{"id":21,"date":"2020-08-12T11:45:27","date_gmt":"2020-08-12T03:45:27","guid":{"rendered":"http:\/\/47.108.144.50\/?p=21"},"modified":"2020-08-13T20:38:12","modified_gmt":"2020-08-13T12:38:12","slug":"llvm%e5%ae%98%e6%96%b9%e4%b8%87%e8%8a%b1%e7%ad%92%e6%95%99%e7%a8%8b1-lexer%ef%bc%8cparse-%ef%bc%8cast","status":"publish","type":"post","link":"http:\/\/pareto.fun\/?p=21","title":{"rendered":"llvm\u524d\u7aef-\u5b98\u65b9\u4e07\u82b1\u7b52\u6559\u7a0b1-Lexer\uff0cParse \uff0cAST"},"content":{"rendered":"\n

\u7531\u4e8e\u5f88\u591a\u6559\u7a0b\u7f16\u5199\u7684pass\u5b58\u5728\u95ee\u9898\uff0c\u81ea\u5df1\u53c8\u4e0d\u4f1a\u6539 \uff0c\u672c\u7cfb\u5217\u6559\u7a0b\u76ee\u7684\u662f\u5b66\u4e60\u7f16\u5199llvm pass \u4e3b\u8981\u8ddf\u7740llvm\u5b98\u7f51\u63d0\u4f9b\u7684\u6559\u7a0b\u6765\u5b66\u4e60\uff0c\u5b98\u65b9\u63d0\u4f9b\u7684\u6559\u7a0b\u4e00\u4e2a\u81ea\u5236\u7684llvm\u524d\u7aef\u7f16\u7a0b\u8bed\u8a00\u4e07\u82b1\u7b52<\/strong>\u4e3a\u4f8b\u3002<\/strong><\/p>\n\n\n\n

\u4e07\u82b1\u7b52\u8bed\u8a00 – The Kaleidoscope Language<\/h1>\n\n\n\n

\u4e07\u82b1\u7b52\u662f\u4e00\u4e2a\u5141\u8bb8\u5b9a\u4e49\u51fd\u6570\uff0c\u4f7f\u7528\u5224\u65ad\uff0c\u6570\u5b66\u7b49\u7684\u7a0b\u5e8f\u8bed\u8a00\uff0c\u6574\u4e2a\u6559\u7a0b\u4e2d\uff0c\u6211\u4eec\u4f1a\u62d3\u5c55\u4e07\u82b1\u7b52\u6765\u652f\u6301 if\/then\/else \u6784\u5efa\uff0c\u5faa\u73af\uff0c\u7528\u6237\u81ea\u5b9a\u4e49\u64cd\u4f5c\u7b26\uff0c\u4e00\u884c\u547d\u4ee4\u63a5\u53e3\u7684JIT \u7f16\u8bd1 \uff0c\u8c03\u8bd5\u4fe1\u606f\u7b49\u3002<\/p>\n\n\n\n

\u6211\u4eec\u60f3\u8ba9\u4e8b\u60c5\u7b80\u5355\uff0c\u6240\u4ee5\u5728\u4e07\u82b1\u7b52\u552f\u4e00\u7684\u6570\u636e\u7c7b\u578b\u662f64\u4f4d\u7684\u6d6e\u70b9\u7c7b\u578b\uff08C\u8bed\u8a00\u4e2d\u7684double\uff09\uff0c\u6240\u6709\u7684\u503c\u9ed8\u8ba4\u90fd\u662f\u53cc\u7cbe\u5ea6\u5e76\u4e14\u4e07\u82b1\u7b52\u4e0d\u9700\u8981\u5b9a\u4e49\u3002\u8fd9\u6837\u8ba9\u4e07\u82b1\u7b52\u975e\u5e38\u53cb\u597d\uff0c\u8bed\u6cd5\u7b80\u5355\u3002\u4f8b\u5982\u4e00\u4e2aFibonacci number \u8ba1\u7b97\u7684\u4f8b\u5b50\u3002<\/p>\n\n\n\n

# Compute the x'th fibonacci number.
def fib(x)
  if x < 3 then
    1
  else
    fib(x-1)+fib(x-2)
\u200b
# This expression will compute the 40th number.
fib(40)<\/pre>\n\n\n\n
LLVM JIT \u4e5f\u8ba9\u4e07\u82b1\u7b52\u652f\u6301\u8c03\u7528\u6807\u51c6\u5e93\u51fd\u6570 \u53d8\u5f97\u7b80\u5355\uff0c\u8fd9\u610f\u5473\u7740\u4f60\u53ef\u4ee5\u7528’extern’ \u5173\u952e\u8bcd\u5728\u4f60\u4f7f\u7528\u524d\u6765\u5b9a\u4e49\u4e2a\u51fd\u6570\u3002\u4f8b\u5982\uff1a<\/p>\n\n\n\n
extern sin(arg);
extern cos(arg);
extern atan2(arg1 arg2);
\u200b
atan2(sin(.4), cos(42))<\/pre>\n\n\n\n
\u66f4\u6709\u8da3\u7684\u4f8b\u5b50\u5728\u7b2c\u516d\u7ae0 \u6211\u4eec\u5199\u4e86\u4e00\u4e9b\u4e07\u82b1\u7b52\u7684\u5e94\u7528\u7528\u6765\u5728\u4e0d\u540c\u7684\u653e\u5927\u500d\u6570\u6765\u5c55\u73b0Mandelbrot \u96c6\u5408\u3002<\/p>\n\n\n\n
\u8bcd\u6cd5\u5206\u6790\u5668 – Lexer<\/h1>\n\n\n\n
\u5b9e\u73b0\u4e00\u4e2a\u7a0b\u5e8f\u8bed\u8a00\u7684\u65f6\u5019\uff0c\u7b2c\u4e00\u4ef6\u4e8b\u5c31\u662f\u9700\u8981\u5904\u7406\u6587\u672c\u5185\u5bb9\u548c\u610f\u8bc6\u5230\u5b83\u5728\u8868\u8fbe\u4ec0\u4e48\u3002\u4f20\u7edf\u5b9e\u73b0\u65b9\u5f0f\u662f\u4f7f\u7528”Lexer” \u8bcd\u6cd5\u5206\u6790\u5668\u628a\u8f93\u5165\u5206\u89e3\u6210 “tokens” \u3002\u6bcf\u4e2a\u8bcd\u6cd5\u5206\u6790\u5668\u8fd4\u56de\u7684token \u5305\u62ectoken\u7801 \u548c\u4e00\u4e9b\u5143\u6570\u636e\u3002<\/p>\n\n\n\n
\/\/ The lexer returns tokens [0-255] if it is an unknown character, otherwise one
\/\/ of these for known things.
enum Token {
  tok_eof = -1,
\u200b
  \/\/ commands
  tok_def = -2,
  tok_extern = -3,
\u200b
  \/\/ primary
  tok_identifier = -4,
  tok_number = -5,
};
\u200b
static std::string IdentifierStr; \/\/ Filled in if tok_identifier
static double NumVal;             \/\/ Filled in if tok_number<\/pre>\n\n\n\n
\u901a\u8fc7\u8bcd\u6cd5\u5206\u6790\u5668\u8fd4\u56de\u7684\u6bcf\u4e2atoken \u4f1a\u662f\u67d0\u4e2aToken \u679a\u4e3e\u503c(Token enum values)\u6216\u8005\u662f\u4e0d\u77e5\u9053(\u672a\u5b9a\u4e49)\u7684\u5b57\u7b26\u50cf “+” \u5c31\u4f1a\u8fd4\u56de\u5b83\u7684Ascii\u503c\u3002\u5982\u679c\u5f53\u524d\u7684token\u662f\u4e00\u4e2a\u6807\u8bc6\u7b26(Token == -4)\uff0c\u90a3\u4e48\u5168\u5c40\u53d8\u91cfIdentifierStr\u5c31\u4f1a\u6807\u8bc6\u7b26\u7684\u540d\u79f0\u3002\u5982\u679c\u5f53\u524dkoken\u662f\u6570\u5b57\uff0c\u90a3\u4e48\u5168\u5c40\u53d8\u91cfNumVal \u5c31\u4f1a\u4fdd\u5b58\u5b83\u7684\u503c\u3002\u6211\u4eec\u56e0\u4e3a\u7b80\u5355\u4f7f\u7528\u5168\u5c40\u53d8\u91cf\uff0c\u4f46\u662f\u5bf9\u4e8e\u771f\u6b63\u7684\u8bed\u8a00\u5b9e\u73b0\u5e76\u4e0d\u662f\u4e00\u4e2a\u597d\u7684\u9009\u62e9\u3002<\/p>\n\n\n\n
\u8bcd\u6cd5\u5206\u6790\u7684\u771f\u6b63\u5b9e\u73b0\u662f\u4e00\u4e2a\u53ebgettok\u7684\u7b80\u5355\u51fd\u6570\uff0cgettok\u51fd\u6570\u88ab\u8c03\u7528\u65f6\u8fd4\u56de\u6807\u51c6\u8f93\u5165\u7684\u4e0b\u4e00\u4e2atoken\u3002\u5b83\u5b9a\u4e49\u7684\u5f00\u59cb\u662f\uff1a<\/p>\n\n\n\n
\/\/ gettok - Return the next token from standard input.
static int gettok() {
  static int LastChar = ' ';
\u200b
  \/\/ Skip any whitespace.
  while (isspace(LastChar))
    LastChar = getchar();<\/pre>\n\n\n\n
gettok \u901a\u8fc7\u8c03\u7528C\u51fd\u6570getchar() \u6765\u4ece\u6807\u51c6\u8f93\u5165\u6bcf\u6b21\u8bfb\u53d6\u4e00\u4e2a\u5b57\u7b26\u3002\u5b83\u5728\u8bc6\u522b\u51fa\u5b83\u4eec\u540e\u5c31\u5403\u6389\u5b83\u4eec\uff0c\u5e76\u5c06\u6700\u540e\u8bfb\u53d6\u4f46\u672a\u5904\u7406\u7684\u5b57\u7b26\u5b58\u50a8\u5728LastChar\u4e2d\u3002\u5b83\u8981\u505a\u7684\u7b2c\u4e00\u4ef6\u4e8b\u662f\u5ffd\u7565\u4ee4\u724c\u4e4b\u95f4\u7684\u7a7a\u683c\u3002\u8fd9\u662f\u901a\u8fc7\u4e0a\u9762\u7684\u5faa\u73af\u5b8c\u6210\u7684\u3002<\/p>\n\n\n\n
gettok\u8981\u505a\u7684\u4e0b\u4e00\u4ef6\u4e8b\u662f\u8bc6\u522b\u6807\u8bc6\u7b26\u548c\u7279\u5b9a\u7684\u5173\u952e\u5b57\u50cf”def” , \u4e07\u82b1\u7b52\u901a\u8fc7\u5b83\u7684\u5faa\u73af\u6765\u5b9e\u73b0\uff1a<\/p>\n\n\n\n
if (isalpha(LastChar)) { \/\/ identifier: [a-zA-Z][a-zA-Z0-9]*
  IdentifierStr = LastChar;
  while (isalnum((LastChar = getchar())))
    IdentifierStr += LastChar;
\u200b
  if (IdentifierStr == \"def\")
    return tok_def;
  if (IdentifierStr == \"extern\")
    return tok_extern;
  return tok_identifier;
}<\/pre>\n\n\n\n
\u6ce8\u610f\uff0c\u6b64\u4ee3\u7801\u5728\u5bf9\u6807\u8bc6\u7b26\u8fdb\u884c\u8bcd\u6cd5\u8bc6\u522b\u65f6\u4f1a\u5168\u5c40\u8bbe\u7f6e\u201c IdentifierStr\u201d\uff0c \u53e6\u5916<\/p>\n\n\n\n
…….\u81ea\u5df1\u770b\u5230\u7b2c\u4e8c\u7ae0\u4e2d\u95f4\u7ec8\u4e8e\u770b\u4e0d\u4e0b\u53bb\u4e86 \uff0c\u76f4\u63a5\u770b\u4ee3\u7801\u624d\u662f\u6211\u7684\u98ce\u683c\u3002<\/p>\n\n\n\n
#include \"llvm\/ADT\/STLExtras.h\"
#include <algorithm>
#include <cctype>
#include <cstdio>
#include <cstdlib>
#include <map>
#include <memory>
#include <string>
#include <vector>
\u200b
\/\/===----------------------------------------------------------------------===\/\/
\/\/ Lexer
\/\/===----------------------------------------------------------------------===\/\/
\u200b
\/\/ The lexer returns tokens [0-255] if it is an unknown character, otherwise one
\/\/ of these for known things.
enum Token {
  tok_eof = -1,
\u200b
  \/\/ commands
  tok_def = -2,
  tok_extern = -3,
\u200b
  \/\/ primary
  tok_identifier = -4,
  tok_number = -5
};
\u200b
static std::string IdentifierStr; \/\/ Filled in if tok_identifier
static double NumVal;             \/\/ Filled in if tok_number
\u200b
\/\/\/ gettok - Return the next token from standard input.
\/\/\u8bfb\u53d6\u8f93\u5165\uff0c\u6807\u8bb0token\u548c\u5bf9\u5e94\u7684\u503c\u3002
static int gettok() {
  static int LastChar = ' ';
\u200b
  \/\/ Skip any whitespace.
  while (isspace(LastChar))
    LastChar = getchar();
\u200b
  if (isalpha(LastChar)) { \/\/ identifier: [a-zA-Z][a-zA-Z0-9]*
    IdentifierStr = LastChar;
    while (isalnum((LastChar = getchar())))
      IdentifierStr += LastChar;
\u200b
    if (IdentifierStr == \"def\")
      return tok_def;
    if (IdentifierStr == \"extern\")
      return tok_extern;
    return tok_identifier;
  }
\u200b
  if (isdigit(LastChar) || LastChar == '.') { \/\/ Number: [0-9.]+
    std::string NumStr;
    do {
      NumStr += LastChar;
      LastChar = getchar();
    } while (isdigit(LastChar) || LastChar == '.');
\u200b
    NumVal = strtod(NumStr.c_str(), nullptr);
    return tok_number;
  }
\u200b
  if (LastChar == '#') {
    \/\/ Comment until end of line.
    do
      LastChar = getchar();
    while (LastChar != EOF && LastChar != '\\n' && LastChar != '\\r');
\u200b
    if (LastChar != EOF)
      return gettok();
  }
\u200b
  \/\/ Check for end of file.  Don't eat the EOF.
  if (LastChar == EOF)
    return tok_eof;
\u200b
  \/\/ Otherwise, just return the character as its ascii value.
  int ThisChar = LastChar;
  LastChar = getchar();
  return ThisChar;
}
\u200b
\/\/===----------------------------------------------------------------------===\/\/
\/\/ Abstract Syntax Tree (aka Parse Tree)
\/\/===----------------------------------------------------------------------===\/\/
\u200b
namespace {
\u200b
\/\/\/ ExprAST - Base class for all expression nodes.
class ExprAST {
public:
  virtual ~ExprAST() = default;
};
\u200b
\/\/\/ NumberExprAST - Expression class for numeric literals like \"1.0\".
class NumberExprAST : public ExprAST {
  double Val;
\u200b
public:
  NumberExprAST(double Val) : Val(Val) {}
};
\u200b
\/\/\/ VariableExprAST - Expression class for referencing a variable, like \"a\".
class VariableExprAST : public ExprAST {
  std::string Name;
\u200b
public:
  VariableExprAST(const std::string &Name) : Name(Name) {}
};
\u200b
\/\/\/ BinaryExprAST - Expression class for a binary operator.
class BinaryExprAST : public ExprAST {
  char Op;
  std::unique_ptr<ExprAST> LHS, RHS;
\u200b
public:
  BinaryExprAST(char Op, std::unique_ptr<ExprAST> LHS,
                std::unique_ptr<ExprAST> RHS)
      : Op(Op), LHS(std::move(LHS)), RHS(std::move(RHS)) {}
};
\u200b
\/\/\/ CallExprAST - Expression class for function calls.
class CallExprAST : public ExprAST {
  std::string Callee;
  std::vector<std::unique_ptr<ExprAST>> Args;
\u200b
public:
  CallExprAST(const std::string &Callee,
              std::vector<std::unique_ptr<ExprAST>> Args)
      : Callee(Callee), Args(std::move(Args)) {}
};
\u200b
\/\/\/ PrototypeAST - This class represents the \"prototype\" for a function,
\/\/\/ which captures its name, and its argument names (thus implicitly the number
\/\/\/ of arguments the function takes).
class PrototypeAST {
  std::string Name;
  std::vector<std::string> Args;
\u200b
public:
  PrototypeAST(const std::string &Name, std::vector<std::string> Args)
      : Name(Name), Args(std::move(Args)) {}
\u200b
  const std::string &getName() const { return Name; }
};
\u200b
\/\/\/ FunctionAST - This class represents a function definition itself.
class FunctionAST {
  std::unique_ptr<PrototypeAST> Proto;
  std::unique_ptr<ExprAST> Body;
\u200b
public:
  FunctionAST(std::unique_ptr<PrototypeAST> Proto,
              std::unique_ptr<ExprAST> Body)
      : Proto(std::move(Proto)), Body(std::move(Body)) {}
};
\u200b
} \/\/ end anonymous namespace
\u200b
\/\/===----------------------------------------------------------------------===\/\/
\/\/ Parser
\/\/===----------------------------------------------------------------------===\/\/
\u200b
\/\/\/ CurTok\/getNextToken - Provide a simple token buffer.  CurTok is the current
\/\/\/ token the parser is looking at.  getNextToken reads another token from the
\/\/\/ lexer and updates CurTok with its results.
static int CurTok;
static int getNextToken() { return CurTok = gettok(); }
\u200b
\/\/\/ BinopPrecedence - This holds the precedence for each binary operator that is
\/\/\/ defined.
static std::map<char, int> BinopPrecedence;
\u200b
\/\/\/ GetTokPrecedence - Get the precedence of the pending binary operator token.
static int GetTokPrecedence() {
  if (!isascii(CurTok))
    return -1;
\u200b
  \/\/ Make sure it's a declared binop.
  int TokPrec = BinopPrecedence[CurTok];
  if (TokPrec <= 0)
    return -1;
  return TokPrec;
}
\u200b
\/\/\/ LogError* - These are little helper functions for error handling.
std::unique_ptr<ExprAST> LogError(const char *Str) {
  fprintf(stderr, \"Error: %s\\n\", Str);
  return nullptr;
}
std::unique_ptr<PrototypeAST> LogErrorP(const char *Str) {
  LogError(Str);
  return nullptr;
}
\u200b
static std::unique_ptr<ExprAST> ParseExpression();
\u200b
\/\/\/ numberexpr ::= number
static std::unique_ptr<ExprAST> ParseNumberExpr() {
  auto Result = std::make_unique<NumberExprAST>(NumVal);
  getNextToken(); \/\/ consume the number
  return std::move(Result);
}
\u200b
\/\/\/ parenexpr ::= '(' expression ')'
static std::unique_ptr<ExprAST> ParseParenExpr() {
  getNextToken(); \/\/ eat (.
  auto V = ParseExpression();
  if (!V)
    return nullptr;
\u200b
  if (CurTok != ')')
    return LogError(\"expected ')'\");
  getNextToken(); \/\/ eat ).
  return V;
}
\u200b
\/\/\/ identifierexpr
\/\/\/   ::= identifier
\/\/\/   ::= identifier '(' expression* ')'
static std::unique_ptr<ExprAST> ParseIdentifierExpr() {
  std::string IdName = IdentifierStr;
\u200b
  getNextToken(); \/\/ eat identifier.
\u200b
  if (CurTok != '(') \/\/ Simple variable ref.
    return std::make_unique<VariableExprAST>(IdName);
\u200b
  \/\/ Call.
  getNextToken(); \/\/ eat (
  std::vector<std::unique_ptr<ExprAST>> Args;
  if (CurTok != ')') {
    while (true) {
      if (auto Arg = ParseExpression())
        Args.push_back(std::move(Arg));
      else
        return nullptr;
\u200b
      if (CurTok == ')')
        break;
\u200b
      if (CurTok != ',')
        return LogError(\"Expected ')' or ',' in argument list\");
      getNextToken();
    }
  }
\u200b
  \/\/ Eat the ')'.
  getNextToken();
\u200b
  return std::make_unique<CallExprAST>(IdName, std::move(Args));
}
\u200b
\/\/\/ primary
\/\/\/   ::= identifierexpr
\/\/\/   ::= numberexpr
\/\/\/   ::= parenexpr
static std::unique_ptr<ExprAST> ParsePrimary() {
  switch (CurTok) {
  default:
    return LogError(\"unknown token when expecting an expression\");
  case tok_identifier:
    return ParseIdentifierExpr();
  case tok_number:
    return ParseNumberExpr();
  case '(':
    return ParseParenExpr();
  }
}
\u200b
\/\/\/ binoprhs
\/\/\/   ::= ('+' primary)*
static std::unique_ptr<ExprAST> ParseBinOpRHS(int ExprPrec,
                                              std::unique_ptr<ExprAST> LHS) {
  \/\/ If this is a binop, find its precedence.
  while (true) {
    int TokPrec = GetTokPrecedence();
\u200b
    \/\/ If this is a binop that binds at least as tightly as the current binop,
    \/\/ consume it, otherwise we are done.
    if (TokPrec < ExprPrec)
      return LHS;
\u200b
    \/\/ Okay, we know this is a binop.
    int BinOp = CurTok;
    getNextToken(); \/\/ eat binop
\u200b
    \/\/ Parse the primary expression after the binary operator.
    auto RHS = ParsePrimary();
    if (!RHS)
      return nullptr;
\u200b
    \/\/ If BinOp binds less tightly with RHS than the operator after RHS, let
    \/\/ the pending operator take RHS as its LHS.
    int NextPrec = GetTokPrecedence();
    if (TokPrec < NextPrec) {
      RHS = ParseBinOpRHS(TokPrec + 1, std::move(RHS));
      if (!RHS)
        return nullptr;
    }
\u200b
    \/\/ Merge LHS\/RHS.
    LHS = std::make_unique<BinaryExprAST>(BinOp, std::move(LHS),
                                           std::move(RHS));
  }
}
\u200b
\/\/\/ expression
\/\/\/   ::= primary binoprhs
\/\/\/
static std::unique_ptr<ExprAST> ParseExpression() {
  auto LHS = ParsePrimary();
  if (!LHS)
    return nullptr;
\u200b
  return ParseBinOpRHS(0, std::move(LHS));
}
\u200b
\/\/\/ prototype
\/\/\/   ::= id '(' id* ')'
static std::unique_ptr<PrototypeAST> ParsePrototype() {
  if (CurTok != tok_identifier)
    return LogErrorP(\"Expected function name in prototype\");
\u200b
  std::string FnName = IdentifierStr;
  getNextToken();
\u200b
  if (CurTok != '(')
    return LogErrorP(\"Expected '(' in prototype\");
\u200b
  std::vector<std::string> ArgNames;
  while (getNextToken() == tok_identifier)
    ArgNames.push_back(IdentifierStr);
  if (CurTok != ')')
    return LogErrorP(\"Expected ')' in prototype\");
\u200b
  \/\/ success.
  getNextToken(); \/\/ eat ')'.
\u200b
  return std::make_unique<PrototypeAST>(FnName, std::move(ArgNames));
}
\u200b
\/\/\/ definition ::= 'def' prototype expression
static std::unique_ptr<FunctionAST> ParseDefinition() {
  fprintf(stdout,\"[+] print %s\\n\" ,IdentifierStr.c_str());
  \/\/ std::cout << IdentifierStr <<std::endl;
  getNextToken(); \/\/ eat def.
  fprintf(stdout,\"[+] print %s\\n\" ,IdentifierStr.c_str());
  auto Proto = ParsePrototype();
  if (!Proto)
    return nullptr;
\u200b
  if (auto E = ParseExpression())
    return std::make_unique<FunctionAST>(std::move(Proto), std::move(E));
  return nullptr;
}
\u200b
\/\/\/ toplevelexpr ::= expression
static std::unique_ptr<FunctionAST> ParseTopLevelExpr() {
  if (auto E = ParseExpression()) {
    \/\/ Make an anonymous proto.
    auto Proto = std::make_unique<PrototypeAST>(\"__anon_expr\",
                                                 std::vector<std::string>());
    return std::make_unique<FunctionAST>(std::move(Proto), std::move(E));
  }
  return nullptr;
}
\u200b
\/\/\/ external ::= 'extern' prototype
static std::unique_ptr<PrototypeAST> ParseExtern() {
  getNextToken(); \/\/ eat extern.
  return ParsePrototype();
}
\u200b
\/\/===----------------------------------------------------------------------===\/\/
\/\/ Top-Level parsing
\/\/===----------------------------------------------------------------------===\/\/
\u200b
static void HandleDefinition() {
  if (ParseDefinition()) {
    fprintf(stderr, \"Parsed a function definition.\\n\");
  } else {
    \/\/ Skip token for error recovery.
    getNextToken();
  }
}
\u200b
static void HandleExtern() {
  if (ParseExtern()) {
    fprintf(stderr, \"Parsed an extern\\n\");
  } else {
    \/\/ Skip token for error recovery.
    getNextToken();
  }
}
\u200b
static void HandleTopLevelExpression() {
  \/\/ Evaluate a top-level expression into an anonymous function.
  if (ParseTopLevelExpr()) {
    fprintf(stderr, \"Parsed a top-level expr\\n\");
  } else {
    \/\/ Skip token for error recovery.
    getNextToken();
  }
}
\u200b
\/\/\/ top ::= definition | external | expression | ';'
static void MainLoop() {
  while (true) {
    fprintf(stderr, \"ready> \");
    switch (CurTok) {
    case tok_eof:
      return;
    case ';': \/\/ ignore top-level semicolons.
      getNextToken();
      break;
    case tok_def:
      HandleDefinition();
      break;
    case tok_extern:
      HandleExtern();
      break;
    default:
      HandleTopLevelExpression();
      break;
    }
  }
}
\u200b
\/\/===----------------------------------------------------------------------===\/\/
\/\/ Main driver code.
\/\/===----------------------------------------------------------------------===\/\/
\u200b
int main() {
  \/\/ Install standard binary operators.
  \/\/ 1 is lowest precedence.
  BinopPrecedence['<'] = 10;
  BinopPrecedence['+'] = 20;
  BinopPrecedence['-'] = 20;
  BinopPrecedence['*'] = 40; \/\/ highest.
\u200b
  \/\/ Prime the first token.
  fprintf(stderr, \"ready> \");
  getNextToken();
\u200b
  \/\/ Run the main \"interpreter loop\" now.
  MainLoop();
\u200b
  return 0;
}<\/pre>\n\n\n\n
\u7f16\u8bd1\u547d\u4ee4\uff08\u81ea\u5df1\u6539\u4e0b\uff09\uff1a$LLVM_HOME\/bin\/clang++ -g -O3 toy.cpp -std=c++14 llvm-config --ldflags --libs<\/code> -lpthread -I$LLVM_HOME\/include -ldl -lz -ltinfo<\/p>\n\n\n\n
\u4ec0\u4e48\u662f\u62bd\u8c61\u8bed\u6cd5\u6811\uff0c\u4e2a\u4eba\u7b80\u5355\u7406\u89e3\u5c31\u662f\u7528\u4e00\u9897\u6811\u6765\u5b58\u4e00\u4e2a\u8868\u8fbe\u5f0f\u6216\u8005\u8bf4\u8bed\u6cd5\u5757\uff08\u4e0d\u4e00\u5b9a\u6b63\u786e\uff09\uff0c\u8868\u8fbe\u5f0f\u8d8a\u590d\u6742\u6811\u8d8a\u590d\u6742\u3002\u901a\u8fc7\u76f4\u63a5\u770b\u4ee3\u7801\u53d1\u73b0\u4e07\u82b1\u7b52\u652f\u6301\u5b9a\u4e49\u4f7f\u7528\u4e86\u56db\u79cdtoken \uff0cdef extern identifier number,\u7528\u4e8e\u8bcd\u6cd5\u5206\u6790\uff0c\u9996\u5148\u5206\u6790\u8bed\u53e5\uff0c\u53ef\u4ee5\u662f\u6807\u8bc6\u7b26\u3001\u6570\u5b57\u6216\u8005\u8868\u8fbe\u5f0f\u3002<\/p>\n\n\n\n
\u4ee5A(1+1)\u4e3a\u4f8b\u5b50\u5206\u6790\uff1a<\/p>\n\n\n\n
\u7a0b\u5e8f\u6267\u884c\u6d41\u7a0b\u4e3a HandleTopLevelExpression -> ParseTopLevelExpr -> ParseExpression->ParsePrimary -> ParseIdentifierExpr ->ParseExpression->ParsePrimary ->ParseNumberExpr ->ParseBinOpRHS- > ParsePrimary ->ParseNumberExpr ->ParseBinOpRHS(\u4e00\u6b21\u8fed\u4ee3) \u5148\u770b\u7a0b\u5e8f\u6267\u884c\u4e0d\u5173\u5fc3\u8bed\u6cd5\u6811\u7ec4\u6210\u3002\u5c06A(1+1)\u770b\u4f5c\u4e00\u4e2a\u8868\u8fbe\u5f0f\uff0cA\u4e3aidentifier \u518d\u5c061+1 \u770b\u4f5c\u4e00\u4e2a\u8868\u8fbe\u5f0f\uff0c\u89e3\u6790 1 \u540e\u9047\u5230”+” \u540e\u5224\u65ad\u64cd\u4f5c\u7b26\u4f18\u5148\u7ea7\u540e\u89e3\u6790\u4e0b\u4e00\u4e2atoken\uff0c\u6b64\u65f6\u4e0b\u4e00\u4e2atoken\u4e3a1 \uff0c\u5728\u5224\u65ad\u518d\u4e0b\u4e00\u4e2atoken\u7684\u4f18\u5148\u7ea7\uff0c\u56e0\u4e3a\u662f”)”\u8fd4\u56de”-1″,\u7136\u540e\u5c06”1 + 1″\u8868\u8fbe\u5f0f\u5408\u5e76\u518d\u8fd4\u56de\uff0c\u5c42\u5c42\u8fd4\u56de\u56de\u5230ParseIDentifierExpr ,\u6210\u4e3a”A”\u6807\u8bc6\u7b26\u7684\u53c2\u6570\uff0c\u8fd4\u56de\u5408\u6210\u7684\u8bed\u6cd5\u6811\uff0c\u8868\u8fbe\u5f0f\u89e3\u6790\u7ed3\u675f\u3002<\/p>\n\n\n\n
\u7136\u540e\u6211\u4eec\u518d\u770b\u5176\u4e2d\u4f7f\u7528\u4e86\u54ea\u4e9b\u8bed\u6cd5\u6811\uff0c\u5728ParseIdentifierExpr\u4f1a\u8fd4\u56de\u4e00\u4e2aCallExprAST \u8bed\u6cd5\u6811\uff0cCallExprAST\u8bed\u6cd5\u6811\u521d\u59cb\u5316\u65f6\u53c2\u6570\u4e3a\u51fd\u6570\u540d\u79f0callee \u548cBinaryExprAST \u8bed\u6cd5\u6811\uff0c\u5728\u8fd9\u4e2a\u4f8b\u5b50\u4e2d\uff0c\u4f7f\u7528\u7684\u7c7bParseNumberExpr\u8be5\u7c7b\u4ec5\u6709\u4e00\u4e2a\u6210\u5458\u8868\u793a\u8fd9\u4e2a\u6570\u503c\uff0c\u563f\u563f \u5176\u5b9e\u518d\u770b\u4e00\u773c\u5c31\u4f1a\u53d1\u73b0”1+1″\u5176\u5b9e\u4e0d\u662f\u8fd4\u56deParseNumberExpr \uff0c\u7531\u4e8e\u9047\u5230”+” \u4e14\u540e\u9762\u7684token\u5408\u6cd5\uff0c\u6240\u4ee5\u8fd4\u56de\u4e86\u4e00\u4e2aBinaryExprAST\u8bed\u6cd5\u6811\uff0c\u6700\u7ec8\u751f\u6210\u7684\u6811\u5c31\u662f:<\/p>\n\n\n\n
\"\"<\/figure>\n\n\n\n
<\/p>\n\n\n\n
https:\/\/llvm.org\/docs\/tutorial\/MyFirstLanguageFrontend\/index.html<\/a> MyfirstLanguagefrontend<\/p><\/blockquote>\n\n\n\n
\u5176\u4e2d\u7684\u5c0f\u63d2\u66f2\uff0c\u6253\u5370std::string \u62a5\u4e86\u4e2a\u9519\uff0c\u5185\u5bb9\u5fd8\u4e86 \u5c31\u662f\u7f3a\u5c11\u4e86stdlibc++ \u7684\u8c03\u8bd5\u4fe1\u606f\u5305\uff0c\u88c5\u4e00\u4e0b\u5c31\u597d\u3002<\/p>\n\n\n\n
sudo apt install libstdc++6-8-dbg<\/pre>\n\n\n\n
<\/p>\n","protected":false},"excerpt":{"rendered":"
\u7531\u4e8e\u5f88\u591a\u6559\u7a0b\u7f16\u5199\u7684pass\u5b58\u5728\u95ee\u9898\uff0c\u81ea\u5df1\u53c8\u4e0d\u4f1a\u6539 \uff0c\u672c\u7cfb\u5217\u6559\u7a0b\u76ee\u7684\u662f\u5b66\u4e60\u7f16\u5199llvm pass \u4e3b\u8981 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[1],"tags":[4],"_links":{"self":[{"href":"http:\/\/pareto.fun\/index.php?rest_route=\/wp\/v2\/posts\/21"}],"collection":[{"href":"http:\/\/pareto.fun\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/pareto.fun\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/pareto.fun\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/pareto.fun\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=21"}],"version-history":[{"count":7,"href":"http:\/\/pareto.fun\/index.php?rest_route=\/wp\/v2\/posts\/21\/revisions"}],"predecessor-version":[{"id":34,"href":"http:\/\/pareto.fun\/index.php?rest_route=\/wp\/v2\/posts\/21\/revisions\/34"}],"wp:attachment":[{"href":"http:\/\/pareto.fun\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=21"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/pareto.fun\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=21"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/pareto.fun\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=21"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}