.2.1 -vectorize-loops -vectorize-slp -analyzer-output=html -analyzer-config stable-report-filename=true -faddrsig -D__GCC_HAVE_DWARF2_CFI_ASM=1 -o /tmp/scan-build-2021-08-28-193554-24367-1 -x c++ /build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp
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//===----------------------------------------------------------------------===//

13#include "llvm/Support/YAMLParser.h"
14#include "llvm/ADT/AllocatorList.h"
15#include "llvm/ADT/ArrayRef.h"
16#include "llvm/ADT/None.h"
17#include "llvm/ADT/STLExtras.h"
18#include "llvm/ADT/SmallString.h"
19#include "llvm/ADT/SmallVector.h"
20#include "llvm/ADT/StringExtras.h"
21#include "llvm/ADT/StringRef.h"
22#include "llvm/ADT/Twine.h"
23#include "llvm/Support/Compiler.h"
24#include "llvm/Support/ErrorHandling.h"
25#include "llvm/Support/MemoryBuffer.h"
26#include "llvm/Support/SMLoc.h"
27#include "llvm/Support/SourceMgr.h"
28#include "llvm/Support/Unicode.h"
29#include "llvm/Support/raw_ostream.h"
30#include <algorithm>
31#include <cassert>
32#include <cstddef>
33#include <cstdint>
34#include <map>
35#include <memory>
36#include <string>
37#include <system_error>
38#include <utility>

40using namespace llvm;
41using namespace yaml;

43enum UnicodeEncodingForm {
UEF_UTF32_LE, ///< UTF-32 Little Endian
UEF_UTF32_BE, ///< UTF-32 Big Endian
UEF_UTF16_LE, ///< UTF-16 Little Endian
UEF_UTF16_BE, ///< UTF-16 Big Endian
UEF_UTF8,     ///< UTF-8 or ascii.
UEF_Unknown   ///< Not a valid Unicode encoding.
50};

52/// EncodingInfo - Holds the encoding type and length of the byte order mark if
53///                it exists. Length is in {0, 2, 3, 4}.
54using EncodingInfo = std::pair<UnicodeEncodingForm, unsigned>;

56/// getUnicodeEncoding - Reads up to the first 4 bytes to determine the Unicode
57///                      encoding form of \a Input.
58///
59/// @param Input A string of length 0 or more.
60/// @returns An EncodingInfo indicating the Unicode encoding form of the input
61///          and how long the byte order mark is if one exists.
62static EncodingInfo getUnicodeEncoding(StringRef Input) {
if (Input.empty())
  return std::make_pair(UEF_Unknown, 0);

switch (uint8_t(Input[0])) {
case 0x00:
  if (Input.size() >= 4) {
    if (  Input[1] == 0
       && uint8_t(Input[2]) == 0xFE
       && uint8_t(Input[3]) == 0xFF)
      return std::make_pair(UEF_UTF32_BE, 4);
    if (Input[1] == 0 && Input[2] == 0 && Input[3] != 0)
      return std::make_pair(UEF_UTF32_BE, 0);
  }

  if (Input.size() >= 2 && Input[1] != 0)
    return std::make_pair(UEF_UTF16_BE, 0);
  return std::make_pair(UEF_Unknown, 0);
case 0xFF:
  if (  Input.size() >= 4
     && uint8_t(Input[1]) == 0xFE
     && Input[2] == 0
     && Input[3] == 0)
    return std::make_pair(UEF_UTF32_LE, 4);

  if (Input.size() >= 2 && uint8_t(Input[1]) == 0xFE)
    return std::make_pair(UEF_UTF16_LE, 2);
  return std::make_pair(UEF_Unknown, 0);
case 0xFE:
  if (Input.size() >= 2 && uint8_t(Input[1]) == 0xFF)
    return std::make_pair(UEF_UTF16_BE, 2);
  return std::make_pair(UEF_Unknown, 0);
case 0xEF:
  if (  Input.size() >= 3
     && uint8_t(Input[1]) == 0xBB
     && uint8_t(Input[2]) == 0xBF)
    return std::make_pair(UEF_UTF8, 3);
  return std::make_pair(UEF_Unknown, 0);
}

// It could still be utf-32 or utf-16.
if (Input.size() >= 4 && Input[1] == 0 && Input[2] == 0 && Input[3] == 0)
  return std::make_pair(UEF_UTF32_LE, 0);

if (Input.size() >= 2 && Input[1] == 0)
  return std::make_pair(UEF_UTF16_LE, 0);

return std::make_pair(UEF_UTF8, 0);
110}

112/// Pin the vtables to this file.
113void Node::anchor() {}
114void NullNode::anchor() {}
115void ScalarNode::anchor() {}
116void BlockScalarNode::anchor() {}
117void KeyValueNode::anchor() {}
118void MappingNode::anchor() {}
119void SequenceNode::anchor() {}
120void AliasNode::anchor() {}

122namespace llvm {
123namespace yaml {

125/// Token - A single YAML token.
126struct Token {
enum TokenKind {
  TK_Error, // Uninitialized token.
  TK_StreamStart,
  TK_StreamEnd,
  TK_VersionDirective,
  TK_TagDirective,
  TK_DocumentStart,
  TK_DocumentEnd,
  TK_BlockEntry,
  TK_BlockEnd,
  TK_BlockSequenceStart,
  TK_BlockMappingStart,
  TK_FlowEntry,
  TK_FlowSequenceStart,
  TK_FlowSequenceEnd,
  TK_FlowMappingStart,
  TK_FlowMappingEnd,
  TK_Key,
  TK_Value,
  TK_Scalar,
  TK_BlockScalar,
  TK_Alias,
  TK_Anchor,
  TK_Tag
} Kind = TK_Error;

/// A string of length 0 or more whose begin() points to the logical location
/// of the token in the input.
StringRef Range;

/// The value of a block scalar node.
std::string Value;

Token() = default;
161};

163} // end namespace yaml
164} // end namespace llvm

166using TokenQueueT = BumpPtrList<Token>;

168namespace {

170/// This struct is used to track simple keys.
171///
172/// Simple keys are handled by creating an entry in SimpleKeys for each Token
173/// which could legally be the start of a simple key. When peekNext is called,
174/// if the Token To be returned is referenced by a SimpleKey, we continue
175/// tokenizing until that potential simple key has either been found to not be
176/// a simple key (we moved on to the next line or went further than 1024 chars).
177/// Or when we run into a Value, and then insert a Key token (and possibly
178/// others) before the SimpleKey's Tok.
179struct SimpleKey {
TokenQueueT::iterator Tok;
unsigned Column = 0;
unsigned Line = 0;
unsigned FlowLevel = 0;
bool IsRequired = false;

bool operator ==(const SimpleKey &Other) {
  return Tok == Other.Tok;
}
189};

191} // end anonymous namespace

193/// The Unicode scalar value of a UTF-8 minimal well-formed code unit
194///        subsequence and the subsequence's length in code units (uint8_t).
195///        A length of 0 represents an error.
196using UTF8Decoded = std::pair<uint32_t, unsigned>;

198static UTF8Decoded decodeUTF8(StringRef Range) {
StringRef::iterator Position= Range.begin();
StringRef::iterator End = Range.end();
// 1 byte: [0x00, 0x7f]
// Bit pattern: 0xxxxxxx
if (Position < End && (*Position & 0x80) == 0) {
  return std::make_pair(*Position, 1);
}
// 2 bytes: [0x80, 0x7ff]
// Bit pattern: 110xxxxx 10xxxxxx
if (Position + 1 < End && ((*Position & 0xE0) == 0xC0) &&
    ((*(Position + 1) & 0xC0) == 0x80)) {
  uint32_t codepoint = ((*Position & 0x1F) << 6) |
                        (*(Position + 1) & 0x3F);
  if (codepoint >= 0x80)
    return std::make_pair(codepoint, 2);
}
// 3 bytes: [0x8000, 0xffff]
// Bit pattern: 1110xxxx 10xxxxxx 10xxxxxx
if (Position + 2 < End && ((*Position & 0xF0) == 0xE0) &&
    ((*(Position + 1) & 0xC0) == 0x80) &&
    ((*(Position + 2) & 0xC0) == 0x80)) {
  uint32_t codepoint = ((*Position & 0x0F) << 12) |
                       ((*(Position + 1) & 0x3F) << 6) |
                        (*(Position + 2) & 0x3F);
  // Codepoints between 0xD800 and 0xDFFF are invalid, as
  // they are high / low surrogate halves used by UTF-16.
  if (codepoint >= 0x800 &&
      (codepoint < 0xD800 || codepoint > 0xDFFF))
    return std::make_pair(codepoint, 3);
}
// 4 bytes: [0x10000, 0x10FFFF]
// Bit pattern: 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
if (Position + 3 < End && ((*Position & 0xF8) == 0xF0) &&
    ((*(Position + 1) & 0xC0) == 0x80) &&
    ((*(Position + 2) & 0xC0) == 0x80) &&
    ((*(Position + 3) & 0xC0) == 0x80)) {
  uint32_t codepoint = ((*Position & 0x07) << 18) |
                       ((*(Position + 1) & 0x3F) << 12) |
                       ((*(Position + 2) & 0x3F) << 6) |
                        (*(Position + 3) & 0x3F);
  if (codepoint >= 0x10000 && codepoint <= 0x10FFFF)
    return std::make_pair(codepoint, 4);
}
return std::make_pair(0, 0);
243}

245namespace llvm {
246namespace yaml {

248/// Scans YAML tokens from a MemoryBuffer.
249class Scanner {
250public:
Scanner(StringRef Input, SourceMgr &SM, bool ShowColors = true,
        std::error_code *EC = nullptr);
Scanner(MemoryBufferRef Buffer, SourceMgr &SM_, bool ShowColors = true,
        std::error_code *EC = nullptr);

/// Parse the next token and return it without popping it.
Token &peekNext();

/// Parse the next token and pop it from the queue.
Token getNext();

void printError(SMLoc Loc, SourceMgr::DiagKind Kind, const Twine &Message,
                ArrayRef<SMRange> Ranges = None) {
  SM.PrintMessage(Loc, Kind, Message, Ranges, /* FixIts= */ None, ShowColors);
}

void setError(const Twine &Message, StringRef::iterator Position) {
  if (Position >= End)
    Position = End - 1;

  // propagate the error if possible
  if (EC)
    *EC = make_error_code(std::errc::invalid_argument);

  // Don't print out more errors after the first one we encounter. The rest
  // are just the result of the first, and have no meaning.
  if (!Failed)
    printError(SMLoc::getFromPointer(Position), SourceMgr::DK_Error, Message);
  Failed = true;
}

/// Returns true if an error occurred while parsing.
bool failed() {
  return Failed;
}

287private:
void init(MemoryBufferRef Buffer);

StringRef currentInput() {
  return StringRef(Current, End - Current);
}

/// Decode a UTF-8 minimal well-formed code unit subsequence starting
///        at \a Position.
///
/// If the UTF-8 code units starting at Position do not form a well-formed
/// code unit subsequence, then the Unicode scalar value is 0, and the length
/// is 0.
UTF8Decoded decodeUTF8(StringRef::iterator Position) {
  return ::decodeUTF8(StringRef(Position, End - Position));
}

// The following functions are based on the gramar rules in the YAML spec. The
// style of the function names it meant to closely match how they are written
// in the spec. The number within the [] is the number of the grammar rule in
// the spec.
//
// See 4.2 [Production Naming Conventions] for the meaning of the prefixes.
//
// c-
//   A production starting and ending with a special character.
// b-
//   A production matching a single line break.
// nb-
//   A production starting and ending with a non-break character.
// s-
//   A production starting and ending with a white space character.
// ns-
//   A production starting and ending with a non-space character.
// l-
//   A production matching complete line(s).

/// Skip a single nb-char[27] starting at Position.
///
/// A nb-char is 0x9 | [0x20-0x7E] | 0x85 | [0xA0-0xD7FF] | [0xE000-0xFEFE]
///                  | [0xFF00-0xFFFD] | [0x10000-0x10FFFF]
///
/// @returns The code unit after the nb-char, or Position if it's not an
///          nb-char.
StringRef::iterator skip_nb_char(StringRef::iterator Position);

/// Skip a single b-break[28] starting at Position.
///
/// A b-break is 0xD 0xA | 0xD | 0xA
///
/// @returns The code unit after the b-break, or Position if it's not a
///          b-break.
StringRef::iterator skip_b_break(StringRef::iterator Position);

/// Skip a single s-space[31] starting at Position.
///
/// An s-space is 0x20
///
/// @returns The code unit after the s-space, or Position if it's not a
///          s-space.
StringRef::iterator skip_s_space(StringRef::iterator Position);

/// Skip a single s-white[33] starting at Position.
///
/// A s-white is 0x20 | 0x9
///
/// @returns The code unit after the s-white, or Position if it's not a
///          s-white.
StringRef::iterator skip_s_white(StringRef::iterator Position);

/// Skip a single ns-char[34] starting at Position.
///
/// A ns-char is nb-char - s-white
///
/// @returns The code unit after the ns-char, or Position if it's not a
///          ns-char.
StringRef::iterator skip_ns_char(StringRef::iterator Position);

using SkipWhileFunc = StringRef::iterator (Scanner::*)(StringRef::iterator);

/// Skip minimal well-formed code unit subsequences until Func
///        returns its input.
///
/// @returns The code unit after the last minimal well-formed code unit
///          subsequence that Func accepted.
StringRef::iterator skip_while( SkipWhileFunc Func
                              , StringRef::iterator Position);

/// Skip minimal well-formed code unit subsequences until Func returns its
/// input.
void advanceWhile(SkipWhileFunc Func);

/// Scan ns-uri-char[39]s starting at Cur.
///
/// This updates Cur and Column while scanning.
void scan_ns_uri_char();

/// Consume a minimal well-formed code unit subsequence starting at
///        \a Cur. Return false if it is not the same Unicode scalar value as
///        \a Expected. This updates \a Column.
bool consume(uint32_t Expected);

/// Skip \a Distance UTF-8 code units. Updates \a Cur and \a Column.
void skip(uint32_t Distance);

/// Return true if the minimal well-formed code unit subsequence at
///        Pos is whitespace or a new line
bool isBlankOrBreak(StringRef::iterator Position);

/// Consume a single b-break[28] if it's present at the current position.
///
/// Return false if the code unit at the current position isn't a line break.
bool consumeLineBreakIfPresent();

/// If IsSimpleKeyAllowed, create and push_back a new SimpleKey.
void saveSimpleKeyCandidate( TokenQueueT::iterator Tok
                           , unsigned AtColumn
                           , bool IsRequired);

/// Remove simple keys that can no longer be valid simple keys.
///
/// Invalid simple keys are not on the current line or are further than 1024
/// columns back.
void removeStaleSimpleKeyCandidates();

/// Remove all simple keys on FlowLevel \a Level.
void removeSimpleKeyCandidatesOnFlowLevel(unsigned Level);

/// Unroll indentation in \a Indents back to \a Col. Creates BlockEnd
///        tokens if needed.
bool unrollIndent(int ToColumn);

/// Increase indent to \a Col. Creates \a Kind token at \a InsertPoint
///        if needed.
bool rollIndent( int ToColumn
               , Token::TokenKind Kind
               , TokenQueueT::iterator InsertPoint);

/// Skip a single-line comment when the comment starts at the current
/// position of the scanner.
void skipComment();

/// Skip whitespace and comments until the start of the next token.
void scanToNextToken();

/// Must be the first token generated.
bool scanStreamStart();

/// Generate tokens needed to close out the stream.
bool scanStreamEnd();

/// Scan a %BLAH directive.
bool scanDirective();

/// Scan a ... or ---.
bool scanDocumentIndicator(bool IsStart);

/// Scan a [ or { and generate the proper flow collection start token.
bool scanFlowCollectionStart(bool IsSequence);

/// Scan a ] or } and generate the proper flow collection end token.
bool scanFlowCollectionEnd(bool IsSequence);

/// Scan the , that separates entries in a flow collection.
bool scanFlowEntry();

/// Scan the - that starts block sequence entries.
bool scanBlockEntry();

/// Scan an explicit ? indicating a key.
bool scanKey();

/// Scan an explicit : indicating a value.
bool scanValue();

/// Scan a quoted scalar.
bool scanFlowScalar(bool IsDoubleQuoted);

/// Scan an unquoted scalar.
bool scanPlainScalar();

/// Scan an Alias or Anchor starting with * or &.
bool scanAliasOrAnchor(bool IsAlias);

/// Scan a block scalar starting with | or >.
bool scanBlockScalar(bool IsLiteral);

/// Scan a chomping indicator in a block scalar header.
char scanBlockChompingIndicator();

/// Scan an indentation indicator in a block scalar header.
unsigned scanBlockIndentationIndicator();

/// Scan a block scalar header.
///
/// Return false if an error occurred.
bool scanBlockScalarHeader(char &ChompingIndicator, unsigned &IndentIndicator,
                           bool &IsDone);

/// Look for the indentation level of a block scalar.
///
/// Return false if an error occurred.
bool findBlockScalarIndent(unsigned &BlockIndent, unsigned BlockExitIndent,
                           unsigned &LineBreaks, bool &IsDone);

/// Scan the indentation of a text line in a block scalar.
///
/// Return false if an error occurred.
bool scanBlockScalarIndent(unsigned BlockIndent, unsigned BlockExitIndent,
                           bool &IsDone);

/// Scan a tag of the form !stuff.
bool scanTag();

/// Dispatch to the next scanning function based on \a *Cur.
bool fetchMoreTokens();

/// The SourceMgr used for diagnostics and buffer management.
SourceMgr &SM;

/// The original input.
MemoryBufferRef InputBuffer;

/// The current position of the scanner.
StringRef::iterator Current;

/// The end of the input (one past the last character).
StringRef::iterator End;

/// Current YAML indentation level in spaces.
int Indent;

/// Current column number in Unicode code points.
unsigned Column;

/// Current line number.
unsigned Line;

/// How deep we are in flow style containers. 0 Means at block level.
unsigned FlowLevel;

/// Are we at the start of the stream?
bool IsStartOfStream;

/// Can the next token be the start of a simple key?
bool IsSimpleKeyAllowed;

/// True if an error has occurred.
bool Failed;

/// Should colors be used when printing out the diagnostic messages?
bool ShowColors;

/// Queue of tokens. This is required to queue up tokens while looking
///        for the end of a simple key. And for cases where a single character
///        can produce multiple tokens (e.g. BlockEnd).
TokenQueueT TokenQueue;

/// Indentation levels.
SmallVector<int, 4> Indents;

/// Potential simple keys.
SmallVector<SimpleKey, 4> SimpleKeys;

std::error_code *EC;
552};

554} // end namespace yaml
555} // end namespace llvm

557/// encodeUTF8 - Encode \a UnicodeScalarValue in UTF-8 and append it to result.
558static void encodeUTF8( uint32_t UnicodeScalarValue
                    , SmallVectorImpl<char> &Result) {
if (UnicodeScalarValue <= 0x7F) {
  Result.push_back(UnicodeScalarValue & 0x7F);
} else if (UnicodeScalarValue <= 0x7FF) {
  uint8_t FirstByte = 0xC0 | ((UnicodeScalarValue & 0x7C0) >> 6);
  uint8_t SecondByte = 0x80 | (UnicodeScalarValue & 0x3F);
  Result.push_back(FirstByte);
  Result.push_back(SecondByte);
} else if (UnicodeScalarValue <= 0xFFFF) {
  uint8_t FirstByte = 0xE0 | ((UnicodeScalarValue & 0xF000) >> 12);
  uint8_t SecondByte = 0x80 | ((UnicodeScalarValue & 0xFC0) >> 6);
  uint8_t ThirdByte = 0x80 | (UnicodeScalarValue & 0x3F);
  Result.push_back(FirstByte);
  Result.push_back(SecondByte);
  Result.push_back(ThirdByte);
} else if (UnicodeScalarValue <= 0x10FFFF) {
  uint8_t FirstByte = 0xF0 | ((UnicodeScalarValue & 0x1F0000) >> 18);
  uint8_t SecondByte = 0x80 | ((UnicodeScalarValue & 0x3F000) >> 12);
  uint8_t ThirdByte = 0x80 | ((UnicodeScalarValue & 0xFC0) >> 6);
  uint8_t FourthByte = 0x80 | (UnicodeScalarValue & 0x3F);
  Result.push_back(FirstByte);
  Result.push_back(SecondByte);
  Result.push_back(ThirdByte);
  Result.push_back(FourthByte);
}
584}

586bool yaml::dumpTokens(StringRef Input, raw_ostream &OS) {
SourceMgr SM;
Scanner scanner(Input, SM);
while (true) {
  Token T = scanner.getNext();
  switch (T.Kind) {
  case Token::TK_StreamStart:
    OS << "Stream-Start: ";
    break;
  case Token::TK_StreamEnd:
    OS << "Stream-End: ";
    break;
  case Token::TK_VersionDirective:
    OS << "Version-Directive: ";
    break;
  case Token::TK_TagDirective:
    OS << "Tag-Directive: ";
    break;
  case Token::TK_DocumentStart:
    OS << "Document-Start: ";
    break;
  case Token::TK_DocumentEnd:
    OS << "Document-End: ";
    break;
  case Token::TK_BlockEntry:
    OS << "Block-Entry: ";
    break;
  case Token::TK_BlockEnd:
    OS << "Block-End: ";
    break;
  case Token::TK_BlockSequenceStart:
    OS << "Block-Sequence-Start: ";
    break;
  case Token::TK_BlockMappingStart:
    OS << "Block-Mapping-Start: ";
    break;
  case Token::TK_FlowEntry:
    OS << "Flow-Entry: ";
    break;
  case Token::TK_FlowSequenceStart:
    OS << "Flow-Sequence-Start: ";
    break;
  case Token::TK_FlowSequenceEnd:
    OS << "Flow-Sequence-End: ";
    break;
  case Token::TK_FlowMappingStart:
    OS << "Flow-Mapping-Start: ";
    break;
  case Token::TK_FlowMappingEnd:
    OS << "Flow-Mapping-End: ";
    break;
  case Token::TK_Key:
    OS << "Key: ";
    break;
  case Token::TK_Value:
    OS << "Value: ";
    break;
  case Token::TK_Scalar:
    OS << "Scalar: ";
    break;
  case Token::TK_BlockScalar:
    OS << "Block Scalar: ";
    break;
  case Token::TK_Alias:
    OS << "Alias: ";
    break;
  case Token::TK_Anchor:
    OS << "Anchor: ";
    break;
  case Token::TK_Tag:
    OS << "Tag: ";
    break;
  case Token::TK_Error:
    break;
  }
  OS << T.Range << "\n";
  if (T.Kind == Token::TK_StreamEnd)
    break;
  else if (T.Kind == Token::TK_Error)
    return false;
}
return true;
668}

670bool yaml::scanTokens(StringRef Input) {
SourceMgr SM;
Scanner scanner(Input, SM);
while (true) {
  Token T = scanner.getNext();
  if (T.Kind == Token::TK_StreamEnd)
    break;
  else if (T.Kind == Token::TK_Error)
    return false;
}
return true;
681}

683std::string yaml::escape(StringRef Input, bool EscapePrintable) {
std::string EscapedInput;
for (StringRef::iterator i = Input.begin(), e = Input.end(); i != e; ++i) {
  if (*i == '\\')
    EscapedInput += "\\\\";
  else if (*i == '"')
    EscapedInput += "\\\"";
  else if (*i == 0)
    EscapedInput += "\\0";
  else if (*i == 0x07)
    EscapedInput += "\\a";
  else if (*i == 0x08)
    EscapedInput += "\\b";
  else if (*i == 0x09)
    EscapedInput += "\\t";
  else if (*i == 0x0A)
    EscapedInput += "\\n";
  else if (*i == 0x0B)
    EscapedInput += "\\v";
  else if (*i == 0x0C)
    EscapedInput += "\\f";
  else if (*i == 0x0D)
    EscapedInput += "\\r";
  else if (*i == 0x1B)
    EscapedInput += "\\e";
  else if ((unsigned char)*i < 0x20) { // Control characters not handled above.
    std::string HexStr = utohexstr(*i);
    EscapedInput += "\\x" + std::string(2 - HexStr.size(), '0') + HexStr;
  } else if (*i & 0x80) { // UTF-8 multiple code unit subsequence.
    UTF8Decoded UnicodeScalarValue
      = decodeUTF8(StringRef(i, Input.end() - i));
    if (UnicodeScalarValue.second == 0) {
      // Found invalid char.
      SmallString<4> Val;
      encodeUTF8(0xFFFD, Val);
      llvm::append_range(EscapedInput, Val);
      // FIXME: Error reporting.
      return EscapedInput;
    }
    if (UnicodeScalarValue.first == 0x85)
      EscapedInput += "\\N";
    else if (UnicodeScalarValue.first == 0xA0)
      EscapedInput += "\\_";
    else if (UnicodeScalarValue.first == 0x2028)
      EscapedInput += "\\L";
    else if (UnicodeScalarValue.first == 0x2029)
      EscapedInput += "\\P";
    else if (!EscapePrintable &&
             sys::unicode::isPrintable(UnicodeScalarValue.first))
      EscapedInput += StringRef(i, UnicodeScalarValue.second);
    else {
      std::string HexStr = utohexstr(UnicodeScalarValue.first);
      if (HexStr.size() <= 2)
        EscapedInput += "\\x" + std::string(2 - HexStr.size(), '0') + HexStr;
      else if (HexStr.size() <= 4)
        EscapedInput += "\\u" + std::string(4 - HexStr.size(), '0') + HexStr;
      else if (HexStr.size() <= 8)
        EscapedInput += "\\U" + std::string(8 - HexStr.size(), '0') + HexStr;
    }
    i += UnicodeScalarValue.second - 1;
  } else
    EscapedInput.push_back(*i);
}
return EscapedInput;
747}

749llvm::Optional<bool> yaml::parseBool(StringRef S) {
switch (S.size()) {
case 1:
  switch (S.front()) {
  case 'y':
  case 'Y':
    return true;
  case 'n':
  case 'N':
    return false;
  default:
    return None;
  }
case 2:
  switch (S.front()) {
  case 'O':
    if (S[1] == 'N') // ON
      return true;
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case 'o':
    if (S[1] == 'n') //[Oo]n
      return true;
    return None;
  case 'N':
    if (S[1] == 'O') // NO
      return false;
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case 'n':
    if (S[1] == 'o') //[Nn]o
      return false;
    return None;
  default:
    return None;
  }
case 3:
  switch (S.front()) {
  case 'O':
    if (S.drop_front() == "FF") // OFF
      return false;
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case 'o':
    if (S.drop_front() == "ff") //[Oo]ff
      return false;
    return None;
  case 'Y':
    if (S.drop_front() == "ES") // YES
      return true;
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case 'y':
    if (S.drop_front() == "es") //[Yy]es
      return true;
    return None;
  default:
    return None;
  }
case 4:
  switch (S.front()) {
  case 'T':
    if (S.drop_front() == "RUE") // TRUE
      return true;
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case 't':
    if (S.drop_front() == "rue") //[Tt]rue
      return true;
    return None;
  default:
    return None;
  }
case 5:
  switch (S.front()) {
  case 'F':
    if (S.drop_front() == "ALSE") // FALSE
      return false;
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case 'f':
    if (S.drop_front() == "alse") //[Ff]alse
      return false;
    return None;
  default:
    return None;
  }
default:
  return None;
}
833}

835Scanner::Scanner(StringRef Input, SourceMgr &sm, bool ShowColors,
               std::error_code *EC)
  : SM(sm), ShowColors(ShowColors), EC(EC) {
init(MemoryBufferRef(Input, "YAML"));
839}

841Scanner::Scanner(MemoryBufferRef Buffer, SourceMgr &SM_, bool ShowColors,
               std::error_code *EC)
  : SM(SM_), ShowColors(ShowColors), EC(EC) {
init(Buffer);
845}

847void Scanner::init(MemoryBufferRef Buffer) {
InputBuffer = Buffer;
Current = InputBuffer.getBufferStart();
End = InputBuffer.getBufferEnd();
Indent = -1;
Column = 0;
Line = 0;
FlowLevel = 0;
IsStartOfStream = true;
IsSimpleKeyAllowed = true;
Failed = false;
std::unique_ptr<MemoryBuffer> InputBufferOwner =
    MemoryBuffer::getMemBuffer(Buffer, /*RequiresNullTerminator=*/false);
SM.AddNewSourceBuffer(std::move(InputBufferOwner), SMLoc());
861}

863Token &Scanner::peekNext() {
// If the current token is a possible simple key, keep parsing until we
// can confirm.
bool NeedMore = false;
while (true) {
  if (TokenQueue.empty() || NeedMore) {
    if (!fetchMoreTokens()) {
      TokenQueue.clear();
      SimpleKeys.clear();
      TokenQueue.push_back(Token());
      return TokenQueue.front();
    }
  }
  assert(!TokenQueue.empty() &&(static_cast <bool> (!TokenQueue.empty() && "fetchMoreTokens lied about getting tokens!"
) ? void (0) : __assert_fail ("!TokenQueue.empty() && \"fetchMoreTokens lied about getting tokens!\""
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 877, __extension__ __PRETTY_FUNCTION__))
          "fetchMoreTokens lied about getting tokens!")(static_cast <bool> (!TokenQueue.empty() && "fetchMoreTokens lied about getting tokens!"
) ? void (0) : __assert_fail ("!TokenQueue.empty() && \"fetchMoreTokens lied about getting tokens!\""
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 877, __extension__ __PRETTY_FUNCTION__));

  removeStaleSimpleKeyCandidates();
  SimpleKey SK;
  SK.Tok = TokenQueue.begin();
  if (!is_contained(SimpleKeys, SK))
    break;
  else
    NeedMore = true;
}
return TokenQueue.front();
888}

890Token Scanner::getNext() {
Token Ret = peekNext();
// TokenQueue can be empty if there was an error getting the next token.
if (!TokenQueue.empty())
  TokenQueue.pop_front();

// There cannot be any referenced Token's if the TokenQueue is empty. So do a
// quick deallocation of them all.
if (TokenQueue.empty())
  TokenQueue.resetAlloc();

return Ret;
902}

904StringRef::iterator Scanner::skip_nb_char(StringRef::iterator Position) {
if (Position == End)
  return Position;
// Check 7 bit c-printable - b-char.
if (   *Position == 0x09
    || (*Position >= 0x20 && *Position <= 0x7E))
  return Position + 1;

// Check for valid UTF-8.
if (uint8_t(*Position) & 0x80) {
  UTF8Decoded u8d = decodeUTF8(Position);
  if (   u8d.second != 0
      && u8d.first != 0xFEFF
      && ( u8d.first == 0x85
        || ( u8d.first >= 0xA0
          && u8d.first <= 0xD7FF)
        || ( u8d.first >= 0xE000
          && u8d.first <= 0xFFFD)
        || ( u8d.first >= 0x10000
          && u8d.first <= 0x10FFFF)))
    return Position + u8d.second;
}
return Position;
927}

929StringRef::iterator Scanner::skip_b_break(StringRef::iterator Position) {
if (Position == End)
  return Position;
if (*Position == 0x0D) {
  if (Position + 1 != End && *(Position + 1) == 0x0A)
    return Position + 2;
  return Position + 1;
}

if (*Position == 0x0A)
  return Position + 1;
return Position;
941}

943StringRef::iterator Scanner::skip_s_space(StringRef::iterator Position) {
if (Position == End)
  return Position;
if (*Position == ' ')
  return Position + 1;
return Position;
949}

951StringRef::iterator Scanner::skip_s_white(StringRef::iterator Position) {
if (Position == End)
  return Position;
if (*Position == ' ' || *Position == '\t')
  return Position + 1;
return Position;
957}

959StringRef::iterator Scanner::skip_ns_char(StringRef::iterator Position) {
if (Position == End)
  return Position;
if (*Position == ' ' || *Position == '\t')
  return Position;
return skip_nb_char(Position);
965}

967StringRef::iterator Scanner::skip_while( SkipWhileFunc Func
                                     , StringRef::iterator Position) {
while (true) {
  StringRef::iterator i = (this->*Func)(Position);
  if (i == Position)
    break;
  Position = i;
}
return Position;
976}

978void Scanner::advanceWhile(SkipWhileFunc Func) {
auto Final = skip_while(Func, Current);
Column += Final - Current;
Current = Final;
982}

984static bool is_ns_hex_digit(const char C) { return isAlnum(C); }

986static bool is_ns_word_char(const char C) { return C == '-' || isAlpha(C); }

988void Scanner::scan_ns_uri_char() {
while (true) {
  if (Current == End)
    break;
  if ((   *Current == '%'
        && Current + 2 < End
        && is_ns_hex_digit(*(Current + 1))
        && is_ns_hex_digit(*(Current + 2)))
      || is_ns_word_char(*Current)
      || StringRef(Current, 1).find_first_of("#;/?:@&=+$,_.!~*'()[]")
        != StringRef::npos) {
    ++Current;
    ++Column;
  } else
    break;
}
1004}

1006bool Scanner::consume(uint32_t Expected) {
if (Expected >= 0x80) {
  setError("Cannot consume non-ascii characters", Current);
  return false;
}
if (Current == End)
  return false;
if (uint8_t(*Current) >= 0x80) {
  setError("Cannot consume non-ascii characters", Current);
  return false;
}
if (uint8_t(*Current) == Expected) {
  ++Current;
  ++Column;
  return true;
}
return false;
1023}

1025void Scanner::skip(uint32_t Distance) {
Current += Distance;
Column += Distance;
assert(Current <= End && "Skipped past the end")(static_cast <bool> (Current <= End && "Skipped past the end"
) ? void (0) : __assert_fail ("Current <= End && \"Skipped past the end\""
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 1028, __extension__ __PRETTY_FUNCTION__));
1029}

1031bool Scanner::isBlankOrBreak(StringRef::iterator Position) {
if (Position == End)
  return false;
return *Position == ' ' || *Position == '\t' || *Position == '\r' ||
       *Position == '\n';
1036}

1038bool Scanner::consumeLineBreakIfPresent() {
auto Next = skip_b_break(Current);
if (Next == Current)
  return false;
Column = 0;
++Line;
Current = Next;
return true;
1046}

1048void Scanner::saveSimpleKeyCandidate( TokenQueueT::iterator Tok
                                  , unsigned AtColumn
                                  , bool IsRequired) {
if (IsSimpleKeyAllowed) {
  SimpleKey SK;
  SK.Tok = Tok;
  SK.Line = Line;
  SK.Column = AtColumn;
  SK.IsRequired = IsRequired;
  SK.FlowLevel = FlowLevel;
  SimpleKeys.push_back(SK);
}
1060}

1062void Scanner::removeStaleSimpleKeyCandidates() {
for (SmallVectorImpl<SimpleKey>::iterator i = SimpleKeys.begin();
                                          i != SimpleKeys.end();) {
  if (i->Line != Line || i->Column + 1024 < Column) {
    if (i->IsRequired)
      setError( "Could not find expected : for simple key"
              , i->Tok->Range.begin());
    i = SimpleKeys.erase(i);
  } else
    ++i;
}
1073}

1075void Scanner::removeSimpleKeyCandidatesOnFlowLevel(unsigned Level) {
if (!SimpleKeys.empty() && (SimpleKeys.end() - 1)->FlowLevel == Level)
  SimpleKeys.pop_back();
1078}

1080bool Scanner::unrollIndent(int ToColumn) {
Token T;
// Indentation is ignored in flow.
if (FlowLevel != 0)
  return true;

while (Indent > ToColumn) {
  T.Kind = Token::TK_BlockEnd;
  T.Range = StringRef(Current, 1);
  TokenQueue.push_back(T);
  Indent = Indents.pop_back_val();
}

return true;
1094}

1096bool Scanner::rollIndent( int ToColumn
                      , Token::TokenKind Kind
                      , TokenQueueT::iterator InsertPoint) {
if (FlowLevel)
  return true;
if (Indent < ToColumn) {
  Indents.push_back(Indent);
  Indent = ToColumn;

  Token T;
  T.Kind = Kind;
  T.Range = StringRef(Current, 0);
  TokenQueue.insert(InsertPoint, T);
}
return true;
1111}

1113void Scanner::skipComment() {
if (Current == End || *Current != '#')
  return;
while (true) {
  // This may skip more than one byte, thus Column is only incremented
  // for code points.
  StringRef::iterator I = skip_nb_char(Current);
  if (I == Current)
    break;
  Current = I;
  ++Column;
}
1125}

1127void Scanner::scanToNextToken() {
while (true) {
  while (Current != End && (*Current == ' ' || *Current == '\t')) {
    skip(1);
  }

  skipComment();

  // Skip EOL.
  StringRef::iterator i = skip_b_break(Current);
  if (i == Current)
    break;
  Current = i;
  ++Line;
  Column = 0;
  // New lines may start a simple key.
  if (!FlowLevel)
    IsSimpleKeyAllowed = true;
}
1146}

1148bool Scanner::scanStreamStart() {
IsStartOfStream = false;

EncodingInfo EI = getUnicodeEncoding(currentInput());

Token T;
T.Kind = Token::TK_StreamStart;
T.Range = StringRef(Current, EI.second);
TokenQueue.push_back(T);
Current += EI.second;
return true;
1159}

1161bool Scanner::scanStreamEnd() {
// Force an ending new line if one isn't present.
if (Column != 0) {
  Column = 0;
  ++Line;
}

unrollIndent(-1);
SimpleKeys.clear();
IsSimpleKeyAllowed = false;

Token T;
T.Kind = Token::TK_StreamEnd;
T.Range = StringRef(Current, 0);
TokenQueue.push_back(T);
return true;
1177}

1179bool Scanner::scanDirective() {
// Reset the indentation level.
unrollIndent(-1);
SimpleKeys.clear();
IsSimpleKeyAllowed = false;

StringRef::iterator Start = Current;
consume('%');
StringRef::iterator NameStart = Current;
Current = skip_while(&Scanner::skip_ns_char, Current);
StringRef Name(NameStart, Current - NameStart);
Current = skip_while(&Scanner::skip_s_white, Current);

Token T;
if (Name == "YAML") {
  Current = skip_while(&Scanner::skip_ns_char, Current);
  T.Kind = Token::TK_VersionDirective;
  T.Range = StringRef(Start, Current - Start);
  TokenQueue.push_back(T);
  return true;
} else if(Name == "TAG") {
  Current = skip_while(&Scanner::skip_ns_char, Current);
  Current = skip_while(&Scanner::skip_s_white, Current);
  Current = skip_while(&Scanner::skip_ns_char, Current);
  T.Kind = Token::TK_TagDirective;
  T.Range = StringRef(Start, Current - Start);
  TokenQueue.push_back(T);
  return true;
}
return false;
1209}

1211bool Scanner::scanDocumentIndicator(bool IsStart) {
unrollIndent(-1);
SimpleKeys.clear();
IsSimpleKeyAllowed = false;

Token T;
T.Kind = IsStart ? Token::TK_DocumentStart : Token::TK_DocumentEnd;
T.Range = StringRef(Current, 3);
skip(3);
TokenQueue.push_back(T);
return true;
1222}

1224bool Scanner::scanFlowCollectionStart(bool IsSequence) {
Token T;
T.Kind = IsSequence ? Token::TK_FlowSequenceStart
                    : Token::TK_FlowMappingStart;
T.Range = StringRef(Current, 1);
skip(1);
TokenQueue.push_back(T);

// [ and { may begin a simple key.
saveSimpleKeyCandidate(--TokenQueue.end(), Column - 1, false);

// And may also be followed by a simple key.
IsSimpleKeyAllowed = true;
++FlowLevel;
return true;
1239}

1241bool Scanner::scanFlowCollectionEnd(bool IsSequence) {
removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
IsSimpleKeyAllowed = false;
Token T;
T.Kind = IsSequence ? Token::TK_FlowSequenceEnd
                    : Token::TK_FlowMappingEnd;
T.Range = StringRef(Current, 1);
skip(1);
TokenQueue.push_back(T);
if (FlowLevel)
  --FlowLevel;
return true;
1253}

1255bool Scanner::scanFlowEntry() {
removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
IsSimpleKeyAllowed = true;
Token T;
T.Kind = Token::TK_FlowEntry;
T.Range = StringRef(Current, 1);
skip(1);
TokenQueue.push_back(T);
return true;
1264}

1266bool Scanner::scanBlockEntry() {
rollIndent(Column, Token::TK_BlockSequenceStart, TokenQueue.end());
removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
IsSimpleKeyAllowed = true;
Token T;
T.Kind = Token::TK_BlockEntry;
T.Range = StringRef(Current, 1);
skip(1);
TokenQueue.push_back(T);
return true;
1276}

1278bool Scanner::scanKey() {
if (!FlowLevel)
  rollIndent(Column, Token::TK_BlockMappingStart, TokenQueue.end());

removeSimpleKeyCandidatesOnFlowLevel(FlowLevel);
IsSimpleKeyAllowed = !FlowLevel;

Token T;
T.Kind = Token::TK_Key;
T.Range = StringRef(Current, 1);
skip(1);
TokenQueue.push_back(T);
return true;
1291}

1293bool Scanner::scanValue() {
// If the previous token could have been a simple key, insert the key token
// into the token queue.
if (!SimpleKeys.empty()) {
  SimpleKey SK = SimpleKeys.pop_back_val();
  Token T;
  T.Kind = Token::TK_Key;
  T.Range = SK.Tok->Range;
  TokenQueueT::iterator i, e;
  for (i = TokenQueue.begin(), e = TokenQueue.end(); i != e; ++i) {
    if (i == SK.Tok)
      break;
  }
  if (i == e) {
    Failed = true;
    return false;
  }
  i = TokenQueue.insert(i, T);

  // We may also need to add a Block-Mapping-Start token.
  rollIndent(SK.Column, Token::TK_BlockMappingStart, i);

  IsSimpleKeyAllowed = false;
} else {
  if (!FlowLevel)
    rollIndent(Column, Token::TK_BlockMappingStart, TokenQueue.end());
  IsSimpleKeyAllowed = !FlowLevel;
}

Token T;
T.Kind = Token::TK_Value;
T.Range = StringRef(Current, 1);
skip(1);
TokenQueue.push_back(T);
return true;
1328}

1330// Forbidding inlining improves performance by roughly 20%.
1331// FIXME: Remove once llvm optimizes this to the faster version without hints.
1332LLVM_ATTRIBUTE_NOINLINE__attribute__((noinline)) static bool
1333wasEscaped(StringRef::iterator First, StringRef::iterator Position);

1335// Returns whether a character at 'Position' was escaped with a leading '\'.
1336// 'First' specifies the position of the first character in the string.
1337static bool wasEscaped(StringRef::iterator First,
                     StringRef::iterator Position) {
assert(Position - 1 >= First)(static_cast <bool> (Position - 1 >= First) ? void (
0) : __assert_fail ("Position - 1 >= First", "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 1339, __extension__ __PRETTY_FUNCTION__));
StringRef::iterator I = Position - 1;
// We calculate the number of consecutive '\'s before the current position
// by iterating backwards through our string.
while (I >= First && *I == '\\') --I;
// (Position - 1 - I) now contains the number of '\'s before the current
// position. If it is odd, the character at 'Position' was escaped.
return (Position - 1 - I) % 2 == 1;
1347}

1349bool Scanner::scanFlowScalar(bool IsDoubleQuoted) {
StringRef::iterator Start = Current;
unsigned ColStart = Column;
if (IsDoubleQuoted) {
  do {
    ++Current;
    while (Current != End && *Current != '"')
      ++Current;
    // Repeat until the previous character was not a '\' or was an escaped
    // backslash.
  } while (   Current != End
           && *(Current - 1) == '\\'
           && wasEscaped(Start + 1, Current));
} else {
  skip(1);
  while (Current != End) {
    // Skip a ' followed by another '.
    if (Current + 1 < End && *Current == '\'' && *(Current + 1) == '\'') {
      skip(2);
      continue;
    } else if (*Current == '\'')
      break;
    StringRef::iterator i = skip_nb_char(Current);
    if (i == Current) {
      i = skip_b_break(Current);
      if (i == Current)
        break;
      Current = i;
      Column = 0;
      ++Line;
    } else {
      if (i == End)
        break;
      Current = i;
      ++Column;
    }
  }
}

if (Current == End) {
  setError("Expected quote at end of scalar", Current);
  return false;
}

skip(1); // Skip ending quote.
Token T;
T.Kind = Token::TK_Scalar;
T.Range = StringRef(Start, Current - Start);
TokenQueue.push_back(T);

saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);

IsSimpleKeyAllowed = false;

return true;
1404}

1406bool Scanner::scanPlainScalar() {
StringRef::iterator Start = Current;
unsigned ColStart = Column;
unsigned LeadingBlanks = 0;
assert(Indent >= -1 && "Indent must be >= -1 !")(static_cast <bool> (Indent >= -1 && "Indent must be >= -1 !"
) ? void (0) : __assert_fail ("Indent >= -1 && \"Indent must be >= -1 !\""
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 1410, __extension__ __PRETTY_FUNCTION__));
unsigned indent = static_cast<unsigned>(Indent + 1);
while (Current != End) {
  if (*Current == '#')
    break;

  while (Current != End && !isBlankOrBreak(Current)) {
    if (FlowLevel && *Current == ':' &&
        (Current + 1 == End ||
         !(isBlankOrBreak(Current + 1) || *(Current + 1) == ','))) {
      setError("Found unexpected ':' while scanning a plain scalar", Current);
      return false;
    }

    // Check for the end of the plain scalar.
    if (  (*Current == ':' && isBlankOrBreak(Current + 1))
        || (  FlowLevel
        && (StringRef(Current, 1).find_first_of(",:?[]{}")
            != StringRef::npos)))
      break;

    StringRef::iterator i = skip_nb_char(Current);
    if (i == Current)
      break;
    Current = i;
    ++Column;
  }

  // Are we at the end?
  if (!isBlankOrBreak(Current))
    break;

  // Eat blanks.
  StringRef::iterator Tmp = Current;
  while (isBlankOrBreak(Tmp)) {
    StringRef::iterator i = skip_s_white(Tmp);
    if (i != Tmp) {
      if (LeadingBlanks && (Column < indent) && *Tmp == '\t') {
        setError("Found invalid tab character in indentation", Tmp);
        return false;
      }
      Tmp = i;
      ++Column;
    } else {
      i = skip_b_break(Tmp);
      if (!LeadingBlanks)
        LeadingBlanks = 1;
      Tmp = i;
      Column = 0;
      ++Line;
    }
  }

  if (!FlowLevel && Column < indent)
    break;

  Current = Tmp;
}
if (Start == Current) {
  setError("Got empty plain scalar", Start);
  return false;
}
Token T;
T.Kind = Token::TK_Scalar;
T.Range = StringRef(Start, Current - Start);
TokenQueue.push_back(T);

// Plain scalars can be simple keys.
saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);

IsSimpleKeyAllowed = false;

return true;
1483}

1485bool Scanner::scanAliasOrAnchor(bool IsAlias) {
StringRef::iterator Start = Current;
unsigned ColStart = Column;
skip(1);
while (Current != End) {
  if (   *Current == '[' || *Current == ']'
      || *Current == '{' || *Current == '}'
      || *Current == ','
      || *Current == ':')
    break;
  StringRef::iterator i = skip_ns_char(Current);
  if (i == Current)
    break;
  Current = i;
  ++Column;
}

if (Start + 1 == Current) {
  setError("Got empty alias or anchor", Start);
  return false;
}

Token T;
T.Kind = IsAlias ? Token::TK_Alias : Token::TK_Anchor;
T.Range = StringRef(Start, Current - Start);
TokenQueue.push_back(T);

// Alias and anchors can be simple keys.
saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);

IsSimpleKeyAllowed = false;

return true;
1518}

1520char Scanner::scanBlockChompingIndicator() {
char Indicator = ' ';
if (Current != End && (*Current == '+' || *Current == '-')) {
  Indicator = *Current;
  skip(1);
}
return Indicator;
1527}

1529/// Get the number of line breaks after chomping.
1530///
1531/// Return the number of trailing line breaks to emit, depending on
1532/// \p ChompingIndicator.
1533static unsigned getChompedLineBreaks(char ChompingIndicator,
                                   unsigned LineBreaks, StringRef Str) {
if (ChompingIndicator == '-') // Strip all line breaks.
  return 0;
if (ChompingIndicator == '+') // Keep all line breaks.
  return LineBreaks;
// Clip trailing lines.
return Str.empty() ? 0 : 1;
1541}

1543unsigned Scanner::scanBlockIndentationIndicator() {
unsigned Indent = 0;
if (Current != End && (*Current >= '1' && *Current <= '9')) {
  Indent = unsigned(*Current - '0');
  skip(1);
}
return Indent;
1550}

1552bool Scanner::scanBlockScalarHeader(char &ChompingIndicator,
                                  unsigned &IndentIndicator, bool &IsDone) {
auto Start = Current;

ChompingIndicator = scanBlockChompingIndicator();
IndentIndicator = scanBlockIndentationIndicator();
// Check for the chomping indicator once again.
if (ChompingIndicator == ' ')
  ChompingIndicator = scanBlockChompingIndicator();
Current = skip_while(&Scanner::skip_s_white, Current);
skipComment();

if (Current == End) { // EOF, we have an empty scalar.
  Token T;
  T.Kind = Token::TK_BlockScalar;
  T.Range = StringRef(Start, Current - Start);
  TokenQueue.push_back(T);
  IsDone = true;
  return true;
}

if (!consumeLineBreakIfPresent()) {
  setError("Expected a line break after block scalar header", Current);
  return false;
}
return true;
1578}

1580bool Scanner::findBlockScalarIndent(unsigned &BlockIndent,
                                  unsigned BlockExitIndent,
                                  unsigned &LineBreaks, bool &IsDone) {
unsigned MaxAllSpaceLineCharacters = 0;
StringRef::iterator LongestAllSpaceLine;
1
'LongestAllSpaceLine' declared without an initial value→

while (true) {
2
←
Loop condition is true.  Entering loop body→
  advanceWhile(&Scanner::skip_s_space);
  if (skip_nb_char(Current) != Current) {
3
←
Taking true branch→
    // This line isn't empty, so try and find the indentation.
    if (Column <= BlockExitIndent) { // End of the block literal.
4
←
Assuming 'BlockExitIndent' is < field 'Column'→
5
←
Taking false branch→
      IsDone = true;
      return true;
    }
    // We found the block's indentation.
    BlockIndent = Column;
    if (MaxAllSpaceLineCharacters > BlockIndent) {
6
←
Assuming 'MaxAllSpaceLineCharacters' is > 'BlockIndent'→
7
←
Taking true branch→
      setError(
8
←
2nd function call argument is an uninitialized value
          "Leading all-spaces line must be smaller than the block indent",
          LongestAllSpaceLine);
      return false;
    }
    return true;
  }
  if (skip_b_break(Current) != Current &&
      Column > MaxAllSpaceLineCharacters) {
    // Record the longest all-space line in case it's longer than the
    // discovered block indent.
    MaxAllSpaceLineCharacters = Column;
    LongestAllSpaceLine = Current;
  }

  // Check for EOF.
  if (Current == End) {
    IsDone = true;
    return true;
  }

  if (!consumeLineBreakIfPresent()) {
    IsDone = true;
    return true;
  }
  ++LineBreaks;
}
return true;
1625}

1627bool Scanner::scanBlockScalarIndent(unsigned BlockIndent,
                                  unsigned BlockExitIndent, bool &IsDone) {
// Skip the indentation.
while (Column < BlockIndent) {
  auto I = skip_s_space(Current);
  if (I == Current)
    break;
  Current = I;
  ++Column;
}

if (skip_nb_char(Current) == Current)
  return true;

if (Column <= BlockExitIndent) { // End of the block literal.
  IsDone = true;
  return true;
}

if (Column < BlockIndent) {
  if (Current != End && *Current == '#') { // Trailing comment.
    IsDone = true;
    return true;
  }
  setError("A text line is less indented than the block scalar", Current);
  return false;
}
return true; // A normal text line.
1655}

1657bool Scanner::scanBlockScalar(bool IsLiteral) {
// Eat '|' or '>'
assert(*Current == '|' || *Current == '>')(static_cast <bool> (*Current == '|' || *Current == '>'
) ? void (0) : __assert_fail ("*Current == '|' || *Current == '>'"
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 1659, __extension__ __PRETTY_FUNCTION__));
skip(1);

char ChompingIndicator;
unsigned BlockIndent;
bool IsDone = false;
if (!scanBlockScalarHeader(ChompingIndicator, BlockIndent, IsDone))
  return false;
if (IsDone)
  return true;

auto Start = Current;
unsigned BlockExitIndent = Indent < 0 ? 0 : (unsigned)Indent;
unsigned LineBreaks = 0;
if (BlockIndent == 0) {
  if (!findBlockScalarIndent(BlockIndent, BlockExitIndent, LineBreaks,
                             IsDone))
    return false;
}

// Scan the block's scalars body.
SmallString<256> Str;
while (!IsDone) {
  if (!scanBlockScalarIndent(BlockIndent, BlockExitIndent, IsDone))
    return false;
  if (IsDone)
    break;

  // Parse the current line.
  auto LineStart = Current;
  advanceWhile(&Scanner::skip_nb_char);
  if (LineStart != Current) {
    Str.append(LineBreaks, '\n');
    Str.append(StringRef(LineStart, Current - LineStart));
    LineBreaks = 0;
  }

  // Check for EOF.
  if (Current == End)
    break;

  if (!consumeLineBreakIfPresent())
    break;
  ++LineBreaks;
}

if (Current == End && !LineBreaks)
  // Ensure that there is at least one line break before the end of file.
  LineBreaks = 1;
Str.append(getChompedLineBreaks(ChompingIndicator, LineBreaks, Str), '\n');

// New lines may start a simple key.
if (!FlowLevel)
  IsSimpleKeyAllowed = true;

Token T;
T.Kind = Token::TK_BlockScalar;
T.Range = StringRef(Start, Current - Start);
T.Value = std::string(Str);
TokenQueue.push_back(T);
return true;
1720}

1722bool Scanner::scanTag() {
StringRef::iterator Start = Current;
unsigned ColStart = Column;
skip(1); // Eat !.
if (Current == End || isBlankOrBreak(Current)); // An empty tag.
else if (*Current == '<') {
  skip(1);
  scan_ns_uri_char();
  if (!consume('>'))
    return false;
} else {
  // FIXME: Actually parse the c-ns-shorthand-tag rule.
  Current = skip_while(&Scanner::skip_ns_char, Current);
}

Token T;
T.Kind = Token::TK_Tag;
T.Range = StringRef(Start, Current - Start);
TokenQueue.push_back(T);

// Tags can be simple keys.
saveSimpleKeyCandidate(--TokenQueue.end(), ColStart, false);

IsSimpleKeyAllowed = false;

return true;
1748}

1750bool Scanner::fetchMoreTokens() {
if (IsStartOfStream)
  return scanStreamStart();

scanToNextToken();

if (Current == End)
  return scanStreamEnd();

removeStaleSimpleKeyCandidates();

unrollIndent(Column);

if (Column == 0 && *Current == '%')
  return scanDirective();

if (Column == 0 && Current + 4 <= End
    && *Current == '-'
    && *(Current + 1) == '-'
    && *(Current + 2) == '-'
    && (Current + 3 == End || isBlankOrBreak(Current + 3)))
  return scanDocumentIndicator(true);

if (Column == 0 && Current + 4 <= End
    && *Current == '.'
    && *(Current + 1) == '.'
    && *(Current + 2) == '.'
    && (Current + 3 == End || isBlankOrBreak(Current + 3)))
  return scanDocumentIndicator(false);

if (*Current == '[')
  return scanFlowCollectionStart(true);

if (*Current == '{')
  return scanFlowCollectionStart(false);

if (*Current == ']')
  return scanFlowCollectionEnd(true);

if (*Current == '}')
  return scanFlowCollectionEnd(false);

if (*Current == ',')
  return scanFlowEntry();

if (*Current == '-' && isBlankOrBreak(Current + 1))
  return scanBlockEntry();

if (*Current == '?' && (FlowLevel || isBlankOrBreak(Current + 1)))
  return scanKey();

if (*Current == ':' && (FlowLevel || isBlankOrBreak(Current + 1)))
  return scanValue();

if (*Current == '*')
  return scanAliasOrAnchor(true);

if (*Current == '&')
  return scanAliasOrAnchor(false);

if (*Current == '!')
  return scanTag();

if (*Current == '|' && !FlowLevel)
  return scanBlockScalar(true);

if (*Current == '>' && !FlowLevel)
  return scanBlockScalar(false);

if (*Current == '\'')
  return scanFlowScalar(false);

if (*Current == '"')
  return scanFlowScalar(true);

// Get a plain scalar.
StringRef FirstChar(Current, 1);
if (!(isBlankOrBreak(Current)
      || FirstChar.find_first_of("-?:,[]{}#&*!|>'\"%@`") != StringRef::npos)
    || (*Current == '-' && !isBlankOrBreak(Current + 1))
    || (!FlowLevel && (*Current == '?' || *Current == ':')
        && isBlankOrBreak(Current + 1))
    || (!FlowLevel && *Current == ':'
                    && Current + 2 < End
                    && *(Current + 1) == ':'
                    && !isBlankOrBreak(Current + 2)))
  return scanPlainScalar();

setError("Unrecognized character while tokenizing.", Current);
return false;
1840}

1842Stream::Stream(StringRef Input, SourceMgr &SM, bool ShowColors,
             std::error_code *EC)
  : scanner(new Scanner(Input, SM, ShowColors, EC)), CurrentDoc() {}

1846Stream::Stream(MemoryBufferRef InputBuffer, SourceMgr &SM, bool ShowColors,
             std::error_code *EC)
  : scanner(new Scanner(InputBuffer, SM, ShowColors, EC)), CurrentDoc() {}

1850Stream::~Stream() = default;

1852bool Stream::failed() { return scanner->failed(); }

1854void Stream::printError(Node *N, const Twine &Msg, SourceMgr::DiagKind Kind) {
printError(N ? N->getSourceRange() : SMRange(), Msg, Kind);
1856}

1858void Stream::printError(const SMRange &Range, const Twine &Msg,
                      SourceMgr::DiagKind Kind) {
scanner->printError(Range.Start, Kind, Msg, Range);
1861}

1863document_iterator Stream::begin() {
if (CurrentDoc)
  report_fatal_error("Can only iterate over the stream once");

// Skip Stream-Start.
scanner->getNext();

CurrentDoc.reset(new Document(*this));
return document_iterator(CurrentDoc);
1872}

1874document_iterator Stream::end() {
return document_iterator();
1876}

1878void Stream::skip() {
for (document_iterator i = begin(), e = end(); i != e; ++i)
  i->skip();
1881}

1883Node::Node(unsigned int Type, std::unique_ptr<Document> &D, StringRef A,
         StringRef T)
  : Doc(D), TypeID(Type), Anchor(A), Tag(T) {
SMLoc Start = SMLoc::getFromPointer(peekNext().Range.begin());
SourceRange = SMRange(Start, Start);
1888}

1890std::string Node::getVerbatimTag() const {
StringRef Raw = getRawTag();
if (!Raw.empty() && Raw != "!") {
  std::string Ret;
  if (Raw.find_last_of('!') == 0) {
    Ret = std::string(Doc->getTagMap().find("!")->second);
    Ret += Raw.substr(1);
    return Ret;
  } else if (Raw.startswith("!!")) {
    Ret = std::string(Doc->getTagMap().find("!!")->second);
    Ret += Raw.substr(2);
    return Ret;
  } else {
    StringRef TagHandle = Raw.substr(0, Raw.find_last_of('!') + 1);
    std::map<StringRef, StringRef>::const_iterator It =
        Doc->getTagMap().find(TagHandle);
    if (It != Doc->getTagMap().end())
      Ret = std::string(It->second);
    else {
      Token T;
      T.Kind = Token::TK_Tag;
      T.Range = TagHandle;
      setError(Twine("Unknown tag handle ") + TagHandle, T);
    }
    Ret += Raw.substr(Raw.find_last_of('!') + 1);
    return Ret;
  }
}

switch (getType()) {
case NK_Null:
  return "tag:yaml.org,2002:null";
case NK_Scalar:
case NK_BlockScalar:
  // TODO: Tag resolution.
  return "tag:yaml.org,2002:str";
case NK_Mapping:
  return "tag:yaml.org,2002:map";
case NK_Sequence:
  return "tag:yaml.org,2002:seq";
}

return "";
1933}

1935Token &Node::peekNext() {
return Doc->peekNext();
1937}

1939Token Node::getNext() {
return Doc->getNext();
1941}

1943Node *Node::parseBlockNode() {
return Doc->parseBlockNode();
1945}

1947BumpPtrAllocator &Node::getAllocator() {
return Doc->NodeAllocator;
1949}

1951void Node::setError(const Twine &Msg, Token &Tok) const {
Doc->setError(Msg, Tok);
1953}

1955bool Node::failed() const {
return Doc->failed();
1957}

1959StringRef ScalarNode::getValue(SmallVectorImpl<char> &Storage) const {
// TODO: Handle newlines properly. We need to remove leading whitespace.
if (Value[0] == '"') { // Double quoted.
  // Pull off the leading and trailing "s.
  StringRef UnquotedValue = Value.substr(1, Value.size() - 2);
  // Search for characters that would require unescaping the value.
  StringRef::size_type i = UnquotedValue.find_first_of("\\\r\n");
  if (i != StringRef::npos)
    return unescapeDoubleQuoted(UnquotedValue, i, Storage);
  return UnquotedValue;
} else if (Value[0] == '\'') { // Single quoted.
  // Pull off the leading and trailing 's.
  StringRef UnquotedValue = Value.substr(1, Value.size() - 2);
  StringRef::size_type i = UnquotedValue.find('\'');
  if (i != StringRef::npos) {
    // We're going to need Storage.
    Storage.clear();
    Storage.reserve(UnquotedValue.size());
    for (; i != StringRef::npos; i = UnquotedValue.find('\'')) {
      StringRef Valid(UnquotedValue.begin(), i);
      llvm::append_range(Storage, Valid);
      Storage.push_back('\'');
      UnquotedValue = UnquotedValue.substr(i + 2);
    }
    llvm::append_range(Storage, UnquotedValue);
    return StringRef(Storage.begin(), Storage.size());
  }
  return UnquotedValue;
}
// Plain or block.
return Value.rtrim(' ');
1990}

1992StringRef ScalarNode::unescapeDoubleQuoted( StringRef UnquotedValue
                                        , StringRef::size_type i
                                        , SmallVectorImpl<char> &Storage)
                                        const {
// Use Storage to build proper value.
Storage.clear();
Storage.reserve(UnquotedValue.size());
for (; i != StringRef::npos; i = UnquotedValue.find_first_of("\\\r\n")) {
  // Insert all previous chars into Storage.
  StringRef Valid(UnquotedValue.begin(), i);
  llvm::append_range(Storage, Valid);
  // Chop off inserted chars.
  UnquotedValue = UnquotedValue.substr(i);

  assert(!UnquotedValue.empty() && "Can't be empty!")(static_cast <bool> (!UnquotedValue.empty() && "Can't be empty!"
) ? void (0) : __assert_fail ("!UnquotedValue.empty() && \"Can't be empty!\""
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 2006, __extension__ __PRETTY_FUNCTION__));

  // Parse escape or line break.
  switch (UnquotedValue[0]) {
  case '\r':
  case '\n':
    Storage.push_back('\n');
    if (   UnquotedValue.size() > 1
        && (UnquotedValue[1] == '\r' || UnquotedValue[1] == '\n'))
      UnquotedValue = UnquotedValue.substr(1);
    UnquotedValue = UnquotedValue.substr(1);
    break;
  default:
    if (UnquotedValue.size() == 1) {
      Token T;
      T.Range = StringRef(UnquotedValue.begin(), 1);
      setError("Unrecognized escape code", T);
      return "";
    }
    UnquotedValue = UnquotedValue.substr(1);
    switch (UnquotedValue[0]) {
    default: {
        Token T;
        T.Range = StringRef(UnquotedValue.begin(), 1);
        setError("Unrecognized escape code", T);
        return "";
      }
    case '\r':
    case '\n':
      // Remove the new line.
      if (   UnquotedValue.size() > 1
          && (UnquotedValue[1] == '\r' || UnquotedValue[1] == '\n'))
        UnquotedValue = UnquotedValue.substr(1);
      // If this was just a single byte newline, it will get skipped
      // below.
      break;
    case '0':
      Storage.push_back(0x00);
      break;
    case 'a':
      Storage.push_back(0x07);
      break;
    case 'b':
      Storage.push_back(0x08);
      break;
    case 't':
    case 0x09:
      Storage.push_back(0x09);
      break;
    case 'n':
      Storage.push_back(0x0A);
      break;
    case 'v':
      Storage.push_back(0x0B);
      break;
    case 'f':
      Storage.push_back(0x0C);
      break;
    case 'r':
      Storage.push_back(0x0D);
      break;
    case 'e':
      Storage.push_back(0x1B);
      break;
    case ' ':
      Storage.push_back(0x20);
      break;
    case '"':
      Storage.push_back(0x22);
      break;
    case '/':
      Storage.push_back(0x2F);
      break;
    case '\\':
      Storage.push_back(0x5C);
      break;
    case 'N':
      encodeUTF8(0x85, Storage);
      break;
    case '_':
      encodeUTF8(0xA0, Storage);
      break;
    case 'L':
      encodeUTF8(0x2028, Storage);
      break;
    case 'P':
      encodeUTF8(0x2029, Storage);
      break;
    case 'x': {
        if (UnquotedValue.size() < 3)
          // TODO: Report error.
          break;
        unsigned int UnicodeScalarValue;
        if (UnquotedValue.substr(1, 2).getAsInteger(16, UnicodeScalarValue))
          // TODO: Report error.
          UnicodeScalarValue = 0xFFFD;
        encodeUTF8(UnicodeScalarValue, Storage);
        UnquotedValue = UnquotedValue.substr(2);
        break;
      }
    case 'u': {
        if (UnquotedValue.size() < 5)
          // TODO: Report error.
          break;
        unsigned int UnicodeScalarValue;
        if (UnquotedValue.substr(1, 4).getAsInteger(16, UnicodeScalarValue))
          // TODO: Report error.
          UnicodeScalarValue = 0xFFFD;
        encodeUTF8(UnicodeScalarValue, Storage);
        UnquotedValue = UnquotedValue.substr(4);
        break;
      }
    case 'U': {
        if (UnquotedValue.size() < 9)
          // TODO: Report error.
          break;
        unsigned int UnicodeScalarValue;
        if (UnquotedValue.substr(1, 8).getAsInteger(16, UnicodeScalarValue))
          // TODO: Report error.
          UnicodeScalarValue = 0xFFFD;
        encodeUTF8(UnicodeScalarValue, Storage);
        UnquotedValue = UnquotedValue.substr(8);
        break;
      }
    }
    UnquotedValue = UnquotedValue.substr(1);
  }
}
llvm::append_range(Storage, UnquotedValue);
return StringRef(Storage.begin(), Storage.size());
2136}

2138Node *KeyValueNode::getKey() {
if (Key)
  return Key;
// Handle implicit null keys.
{
  Token &t = peekNext();
  if (   t.Kind == Token::TK_BlockEnd
      || t.Kind == Token::TK_Value
      || t.Kind == Token::TK_Error) {
    return Key = new (getAllocator()) NullNode(Doc);
  }
  if (t.Kind == Token::TK_Key)
    getNext(); // skip TK_Key.
}

// Handle explicit null keys.
Token &t = peekNext();
if (t.Kind == Token::TK_BlockEnd || t.Kind == Token::TK_Value) {
  return Key = new (getAllocator()) NullNode(Doc);
}

// We've got a normal key.
return Key = parseBlockNode();
2161}

2163Node *KeyValueNode::getValue() {
if (Value)
  return Value;

if (Node* Key = getKey())
  Key->skip();
else {
  setError("Null key in Key Value.", peekNext());
  return Value = new (getAllocator()) NullNode(Doc);
}

if (failed())
  return Value = new (getAllocator()) NullNode(Doc);

// Handle implicit null values.
{
  Token &t = peekNext();
  if (   t.Kind == Token::TK_BlockEnd
      || t.Kind == Token::TK_FlowMappingEnd
      || t.Kind == Token::TK_Key
      || t.Kind == Token::TK_FlowEntry
      || t.Kind == Token::TK_Error) {
    return Value = new (getAllocator()) NullNode(Doc);
  }

  if (t.Kind != Token::TK_Value) {
    setError("Unexpected token in Key Value.", t);
    return Value = new (getAllocator()) NullNode(Doc);
  }
  getNext(); // skip TK_Value.
}

// Handle explicit null values.
Token &t = peekNext();
if (t.Kind == Token::TK_BlockEnd || t.Kind == Token::TK_Key) {
  return Value = new (getAllocator()) NullNode(Doc);
}

// We got a normal value.
return Value = parseBlockNode();
2203}

2205void MappingNode::increment() {
if (failed()) {
  IsAtEnd = true;
  CurrentEntry = nullptr;
  return;
}
if (CurrentEntry) {
  CurrentEntry->skip();
  if (Type == MT_Inline) {
    IsAtEnd = true;
    CurrentEntry = nullptr;
    return;
  }
}
Token T = peekNext();
if (T.Kind == Token::TK_Key || T.Kind == Token::TK_Scalar) {
  // KeyValueNode eats the TK_Key. That way it can detect null keys.
  CurrentEntry = new (getAllocator()) KeyValueNode(Doc);
} else if (Type == MT_Block) {
  switch (T.Kind) {
  case Token::TK_BlockEnd:
    getNext();
    IsAtEnd = true;
    CurrentEntry = nullptr;
    break;
  default:
    setError("Unexpected token. Expected Key or Block End", T);
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case Token::TK_Error:
    IsAtEnd = true;
    CurrentEntry = nullptr;
  }
} else {
  switch (T.Kind) {
  case Token::TK_FlowEntry:
    // Eat the flow entry and recurse.
    getNext();
    return increment();
  case Token::TK_FlowMappingEnd:
    getNext();
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case Token::TK_Error:
    // Set this to end iterator.
    IsAtEnd = true;
    CurrentEntry = nullptr;
    break;
  default:
    setError( "Unexpected token. Expected Key, Flow Entry, or Flow "
              "Mapping End."
            , T);
    IsAtEnd = true;
    CurrentEntry = nullptr;
  }
}
2259}

2261void SequenceNode::increment() {
if (failed()) {
  IsAtEnd = true;
  CurrentEntry = nullptr;
  return;
}
if (CurrentEntry)
  CurrentEntry->skip();
Token T = peekNext();
if (SeqType == ST_Block) {
  switch (T.Kind) {
  case Token::TK_BlockEntry:
    getNext();
    CurrentEntry = parseBlockNode();
    if (!CurrentEntry) { // An error occurred.
      IsAtEnd = true;
      CurrentEntry = nullptr;
    }
    break;
  case Token::TK_BlockEnd:
    getNext();
    IsAtEnd = true;
    CurrentEntry = nullptr;
    break;
  default:
    setError( "Unexpected token. Expected Block Entry or Block End."
            , T);
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case Token::TK_Error:
    IsAtEnd = true;
    CurrentEntry = nullptr;
  }
} else if (SeqType == ST_Indentless) {
  switch (T.Kind) {
  case Token::TK_BlockEntry:
    getNext();
    CurrentEntry = parseBlockNode();
    if (!CurrentEntry) { // An error occurred.
      IsAtEnd = true;
      CurrentEntry = nullptr;
    }
    break;
  default:
  case Token::TK_Error:
    IsAtEnd = true;
    CurrentEntry = nullptr;
  }
} else if (SeqType == ST_Flow) {
  switch (T.Kind) {
  case Token::TK_FlowEntry:
    // Eat the flow entry and recurse.
    getNext();
    WasPreviousTokenFlowEntry = true;
    return increment();
  case Token::TK_FlowSequenceEnd:
    getNext();
    LLVM_FALLTHROUGH[[gnu::fallthrough]];
  case Token::TK_Error:
    // Set this to end iterator.
    IsAtEnd = true;
    CurrentEntry = nullptr;
    break;
  case Token::TK_StreamEnd:
  case Token::TK_DocumentEnd:
  case Token::TK_DocumentStart:
    setError("Could not find closing ]!", T);
    // Set this to end iterator.
    IsAtEnd = true;
    CurrentEntry = nullptr;
    break;
  default:
    if (!WasPreviousTokenFlowEntry) {
      setError("Expected , between entries!", T);
      IsAtEnd = true;
      CurrentEntry = nullptr;
      break;
    }
    // Otherwise it must be a flow entry.
    CurrentEntry = parseBlockNode();
    if (!CurrentEntry) {
      IsAtEnd = true;
    }
    WasPreviousTokenFlowEntry = false;
    break;
  }
}
2347}

2349Document::Document(Stream &S) : stream(S), Root(nullptr) {
// Tag maps starts with two default mappings.
TagMap["!"] = "!";
TagMap["!!"] = "tag:yaml.org,2002:";

if (parseDirectives())
  expectToken(Token::TK_DocumentStart);
Token &T = peekNext();
if (T.Kind == Token::TK_DocumentStart)
  getNext();
2359}

2361bool Document::skip()  {
if (stream.scanner->failed())
  return false;
if (!Root && !getRoot())
  return false;
Root->skip();
Token &T = peekNext();
if (T.Kind == Token::TK_StreamEnd)
  return false;
if (T.Kind == Token::TK_DocumentEnd) {
  getNext();
  return skip();
}
return true;
2375}

2377Token &Document::peekNext() {
return stream.scanner->peekNext();
2379}

2381Token Document::getNext() {
return stream.scanner->getNext();
2383}

2385void Document::setError(const Twine &Message, Token &Location) const {
stream.scanner->setError(Message, Location.Range.begin());
2387}

2389bool Document::failed() const {
return stream.scanner->failed();
2391}

2393Node *Document::parseBlockNode() {
Token T = peekNext();
// Handle properties.
Token AnchorInfo;
Token TagInfo;
2398parse_property:
switch (T.Kind) {
case Token::TK_Alias:
  getNext();
  return new (NodeAllocator) AliasNode(stream.CurrentDoc, T.Range.substr(1));
case Token::TK_Anchor:
  if (AnchorInfo.Kind == Token::TK_Anchor) {
    setError("Already encountered an anchor for this node!", T);
    return nullptr;
  }
  AnchorInfo = getNext(); // Consume TK_Anchor.
  T = peekNext();
  goto parse_property;
case Token::TK_Tag:
  if (TagInfo.Kind == Token::TK_Tag) {
    setError("Already encountered a tag for this node!", T);
    return nullptr;
  }
  TagInfo = getNext(); // Consume TK_Tag.
  T = peekNext();
  goto parse_property;
default:
  break;
}

switch (T.Kind) {
case Token::TK_BlockEntry:
  // We got an unindented BlockEntry sequence. This is not terminated with
  // a BlockEnd.
  // Don't eat the TK_BlockEntry, SequenceNode needs it.
  return new (NodeAllocator) SequenceNode( stream.CurrentDoc
                                         , AnchorInfo.Range.substr(1)
                                         , TagInfo.Range
                                         , SequenceNode::ST_Indentless);
case Token::TK_BlockSequenceStart:
  getNext();
  return new (NodeAllocator)
    SequenceNode( stream.CurrentDoc
                , AnchorInfo.Range.substr(1)
                , TagInfo.Range
                , SequenceNode::ST_Block);
case Token::TK_BlockMappingStart:
  getNext();
  return new (NodeAllocator)
    MappingNode( stream.CurrentDoc
               , AnchorInfo.Range.substr(1)
               , TagInfo.Range
               , MappingNode::MT_Block);
case Token::TK_FlowSequenceStart:
  getNext();
  return new (NodeAllocator)
    SequenceNode( stream.CurrentDoc
                , AnchorInfo.Range.substr(1)
                , TagInfo.Range
                , SequenceNode::ST_Flow);
case Token::TK_FlowMappingStart:
  getNext();
  return new (NodeAllocator)
    MappingNode( stream.CurrentDoc
               , AnchorInfo.Range.substr(1)
               , TagInfo.Range
               , MappingNode::MT_Flow);
case Token::TK_Scalar:
  getNext();
  return new (NodeAllocator)
    ScalarNode( stream.CurrentDoc
              , AnchorInfo.Range.substr(1)
              , TagInfo.Range
              , T.Range);
case Token::TK_BlockScalar: {
  getNext();
  StringRef NullTerminatedStr(T.Value.c_str(), T.Value.length() + 1);
  StringRef StrCopy = NullTerminatedStr.copy(NodeAllocator).drop_back();
  return new (NodeAllocator)
      BlockScalarNode(stream.CurrentDoc, AnchorInfo.Range.substr(1),
                      TagInfo.Range, StrCopy, T.Range);
}
case Token::TK_Key:
  // Don't eat the TK_Key, KeyValueNode expects it.
  return new (NodeAllocator)
    MappingNode( stream.CurrentDoc
               , AnchorInfo.Range.substr(1)
               , TagInfo.Range
               , MappingNode::MT_Inline);
case Token::TK_DocumentStart:
case Token::TK_DocumentEnd:
case Token::TK_StreamEnd:
default:
  // TODO: Properly handle tags. "[!!str ]" should resolve to !!str "", not
  //       !!null null.
  return new (NodeAllocator) NullNode(stream.CurrentDoc);
case Token::TK_FlowMappingEnd:
case Token::TK_FlowSequenceEnd:
case Token::TK_FlowEntry: {
  if (Root && (isa<MappingNode>(Root) || isa<SequenceNode>(Root)))
    return new (NodeAllocator) NullNode(stream.CurrentDoc);

  setError("Unexpected token", T);
  return nullptr;
}
case Token::TK_Error:
  return nullptr;
}
llvm_unreachable("Control flow shouldn't reach here.")::llvm::llvm_unreachable_internal("Control flow shouldn't reach here."
, "/build/llvm-toolchain-snapshot-14~++20210828111110+16086d47c0d0/llvm/lib/Support/YAMLParser.cpp"
, 2501);
return nullptr;
2503}

2505bool Document::parseDirectives() {
bool isDirective = false;
while (true) {
  Token T = peekNext();
  if (T.Kind == Token::TK_TagDirective) {
    parseTAGDirective();
    isDirective = true;
  } else if (T.Kind == Token::TK_VersionDirective) {
    parseYAMLDirective();
    isDirective = true;
  } else
    break;
}
return isDirective;
2519}

2521void Document::parseYAMLDirective() {
getNext(); // Eat %YAML <version>
2523}

2525void Document::parseTAGDirective() {
Token Tag = getNext(); // %TAG <handle> <prefix>
StringRef T = Tag.Range;
// Strip %TAG
T = T.substr(T.find_first_of(" \t")).ltrim(" \t");
std::size_t HandleEnd = T.find_first_of(" \t");
StringRef TagHandle = T.substr(0, HandleEnd);
StringRef TagPrefix = T.substr(HandleEnd).ltrim(" \t");
TagMap[TagHandle] = TagPrefix;
2534}

2536bool Document::expectToken(int TK) {
Token T = getNext();
if (T.Kind != TK) {
  setError("Unexpected token", T);
  return false;
}
return true;
2543}