BaseTools/Source/Python/Common/Expression.py

   1 ## @file
   2 # This file is used to parse and evaluate expression in directive or PCD value.
   3 #
   4 # Copyright (c) 2011 - 2018, Intel Corporation. All rights reserved.<BR>
   5 # SPDX-License-Identifier: BSD-2-Clause-Patent
   6
   7 ## Import Modules
   8 #
   9 from __future__ import print_function
  10 from __future__ import absolute_import
  11 from Common.GlobalData import *
  12 from CommonDataClass.Exceptions import BadExpression
  13 from CommonDataClass.Exceptions import WrnExpression
  14 from .Misc import GuidStringToGuidStructureString, ParseFieldValue,CopyDict
  15 import Common.EdkLogger as EdkLogger
  16 import copy
  17 from Common.DataType import *
  18 import sys
  19 from random import sample
  20 import string
  21
  22 ERR_STRING_EXPR         = 'This operator cannot be used in string expression: [%s].'
  23 ERR_SNYTAX              = 'Syntax error, the rest of expression cannot be evaluated: [%s].'
  24 ERR_MATCH               = 'No matching right parenthesis.'
  25 ERR_STRING_TOKEN        = 'Bad string token: [%s].'
  26 ERR_MACRO_TOKEN         = 'Bad macro token: [%s].'
  27 ERR_EMPTY_TOKEN         = 'Empty token is not allowed.'
  28 ERR_PCD_RESOLVE         = 'The PCD should be FeatureFlag type or FixedAtBuild type: [%s].'
  29 ERR_VALID_TOKEN         = 'No more valid token found from rest of string: [%s].'
  30 ERR_EXPR_TYPE           = 'Different types found in expression.'
  31 ERR_OPERATOR_UNSUPPORT  = 'Unsupported operator: [%s]'
  32 ERR_REL_NOT_IN          = 'Expect "IN" after "not" operator.'
  33 WRN_BOOL_EXPR           = 'Operand of boolean type cannot be used in arithmetic expression.'
  34 WRN_EQCMP_STR_OTHERS    = '== Comparison between Operand of string type and Boolean/Number Type always return False.'
  35 WRN_NECMP_STR_OTHERS    = '!= Comparison between Operand of string type and Boolean/Number Type always return True.'
  36 ERR_RELCMP_STR_OTHERS   = 'Operator taking Operand of string type and Boolean/Number Type is not allowed: [%s].'
  37 ERR_STRING_CMP          = 'Unicode string and general string cannot be compared: [%s %s %s]'
  38 ERR_ARRAY_TOKEN         = 'Bad C array or C format GUID token: [%s].'
  39 ERR_ARRAY_ELE           = 'This must be HEX value for NList or Array: [%s].'
  40 ERR_EMPTY_EXPR          = 'Empty expression is not allowed.'
  41 ERR_IN_OPERAND          = 'Macro after IN operator can only be: $(FAMILY), $(ARCH), $(TOOL_CHAIN_TAG) and $(TARGET).'
  42
  43 __ValidString = re.compile(r'[_a-zA-Z][_0-9a-zA-Z]*$')
  44 _ReLabel = re.compile('LABEL\((\w+)\)')
  45 _ReOffset = re.compile('OFFSET_OF\((\w+)\)')
  46 PcdPattern = re.compile(r'^[_a-zA-Z][0-9A-Za-z_]*\.[_a-zA-Z][0-9A-Za-z_]*$')
  47
  48 ## SplitString
  49 #  Split string to list according double quote
  50 #  For example: abc"de\"f"ghi"jkl"mn will be: ['abc', '"de\"f"', 'ghi', '"jkl"', 'mn']
  51 #
  52 def SplitString(String):
  53     # There might be escaped quote: "abc\"def\\\"ghi", 'abc\'def\\\'ghi'
  54     RanStr = ''.join(sample(string.ascii_letters + string.digits, 8))
  55     String = String.replace('\\\\', RanStr).strip()
  56     RetList = []
  57     InSingleQuote = False
  58     InDoubleQuote = False
  59     Item = ''
  60     for i, ch in enumerate(String):
  61         if ch == '"' and not InSingleQuote:
  62             if String[i - 1] != '\\':
  63                 InDoubleQuote = not InDoubleQuote
  64             if not InDoubleQuote:
  65                 Item += String[i]
  66                 RetList.append(Item)
  67                 Item = ''
  68                 continue
  69             if Item:
  70                 RetList.append(Item)
  71                 Item = ''
  72         elif ch == "'" and not InDoubleQuote:
  73             if String[i - 1] != '\\':
  74                 InSingleQuote = not InSingleQuote
  75             if not InSingleQuote:
  76                 Item += String[i]
  77                 RetList.append(Item)
  78                 Item = ''
  79                 continue
  80             if Item:
  81                 RetList.append(Item)
  82                 Item = ''
  83         Item += String[i]
  84     if InSingleQuote or InDoubleQuote:
  85         raise BadExpression(ERR_STRING_TOKEN % Item)
  86     if Item:
  87         RetList.append(Item)
  88     for i, ch in enumerate(RetList):
  89         if RanStr in ch:
  90             RetList[i] = ch.replace(RanStr,'\\\\')
  91     return RetList
  92
  93 def SplitPcdValueString(String):
  94     # There might be escaped comma in GUID() or DEVICE_PATH() or " "
  95     # or ' ' or L' ' or L" "
  96     RanStr = ''.join(sample(string.ascii_letters + string.digits, 8))
  97     String = String.replace('\\\\', RanStr).strip()
  98     RetList = []
  99     InParenthesis = 0
 100     InSingleQuote = False
 101     InDoubleQuote = False
 102     Item = ''
 103     for i, ch in enumerate(String):
 104         if ch == '(':
 105             InParenthesis += 1
 106         elif ch == ')':
 107             if InParenthesis:
 108                 InParenthesis -= 1
 109             else:
 110                 raise BadExpression(ERR_STRING_TOKEN % Item)
 111         elif ch == '"' and not InSingleQuote:
 112             if String[i-1] != '\\':
 113                 InDoubleQuote = not InDoubleQuote
 114         elif ch == "'" and not InDoubleQuote:
 115             if String[i-1] != '\\':
 116                 InSingleQuote = not InSingleQuote
 117         elif ch == ',':
 118             if InParenthesis or InSingleQuote or InDoubleQuote:
 119                 Item += String[i]
 120                 continue
 121             elif Item:
 122                 RetList.append(Item)
 123                 Item = ''
 124             continue
 125         Item += String[i]
 126     if InSingleQuote or InDoubleQuote or InParenthesis:
 127         raise BadExpression(ERR_STRING_TOKEN % Item)
 128     if Item:
 129         RetList.append(Item)
 130     for i, ch in enumerate(RetList):
 131         if RanStr in ch:
 132             RetList[i] = ch.replace(RanStr,'\\\\')
 133     return RetList
 134
 135 def IsValidCName(Str):
 136     return True if __ValidString.match(Str) else False
 137
 138 def BuildOptionValue(PcdValue, GuidDict):
 139     if PcdValue.startswith('H'):
 140         InputValue = PcdValue[1:]
 141     elif PcdValue.startswith("L'") or PcdValue.startswith("'"):
 142         InputValue = PcdValue
 143     elif PcdValue.startswith('L'):
 144         InputValue = 'L"' + PcdValue[1:] + '"'
 145     else:
 146         InputValue = PcdValue
 147     try:
 148         PcdValue = ValueExpressionEx(InputValue, TAB_VOID, GuidDict)(True)
 149     except:
 150         pass
 151
 152     return PcdValue
 153
 154 ## ReplaceExprMacro
 155 #
 156 def ReplaceExprMacro(String, Macros, ExceptionList = None):
 157     StrList = SplitString(String)
 158     for i, String in enumerate(StrList):
 159         InQuote = False
 160         if String.startswith('"'):
 161             InQuote = True
 162         MacroStartPos = String.find('$(')
 163         if MacroStartPos < 0:
 164             for Pcd in gPlatformPcds:
 165                 if Pcd in String:
 166                     if Pcd not in gConditionalPcds:
 167                         gConditionalPcds.append(Pcd)
 168             continue
 169         RetStr = ''
 170         while MacroStartPos >= 0:
 171             RetStr = String[0:MacroStartPos]
 172             MacroEndPos = String.find(')', MacroStartPos)
 173             if MacroEndPos < 0:
 174                 raise BadExpression(ERR_MACRO_TOKEN % String[MacroStartPos:])
 175             Macro = String[MacroStartPos+2:MacroEndPos]
 176             if Macro not in Macros:
 177                 # From C reference manual:
 178                 # If an undefined macro name appears in the constant-expression of
 179                 # !if or !elif, it is replaced by the integer constant 0.
 180                 RetStr += '0'
 181             elif not InQuote:
 182                 Tklst = RetStr.split()
 183                 if Tklst and Tklst[-1] in {'IN', 'in'} and ExceptionList and Macro not in ExceptionList:
 184                     raise BadExpression(ERR_IN_OPERAND)
 185                 # Make sure the macro in exception list is encapsulated by double quote
 186                 # For example: DEFINE ARCH = IA32 X64
 187                 # $(ARCH) is replaced with "IA32 X64"
 188                 if ExceptionList and Macro in ExceptionList:
 189                     RetStr += '"' + Macros[Macro] + '"'
 190                 elif Macros[Macro].strip():
 191                     RetStr += Macros[Macro]
 192                 else:
 193                     RetStr += '""'
 194             else:
 195                 RetStr += Macros[Macro]
 196             RetStr += String[MacroEndPos+1:]
 197             String = RetStr
 198             MacroStartPos = String.find('$(')
 199         StrList[i] = RetStr
 200     return ''.join(StrList)
 201
 202 # transfer int to string for in/not in expression
 203 def IntToStr(Value):
 204     StrList = []
 205     while Value > 0:
 206         StrList.append(chr(Value & 0xff))
 207         Value = Value >> 8
 208     Value = '"' + ''.join(StrList) + '"'
 209     return Value
 210
 211 SupportedInMacroList = ['TARGET', 'TOOL_CHAIN_TAG', 'ARCH', 'FAMILY']
 212
 213 class BaseExpression(object):
 214     def __init__(self, *args, **kwargs):
 215         super(BaseExpression, self).__init__()
 216
 217     # Check if current token matches the operators given from parameter
 218     def _IsOperator(self, OpSet):
 219         Idx = self._Idx
 220         self._GetOperator()
 221         if self._Token in OpSet:
 222             if self._Token in self.LogicalOperators:
 223                 self._Token = self.LogicalOperators[self._Token]
 224             return True
 225         self._Idx = Idx
 226         return False
 227
 228 class ValueExpression(BaseExpression):
 229     # Logical operator mapping
 230     LogicalOperators = {
 231         '&&' : 'and', '||' : 'or',
 232         '!'  : 'not', 'AND': 'and',
 233         'OR' : 'or' , 'NOT': 'not',
 234         'XOR': '^'  , 'xor': '^',
 235         'EQ' : '==' , 'NE' : '!=',
 236         'GT' : '>'  , 'LT' : '<',
 237         'GE' : '>=' , 'LE' : '<=',
 238         'IN' : 'in'
 239     }
 240
 241     NonLetterOpLst = ['+', '-', TAB_STAR, '/', '%', '&', '|', '^', '~', '<<', '>>', '!', '=', '>', '<', '?', ':']
 242
 243
 244     SymbolPattern = re.compile("("
 245                                  "\$\([A-Z][A-Z0-9_]*\)|\$\(\w+\.\w+\)|\w+\.\w+|"
 246                                  "&&|\|\||!(?!=)|"
 247                                  "(?<=\W)AND(?=\W)|(?<=\W)OR(?=\W)|(?<=\W)NOT(?=\W)|(?<=\W)XOR(?=\W)|"
 248                                  "(?<=\W)EQ(?=\W)|(?<=\W)NE(?=\W)|(?<=\W)GT(?=\W)|(?<=\W)LT(?=\W)|(?<=\W)GE(?=\W)|(?<=\W)LE(?=\W)"
 249                                ")")
 250
 251     @staticmethod
 252     def Eval(Operator, Oprand1, Oprand2 = None):
 253         WrnExp = None
 254
 255         if Operator not in {"==", "!=", ">=", "<=", ">", "<", "in", "not in"} and \
 256             (isinstance(Oprand1, type('')) or isinstance(Oprand2, type(''))):
 257             raise BadExpression(ERR_STRING_EXPR % Operator)
 258         if Operator in {'in', 'not in'}:
 259             if not isinstance(Oprand1, type('')):
 260                 Oprand1 = IntToStr(Oprand1)
 261             if not isinstance(Oprand2, type('')):
 262                 Oprand2 = IntToStr(Oprand2)
 263         TypeDict = {
 264             type(0)  : 0,
 265             # For python2 long type
 266             type(sys.maxsize + 1) : 0,
 267             type('') : 1,
 268             type(True) : 2
 269         }
 270
 271         EvalStr = ''
 272         if Operator in {"!", "NOT", "not"}:
 273             if isinstance(Oprand1, type('')):
 274                 raise BadExpression(ERR_STRING_EXPR % Operator)
 275             EvalStr = 'not Oprand1'
 276         elif Operator in {"~"}:
 277             if isinstance(Oprand1, type('')):
 278                 raise BadExpression(ERR_STRING_EXPR % Operator)
 279             EvalStr = '~ Oprand1'
 280         else:
 281             if Operator in {"+", "-"} and (type(True) in {type(Oprand1), type(Oprand2)}):
 282                 # Boolean in '+'/'-' will be evaluated but raise warning
 283                 WrnExp = WrnExpression(WRN_BOOL_EXPR)
 284             elif type('') in {type(Oprand1), type(Oprand2)} and not isinstance(Oprand1, type(Oprand2)):
 285                 # == between string and number/boolean will always return False, != return True
 286                 if Operator == "==":
 287                     WrnExp = WrnExpression(WRN_EQCMP_STR_OTHERS)
 288                     WrnExp.result = False
 289                     raise WrnExp
 290                 elif Operator == "!=":
 291                     WrnExp = WrnExpression(WRN_NECMP_STR_OTHERS)
 292                     WrnExp.result = True
 293                     raise WrnExp
 294                 else:
 295                     raise BadExpression(ERR_RELCMP_STR_OTHERS % Operator)
 296             elif TypeDict[type(Oprand1)] != TypeDict[type(Oprand2)]:
 297                 if Operator in {"==", "!=", ">=", "<=", ">", "<"} and set((TypeDict[type(Oprand1)], TypeDict[type(Oprand2)])) == set((TypeDict[type(True)], TypeDict[type(0)])):
 298                     # comparison between number and boolean is allowed
 299                     pass
 300                 elif Operator in {'&', '|', '^', "and", "or"} and set((TypeDict[type(Oprand1)], TypeDict[type(Oprand2)])) == set((TypeDict[type(True)], TypeDict[type(0)])):
 301                     # bitwise and logical operation between number and boolean is allowed
 302                     pass
 303                 else:
 304                     raise BadExpression(ERR_EXPR_TYPE)
 305             if isinstance(Oprand1, type('')) and isinstance(Oprand2, type('')):
 306                 if ((Oprand1.startswith('L"') or Oprand1.startswith("L'")) and (not Oprand2.startswith('L"')) and (not Oprand2.startswith("L'"))) or \
 307                         (((not Oprand1.startswith('L"')) and (not Oprand1.startswith("L'"))) and (Oprand2.startswith('L"') or Oprand2.startswith("L'"))):
 308                     raise BadExpression(ERR_STRING_CMP % (Oprand1, Operator, Oprand2))
 309             if 'in' in Operator and isinstance(Oprand2, type('')):
 310                 Oprand2 = Oprand2.split()
 311             EvalStr = 'Oprand1 ' + Operator + ' Oprand2'
 312
 313         # Local symbols used by built in eval function
 314         Dict = {
 315             'Oprand1' : Oprand1,
 316             'Oprand2' : Oprand2
 317         }
 318         try:
 319             Val = eval(EvalStr, {}, Dict)
 320         except Exception as Excpt:
 321             raise BadExpression(str(Excpt))
 322
 323         if Operator in {'and', 'or'}:
 324             if Val:
 325                 Val = True
 326             else:
 327                 Val = False
 328
 329         if WrnExp:
 330             WrnExp.result = Val
 331             raise WrnExp
 332         return Val
 333
 334     def __init__(self, Expression, SymbolTable={}):
 335         super(ValueExpression, self).__init__(self, Expression, SymbolTable)
 336         self._NoProcess = False
 337         if not isinstance(Expression, type('')):
 338             self._Expr = Expression
 339             self._NoProcess = True
 340             return
 341
 342         self._Expr = ReplaceExprMacro(Expression.strip(),
 343                                   SymbolTable,
 344                                   SupportedInMacroList)
 345
 346         if not self._Expr.strip():
 347             raise BadExpression(ERR_EMPTY_EXPR)
 348
 349         #
 350         # The symbol table including PCD and macro mapping
 351         #
 352         self._Symb = CopyDict(SymbolTable)
 353         self._Symb.update(self.LogicalOperators)
 354         self._Idx = 0
 355         self._Len = len(self._Expr)
 356         self._Token = ''
 357         self._WarnExcept = None
 358
 359         # Literal token without any conversion
 360         self._LiteralToken = ''
 361
 362     # Public entry for this class
 363     #   @param RealValue: False: only evaluate if the expression is true or false, used for conditional expression
 364     #                     True : return the evaluated str(value), used for PCD value
 365     #
 366     #   @return: True or False if RealValue is False
 367     #            Evaluated value of string format if RealValue is True
 368     #
 369     def __call__(self, RealValue=False, Depth=0):
 370         if self._NoProcess:
 371             return self._Expr
 372
 373         self._Depth = Depth
 374
 375         self._Expr = self._Expr.strip()
 376         if RealValue and Depth == 0:
 377             self._Token = self._Expr
 378             if self.__IsNumberToken():
 379                 return self._Expr
 380             Token = ''
 381             try:
 382                 Token = self._GetToken()
 383             except BadExpression:
 384                 pass
 385             if isinstance(Token, type('')) and Token.startswith('{') and Token.endswith('}') and self._Idx >= self._Len:
 386                 return self._Expr
 387
 388             self._Idx = 0
 389             self._Token = ''
 390
 391         Val = self._ConExpr()
 392         RealVal = Val
 393         if isinstance(Val, type('')):
 394             if Val == 'L""':
 395                 Val = False
 396             elif not Val:
 397                 Val = False
 398                 RealVal = '""'
 399             elif not Val.startswith('L"') and not Val.startswith('{') and not Val.startswith("L'") and not Val.startswith("'"):
 400                 Val = True
 401                 RealVal = '"' + RealVal + '"'
 402
 403         # The expression has been parsed, but the end of expression is not reached
 404         # It means the rest does not comply EBNF of <Expression>
 405         if self._Idx != self._Len:
 406             raise BadExpression(ERR_SNYTAX % self._Expr[self._Idx:])
 407
 408         if RealValue:
 409             RetVal = str(RealVal)
 410         elif Val:
 411             RetVal = True
 412         else:
 413             RetVal = False
 414
 415         if self._WarnExcept:
 416             self._WarnExcept.result = RetVal
 417             raise self._WarnExcept
 418         else:
 419             return RetVal
 420
 421     # Template function to parse binary operators which have same precedence
 422     # Expr [Operator Expr]*
 423     def _ExprFuncTemplate(self, EvalFunc, OpSet):
 424         Val = EvalFunc()
 425         while self._IsOperator(OpSet):
 426             Op = self._Token
 427             if Op == '?':
 428                 Val2 = EvalFunc()
 429                 if self._IsOperator({':'}):
 430                     Val3 = EvalFunc()
 431                 if Val:
 432                     Val = Val2
 433                 else:
 434                     Val = Val3
 435                 continue
 436             #
 437             # PEP 238 -- Changing the Division Operator
 438             # x/y to return a reasonable approximation of the mathematical result of the division ("true division")
 439             # x//y to return the floor ("floor division")
 440             #
 441             if Op == '/':
 442                 Op = '//'
 443             try:
 444                 Val = self.Eval(Op, Val, EvalFunc())
 445             except WrnExpression as Warn:
 446                 self._WarnExcept = Warn
 447                 Val = Warn.result
 448         return Val
 449     # A [? B]*
 450     def _ConExpr(self):
 451         return self._ExprFuncTemplate(self._OrExpr, {'?', ':'})
 452
 453     # A [|| B]*
 454     def _OrExpr(self):
 455         return self._ExprFuncTemplate(self._AndExpr, {"OR", "or", "||"})
 456
 457     # A [&& B]*
 458     def _AndExpr(self):
 459         return self._ExprFuncTemplate(self._BitOr, {"AND", "and", "&&"})
 460
 461     # A [ | B]*
 462     def _BitOr(self):
 463         return self._ExprFuncTemplate(self._BitXor, {"|"})
 464
 465     # A [ ^ B]*
 466     def _BitXor(self):
 467         return self._ExprFuncTemplate(self._BitAnd, {"XOR", "xor", "^"})
 468
 469     # A [ & B]*
 470     def _BitAnd(self):
 471         return self._ExprFuncTemplate(self._EqExpr, {"&"})
 472
 473     # A [ == B]*
 474     def _EqExpr(self):
 475         Val = self._RelExpr()
 476         while self._IsOperator({"==", "!=", "EQ", "NE", "IN", "in", "!", "NOT", "not"}):
 477             Op = self._Token
 478             if Op in {"!", "NOT", "not"}:
 479                 if not self._IsOperator({"IN", "in"}):
 480                     raise BadExpression(ERR_REL_NOT_IN)
 481                 Op += ' ' + self._Token
 482             try:
 483                 Val = self.Eval(Op, Val, self._RelExpr())
 484             except WrnExpression as Warn:
 485                 self._WarnExcept = Warn
 486                 Val = Warn.result
 487         return Val
 488
 489     # A [ > B]*
 490     def _RelExpr(self):
 491         return self._ExprFuncTemplate(self._ShiftExpr, {"<=", ">=", "<", ">", "LE", "GE", "LT", "GT"})
 492
 493     def _ShiftExpr(self):
 494         return self._ExprFuncTemplate(self._AddExpr, {"<<", ">>"})
 495
 496     # A [ + B]*
 497     def _AddExpr(self):
 498         return self._ExprFuncTemplate(self._MulExpr, {"+", "-"})
 499
 500     # A [ * B]*
 501     def _MulExpr(self):
 502         return self._ExprFuncTemplate(self._UnaryExpr, {TAB_STAR, "/", "%"})
 503
 504     # [!]*A
 505     def _UnaryExpr(self):
 506         if self._IsOperator({"!", "NOT", "not"}):
 507             Val = self._UnaryExpr()
 508             try:
 509                 return self.Eval('not', Val)
 510             except WrnExpression as Warn:
 511                 self._WarnExcept = Warn
 512                 return Warn.result
 513         if self._IsOperator({"~"}):
 514             Val = self._UnaryExpr()
 515             try:
 516                 return self.Eval('~', Val)
 517             except WrnExpression as Warn:
 518                 self._WarnExcept = Warn
 519                 return Warn.result
 520         return self._IdenExpr()
 521
 522     # Parse identifier or encapsulated expression
 523     def _IdenExpr(self):
 524         Tk = self._GetToken()
 525         if Tk == '(':
 526             Val = self._ConExpr()
 527             try:
 528                 # _GetToken may also raise BadExpression
 529                 if self._GetToken() != ')':
 530                     raise BadExpression(ERR_MATCH)
 531             except BadExpression:
 532                 raise BadExpression(ERR_MATCH)
 533             return Val
 534         return Tk
 535
 536     # Skip whitespace or tab
 537     def __SkipWS(self):
 538         for Char in self._Expr[self._Idx:]:
 539             if Char not in ' \t':
 540                 break
 541             self._Idx += 1
 542
 543     # Try to convert string to number
 544     def __IsNumberToken(self):
 545         Radix = 10
 546         if self._Token.lower()[0:2] == '0x' and len(self._Token) > 2:
 547             Radix = 16
 548         if self._Token.startswith('"') or self._Token.startswith('L"'):
 549             Flag = 0
 550             for Index in range(len(self._Token)):
 551                 if self._Token[Index] in {'"'}:
 552                     if self._Token[Index - 1] == '\\':
 553                         continue
 554                     Flag += 1
 555             if Flag == 2 and self._Token.endswith('"'):
 556                 return True
 557         if self._Token.startswith("'") or self._Token.startswith("L'"):
 558             Flag = 0
 559             for Index in range(len(self._Token)):
 560                 if self._Token[Index] in {"'"}:
 561                     if self._Token[Index - 1] == '\\':
 562                         continue
 563                     Flag += 1
 564             if Flag == 2 and self._Token.endswith("'"):
 565                 return True
 566         try:
 567             self._Token = int(self._Token, Radix)
 568             return True
 569         except ValueError:
 570             return False
 571         except TypeError:
 572             return False
 573
 574     # Parse array: {...}
 575     def __GetArray(self):
 576         Token = '{'
 577         self._Idx += 1
 578         self.__GetNList(True)
 579         Token += self._LiteralToken
 580         if self._Idx >= self._Len or self._Expr[self._Idx] != '}':
 581             raise BadExpression(ERR_ARRAY_TOKEN % Token)
 582         Token += '}'
 583
 584         # All whitespace and tabs in array are already stripped.
 585         IsArray = IsGuid = False
 586         if len(Token.split(',')) == 11 and len(Token.split(',{')) == 2 \
 587             and len(Token.split('},')) == 1:
 588             HexLen = [11, 6, 6, 5, 4, 4, 4, 4, 4, 4, 6]
 589             HexList= Token.split(',')
 590             if HexList[3].startswith('{') and \
 591                 not [Index for Index, Hex in enumerate(HexList) if len(Hex) > HexLen[Index]]:
 592                 IsGuid = True
 593         if Token.lstrip('{').rstrip('}').find('{') == -1:
 594             if not [Hex for Hex in Token.lstrip('{').rstrip('}').split(',') if len(Hex) > 4]:
 595                 IsArray = True
 596         if not IsArray and not IsGuid:
 597             raise BadExpression(ERR_ARRAY_TOKEN % Token)
 598         self._Idx += 1
 599         self._Token = self._LiteralToken = Token
 600         return self._Token
 601
 602     # Parse string, the format must be: "..."
 603     def __GetString(self):
 604         Idx = self._Idx
 605
 606         # Skip left quote
 607         self._Idx += 1
 608
 609         # Replace escape \\\", \"
 610         if self._Expr[Idx] == '"':
 611             Expr = self._Expr[self._Idx:].replace('\\\\', '//').replace('\\\"', '\\\'')
 612             for Ch in Expr:
 613                 self._Idx += 1
 614                 if Ch == '"':
 615                     break
 616             self._Token = self._LiteralToken = self._Expr[Idx:self._Idx]
 617             if not self._Token.endswith('"'):
 618                 raise BadExpression(ERR_STRING_TOKEN % self._Token)
 619         #Replace escape \\\', \'
 620         elif self._Expr[Idx] == "'":
 621             Expr = self._Expr[self._Idx:].replace('\\\\', '//').replace("\\\'", "\\\"")
 622             for Ch in Expr:
 623                 self._Idx += 1
 624                 if Ch == "'":
 625                     break
 626             self._Token = self._LiteralToken = self._Expr[Idx:self._Idx]
 627             if not self._Token.endswith("'"):
 628                 raise BadExpression(ERR_STRING_TOKEN % self._Token)
 629         self._Token = self._Token[1:-1]
 630         return self._Token
 631
 632     # Get token that is comprised by alphanumeric, underscore or dot(used by PCD)
 633     # @param IsAlphaOp: Indicate if parsing general token or script operator(EQ, NE...)
 634     def __GetIdToken(self, IsAlphaOp = False):
 635         IdToken = ''
 636         for Ch in self._Expr[self._Idx:]:
 637             if not self.__IsIdChar(Ch) or ('?' in self._Expr and Ch == ':'):
 638                 break
 639             self._Idx += 1
 640             IdToken += Ch
 641
 642         self._Token = self._LiteralToken = IdToken
 643         if not IsAlphaOp:
 644             self.__ResolveToken()
 645         return self._Token
 646
 647     # Try to resolve token
 648     def __ResolveToken(self):
 649         if not self._Token:
 650             raise BadExpression(ERR_EMPTY_TOKEN)
 651
 652         # PCD token
 653         if PcdPattern.match(self._Token):
 654             if self._Token not in self._Symb:
 655                 Ex = BadExpression(ERR_PCD_RESOLVE % self._Token)
 656                 Ex.Pcd = self._Token
 657                 raise Ex
 658             self._Token = ValueExpression(self._Symb[self._Token], self._Symb)(True, self._Depth+1)
 659             if not isinstance(self._Token, type('')):
 660                 self._LiteralToken = hex(self._Token)
 661                 return
 662
 663         if self._Token.startswith('"'):
 664             self._Token = self._Token[1:-1]
 665         elif self._Token in {"FALSE", "false", "False"}:
 666             self._Token = False
 667         elif self._Token in {"TRUE", "true", "True"}:
 668             self._Token = True
 669         else:
 670             self.__IsNumberToken()
 671
 672     def __GetNList(self, InArray=False):
 673         self._GetSingleToken()
 674         if not self.__IsHexLiteral():
 675             if InArray:
 676                 raise BadExpression(ERR_ARRAY_ELE % self._Token)
 677             return self._Token
 678
 679         self.__SkipWS()
 680         Expr = self._Expr[self._Idx:]
 681         if not Expr.startswith(','):
 682             return self._Token
 683
 684         NList = self._LiteralToken
 685         while Expr.startswith(','):
 686             NList += ','
 687             self._Idx += 1
 688             self.__SkipWS()
 689             self._GetSingleToken()
 690             if not self.__IsHexLiteral():
 691                 raise BadExpression(ERR_ARRAY_ELE % self._Token)
 692             NList += self._LiteralToken
 693             self.__SkipWS()
 694             Expr = self._Expr[self._Idx:]
 695         self._Token = self._LiteralToken = NList
 696         return self._Token
 697
 698     def __IsHexLiteral(self):
 699         if self._LiteralToken.startswith('{') and \
 700             self._LiteralToken.endswith('}'):
 701             return True
 702
 703         if gHexPattern.match(self._LiteralToken):
 704             Token = self._LiteralToken[2:]
 705             if not Token:
 706                 self._LiteralToken = '0x0'
 707             else:
 708                 self._LiteralToken = '0x' + Token
 709             return True
 710         return False
 711
 712     def _GetToken(self):
 713         return self.__GetNList()
 714
 715     @staticmethod
 716     def __IsIdChar(Ch):
 717         return Ch in '._:' or Ch.isalnum()
 718
 719     # Parse operand
 720     def _GetSingleToken(self):
 721         self.__SkipWS()
 722         Expr = self._Expr[self._Idx:]
 723         if Expr.startswith('L"'):
 724             # Skip L
 725             self._Idx += 1
 726             UStr = self.__GetString()
 727             self._Token = 'L"' + UStr + '"'
 728             return self._Token
 729         elif Expr.startswith("L'"):
 730             # Skip L
 731             self._Idx += 1
 732             UStr = self.__GetString()
 733             self._Token = "L'" + UStr + "'"
 734             return self._Token
 735         elif Expr.startswith("'"):
 736             UStr = self.__GetString()
 737             self._Token = "'" + UStr + "'"
 738             return self._Token
 739         elif Expr.startswith('UINT'):
 740             Re = re.compile('(?:UINT8|UINT16|UINT32|UINT64)\((.+)\)')
 741             try:
 742                 RetValue = Re.search(Expr).group(1)
 743             except:
 744                  raise BadExpression('Invalid Expression %s' % Expr)
 745             Idx = self._Idx
 746             for Ch in Expr:
 747                 self._Idx += 1
 748                 if Ch == '(':
 749                     Prefix = self._Expr[Idx:self._Idx - 1]
 750                     Idx = self._Idx
 751                 if Ch == ')':
 752                     TmpValue = self._Expr[Idx :self._Idx - 1]
 753                     TmpValue = ValueExpression(TmpValue)(True)
 754                     TmpValue = '0x%x' % int(TmpValue) if not isinstance(TmpValue, type('')) else TmpValue
 755                     break
 756             self._Token, Size = ParseFieldValue(Prefix + '(' + TmpValue + ')')
 757             return  self._Token
 758
 759         self._Token = ''
 760         if Expr:
 761             Ch = Expr[0]
 762             Match = gGuidPattern.match(Expr)
 763             if Match and not Expr[Match.end():Match.end()+1].isalnum() \
 764                 and Expr[Match.end():Match.end()+1] != '_':
 765                 self._Idx += Match.end()
 766                 self._Token = ValueExpression(GuidStringToGuidStructureString(Expr[0:Match.end()]))(True, self._Depth+1)
 767                 return self._Token
 768             elif self.__IsIdChar(Ch):
 769                 return self.__GetIdToken()
 770             elif Ch == '"':
 771                 return self.__GetString()
 772             elif Ch == '{':
 773                 return self.__GetArray()
 774             elif Ch == '(' or Ch == ')':
 775                 self._Idx += 1
 776                 self._Token = Ch
 777                 return self._Token
 778
 779         raise BadExpression(ERR_VALID_TOKEN % Expr)
 780
 781     # Parse operator
 782     def _GetOperator(self):
 783         self.__SkipWS()
 784         LegalOpLst = ['&&', '||', '!=', '==', '>=', '<='] + self.NonLetterOpLst + ['?', ':']
 785
 786         self._Token = ''
 787         Expr = self._Expr[self._Idx:]
 788
 789         # Reach end of expression
 790         if not Expr:
 791             return ''
 792
 793         # Script operator: LT, GT, LE, GE, EQ, NE, and, or, xor, not
 794         if Expr[0].isalpha():
 795             return self.__GetIdToken(True)
 796
 797         # Start to get regular operator: +, -, <, > ...
 798         if Expr[0] not in self.NonLetterOpLst:
 799             return ''
 800
 801         OpToken = ''
 802         for Ch in Expr:
 803             if Ch in self.NonLetterOpLst:
 804                 if Ch in ['!', '~'] and OpToken:
 805                     break
 806                 self._Idx += 1
 807                 OpToken += Ch
 808             else:
 809                 break
 810
 811         if OpToken not in LegalOpLst:
 812             raise BadExpression(ERR_OPERATOR_UNSUPPORT % OpToken)
 813         self._Token = OpToken
 814         return OpToken
 815
 816 class ValueExpressionEx(ValueExpression):
 817     def __init__(self, PcdValue, PcdType, SymbolTable={}):
 818         ValueExpression.__init__(self, PcdValue, SymbolTable)
 819         self.PcdValue = PcdValue
 820         self.PcdType = PcdType
 821
 822     def __call__(self, RealValue=False, Depth=0):
 823         PcdValue = self.PcdValue
 824         if "{CODE(" not in PcdValue:
 825             try:
 826                 PcdValue = ValueExpression.__call__(self, RealValue, Depth)
 827                 if self.PcdType == TAB_VOID and (PcdValue.startswith("'") or PcdValue.startswith("L'")):
 828                     PcdValue, Size = ParseFieldValue(PcdValue)
 829                     PcdValueList = []
 830                     for I in range(Size):
 831                         PcdValueList.append('0x%02X'%(PcdValue & 0xff))
 832                         PcdValue = PcdValue >> 8
 833                     PcdValue = '{' + ','.join(PcdValueList) + '}'
 834                 elif self.PcdType in TAB_PCD_NUMERIC_TYPES and (PcdValue.startswith("'") or \
 835                           PcdValue.startswith('"') or PcdValue.startswith("L'") or PcdValue.startswith('L"') or PcdValue.startswith('{')):
 836                     raise BadExpression
 837             except WrnExpression as Value:
 838                 PcdValue = Value.result
 839             except BadExpression as Value:
 840                 if self.PcdType in TAB_PCD_NUMERIC_TYPES:
 841                     PcdValue = PcdValue.strip()
 842                     if PcdValue.startswith('{') and PcdValue.endswith('}'):
 843                         PcdValue = SplitPcdValueString(PcdValue[1:-1])
 844                     if isinstance(PcdValue, type([])):
 845                         TmpValue = 0
 846                         Size = 0
 847                         ValueType = ''
 848                         for Item in PcdValue:
 849                             Item = Item.strip()
 850                             if Item.startswith(TAB_UINT8):
 851                                 ItemSize = 1
 852                                 ValueType = TAB_UINT8
 853                             elif Item.startswith(TAB_UINT16):
 854                                 ItemSize = 2
 855                                 ValueType = TAB_UINT16
 856                             elif Item.startswith(TAB_UINT32):
 857                                 ItemSize = 4
 858                                 ValueType = TAB_UINT32
 859                             elif Item.startswith(TAB_UINT64):
 860                                 ItemSize = 8
 861                                 ValueType = TAB_UINT64
 862                             elif Item[0] in {'"', "'", 'L'}:
 863                                 ItemSize = 0
 864                                 ValueType = TAB_VOID
 865                             else:
 866                                 ItemSize = 0
 867                                 ValueType = TAB_UINT8
 868                             Item = ValueExpressionEx(Item, ValueType, self._Symb)(True)
 869                             if ItemSize == 0:
 870                                 try:
 871                                     tmpValue = int(Item, 0)
 872                                     if tmpValue > 255:
 873                                         raise BadExpression("Byte  array number %s should less than 0xFF." % Item)
 874                                 except BadExpression as Value:
 875                                     raise BadExpression(Value)
 876                                 except ValueError:
 877                                     pass
 878                                 ItemValue, ItemSize = ParseFieldValue(Item)
 879                             else:
 880                                 ItemValue = ParseFieldValue(Item)[0]
 881
 882                             if isinstance(ItemValue, type('')):
 883                                 ItemValue = int(ItemValue, 0)
 884
 885                             TmpValue = (ItemValue << (Size * 8)) | TmpValue
 886                             Size = Size + ItemSize
 887                     else:
 888                         try:
 889                             TmpValue, Size = ParseFieldValue(PcdValue)
 890                         except BadExpression as Value:
 891                             raise BadExpression("Type: %s, Value: %s, %s" % (self.PcdType, PcdValue, Value))
 892                     if isinstance(TmpValue, type('')):
 893                         try:
 894                             TmpValue = int(TmpValue)
 895                         except:
 896                             raise  BadExpression(Value)
 897                     else:
 898                         PcdValue = '0x%0{}X'.format(Size) % (TmpValue)
 899                     if TmpValue < 0:
 900                         raise  BadExpression('Type %s PCD Value is negative' % self.PcdType)
 901                     if self.PcdType == TAB_UINT8 and Size > 1:
 902                         raise BadExpression('Type %s PCD Value Size is Larger than 1 byte' % self.PcdType)
 903                     if self.PcdType == TAB_UINT16 and Size > 2:
 904                         raise BadExpression('Type %s PCD Value Size is Larger than 2 byte' % self.PcdType)
 905                     if self.PcdType == TAB_UINT32 and Size > 4:
 906                         raise BadExpression('Type %s PCD Value Size is Larger than 4 byte' % self.PcdType)
 907                     if self.PcdType == TAB_UINT64 and Size > 8:
 908                         raise BadExpression('Type %s PCD Value Size is Larger than 8 byte' % self.PcdType)
 909                 else:
 910                     try:
 911                         TmpValue = int(PcdValue)
 912                         TmpList = []
 913                         if TmpValue.bit_length() == 0:
 914                             PcdValue = '{0x00}'
 915                         else:
 916                             for I in range((TmpValue.bit_length() + 7) // 8):
 917                                 TmpList.append('0x%02x' % ((TmpValue >> I * 8) & 0xff))
 918                             PcdValue = '{' + ', '.join(TmpList) + '}'
 919                     except:
 920                         if PcdValue.strip().startswith('{'):
 921                             PcdValueList = SplitPcdValueString(PcdValue.strip()[1:-1])
 922                             LabelDict = {}
 923                             NewPcdValueList = []
 924                             LabelOffset = 0
 925                             for Item in PcdValueList:
 926                                 # compute byte offset of every LABEL
 927                                 LabelList = _ReLabel.findall(Item)
 928                                 Item = _ReLabel.sub('', Item)
 929                                 Item = Item.strip()
 930                                 if LabelList:
 931                                     for Label in LabelList:
 932                                         if not IsValidCName(Label):
 933                                             raise BadExpression('%s is not a valid c variable name' % Label)
 934                                         if Label not in LabelDict:
 935                                             LabelDict[Label] = str(LabelOffset)
 936                                 if Item.startswith(TAB_UINT8):
 937                                     LabelOffset = LabelOffset + 1
 938                                 elif Item.startswith(TAB_UINT16):
 939                                     LabelOffset = LabelOffset + 2
 940                                 elif Item.startswith(TAB_UINT32):
 941                                     LabelOffset = LabelOffset + 4
 942                                 elif Item.startswith(TAB_UINT64):
 943                                     LabelOffset = LabelOffset + 8
 944                                 else:
 945                                     try:
 946                                         ItemValue, ItemSize = ParseFieldValue(Item)
 947                                         LabelOffset = LabelOffset + ItemSize
 948                                     except:
 949                                         LabelOffset = LabelOffset + 1
 950
 951                             for Item in PcdValueList:
 952                                 # for LABEL parse
 953                                 Item = Item.strip()
 954                                 try:
 955                                     Item = _ReLabel.sub('', Item)
 956                                 except:
 957                                     pass
 958                                 try:
 959                                     OffsetList = _ReOffset.findall(Item)
 960                                 except:
 961                                     pass
 962                                 # replace each offset, except errors
 963                                 for Offset in OffsetList:
 964                                     try:
 965                                         Item = Item.replace('OFFSET_OF({})'.format(Offset), LabelDict[Offset])
 966                                     except:
 967                                         raise BadExpression('%s not defined' % Offset)
 968
 969                                 NewPcdValueList.append(Item)
 970
 971                             AllPcdValueList = []
 972                             for Item in NewPcdValueList:
 973                                 Size = 0
 974                                 ValueStr = ''
 975                                 TokenSpaceGuidName = ''
 976                                 if Item.startswith(TAB_GUID) and Item.endswith(')'):
 977                                     try:
 978                                         TokenSpaceGuidName = re.search('GUID\((\w+)\)', Item).group(1)
 979                                     except:
 980                                         pass
 981                                     if TokenSpaceGuidName and TokenSpaceGuidName in self._Symb:
 982                                         Item = 'GUID(' + self._Symb[TokenSpaceGuidName] + ')'
 983                                     elif TokenSpaceGuidName:
 984                                         raise BadExpression('%s not found in DEC file' % TokenSpaceGuidName)
 985                                     Item, Size = ParseFieldValue(Item)
 986                                     for Index in range(0, Size):
 987                                         ValueStr = '0x%02X' % (int(Item) & 255)
 988                                         Item >>= 8
 989                                         AllPcdValueList.append(ValueStr)
 990                                     continue
 991                                 elif Item.startswith('DEVICE_PATH') and Item.endswith(')'):
 992                                     Item, Size = ParseFieldValue(Item)
 993                                     AllPcdValueList.append(Item[1:-1])
 994                                     continue
 995                                 else:
 996                                     ValueType = ""
 997                                     if Item.startswith(TAB_UINT8):
 998                                         ItemSize = 1
 999                                         ValueType = TAB_UINT8
1000                                     elif Item.startswith(TAB_UINT16):
1001                                         ItemSize = 2
1002                                         ValueType = TAB_UINT16
1003                                     elif Item.startswith(TAB_UINT32):
1004                                         ItemSize = 4
1005                                         ValueType = TAB_UINT32
1006                                     elif Item.startswith(TAB_UINT64):
1007                                         ItemSize = 8
1008                                         ValueType = TAB_UINT64
1009                                     else:
1010                                         ItemSize = 0
1011                                     if ValueType:
1012                                         TmpValue = ValueExpressionEx(Item, ValueType, self._Symb)(True)
1013                                     else:
1014                                         TmpValue = ValueExpressionEx(Item, self.PcdType, self._Symb)(True)
1015                                     Item = '0x%x' % TmpValue if not isinstance(TmpValue, type('')) else TmpValue
1016                                     if ItemSize == 0:
1017                                         ItemValue, ItemSize = ParseFieldValue(Item)
1018                                         if Item[0] not in {'"', 'L', '{'} and ItemSize > 1:
1019                                             raise BadExpression("Byte  array number %s should less than 0xFF." % Item)
1020                                     else:
1021                                         ItemValue = ParseFieldValue(Item)[0]
1022                                     for I in range(0, ItemSize):
1023                                         ValueStr = '0x%02X' % (int(ItemValue) & 255)
1024                                         ItemValue >>= 8
1025                                         AllPcdValueList.append(ValueStr)
1026                                     Size += ItemSize
1027
1028                             if Size > 0:
1029                                 PcdValue = '{' + ','.join(AllPcdValueList) + '}'
1030                         else:
1031                             raise  BadExpression("Type: %s, Value: %s, %s"%(self.PcdType, PcdValue, Value))
1032
1033             if PcdValue == 'True':
1034                 PcdValue = '1'
1035             if PcdValue == 'False':
1036                 PcdValue = '0'
1037
1038         if RealValue:
1039             return PcdValue
1040
1041 if __name__ == '__main__':
1042     pass
1043     while True:
1044         input = raw_input('Input expr: ')
1045         if input in 'qQ':
1046             break
1047         try:
1048             print(ValueExpression(input)(True))
1049             print(ValueExpression(input)(False))
1050         except WrnExpression as Ex:
1051             print(Ex.result)
1052             print(str(Ex))
1053         except Exception as Ex:
1054             print(str(Ex))