From: Jaben Carsey <jaben.carsey@intel.com>
To: edk2-devel@lists.01.org
Cc: Yonghong Zhu <yonghong.zhu@intel.com>,
Liming Gao <liming.gao@intel.com>,
Bob C Feng <bob.c.feng@intel.com>
Subject: [PATCH v2 1/7] BaseTools/GenFds: cleanup GenFds
Date: Fri, 28 Sep 2018 14:54:26 -0700 [thread overview]
Message-ID: <cec710b033a13b8ddd2865b9548fc19f60898797.1538171557.git.jaben.carsey@intel.com> (raw)
In-Reply-To: <cover.1538171557.git.jaben.carsey@intel.com>
In-Reply-To: <cover.1538171557.git.jaben.carsey@intel.com>
1) remove wildcard imports and use explicit imports
2) refactor to use shared variables from Common/DataType
3) rename to not shadow imports
4) don't assign a variable in a loop (just do final assignment)
5) remove spaces, parens, unused or commented out code, etc.
6) merge unnecessary parent classes into child
7) refactor to share DXE and PEI apriori GUIDs from one place
this includes changes to Build and EOT files
8) for PEP8, dont use __ for custom methods.
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Bob C Feng <bob.c.feng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jaben Carsey <jaben.carsey@intel.com>
---
BaseTools/Source/Python/CommonDataClass/FdfClass.py | 73 -
BaseTools/Source/Python/Eot/Eot.py | 11 +-
BaseTools/Source/Python/GenFds/AprioriSection.py | 45 +-
BaseTools/Source/Python/GenFds/Capsule.py | 26 +-
BaseTools/Source/Python/GenFds/CapsuleData.py | 1 -
BaseTools/Source/Python/GenFds/CompressSection.py | 4 +-
BaseTools/Source/Python/GenFds/DataSection.py | 4 +-
BaseTools/Source/Python/GenFds/DepexSection.py | 5 +-
BaseTools/Source/Python/GenFds/EfiSection.py | 16 +-
BaseTools/Source/Python/GenFds/FdfParser.py | 2990 ++++++++++----------
BaseTools/Source/Python/GenFds/Ffs.py | 82 +-
BaseTools/Source/Python/GenFds/FfsFileStatement.py | 37 +-
BaseTools/Source/Python/GenFds/FfsInfStatement.py | 10 +-
BaseTools/Source/Python/GenFds/Fv.py | 47 +-
BaseTools/Source/Python/GenFds/FvImageSection.py | 6 +-
BaseTools/Source/Python/GenFds/GenFds.py | 160 +-
BaseTools/Source/Python/GenFds/GenFdsGlobalVariable.py | 208 +-
BaseTools/Source/Python/GenFds/GuidSection.py | 4 +-
BaseTools/Source/Python/GenFds/OptionRom.py | 6 +-
BaseTools/Source/Python/GenFds/Region.py | 12 +-
BaseTools/Source/Python/GenFds/UiSection.py | 4 +-
BaseTools/Source/Python/GenFds/VerSection.py | 16 +-
BaseTools/Source/Python/GenFds/Vtf.py | 48 +-
BaseTools/Source/Python/build/BuildReport.py | 5 +-
24 files changed, 1810 insertions(+), 2010 deletions(-)
diff --git a/BaseTools/Source/Python/CommonDataClass/FdfClass.py b/BaseTools/Source/Python/CommonDataClass/FdfClass.py
index 64b58c2078a3..3df82082457f 100644
--- a/BaseTools/Source/Python/CommonDataClass/FdfClass.py
+++ b/BaseTools/Source/Python/CommonDataClass/FdfClass.py
@@ -37,50 +37,6 @@ class FDClassObject:
self.RegionList = []
self.vtfRawDict = {}
-## FV data in FDF
-#
-#
-class FvClassObject:
- ## The constructor
- #
- # @param self The object pointer
- #
- def __init__(self):
- self.UiFvName = None
- self.CreateFileName = None
- # 3-tuple list (blockSize, numBlocks, pcd)
- self.BlockSizeList = []
- # DefineVarDict[var] = value
- self.DefineVarDict = {}
- # SetVarDict[var] = value
- self.SetVarDict = {}
- self.FvAlignment = None
- # FvAttributeDict[attribute] = TRUE/FALSE (1/0)
- self.FvAttributeDict = {}
- self.FvNameGuid = None
- self.FvNameString = None
- self.AprioriSectionList = []
- self.FfsList = []
- self.BsBaseAddress = None
- self.RtBaseAddress = None
-
-## Region data in FDF
-#
-#
-class RegionClassObject:
- ## The constructor
- #
- # @param self The object pointer
- #
- def __init__(self):
- self.Offset = None # The begin position of the Region
- self.Size = None # The Size of the Region
- self.PcdOffset = None
- self.PcdSize = None
- self.SetVarDict = {}
- self.RegionType = None
- self.RegionDataList = []
-
## FFS data in FDF
#
#
@@ -112,7 +68,6 @@ class FileStatementClassObject (FfsClassObject) :
self.FvName = None
self.FdName = None
self.DefineVarDict = {}
- self.AprioriSection = None
self.KeepReloc = None
## INF statement data in FDF
@@ -134,19 +89,6 @@ class FfsInfStatementClassObject(FfsClassObject):
self.KeepReloc = None
self.UseArch = None
-## APRIORI section data in FDF
-#
-#
-class AprioriSectionClassObject:
- ## The constructor
- #
- # @param self The object pointer
- #
- def __init__(self):
- # DefineVarDict[var] = value
- self.DefineVarDict = {}
- self.FfsList = []
-
## section data in FDF
#
#
@@ -363,21 +305,6 @@ class CapsuleClassObject :
self.CapsuleDataList = []
self.FmpPayloadList = []
-## VTF data in FDF
-#
-#
-class VtfClassObject :
- ## The constructor
- #
- # @param self The object pointer
- #
- def __init__(self):
- self.KeyArch = None
- self.ArchList = None
- self.UiName = None
- self.ResetBin = None
- self.ComponentStatementList = []
-
## VTF component data in FDF
#
#
diff --git a/BaseTools/Source/Python/Eot/Eot.py b/BaseTools/Source/Python/Eot/Eot.py
index ce83da14957b..1bf00a9b43d3 100644
--- a/BaseTools/Source/Python/Eot/Eot.py
+++ b/BaseTools/Source/Python/Eot/Eot.py
@@ -34,10 +34,9 @@ import struct
import uuid
import copy
import codecs
+from GenFds.AprioriSection import DXE_APRIORI_GUID, PEI_APRIORI_GUID
gGuidStringFormat = "%08X-%04X-%04X-%02X%02X-%02X%02X%02X%02X%02X%02X"
-gPeiAprioriFileNameGuid = '1b45cc0a-156a-428a-af62-49864da0e6e6'
-gAprioriGuid = 'fc510ee7-ffdc-11d4-bd41-0080c73c8881'
gIndention = -4
class Image(array):
@@ -408,10 +407,10 @@ class FirmwareVolume(Image):
if Ffs.Type == 0x05:
FfsDxeCoreGuid = FfsID
continue
- if Ffs.Guid.lower() == gPeiAprioriFileNameGuid:
+ if Ffs.Guid.lower() == PEI_APRIORI_GUID:
FfsPeiPrioriGuid = FfsID
continue
- if Ffs.Guid.lower() == gAprioriGuid:
+ if Ffs.Guid.lower() == DXE_APRIORI_GUID:
FfsDxePrioriGuid = FfsID
continue
@@ -1198,10 +1197,10 @@ class FirmwareVolume(Image):
if Ffs.Type == 0x05:
FfsDxeCoreGuid = FfsID
continue
- if Ffs.Guid.lower() == gPeiAprioriFileNameGuid:
+ if Ffs.Guid.lower() == PEI_APRIORI_GUID:
FfsPeiPrioriGuid = FfsID
continue
- if Ffs.Guid.lower() == gAprioriGuid:
+ if Ffs.Guid.lower() == DXE_APRIORI_GUID:
FfsDxePrioriGuid = FfsID
continue
diff --git a/BaseTools/Source/Python/GenFds/AprioriSection.py b/BaseTools/Source/Python/GenFds/AprioriSection.py
index 7196f7f2c753..e7b856a1150a 100644
--- a/BaseTools/Source/Python/GenFds/AprioriSection.py
+++ b/BaseTools/Source/Python/GenFds/AprioriSection.py
@@ -16,28 +16,31 @@
# Import Modules
#
from __future__ import absolute_import
-from struct import *
+from struct import pack
import Common.LongFilePathOs as os
from io import BytesIO
-from . import FfsFileStatement
+from .FfsFileStatement import FileStatement
from .GenFdsGlobalVariable import GenFdsGlobalVariable
-from CommonDataClass.FdfClass import AprioriSectionClassObject
-from Common.StringUtils import *
+from Common.StringUtils import NormPath
from Common.Misc import SaveFileOnChange, PathClass
-from Common import EdkLogger
-from Common.BuildToolError import *
+from Common.EdkLogger import error as EdkLoggerError
+from Common.BuildToolError import RESOURCE_NOT_AVAILABLE
from Common.DataType import TAB_COMMON
+DXE_APRIORI_GUID = "FC510EE7-FFDC-11D4-BD41-0080C73C8881"
+PEI_APRIORI_GUID = "1B45CC0A-156A-428A-AF62-49864DA0E6E6"
+
## process APRIORI file data and generate PEI/DXE APRIORI file
#
#
-class AprioriSection (AprioriSectionClassObject):
+class AprioriSection (object):
## The constructor
#
# @param self The object pointer
#
def __init__(self):
- AprioriSectionClassObject.__init__(self)
+ self.DefineVarDict = {}
+ self.FfsList = []
self.AprioriType = ""
## GenFfs() method
@@ -50,16 +53,16 @@ class AprioriSection (AprioriSectionClassObject):
# @retval string Generated file name
#
def GenFfs (self, FvName, Dict = {}, IsMakefile = False):
- DXE_GUID = "FC510EE7-FFDC-11D4-BD41-0080C73C8881"
- PEI_GUID = "1B45CC0A-156A-428A-AF62-49864DA0E6E6"
Buffer = BytesIO('')
- AprioriFileGuid = DXE_GUID
if self.AprioriType == "PEI":
- AprioriFileGuid = PEI_GUID
+ AprioriFileGuid = PEI_APRIORI_GUID
+ else:
+ AprioriFileGuid = DXE_APRIORI_GUID
+
OutputAprFilePath = os.path.join (GenFdsGlobalVariable.WorkSpaceDir, \
GenFdsGlobalVariable.FfsDir,\
AprioriFileGuid + FvName)
- if not os.path.exists(OutputAprFilePath) :
+ if not os.path.exists(OutputAprFilePath):
os.makedirs(OutputAprFilePath)
OutputAprFileName = os.path.join( OutputAprFilePath, \
@@ -69,33 +72,30 @@ class AprioriSection (AprioriSectionClassObject):
Dict.update(self.DefineVarDict)
InfFileName = None
- for FfsObj in self.FfsList :
+ for FfsObj in self.FfsList:
Guid = ""
- if isinstance(FfsObj, FfsFileStatement.FileStatement):
+ if isinstance(FfsObj, FileStatement):
Guid = FfsObj.NameGuid
else:
InfFileName = NormPath(FfsObj.InfFileName)
Arch = FfsObj.GetCurrentArch()
- if Arch is not None:
+ if Arch:
Dict['$(ARCH)'] = Arch
InfFileName = GenFdsGlobalVariable.MacroExtend(InfFileName, Dict, Arch)
- if Arch is not None:
+ if Arch:
Inf = GenFdsGlobalVariable.WorkSpace.BuildObject[PathClass(InfFileName, GenFdsGlobalVariable.WorkSpaceDir), Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag]
Guid = Inf.Guid
-
else:
Inf = GenFdsGlobalVariable.WorkSpace.BuildObject[PathClass(InfFileName, GenFdsGlobalVariable.WorkSpaceDir), TAB_COMMON, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag]
Guid = Inf.Guid
- self.BinFileList = Inf.Module.Binaries
- if self.BinFileList == []:
- EdkLogger.error("GenFds", RESOURCE_NOT_AVAILABLE,
+ if not Inf.Module.Binaries:
+ EdkLoggerError("GenFds", RESOURCE_NOT_AVAILABLE,
"INF %s not found in build ARCH %s!" \
% (InfFileName, GenFdsGlobalVariable.ArchList))
-
GuidPart = Guid.split('-')
Buffer.write(pack('I', long(GuidPart[0], 16)))
Buffer.write(pack('H', int(GuidPart[1], 16)))
@@ -123,4 +123,3 @@ class AprioriSection (AprioriSectionClassObject):
'EFI_FV_FILETYPE_FREEFORM', AprioriFileGuid, MakefilePath=MakefilePath)
return AprFfsFileName
-
diff --git a/BaseTools/Source/Python/GenFds/Capsule.py b/BaseTools/Source/Python/GenFds/Capsule.py
index baa6e1dfa2ab..df29c40dbd20 100644
--- a/BaseTools/Source/Python/GenFds/Capsule.py
+++ b/BaseTools/Source/Python/GenFds/Capsule.py
@@ -16,22 +16,18 @@
# Import Modules
#
from __future__ import absolute_import
-from .GenFdsGlobalVariable import GenFdsGlobalVariable
-from .GenFdsGlobalVariable import FindExtendTool
+from .GenFdsGlobalVariable import GenFdsGlobalVariable, FindExtendTool
from CommonDataClass.FdfClass import CapsuleClassObject
import Common.LongFilePathOs as os
-import subprocess
from io import BytesIO
-from Common.Misc import SaveFileOnChange
-from Common.Misc import PackRegistryFormatGuid
+from Common.Misc import SaveFileOnChange, PackRegistryFormatGuid
import uuid
from struct import pack
from Common import EdkLogger
-from Common.BuildToolError import *
+from Common.BuildToolError import GENFDS_ERROR
+from Common.DataType import TAB_LINE_BREAK
-
-T_CHAR_LF = '\n'
-WIN_CERT_REVISION = 0x0200
+WIN_CERT_REVISION = 0x0200
WIN_CERT_TYPE_EFI_GUID = 0x0EF1
EFI_CERT_TYPE_PKCS7_GUID = uuid.UUID('{4aafd29d-68df-49ee-8aa9-347d375665a7}')
EFI_CERT_TYPE_RSA2048_SHA256_GUID = uuid.UUID('{a7717414-c616-4977-9420-844712a735bf}')
@@ -39,7 +35,7 @@ EFI_CERT_TYPE_RSA2048_SHA256_GUID = uuid.UUID('{a7717414-c616-4977-9420-844712a7
## create inf file describes what goes into capsule and call GenFv to generate capsule
#
#
-class Capsule (CapsuleClassObject) :
+class Capsule (CapsuleClassObject):
## The constructor
#
# @param self The object pointer
@@ -210,16 +206,16 @@ class Capsule (CapsuleClassObject) :
return self.GenFmpCapsule()
CapInfFile = self.GenCapInf()
- CapInfFile.writelines("[files]" + T_CHAR_LF)
+ CapInfFile.writelines("[files]" + TAB_LINE_BREAK)
CapFileList = []
- for CapsuleDataObj in self.CapsuleDataList :
+ for CapsuleDataObj in self.CapsuleDataList:
CapsuleDataObj.CapsuleName = self.CapsuleName
FileName = CapsuleDataObj.GenCapsuleSubItem()
CapsuleDataObj.CapsuleName = None
CapFileList.append(FileName)
CapInfFile.writelines("EFI_FILE_NAME = " + \
FileName + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
SaveFileOnChange(self.CapInfFileName, CapInfFile.getvalue(), False)
CapInfFile.close()
#
@@ -249,13 +245,13 @@ class Capsule (CapsuleClassObject) :
self.UiCapsuleName + "_Cap" + '.inf')
CapInfFile = BytesIO() #open (self.CapInfFileName , 'w+')
- CapInfFile.writelines("[options]" + T_CHAR_LF)
+ CapInfFile.writelines("[options]" + TAB_LINE_BREAK)
for Item in self.TokensDict:
CapInfFile.writelines("EFI_" + \
Item + \
' = ' + \
self.TokensDict[Item] + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
return CapInfFile
diff --git a/BaseTools/Source/Python/GenFds/CapsuleData.py b/BaseTools/Source/Python/GenFds/CapsuleData.py
index 0caba8983d2e..6071e9f4a545 100644
--- a/BaseTools/Source/Python/GenFds/CapsuleData.py
+++ b/BaseTools/Source/Python/GenFds/CapsuleData.py
@@ -16,7 +16,6 @@
# Import Modules
#
from __future__ import absolute_import
-from . import Ffs
from .GenFdsGlobalVariable import GenFdsGlobalVariable
from io import BytesIO
from struct import pack
diff --git a/BaseTools/Source/Python/GenFds/CompressSection.py b/BaseTools/Source/Python/GenFds/CompressSection.py
index aaaabf84dca8..51f6791fab19 100644
--- a/BaseTools/Source/Python/GenFds/CompressSection.py
+++ b/BaseTools/Source/Python/GenFds/CompressSection.py
@@ -16,7 +16,7 @@
# Import Modules
#
from __future__ import absolute_import
-from .Ffs import Ffs
+from .Ffs import SectionSuffix
from . import Section
import subprocess
import Common.LongFilePathOs as os
@@ -86,7 +86,7 @@ class CompressSection (CompressSectionClassObject) :
ModuleName + \
SUP_MODULE_SEC + \
SecNum + \
- Ffs.SectionSuffix['COMPRESS']
+ SectionSuffix['COMPRESS']
OutputFile = os.path.normpath(OutputFile)
DummyFile = OutputFile + '.dummy'
GenFdsGlobalVariable.GenerateSection(DummyFile, SectFiles, InputAlign=SectAlign, IsMakefile=IsMakefile)
diff --git a/BaseTools/Source/Python/GenFds/DataSection.py b/BaseTools/Source/Python/GenFds/DataSection.py
index a6387b07c582..28f9b931ca4b 100644
--- a/BaseTools/Source/Python/GenFds/DataSection.py
+++ b/BaseTools/Source/Python/GenFds/DataSection.py
@@ -19,7 +19,7 @@ from __future__ import absolute_import
from . import Section
from .GenFdsGlobalVariable import GenFdsGlobalVariable
import subprocess
-from .Ffs import Ffs
+from .Ffs import SectionSuffix
import Common.LongFilePathOs as os
from CommonDataClass.FdfClass import DataSectionClassObject
from Common.Misc import PeImageClass
@@ -121,7 +121,7 @@ class DataSection (DataSectionClassObject):
)
self.SectFileName = TeFile
- OutputFile = os.path.join (OutputPath, ModuleName + SUP_MODULE_SEC + SecNum + Ffs.SectionSuffix.get(self.SecType))
+ OutputFile = os.path.join (OutputPath, ModuleName + SUP_MODULE_SEC + SecNum + SectionSuffix.get(self.SecType))
OutputFile = os.path.normpath(OutputFile)
GenFdsGlobalVariable.GenerateSection(OutputFile, [self.SectFileName], Section.Section.SectionType.get(self.SecType), IsMakefile = IsMakefile)
FileList = [OutputFile]
diff --git a/BaseTools/Source/Python/GenFds/DepexSection.py b/BaseTools/Source/Python/GenFds/DepexSection.py
index b2d123bfc045..78efc18816e8 100644
--- a/BaseTools/Source/Python/GenFds/DepexSection.py
+++ b/BaseTools/Source/Python/GenFds/DepexSection.py
@@ -18,8 +18,6 @@
from __future__ import absolute_import
from . import Section
from .GenFdsGlobalVariable import GenFdsGlobalVariable
-import subprocess
-from .Ffs import Ffs
import Common.LongFilePathOs as os
from CommonDataClass.FdfClass import DepexSectionClassObject
from AutoGen.GenDepex import DependencyExpression
@@ -46,9 +44,8 @@ class DepexSection (DepexSectionClassObject):
GenFdsGlobalVariable.TargetName,
GenFdsGlobalVariable.ToolChainTag)
for Inf in GenFdsGlobalVariable.FdfParser.Profile.InfList:
- ModuleFile = PathClass(Inf, GenFdsGlobalVariable.WorkSpaceDir)
ModuleData = GenFdsGlobalVariable.WorkSpace.BuildObject[
- ModuleFile,
+ PathClass(Inf, GenFdsGlobalVariable.WorkSpaceDir),
Arch,
GenFdsGlobalVariable.TargetName,
GenFdsGlobalVariable.ToolChainTag
diff --git a/BaseTools/Source/Python/GenFds/EfiSection.py b/BaseTools/Source/Python/GenFds/EfiSection.py
index 623b77d27427..f8573b5c7d1b 100644
--- a/BaseTools/Source/Python/GenFds/EfiSection.py
+++ b/BaseTools/Source/Python/GenFds/EfiSection.py
@@ -20,7 +20,7 @@ from struct import *
from . import Section
from .GenFdsGlobalVariable import GenFdsGlobalVariable
import subprocess
-from .Ffs import Ffs
+from .Ffs import SectionSuffix
import Common.LongFilePathOs as os
from CommonDataClass.FdfClass import EfiSectionClassObject
from Common import EdkLogger
@@ -124,7 +124,7 @@ class EfiSection (EfiSectionClassObject):
BuildNumTuple = tuple()
Num = SecNum
- OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + SectionSuffix.get(SectionType))
GenFdsGlobalVariable.GenerateSection(OutputFile, [], 'EFI_SECTION_VERSION',
#Ui=StringData,
Ver=BuildNum,
@@ -135,7 +135,7 @@ class EfiSection (EfiSectionClassObject):
for File in FileList:
Index = Index + 1
Num = '%s.%d' %(SecNum, Index)
- OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + Num + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + Num + SectionSuffix.get(SectionType))
f = open(File, 'r')
VerString = f.read()
f.close()
@@ -164,7 +164,7 @@ class EfiSection (EfiSectionClassObject):
else:
EdkLogger.error("GenFds", GENFDS_ERROR, "File: %s miss Version Section value" %InfFileName)
Num = SecNum
- OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + SectionSuffix.get(SectionType))
GenFdsGlobalVariable.GenerateSection(OutputFile, [], 'EFI_SECTION_VERSION',
#Ui=VerString,
Ver=BuildNum,
@@ -185,7 +185,7 @@ class EfiSection (EfiSectionClassObject):
Num = SecNum
if IsMakefile and StringData == ModuleNameStr:
StringData = "$(MODULE_NAME)"
- OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + SectionSuffix.get(SectionType))
GenFdsGlobalVariable.GenerateSection(OutputFile, [], 'EFI_SECTION_USER_INTERFACE',
Ui=StringData, IsMakefile=IsMakefile)
OutputFileList.append(OutputFile)
@@ -194,7 +194,7 @@ class EfiSection (EfiSectionClassObject):
for File in FileList:
Index = Index + 1
Num = '%s.%d' %(SecNum, Index)
- OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + Num + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + Num + SectionSuffix.get(SectionType))
f = open(File, 'r')
UiString = f.read()
f.close()
@@ -218,7 +218,7 @@ class EfiSection (EfiSectionClassObject):
Num = SecNum
if IsMakefile and StringData == ModuleNameStr:
StringData = "$(MODULE_NAME)"
- OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + str(Num) + SectionSuffix.get(SectionType))
GenFdsGlobalVariable.GenerateSection(OutputFile, [], 'EFI_SECTION_USER_INTERFACE',
Ui=StringData, IsMakefile=IsMakefile)
OutputFileList.append(OutputFile)
@@ -239,7 +239,7 @@ class EfiSection (EfiSectionClassObject):
""" Copy Map file to FFS output path """
Index = Index + 1
Num = '%s.%d' %(SecNum, Index)
- OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + Num + Ffs.SectionSuffix.get(SectionType))
+ OutputFile = os.path.join( OutputPath, ModuleName + SUP_MODULE_SEC + Num + SectionSuffix.get(SectionType))
File = GenFdsGlobalVariable.MacroExtend(File, Dict)
#Get PE Section alignment when align is set to AUTO
diff --git a/BaseTools/Source/Python/GenFds/FdfParser.py b/BaseTools/Source/Python/GenFds/FdfParser.py
index 63687e98bb31..814874d93997 100644
--- a/BaseTools/Source/Python/GenFds/FdfParser.py
+++ b/BaseTools/Source/Python/GenFds/FdfParser.py
@@ -18,79 +18,66 @@
#
from __future__ import print_function
from __future__ import absolute_import
-import re
+from re import compile, DOTALL
+from string import hexdigits
+from uuid import UUID
-from . import Fd
-from . import Region
-from . import Fv
-from . import AprioriSection
-from . import FfsInfStatement
-from . import FfsFileStatement
-from . import VerSection
-from . import UiSection
-from . import FvImageSection
-from . import DataSection
-from . import DepexSection
-from . import CompressSection
-from . import GuidSection
-from . import Capsule
-from . import CapsuleData
-from . import Rule
-from . import RuleComplexFile
-from . import RuleSimpleFile
-from . import EfiSection
-from . import Vtf
-from . import ComponentStatement
-from . import OptionRom
-from . import OptRomInfStatement
-from . import OptRomFileStatement
-import string
-
-from .GenFdsGlobalVariable import GenFdsGlobalVariable
from Common.BuildToolError import *
from Common import EdkLogger
-from Common.Misc import PathClass
-from Common.StringUtils import NormPath
-import Common.GlobalData as GlobalData
-from Common.Expression import *
+from Common.Misc import PathClass, tdict
+from Common.StringUtils import NormPath, ReplaceMacro
from Common import GlobalData
+from Common.Expression import *
from Common.DataType import *
-from Common.StringUtils import ReplaceMacro
-import uuid
-from Common.Misc import tdict
from Common.MultipleWorkspace import MultipleWorkspace as mws
import Common.LongFilePathOs as os
from Common.LongFilePathSupport import OpenLongFilePath as open
-from .Capsule import EFI_CERT_TYPE_PKCS7_GUID
-from .Capsule import EFI_CERT_TYPE_RSA2048_SHA256_GUID
from Common.RangeExpression import RangeExpression
from collections import OrderedDict
-##define T_CHAR_SPACE ' '
-##define T_CHAR_NULL '\0'
-##define T_CHAR_CR '\r'
-##define T_CHAR_TAB '\t'
-##define T_CHAR_LF '\n'
-##define T_CHAR_SLASH '/'
-##define T_CHAR_BACKSLASH '\\'
-##define T_CHAR_DOUBLE_QUOTE '\"'
-##define T_CHAR_SINGLE_QUOTE '\''
-##define T_CHAR_STAR '*'
-##define T_CHAR_HASH '#'
+from .Fd import FD
+from .Region import Region
+from .Fv import FV
+from .AprioriSection import AprioriSection
+from .FfsInfStatement import FfsInfStatement
+from .FfsFileStatement import FileStatement
+from .VerSection import VerSection
+from .UiSection import UiSection
+from .FvImageSection import FvImageSection
+from .DataSection import DataSection
+from .DepexSection import DepexSection
+from .CompressSection import CompressSection
+from .GuidSection import GuidSection
+from .Capsule import EFI_CERT_TYPE_PKCS7_GUID, EFI_CERT_TYPE_RSA2048_SHA256_GUID, Capsule
+from .CapsuleData import CapsuleFfs, CapsulePayload, CapsuleFv, CapsuleFd, CapsuleAnyFile, CapsuleAfile
+from .RuleComplexFile import RuleComplexFile
+from .RuleSimpleFile import RuleSimpleFile
+from .EfiSection import EfiSection
+from .Vtf import Vtf
+from .ComponentStatement import ComponentStatement
+from .OptionRom import OPTIONROM
+from .OptRomInfStatement import OptRomInfStatement, OverrideAttribs
+from .OptRomFileStatement import OptRomFileStatement
+from .GenFdsGlobalVariable import GenFdsGlobalVariable
-(T_CHAR_SPACE, T_CHAR_NULL, T_CHAR_CR, T_CHAR_TAB, T_CHAR_LF, T_CHAR_SLASH, \
-T_CHAR_BACKSLASH, T_CHAR_DOUBLE_QUOTE, T_CHAR_SINGLE_QUOTE, T_CHAR_STAR, T_CHAR_HASH) = \
-(' ', '\0', '\r', '\t', '\n', '/', '\\', '\"', '\'', '*', '#')
+T_CHAR_CR = '\r'
+T_CHAR_TAB = '\t'
+T_CHAR_DOUBLE_QUOTE = '\"'
+T_CHAR_SINGLE_QUOTE = '\''
+T_CHAR_STAR = '*'
-SEPERATOR_TUPLE = ('=', '|', ',', '{', '}')
+SEPARATORS = {TAB_EQUAL_SPLIT, TAB_VALUE_SPLIT, TAB_COMMA_SPLIT, '{', '}'}
+ALIGNMENTS = {"Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
+ "256K", "512K", "1M", "2M", "4M", "8M", "16M"}
+ALIGNMENT_NOAUTO = ALIGNMENTS - {"Auto"}
-RegionSizePattern = re.compile("\s*(?P<base>(?:0x|0X)?[a-fA-F0-9]+)\s*\|\s*(?P<size>(?:0x|0X)?[a-fA-F0-9]+)\s*")
-RegionSizeGuidPattern = re.compile("\s*(?P<base>\w+\.\w+[\.\w\[\]]*)\s*\|\s*(?P<size>\w+\.\w+[\.\w\[\]]*)\s*")
-RegionOffsetPcdPattern = re.compile("\s*(?P<base>\w+\.\w+[\.\w\[\]]*)\s*$")
-ShortcutPcdPattern = re.compile("\s*\w+\s*=\s*(?P<value>(?:0x|0X)?[a-fA-F0-9]+)\s*\|\s*(?P<name>\w+\.\w+)\s*")
-BaseAddrValuePattern = re.compile('^0[xX][0-9a-fA-F]+')
-FileExtensionPattern = re.compile(r'([a-zA-Z][a-zA-Z0-9]*)')
-TokenFindPattern = re.compile(r'([a-zA-Z0-9\-]+|\$\(TARGET\)|\*)_([a-zA-Z0-9\-]+|\$\(TOOL_CHAIN_TAG\)|\*)_([a-zA-Z0-9\-]+|\$\(ARCH\)|\*)')
+RegionSizePattern = compile("\s*(?P<base>(?:0x|0X)?[a-fA-F0-9]+)\s*\|\s*(?P<size>(?:0x|0X)?[a-fA-F0-9]+)\s*")
+RegionSizeGuidPattern = compile("\s*(?P<base>\w+\.\w+[\.\w\[\]]*)\s*\|\s*(?P<size>\w+\.\w+[\.\w\[\]]*)\s*")
+RegionOffsetPcdPattern = compile("\s*(?P<base>\w+\.\w+[\.\w\[\]]*)\s*$")
+ShortcutPcdPattern = compile("\s*\w+\s*=\s*(?P<value>(?:0x|0X)?[a-fA-F0-9]+)\s*\|\s*(?P<name>\w+\.\w+)\s*")
+BaseAddrValuePattern = compile('^0[xX][0-9a-fA-F]+')
+FileExtensionPattern = compile(r'([a-zA-Z][a-zA-Z0-9]*)')
+TokenFindPattern = compile(r'([a-zA-Z0-9\-]+|\$\(TARGET\)|\*)_([a-zA-Z0-9\-]+|\$\(TOOL_CHAIN_TAG\)|\*)_([a-zA-Z0-9\-]+|\$\(ARCH\)|\*)')
AllIncludeFileList = []
# Get the closest parent
@@ -109,7 +96,6 @@ def IsValidInclude (File, Line):
return True
def GetRealFileLine (File, Line):
-
InsertedLines = 0
for Profile in AllIncludeFileList:
if Profile.IsLineInFile(Line):
@@ -121,7 +107,7 @@ def GetRealFileLine (File, Line):
## The exception class that used to report error messages when parsing FDF
#
-# Currently the "ToolName" is set to be "FDF Parser".
+# Currently the "ToolName" is set to be "FdfParser".
#
class Warning (Exception):
## The constructor
@@ -132,7 +118,6 @@ class Warning (Exception):
# @param Line The Line number that error occurs
#
def __init__(self, Str, File = None, Line = None):
-
FileLineTuple = GetRealFileLine(File, Line)
self.FileName = FileLineTuple[0]
self.LineNumber = FileLineTuple[1]
@@ -147,7 +132,7 @@ class Warning (Exception):
#
# May raise Exception when opening file.
#
-class IncludeFileProfile :
+class IncludeFileProfile:
## The constructor
#
# @param self The object pointer
@@ -157,16 +142,11 @@ class IncludeFileProfile :
self.FileName = FileName
self.FileLinesList = []
try:
- fsock = open(FileName, "rb", 0)
- try:
+ with open(FileName, "rb", 0) as fsock:
self.FileLinesList = fsock.readlines()
for index, line in enumerate(self.FileLinesList):
- if not line.endswith('\n'):
- self.FileLinesList[index] += '\n'
-
- finally:
- fsock.close()
-
+ if not line.endswith(TAB_LINE_BREAK):
+ self.FileLinesList[index] += TAB_LINE_BREAK
except:
EdkLogger.error("FdfParser", FILE_OPEN_FAILURE, ExtraData=FileName)
@@ -209,7 +189,7 @@ class IncludeFileProfile :
#
# May raise Exception when opening file.
#
-class FileProfile :
+class FileProfile:
## The constructor
#
# @param self The object pointer
@@ -218,11 +198,8 @@ class FileProfile :
def __init__(self, FileName):
self.FileLinesList = []
try:
- fsock = open(FileName, "rb", 0)
- try:
+ with open(FileName, "rb", 0) as fsock:
self.FileLinesList = fsock.readlines()
- finally:
- fsock.close()
except:
EdkLogger.error("FdfParser", FILE_OPEN_FAILURE, ExtraData=FileName)
@@ -266,41 +243,37 @@ class FdfParser:
self.CurrentOffsetWithinLine = 0
self.CurrentFdName = None
self.CurrentFvName = None
- self.__Token = ""
- self.__SkippedChars = ""
+ self._Token = ""
+ self._SkippedChars = ""
GlobalData.gFdfParser = self
# Used to section info
- self.__CurSection = []
+ self._CurSection = []
# Key: [section name, UI name, arch]
- # Value: {MACRO_NAME : MACRO_VALUE}
- self.__MacroDict = tdict(True, 3)
- self.__PcdDict = OrderedDict()
+ # Value: {MACRO_NAME: MACRO_VALUE}
+ self._MacroDict = tdict(True, 3)
+ self._PcdDict = OrderedDict()
- self.__WipeOffArea = []
+ self._WipeOffArea = []
if GenFdsGlobalVariable.WorkSpaceDir == '':
GenFdsGlobalVariable.WorkSpaceDir = os.getenv("WORKSPACE")
- ## __SkipWhiteSpace() method
+ ## _SkipWhiteSpace() method
#
- # Skip white spaces from current char, return number of chars skipped
+ # Skip white spaces from current char.
#
# @param self The object pointer
- # @retval Count The number of chars skipped
#
- def __SkipWhiteSpace(self):
- Count = 0
- while not self.__EndOfFile():
- Count += 1
- if self.__CurrentChar() in (T_CHAR_NULL, T_CHAR_CR, T_CHAR_LF, T_CHAR_SPACE, T_CHAR_TAB):
- self.__SkippedChars += str(self.__CurrentChar())
- self.__GetOneChar()
-
+ def _SkipWhiteSpace(self):
+ while not self._EndOfFile():
+ if self._CurrentChar() in (TAB_PRINTCHAR_NUL, T_CHAR_CR, TAB_LINE_BREAK, TAB_SPACE_SPLIT, T_CHAR_TAB):
+ self._SkippedChars += str(self._CurrentChar())
+ self._GetOneChar()
else:
- Count = Count - 1
- return Count
+ return
+ return
- ## __EndOfFile() method
+ ## _EndOfFile() method
#
# Judge current buffer pos is at file end
#
@@ -308,17 +281,16 @@ class FdfParser:
# @retval True Current File buffer position is at file end
# @retval False Current File buffer position is NOT at file end
#
- def __EndOfFile(self):
+ def _EndOfFile(self):
NumberOfLines = len(self.Profile.FileLinesList)
SizeOfLastLine = len(self.Profile.FileLinesList[-1])
if self.CurrentLineNumber == NumberOfLines and self.CurrentOffsetWithinLine >= SizeOfLastLine - 1:
return True
- elif self.CurrentLineNumber > NumberOfLines:
+ if self.CurrentLineNumber > NumberOfLines:
return True
- else:
- return False
+ return False
- ## __EndOfLine() method
+ ## _EndOfLine() method
#
# Judge current buffer pos is at line end
#
@@ -326,14 +298,13 @@ class FdfParser:
# @retval True Current File buffer position is at line end
# @retval False Current File buffer position is NOT at line end
#
- def __EndOfLine(self):
+ def _EndOfLine(self):
if self.CurrentLineNumber > len(self.Profile.FileLinesList):
return True
SizeOfCurrentLine = len(self.Profile.FileLinesList[self.CurrentLineNumber - 1])
if self.CurrentOffsetWithinLine >= SizeOfCurrentLine:
return True
- else:
- return False
+ return False
## Rewind() method
#
@@ -347,7 +318,7 @@ class FdfParser:
self.CurrentLineNumber = DestLine
self.CurrentOffsetWithinLine = DestOffset
- ## __UndoOneChar() method
+ ## _UndoOneChar() method
#
# Go back one char in the file buffer
#
@@ -355,80 +326,78 @@ class FdfParser:
# @retval True Successfully go back one char
# @retval False Not able to go back one char as file beginning reached
#
- def __UndoOneChar(self):
-
+ def _UndoOneChar(self):
if self.CurrentLineNumber == 1 and self.CurrentOffsetWithinLine == 0:
return False
elif self.CurrentOffsetWithinLine == 0:
self.CurrentLineNumber -= 1
- self.CurrentOffsetWithinLine = len(self.__CurrentLine()) - 1
+ self.CurrentOffsetWithinLine = len(self._CurrentLine()) - 1
else:
self.CurrentOffsetWithinLine -= 1
return True
- ## __GetOneChar() method
+ ## _GetOneChar() method
#
# Move forward one char in the file buffer
#
# @param self The object pointer
#
- def __GetOneChar(self):
+ def _GetOneChar(self):
if self.CurrentOffsetWithinLine == len(self.Profile.FileLinesList[self.CurrentLineNumber - 1]) - 1:
self.CurrentLineNumber += 1
self.CurrentOffsetWithinLine = 0
else:
self.CurrentOffsetWithinLine += 1
- ## __CurrentChar() method
+ ## _CurrentChar() method
#
# Get the char pointed to by the file buffer pointer
#
# @param self The object pointer
# @retval Char Current char
#
- def __CurrentChar(self):
+ def _CurrentChar(self):
return self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine]
- ## __NextChar() method
+ ## _NextChar() method
#
# Get the one char pass the char pointed to by the file buffer pointer
#
# @param self The object pointer
# @retval Char Next char
#
- def __NextChar(self):
+ def _NextChar(self):
if self.CurrentOffsetWithinLine == len(self.Profile.FileLinesList[self.CurrentLineNumber - 1]) - 1:
return self.Profile.FileLinesList[self.CurrentLineNumber][0]
- else:
- return self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine + 1]
+ return self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine + 1]
- ## __SetCurrentCharValue() method
+ ## _SetCurrentCharValue() method
#
# Modify the value of current char
#
# @param self The object pointer
# @param Value The new value of current char
#
- def __SetCurrentCharValue(self, Value):
+ def _SetCurrentCharValue(self, Value):
self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine] = Value
- ## __CurrentLine() method
+ ## _CurrentLine() method
#
# Get the list that contains current line contents
#
# @param self The object pointer
# @retval List current line contents
#
- def __CurrentLine(self):
+ def _CurrentLine(self):
return self.Profile.FileLinesList[self.CurrentLineNumber - 1]
- def __StringToList(self):
+ def _StringToList(self):
self.Profile.FileLinesList = [list(s) for s in self.Profile.FileLinesList]
if not self.Profile.FileLinesList:
EdkLogger.error('FdfParser', FILE_READ_FAILURE, 'The file is empty!', File=self.FileName)
self.Profile.FileLinesList[-1].append(' ')
- def __ReplaceFragment(self, StartPos, EndPos, Value = ' '):
+ def _ReplaceFragment(self, StartPos, EndPos, Value = ' '):
if StartPos[0] == EndPos[0]:
Offset = StartPos[1]
while Offset <= EndPos[1]:
@@ -437,14 +406,14 @@ class FdfParser:
return
Offset = StartPos[1]
- while self.Profile.FileLinesList[StartPos[0]][Offset] not in ('\r', '\n'):
+ while self.Profile.FileLinesList[StartPos[0]][Offset] not in (T_CHAR_CR, TAB_LINE_BREAK):
self.Profile.FileLinesList[StartPos[0]][Offset] = Value
Offset += 1
Line = StartPos[0]
while Line < EndPos[0]:
Offset = 0
- while self.Profile.FileLinesList[Line][Offset] not in ('\r', '\n'):
+ while self.Profile.FileLinesList[Line][Offset] not in (T_CHAR_CR, TAB_LINE_BREAK):
self.Profile.FileLinesList[Line][Offset] = Value
Offset += 1
Line += 1
@@ -454,45 +423,29 @@ class FdfParser:
self.Profile.FileLinesList[EndPos[0]][Offset] = Value
Offset += 1
-
- def __GetMacroName(self):
- if not self.__GetNextToken():
- raise Warning("expected Macro name", self.FileName, self.CurrentLineNumber)
- MacroName = self.__Token
- NotFlag = False
- if MacroName.startswith('!'):
- NotFlag = True
- MacroName = MacroName[1:].strip()
-
- if not MacroName.startswith('$(') or not MacroName.endswith(')'):
- raise Warning("Macro name expected(Please use '$(%(Token)s)' if '%(Token)s' is a macro.)" % {"Token" : MacroName},
- self.FileName, self.CurrentLineNumber)
- MacroName = MacroName[2:-1]
- return MacroName, NotFlag
-
- def __SetMacroValue(self, Macro, Value):
- if not self.__CurSection:
+ def _SetMacroValue(self, Macro, Value):
+ if not self._CurSection:
return
MacroDict = {}
- if not self.__MacroDict[self.__CurSection[0], self.__CurSection[1], self.__CurSection[2]]:
- self.__MacroDict[self.__CurSection[0], self.__CurSection[1], self.__CurSection[2]] = MacroDict
+ if not self._MacroDict[self._CurSection[0], self._CurSection[1], self._CurSection[2]]:
+ self._MacroDict[self._CurSection[0], self._CurSection[1], self._CurSection[2]] = MacroDict
else:
- MacroDict = self.__MacroDict[self.__CurSection[0], self.__CurSection[1], self.__CurSection[2]]
+ MacroDict = self._MacroDict[self._CurSection[0], self._CurSection[1], self._CurSection[2]]
MacroDict[Macro] = Value
- def __GetMacroValue(self, Macro):
+ def _GetMacroValue(self, Macro):
# Highest priority
if Macro in GlobalData.gCommandLineDefines:
return GlobalData.gCommandLineDefines[Macro]
if Macro in GlobalData.gGlobalDefines:
return GlobalData.gGlobalDefines[Macro]
- if self.__CurSection:
- MacroDict = self.__MacroDict[
- self.__CurSection[0],
- self.__CurSection[1],
- self.__CurSection[2]
+ if self._CurSection:
+ MacroDict = self._MacroDict[
+ self._CurSection[0],
+ self._CurSection[1],
+ self._CurSection[2]
]
if MacroDict and Macro in MacroDict:
return MacroDict[Macro]
@@ -502,7 +455,7 @@ class FdfParser:
return GlobalData.gPlatformDefines[Macro]
return None
- def __SectionHeaderParser(self, Section):
+ def _SectionHeaderParser(self, Section):
# [Defines]
# [FD.UiName]: use dummy instead if UI name is optional
# [FV.UiName]
@@ -510,25 +463,25 @@ class FdfParser:
# [Rule]: don't take rule section into account, macro is not allowed in this section
# [VTF.arch.UiName, arch]
# [OptionRom.DriverName]
- self.__CurSection = []
- Section = Section.strip()[1:-1].upper().replace(' ', '').strip('.')
- ItemList = Section.split('.')
+ self._CurSection = []
+ Section = Section.strip()[1:-1].upper().replace(' ', '').strip(TAB_SPLIT)
+ ItemList = Section.split(TAB_SPLIT)
Item = ItemList[0]
if Item == '' or Item == 'RULE':
return
if Item == TAB_COMMON_DEFINES.upper():
- self.__CurSection = [TAB_COMMON, TAB_COMMON, TAB_COMMON]
+ self._CurSection = [TAB_COMMON, TAB_COMMON, TAB_COMMON]
elif Item == 'VTF' and len(ItemList) == 3:
UiName = ItemList[2]
- Pos = UiName.find(',')
+ Pos = UiName.find(TAB_COMMA_SPLIT)
if Pos != -1:
UiName = UiName[:Pos]
- self.__CurSection = ['VTF', UiName, ItemList[1]]
+ self._CurSection = ['VTF', UiName, ItemList[1]]
elif len(ItemList) > 1:
- self.__CurSection = [ItemList[0], ItemList[1], TAB_COMMON]
+ self._CurSection = [ItemList[0], ItemList[1], TAB_COMMON]
elif len(ItemList) > 0:
- self.__CurSection = [ItemList[0], 'DUMMY', TAB_COMMON]
+ self._CurSection = [ItemList[0], 'DUMMY', TAB_COMMON]
## PreprocessFile() method
#
@@ -540,7 +493,6 @@ class FdfParser:
# @param self The object pointer
#
def PreprocessFile(self):
-
self.Rewind()
InComment = False
DoubleSlashComment = False
@@ -548,12 +500,12 @@ class FdfParser:
# HashComment in quoted string " " is ignored.
InString = False
- while not self.__EndOfFile():
+ while not self._EndOfFile():
- if self.__CurrentChar() == T_CHAR_DOUBLE_QUOTE and not InComment:
+ if self._CurrentChar() == T_CHAR_DOUBLE_QUOTE and not InComment:
InString = not InString
# meet new line, then no longer in a comment for // and '#'
- if self.__CurrentChar() == T_CHAR_LF:
+ if self._CurrentChar() == TAB_LINE_BREAK:
self.CurrentLineNumber += 1
self.CurrentOffsetWithinLine = 0
if InComment and DoubleSlashComment:
@@ -563,33 +515,33 @@ class FdfParser:
InComment = False
HashComment = False
# check for */ comment end
- elif InComment and not DoubleSlashComment and not HashComment and self.__CurrentChar() == T_CHAR_STAR and self.__NextChar() == T_CHAR_SLASH:
- self.__SetCurrentCharValue(T_CHAR_SPACE)
- self.__GetOneChar()
- self.__SetCurrentCharValue(T_CHAR_SPACE)
- self.__GetOneChar()
+ elif InComment and not DoubleSlashComment and not HashComment and self._CurrentChar() == T_CHAR_STAR and self._NextChar() == TAB_BACK_SLASH:
+ self._SetCurrentCharValue(TAB_SPACE_SPLIT)
+ self._GetOneChar()
+ self._SetCurrentCharValue(TAB_SPACE_SPLIT)
+ self._GetOneChar()
InComment = False
# set comments to spaces
elif InComment:
- self.__SetCurrentCharValue(T_CHAR_SPACE)
- self.__GetOneChar()
+ self._SetCurrentCharValue(TAB_SPACE_SPLIT)
+ self._GetOneChar()
# check for // comment
- elif self.__CurrentChar() == T_CHAR_SLASH and self.__NextChar() == T_CHAR_SLASH and not self.__EndOfLine():
+ elif self._CurrentChar() == TAB_BACK_SLASH and self._NextChar() == TAB_BACK_SLASH and not self._EndOfLine():
InComment = True
DoubleSlashComment = True
# check for '#' comment
- elif self.__CurrentChar() == T_CHAR_HASH and not self.__EndOfLine() and not InString:
+ elif self._CurrentChar() == TAB_COMMENT_SPLIT and not self._EndOfLine() and not InString:
InComment = True
HashComment = True
# check for /* comment start
- elif self.__CurrentChar() == T_CHAR_SLASH and self.__NextChar() == T_CHAR_STAR:
- self.__SetCurrentCharValue( T_CHAR_SPACE)
- self.__GetOneChar()
- self.__SetCurrentCharValue( T_CHAR_SPACE)
- self.__GetOneChar()
+ elif self._CurrentChar() == TAB_BACK_SLASH and self._NextChar() == T_CHAR_STAR:
+ self._SetCurrentCharValue(TAB_SPACE_SPLIT)
+ self._GetOneChar()
+ self._SetCurrentCharValue(TAB_SPACE_SPLIT)
+ self._GetOneChar()
InComment = True
else:
- self.__GetOneChar()
+ self._GetOneChar()
# restore from ListOfList to ListOfString
self.Profile.FileLinesList = ["".join(list) for list in self.Profile.FileLinesList]
@@ -606,30 +558,30 @@ class FdfParser:
# nested include support
Processed = False
MacroDict = {}
- while self.__GetNextToken():
+ while self._GetNextToken():
- if self.__Token == 'DEFINE':
- if not self.__GetNextToken():
+ if self._Token == TAB_DEFINE:
+ if not self._GetNextToken():
raise Warning("expected Macro name", self.FileName, self.CurrentLineNumber)
- Macro = self.__Token
- if not self.__IsToken( "="):
+ Macro = self._Token
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- Value = self.__GetExpression()
+ Value = self._GetExpression()
MacroDict[Macro] = Value
- elif self.__Token == '!include':
+ elif self._Token == TAB_INCLUDE:
Processed = True
IncludeLine = self.CurrentLineNumber
- IncludeOffset = self.CurrentOffsetWithinLine - len('!include')
- if not self.__GetNextToken():
+ IncludeOffset = self.CurrentOffsetWithinLine - len(TAB_INCLUDE)
+ if not self._GetNextToken():
raise Warning("expected include file name", self.FileName, self.CurrentLineNumber)
- IncFileName = self.__Token
+ IncFileName = self._Token
PreIndex = 0
StartPos = IncFileName.find('$(', PreIndex)
EndPos = IncFileName.find(')', StartPos+2)
while StartPos != -1 and EndPos != -1:
- Macro = IncFileName[StartPos+2 : EndPos]
- MacroVal = self.__GetMacroValue(Macro)
+ Macro = IncFileName[StartPos+2: EndPos]
+ MacroVal = self._GetMacroValue(Macro)
if not MacroVal:
if Macro in MacroDict:
MacroVal = MacroDict[Macro]
@@ -686,9 +638,9 @@ class FdfParser:
IncFileProfile.Level = ParentProfile.Level + 1
IncFileProfile.InsertStartLineNumber = InsertAtLine + 1
# deal with remaining portions after "!include filename", if exists.
- if self.__GetNextToken():
+ if self._GetNextToken():
if self.CurrentLineNumber == CurrentLine:
- RemainingLine = self.__CurrentLine()[CurrentOffset:]
+ RemainingLine = self._CurrentLine()[CurrentOffset:]
self.Profile.FileLinesList.insert(self.CurrentLineNumber, RemainingLine)
IncFileProfile.InsertAdjust += 1
self.CurrentLineNumber += 1
@@ -704,7 +656,7 @@ class FdfParser:
# comment out the processed include file statement
TempList = list(self.Profile.FileLinesList[IncludeLine - 1])
- TempList.insert(IncludeOffset, '#')
+ TempList.insert(IncludeOffset, TAB_COMMENT_SPLIT)
self.Profile.FileLinesList[IncludeLine - 1] = ''.join(TempList)
if Processed: # Nested and back-to-back support
self.Rewind(DestLine = IncFileProfile.InsertStartLineNumber - 1)
@@ -713,7 +665,7 @@ class FdfParser:
self.Rewind()
@staticmethod
- def __GetIfListCurrentItemStat(IfList):
+ def _GetIfListCurrentItemStat(IfList):
if len(IfList) == 0:
return True
@@ -735,29 +687,29 @@ class FdfParser:
IfList = []
RegionLayoutLine = 0
ReplacedLine = -1
- while self.__GetNextToken():
+ while self._GetNextToken():
# Determine section name and the location dependent macro
- if self.__GetIfListCurrentItemStat(IfList):
- if self.__Token.startswith('['):
- Header = self.__Token
- if not self.__Token.endswith(']'):
- self.__SkipToToken(']')
- Header += self.__SkippedChars
+ if self._GetIfListCurrentItemStat(IfList):
+ if self._Token.startswith(TAB_SECTION_START):
+ Header = self._Token
+ if not self._Token.endswith(TAB_SECTION_END):
+ self._SkipToToken(TAB_SECTION_END)
+ Header += self._SkippedChars
if Header.find('$(') != -1:
raise Warning("macro cannot be used in section header", self.FileName, self.CurrentLineNumber)
- self.__SectionHeaderParser(Header)
+ self._SectionHeaderParser(Header)
continue
# Replace macros except in RULE section or out of section
- elif self.__CurSection and ReplacedLine != self.CurrentLineNumber:
+ elif self._CurSection and ReplacedLine != self.CurrentLineNumber:
ReplacedLine = self.CurrentLineNumber
- self.__UndoToken()
+ self._UndoToken()
CurLine = self.Profile.FileLinesList[ReplacedLine - 1]
PreIndex = 0
StartPos = CurLine.find('$(', PreIndex)
EndPos = CurLine.find(')', StartPos+2)
- while StartPos != -1 and EndPos != -1 and self.__Token not in ['!ifdef', '!ifndef', '!if', '!elseif']:
- MacroName = CurLine[StartPos+2 : EndPos]
- MacorValue = self.__GetMacroValue(MacroName)
+ while StartPos != -1 and EndPos != -1 and self._Token not in [TAB_IF_DEF, TAB_IF_N_DEF, TAB_IF, TAB_ELSE_IF]:
+ MacroName = CurLine[StartPos+2: EndPos]
+ MacorValue = self._GetMacroValue(MacroName)
if MacorValue is not None:
CurLine = CurLine.replace('$(' + MacroName + ')', MacorValue, 1)
if MacorValue.find('$(') != -1:
@@ -771,74 +723,74 @@ class FdfParser:
self.Profile.FileLinesList[ReplacedLine - 1] = CurLine
continue
- if self.__Token == 'DEFINE':
- if self.__GetIfListCurrentItemStat(IfList):
- if not self.__CurSection:
+ if self._Token == TAB_DEFINE:
+ if self._GetIfListCurrentItemStat(IfList):
+ if not self._CurSection:
raise Warning("macro cannot be defined in Rule section or out of section", self.FileName, self.CurrentLineNumber)
DefineLine = self.CurrentLineNumber - 1
- DefineOffset = self.CurrentOffsetWithinLine - len('DEFINE')
- if not self.__GetNextToken():
+ DefineOffset = self.CurrentOffsetWithinLine - len(TAB_DEFINE)
+ if not self._GetNextToken():
raise Warning("expected Macro name", self.FileName, self.CurrentLineNumber)
- Macro = self.__Token
- if not self.__IsToken( "="):
+ Macro = self._Token
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- Value = self.__GetExpression()
- self.__SetMacroValue(Macro, Value)
- self.__WipeOffArea.append(((DefineLine, DefineOffset), (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
- elif self.__Token == 'SET':
- if not self.__GetIfListCurrentItemStat(IfList):
+ Value = self._GetExpression()
+ self._SetMacroValue(Macro, Value)
+ self._WipeOffArea.append(((DefineLine, DefineOffset), (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ elif self._Token == 'SET':
+ if not self._GetIfListCurrentItemStat(IfList):
continue
SetLine = self.CurrentLineNumber - 1
SetOffset = self.CurrentOffsetWithinLine - len('SET')
- PcdPair = self.__GetNextPcdSettings()
+ PcdPair = self._GetNextPcdSettings()
PcdName = "%s.%s" % (PcdPair[1], PcdPair[0])
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- Value = self.__GetExpression()
- Value = self.__EvaluateConditional(Value, self.CurrentLineNumber, 'eval', True)
+ Value = self._GetExpression()
+ Value = self._EvaluateConditional(Value, self.CurrentLineNumber, 'eval', True)
- self.__PcdDict[PcdName] = Value
+ self._PcdDict[PcdName] = Value
self.Profile.PcdDict[PcdPair] = Value
self.SetPcdLocalation(PcdPair)
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
self.Profile.PcdFileLineDict[PcdPair] = FileLineTuple
- self.__WipeOffArea.append(((SetLine, SetOffset), (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
- elif self.__Token in ('!ifdef', '!ifndef', '!if'):
- IfStartPos = (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - len(self.__Token))
+ self._WipeOffArea.append(((SetLine, SetOffset), (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ elif self._Token in (TAB_IF_DEF, TAB_IF_N_DEF, TAB_IF):
+ IfStartPos = (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - len(self._Token))
IfList.append([IfStartPos, None, None])
- CondLabel = self.__Token
- Expression = self.__GetExpression()
+ CondLabel = self._Token
+ Expression = self._GetExpression()
- if CondLabel == '!if':
- ConditionSatisfied = self.__EvaluateConditional(Expression, IfList[-1][0][0] + 1, 'eval')
+ if CondLabel == TAB_IF:
+ ConditionSatisfied = self._EvaluateConditional(Expression, IfList[-1][0][0] + 1, 'eval')
else:
- ConditionSatisfied = self.__EvaluateConditional(Expression, IfList[-1][0][0] + 1, 'in')
- if CondLabel == '!ifndef':
+ ConditionSatisfied = self._EvaluateConditional(Expression, IfList[-1][0][0] + 1, 'in')
+ if CondLabel == TAB_IF_N_DEF:
ConditionSatisfied = not ConditionSatisfied
BranchDetermined = ConditionSatisfied
IfList[-1] = [IfList[-1][0], ConditionSatisfied, BranchDetermined]
if ConditionSatisfied:
- self.__WipeOffArea.append((IfList[-1][0], (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
- elif self.__Token in ('!elseif', '!else'):
- ElseStartPos = (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - len(self.__Token))
+ self._WipeOffArea.append((IfList[-1][0], (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ elif self._Token in (TAB_ELSE_IF, '!else'):
+ ElseStartPos = (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - len(self._Token))
if len(IfList) <= 0:
raise Warning("Missing !if statement", self.FileName, self.CurrentLineNumber)
if IfList[-1][1]:
IfList[-1] = [ElseStartPos, False, True]
- self.__WipeOffArea.append((ElseStartPos, (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ self._WipeOffArea.append((ElseStartPos, (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
else:
- self.__WipeOffArea.append((IfList[-1][0], ElseStartPos))
+ self._WipeOffArea.append((IfList[-1][0], ElseStartPos))
IfList[-1] = [ElseStartPos, True, IfList[-1][2]]
- if self.__Token == '!elseif':
- Expression = self.__GetExpression()
- ConditionSatisfied = self.__EvaluateConditional(Expression, IfList[-1][0][0] + 1, 'eval')
+ if self._Token == TAB_ELSE_IF:
+ Expression = self._GetExpression()
+ ConditionSatisfied = self._EvaluateConditional(Expression, IfList[-1][0][0] + 1, 'eval')
IfList[-1] = [IfList[-1][0], ConditionSatisfied, IfList[-1][2]]
if IfList[-1][1]:
@@ -846,14 +798,14 @@ class FdfParser:
IfList[-1][1] = False
else:
IfList[-1][2] = True
- self.__WipeOffArea.append((IfList[-1][0], (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
- elif self.__Token == '!endif':
+ self._WipeOffArea.append((IfList[-1][0], (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ elif self._Token == '!endif':
if len(IfList) <= 0:
raise Warning("Missing !if statement", self.FileName, self.CurrentLineNumber)
if IfList[-1][1]:
- self.__WipeOffArea.append(((self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - len('!endif')), (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ self._WipeOffArea.append(((self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - len('!endif')), (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
else:
- self.__WipeOffArea.append((IfList[-1][0], (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
+ self._WipeOffArea.append((IfList[-1][0], (self.CurrentLineNumber - 1, self.CurrentOffsetWithinLine - 1)))
IfList.pop()
elif not IfList: # Don't use PCDs inside conditional directive
@@ -862,7 +814,7 @@ class FdfParser:
continue
SetPcd = ShortcutPcdPattern.match(self.Profile.FileLinesList[self.CurrentLineNumber - 1])
if SetPcd:
- self.__PcdDict[SetPcd.group('name')] = SetPcd.group('value')
+ self._PcdDict[SetPcd.group('name')] = SetPcd.group('value')
RegionLayoutLine = self.CurrentLineNumber
continue
RegionSize = RegionSizePattern.match(self.Profile.FileLinesList[self.CurrentLineNumber - 1])
@@ -873,35 +825,35 @@ class FdfParser:
if not RegionSizeGuid:
RegionLayoutLine = self.CurrentLineNumber + 1
continue
- self.__PcdDict[RegionSizeGuid.group('base')] = RegionSize.group('base')
- self.__PcdDict[RegionSizeGuid.group('size')] = RegionSize.group('size')
+ self._PcdDict[RegionSizeGuid.group('base')] = RegionSize.group('base')
+ self._PcdDict[RegionSizeGuid.group('size')] = RegionSize.group('size')
RegionLayoutLine = self.CurrentLineNumber + 1
if IfList:
raise Warning("Missing !endif", self.FileName, self.CurrentLineNumber)
self.Rewind()
- def __CollectMacroPcd(self):
+ def _CollectMacroPcd(self):
MacroDict = {}
# PCD macro
MacroDict.update(GlobalData.gPlatformPcds)
- MacroDict.update(self.__PcdDict)
+ MacroDict.update(self._PcdDict)
# Lowest priority
MacroDict.update(GlobalData.gPlatformDefines)
- if self.__CurSection:
+ if self._CurSection:
# Defines macro
- ScopeMacro = self.__MacroDict[TAB_COMMON, TAB_COMMON, TAB_COMMON]
+ ScopeMacro = self._MacroDict[TAB_COMMON, TAB_COMMON, TAB_COMMON]
if ScopeMacro:
MacroDict.update(ScopeMacro)
# Section macro
- ScopeMacro = self.__MacroDict[
- self.__CurSection[0],
- self.__CurSection[1],
- self.__CurSection[2]
+ ScopeMacro = self._MacroDict[
+ self._CurSection[0],
+ self._CurSection[1],
+ self._CurSection[2]
]
if ScopeMacro:
MacroDict.update(ScopeMacro)
@@ -912,15 +864,15 @@ class FdfParser:
for Item in GlobalData.BuildOptionPcd:
if isinstance(Item, tuple):
continue
- PcdName, TmpValue = Item.split("=")
+ PcdName, TmpValue = Item.split(TAB_EQUAL_SPLIT)
TmpValue = BuildOptionValue(TmpValue, {})
MacroDict[PcdName.strip()] = TmpValue
# Highest priority
return MacroDict
- def __EvaluateConditional(self, Expression, Line, Op = None, Value = None):
- MacroPcdDict = self.__CollectMacroPcd()
+ def _EvaluateConditional(self, Expression, Line, Op = None, Value = None):
+ MacroPcdDict = self._CollectMacroPcd()
if Op == 'eval':
try:
if Value:
@@ -933,7 +885,7 @@ class FdfParser:
# the precise number of line and return the evaluation result
#
EdkLogger.warn('Parser', "Suspicious expression: %s" % str(Excpt),
- File=self.FileName, ExtraData=self.__CurrentLine(),
+ File=self.FileName, ExtraData=self._CurrentLine(),
Line=Line)
return Excpt.result
except Exception as Excpt:
@@ -955,10 +907,10 @@ class FdfParser:
Expression = Expression[2:-1]
return Expression in MacroPcdDict
- ## __IsToken() method
+ ## _IsToken() method
#
# Check whether input string is found from current char position along
- # If found, the string value is put into self.__Token
+ # If found, the string value is put into self._Token
#
# @param self The object pointer
# @param String The string to search
@@ -966,26 +918,26 @@ class FdfParser:
# @retval True Successfully find string, file buffer pointer moved forward
# @retval False Not able to find string, file buffer pointer not changed
#
- def __IsToken(self, String, IgnoreCase = False):
- self.__SkipWhiteSpace()
+ def _IsToken(self, String, IgnoreCase = False):
+ self._SkipWhiteSpace()
# Only consider the same line, no multi-line token allowed
StartPos = self.CurrentOffsetWithinLine
index = -1
if IgnoreCase:
- index = self.__CurrentLine()[self.CurrentOffsetWithinLine : ].upper().find(String.upper())
+ index = self._CurrentLine()[self.CurrentOffsetWithinLine: ].upper().find(String.upper())
else:
- index = self.__CurrentLine()[self.CurrentOffsetWithinLine : ].find(String)
+ index = self._CurrentLine()[self.CurrentOffsetWithinLine: ].find(String)
if index == 0:
self.CurrentOffsetWithinLine += len(String)
- self.__Token = self.__CurrentLine()[StartPos : self.CurrentOffsetWithinLine]
+ self._Token = self._CurrentLine()[StartPos: self.CurrentOffsetWithinLine]
return True
return False
- ## __IsKeyword() method
+ ## _IsKeyword() method
#
# Check whether input keyword is found from current char position along, whole word only!
- # If found, the string value is put into self.__Token
+ # If found, the string value is put into self._Token
#
# @param self The object pointer
# @param Keyword The string to search
@@ -993,117 +945,117 @@ class FdfParser:
# @retval True Successfully find string, file buffer pointer moved forward
# @retval False Not able to find string, file buffer pointer not changed
#
- def __IsKeyword(self, KeyWord, IgnoreCase = False):
- self.__SkipWhiteSpace()
+ def _IsKeyword(self, KeyWord, IgnoreCase = False):
+ self._SkipWhiteSpace()
# Only consider the same line, no multi-line token allowed
StartPos = self.CurrentOffsetWithinLine
index = -1
if IgnoreCase:
- index = self.__CurrentLine()[self.CurrentOffsetWithinLine : ].upper().find(KeyWord.upper())
+ index = self._CurrentLine()[self.CurrentOffsetWithinLine: ].upper().find(KeyWord.upper())
else:
- index = self.__CurrentLine()[self.CurrentOffsetWithinLine : ].find(KeyWord)
+ index = self._CurrentLine()[self.CurrentOffsetWithinLine: ].find(KeyWord)
if index == 0:
- followingChar = self.__CurrentLine()[self.CurrentOffsetWithinLine + len(KeyWord)]
- if not str(followingChar).isspace() and followingChar not in SEPERATOR_TUPLE:
+ followingChar = self._CurrentLine()[self.CurrentOffsetWithinLine + len(KeyWord)]
+ if not str(followingChar).isspace() and followingChar not in SEPARATORS:
return False
self.CurrentOffsetWithinLine += len(KeyWord)
- self.__Token = self.__CurrentLine()[StartPos : self.CurrentOffsetWithinLine]
+ self._Token = self._CurrentLine()[StartPos: self.CurrentOffsetWithinLine]
return True
return False
- def __GetExpression(self):
+ def _GetExpression(self):
Line = self.Profile.FileLinesList[self.CurrentLineNumber - 1]
Index = len(Line) - 1
- while Line[Index] in ['\r', '\n']:
+ while Line[Index] in [T_CHAR_CR, TAB_LINE_BREAK]:
Index -= 1
ExpressionString = self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:Index+1]
self.CurrentOffsetWithinLine += len(ExpressionString)
ExpressionString = ExpressionString.strip()
return ExpressionString
- ## __GetNextWord() method
+ ## _GetNextWord() method
#
# Get next C name from file lines
- # If found, the string value is put into self.__Token
+ # If found, the string value is put into self._Token
#
# @param self The object pointer
# @retval True Successfully find a C name string, file buffer pointer moved forward
# @retval False Not able to find a C name string, file buffer pointer not changed
#
- def __GetNextWord(self):
- self.__SkipWhiteSpace()
- if self.__EndOfFile():
+ def _GetNextWord(self):
+ self._SkipWhiteSpace()
+ if self._EndOfFile():
return False
- TempChar = self.__CurrentChar()
+ TempChar = self._CurrentChar()
StartPos = self.CurrentOffsetWithinLine
if (TempChar >= 'a' and TempChar <= 'z') or (TempChar >= 'A' and TempChar <= 'Z') or TempChar == '_':
- self.__GetOneChar()
- while not self.__EndOfLine():
- TempChar = self.__CurrentChar()
+ self._GetOneChar()
+ while not self._EndOfLine():
+ TempChar = self._CurrentChar()
if (TempChar >= 'a' and TempChar <= 'z') or (TempChar >= 'A' and TempChar <= 'Z') \
or (TempChar >= '0' and TempChar <= '9') or TempChar == '_' or TempChar == '-':
- self.__GetOneChar()
+ self._GetOneChar()
else:
break
- self.__Token = self.__CurrentLine()[StartPos : self.CurrentOffsetWithinLine]
+ self._Token = self._CurrentLine()[StartPos: self.CurrentOffsetWithinLine]
return True
return False
- def __GetNextPcdWord(self):
- self.__SkipWhiteSpace()
- if self.__EndOfFile():
+ def _GetNextPcdWord(self):
+ self._SkipWhiteSpace()
+ if self._EndOfFile():
return False
- TempChar = self.__CurrentChar()
+ TempChar = self._CurrentChar()
StartPos = self.CurrentOffsetWithinLine
- if (TempChar >= 'a' and TempChar <= 'z') or (TempChar >= 'A' and TempChar <= 'Z') or TempChar == '_' or TempChar == '[' or TempChar == ']':
- self.__GetOneChar()
- while not self.__EndOfLine():
- TempChar = self.__CurrentChar()
+ if (TempChar >= 'a' and TempChar <= 'z') or (TempChar >= 'A' and TempChar <= 'Z') or TempChar == '_' or TempChar == TAB_SECTION_START or TempChar == TAB_SECTION_END:
+ self._GetOneChar()
+ while not self._EndOfLine():
+ TempChar = self._CurrentChar()
if (TempChar >= 'a' and TempChar <= 'z') or (TempChar >= 'A' and TempChar <= 'Z') \
- or (TempChar >= '0' and TempChar <= '9') or TempChar == '_' or TempChar == '-' or TempChar == '[' or TempChar == ']':
- self.__GetOneChar()
+ or (TempChar >= '0' and TempChar <= '9') or TempChar == '_' or TempChar == '-' or TempChar == TAB_SECTION_START or TempChar == TAB_SECTION_END:
+ self._GetOneChar()
else:
break
- self.__Token = self.__CurrentLine()[StartPos : self.CurrentOffsetWithinLine]
+ self._Token = self._CurrentLine()[StartPos: self.CurrentOffsetWithinLine]
return True
return False
- ## __GetNextToken() method
+ ## _GetNextToken() method
#
# Get next token unit before a seperator
- # If found, the string value is put into self.__Token
+ # If found, the string value is put into self._Token
#
# @param self The object pointer
# @retval True Successfully find a token unit, file buffer pointer moved forward
# @retval False Not able to find a token unit, file buffer pointer not changed
#
- def __GetNextToken(self):
+ def _GetNextToken(self):
# Skip leading spaces, if exist.
- self.__SkipWhiteSpace()
- if self.__EndOfFile():
+ self._SkipWhiteSpace()
+ if self._EndOfFile():
return False
# Record the token start position, the position of the first non-space char.
StartPos = self.CurrentOffsetWithinLine
StartLine = self.CurrentLineNumber
while StartLine == self.CurrentLineNumber:
- TempChar = self.__CurrentChar()
+ TempChar = self._CurrentChar()
# Try to find the end char that is not a space and not in seperator tuple.
# That is, when we got a space or any char in the tuple, we got the end of token.
- if not str(TempChar).isspace() and TempChar not in SEPERATOR_TUPLE:
- self.__GetOneChar()
+ if not str(TempChar).isspace() and TempChar not in SEPARATORS:
+ self._GetOneChar()
# if we happen to meet a seperator as the first char, we must proceed to get it.
# That is, we get a token that is a seperator char. nomally it is the boundary of other tokens.
- elif StartPos == self.CurrentOffsetWithinLine and TempChar in SEPERATOR_TUPLE:
- self.__GetOneChar()
+ elif StartPos == self.CurrentOffsetWithinLine and TempChar in SEPARATORS:
+ self._GetOneChar()
break
else:
break
@@ -1113,35 +1065,34 @@ class FdfParser:
EndPos = self.CurrentOffsetWithinLine
if self.CurrentLineNumber != StartLine:
EndPos = len(self.Profile.FileLinesList[StartLine-1])
- self.__Token = self.Profile.FileLinesList[StartLine-1][StartPos : EndPos]
- if self.__Token.lower() in [TAB_IF, TAB_END_IF, TAB_ELSE_IF, TAB_ELSE, TAB_IF_DEF, TAB_IF_N_DEF, TAB_ERROR, TAB_INCLUDE]:
- self.__Token = self.__Token.lower()
+ self._Token = self.Profile.FileLinesList[StartLine-1][StartPos: EndPos]
+ if self._Token.lower() in [TAB_IF, TAB_END_IF, TAB_ELSE_IF, TAB_ELSE, TAB_IF_DEF, TAB_IF_N_DEF, TAB_ERROR, TAB_INCLUDE]:
+ self._Token = self._Token.lower()
if StartPos != self.CurrentOffsetWithinLine:
return True
else:
return False
- ## __GetNextGuid() method
+ ## _GetNextGuid() method
#
# Get next token unit before a seperator
- # If found, the GUID string is put into self.__Token
+ # If found, the GUID string is put into self._Token
#
# @param self The object pointer
# @retval True Successfully find a registry format GUID, file buffer pointer moved forward
# @retval False Not able to find a registry format GUID, file buffer pointer not changed
#
- def __GetNextGuid(self):
-
- if not self.__GetNextToken():
+ def _GetNextGuid(self):
+ if not self._GetNextToken():
return False
- if gGuidPattern.match(self.__Token) is not None:
+ if gGuidPattern.match(self._Token) is not None:
return True
else:
- self.__UndoToken()
+ self._UndoToken()
return False
@staticmethod
- def __Verify(Name, Value, Scope):
+ def _Verify(Name, Value, Scope):
# value verification only applies to numeric values.
if Scope not in TAB_PCD_NUMERIC_TYPES:
return
@@ -1157,17 +1108,17 @@ class FdfParser:
EdkLogger.error("FdfParser", FORMAT_INVALID, "Too large value for %s." % Name)
return True
- ## __UndoToken() method
+ ## _UndoToken() method
#
# Go back one token unit in file buffer
#
# @param self The object pointer
#
- def __UndoToken(self):
- self.__UndoOneChar()
- while self.__CurrentChar().isspace():
- if not self.__UndoOneChar():
- self.__GetOneChar()
+ def _UndoToken(self):
+ self._UndoOneChar()
+ while self._CurrentChar().isspace():
+ if not self._UndoOneChar():
+ self._GetOneChar()
return
@@ -1175,118 +1126,118 @@ class FdfParser:
CurrentLine = self.CurrentLineNumber
while CurrentLine == self.CurrentLineNumber:
- TempChar = self.__CurrentChar()
+ TempChar = self._CurrentChar()
# Try to find the end char that is not a space and not in seperator tuple.
# That is, when we got a space or any char in the tuple, we got the end of token.
- if not str(TempChar).isspace() and not TempChar in SEPERATOR_TUPLE:
- if not self.__UndoOneChar():
+ if not str(TempChar).isspace() and not TempChar in SEPARATORS:
+ if not self._UndoOneChar():
return
# if we happen to meet a seperator as the first char, we must proceed to get it.
# That is, we get a token that is a seperator char. nomally it is the boundary of other tokens.
- elif StartPos == self.CurrentOffsetWithinLine and TempChar in SEPERATOR_TUPLE:
+ elif StartPos == self.CurrentOffsetWithinLine and TempChar in SEPARATORS:
return
else:
break
- self.__GetOneChar()
+ self._GetOneChar()
- ## __GetNextHexNumber() method
+ ## _GetNextHexNumber() method
#
# Get next HEX data before a seperator
- # If found, the HEX data is put into self.__Token
+ # If found, the HEX data is put into self._Token
#
# @param self The object pointer
# @retval True Successfully find a HEX data, file buffer pointer moved forward
# @retval False Not able to find a HEX data, file buffer pointer not changed
#
- def __GetNextHexNumber(self):
- if not self.__GetNextToken():
+ def _GetNextHexNumber(self):
+ if not self._GetNextToken():
return False
- if gHexPatternAll.match(self.__Token):
+ if gHexPatternAll.match(self._Token):
return True
else:
- self.__UndoToken()
+ self._UndoToken()
return False
- ## __GetNextDecimalNumber() method
+ ## _GetNextDecimalNumber() method
#
# Get next decimal data before a seperator
- # If found, the decimal data is put into self.__Token
+ # If found, the decimal data is put into self._Token
#
# @param self The object pointer
# @retval True Successfully find a decimal data, file buffer pointer moved forward
# @retval False Not able to find a decimal data, file buffer pointer not changed
#
- def __GetNextDecimalNumber(self):
- if not self.__GetNextToken():
+ def _GetNextDecimalNumber(self):
+ if not self._GetNextToken():
return False
- if self.__Token.isdigit():
+ if self._Token.isdigit():
return True
else:
- self.__UndoToken()
+ self._UndoToken()
return False
- def __GetNextPcdSettings(self):
- if not self.__GetNextWord():
+ def _GetNextPcdSettings(self):
+ if not self._GetNextWord():
raise Warning("expected format of <PcdTokenSpaceCName>.<PcdCName>", self.FileName, self.CurrentLineNumber)
- pcdTokenSpaceCName = self.__Token
+ pcdTokenSpaceCName = self._Token
- if not self.__IsToken( "."):
+ if not self._IsToken(TAB_SPLIT):
raise Warning("expected format of <PcdTokenSpaceCName>.<PcdCName>", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected format of <PcdTokenSpaceCName>.<PcdCName>", self.FileName, self.CurrentLineNumber)
- pcdCName = self.__Token
+ pcdCName = self._Token
Fields = []
- while self.__IsToken("."):
- if not self.__GetNextPcdWord():
+ while self._IsToken(TAB_SPLIT):
+ if not self._GetNextPcdWord():
raise Warning("expected format of <PcdTokenSpaceCName>.<PcdCName>", self.FileName, self.CurrentLineNumber)
- Fields.append(self.__Token)
+ Fields.append(self._Token)
- return (pcdCName, pcdTokenSpaceCName,".".join(Fields))
+ return (pcdCName, pcdTokenSpaceCName,TAB_SPLIT.join(Fields))
- ## __GetStringData() method
+ ## _GetStringData() method
#
# Get string contents quoted in ""
- # If found, the decimal data is put into self.__Token
+ # If found, the decimal data is put into self._Token
#
# @param self The object pointer
# @retval True Successfully find a string data, file buffer pointer moved forward
# @retval False Not able to find a string data, file buffer pointer not changed
#
- def __GetStringData(self):
- if self.__Token.startswith("\"") or self.__Token.startswith("L\""):
- self.__UndoToken()
- self.__SkipToToken("\"")
+ def _GetStringData(self):
+ if self._Token.startswith(T_CHAR_DOUBLE_QUOTE) or self._Token.startswith("L\""):
+ self._UndoToken()
+ self._SkipToToken(T_CHAR_DOUBLE_QUOTE)
currentLineNumber = self.CurrentLineNumber
- if not self.__SkipToToken("\""):
+ if not self._SkipToToken(T_CHAR_DOUBLE_QUOTE):
raise Warning("Missing Quote \" for String", self.FileName, self.CurrentLineNumber)
if currentLineNumber != self.CurrentLineNumber:
raise Warning("Missing Quote \" for String", self.FileName, self.CurrentLineNumber)
- self.__Token = self.__SkippedChars.rstrip('\"')
+ self._Token = self._SkippedChars.rstrip(T_CHAR_DOUBLE_QUOTE)
return True
- elif self.__Token.startswith("\'") or self.__Token.startswith("L\'"):
- self.__UndoToken()
- self.__SkipToToken("\'")
+ elif self._Token.startswith(T_CHAR_SINGLE_QUOTE) or self._Token.startswith("L\'"):
+ self._UndoToken()
+ self._SkipToToken(T_CHAR_SINGLE_QUOTE)
currentLineNumber = self.CurrentLineNumber
- if not self.__SkipToToken("\'"):
+ if not self._SkipToToken(T_CHAR_SINGLE_QUOTE):
raise Warning("Missing Quote \' for String", self.FileName, self.CurrentLineNumber)
if currentLineNumber != self.CurrentLineNumber:
raise Warning("Missing Quote \' for String", self.FileName, self.CurrentLineNumber)
- self.__Token = self.__SkippedChars.rstrip('\'')
+ self._Token = self._SkippedChars.rstrip(T_CHAR_SINGLE_QUOTE)
return True
else:
return False
- ## __SkipToToken() method
+ ## _SkipToToken() method
#
# Search forward in file buffer for the string
- # The skipped chars are put into self.__SkippedChars
+ # The skipped chars are put into self._SkippedChars
#
# @param self The object pointer
# @param String The string to search
@@ -1294,25 +1245,25 @@ class FdfParser:
# @retval True Successfully find the string, file buffer pointer moved forward
# @retval False Not able to find the string, file buffer pointer not changed
#
- def __SkipToToken(self, String, IgnoreCase = False):
+ def _SkipToToken(self, String, IgnoreCase = False):
StartPos = self.GetFileBufferPos()
- self.__SkippedChars = ""
- while not self.__EndOfFile():
+ self._SkippedChars = ""
+ while not self._EndOfFile():
index = -1
if IgnoreCase:
- index = self.__CurrentLine()[self.CurrentOffsetWithinLine : ].upper().find(String.upper())
+ index = self._CurrentLine()[self.CurrentOffsetWithinLine: ].upper().find(String.upper())
else:
- index = self.__CurrentLine()[self.CurrentOffsetWithinLine : ].find(String)
+ index = self._CurrentLine()[self.CurrentOffsetWithinLine: ].find(String)
if index == 0:
self.CurrentOffsetWithinLine += len(String)
- self.__SkippedChars += String
+ self._SkippedChars += String
return True
- self.__SkippedChars += str(self.__CurrentChar())
- self.__GetOneChar()
+ self._SkippedChars += str(self._CurrentChar())
+ self._GetOneChar()
- self.SetFileBufferPos( StartPos)
- self.__SkippedChars = ""
+ self.SetFileBufferPos(StartPos)
+ self._SkippedChars = ""
return False
## GetFileBufferPos() method
@@ -1343,18 +1294,18 @@ class FdfParser:
# @param self The object pointer
#
def Preprocess(self):
- self.__StringToList()
+ self._StringToList()
self.PreprocessFile()
self.PreprocessIncludeFile()
- self.__StringToList()
+ self._StringToList()
self.PreprocessFile()
self.PreprocessConditionalStatement()
- self.__StringToList()
- for Pos in self.__WipeOffArea:
- self.__ReplaceFragment(Pos[0], Pos[1])
+ self._StringToList()
+ for Pos in self._WipeOffArea:
+ self._ReplaceFragment(Pos[0], Pos[1])
self.Profile.FileLinesList = ["".join(list) for list in self.Profile.FileLinesList]
- while self.__GetDefines():
+ while self._GetDefines():
pass
## ParseFile() method
@@ -1365,19 +1316,18 @@ class FdfParser:
# @param self The object pointer
#
def ParseFile(self):
-
try:
self.Preprocess()
- self.__GetError()
+ self._GetError()
#
# Keep processing sections of the FDF until no new sections or a syntax error is found
#
- while self.__GetFd() or self.__GetFv() or self.__GetFmp() or self.__GetCapsule() or self.__GetVtf() or self.__GetRule() or self.__GetOptionRom():
+ while self._GetFd() or self._GetFv() or self._GetFmp() or self._GetCapsule() or self._GetVtf() or self._GetRule() or self._GetOptionRom():
pass
except Warning as X:
- self.__UndoToken()
- #'\n\tGot Token: \"%s\" from File %s\n' % (self.__Token, FileLineTuple[0]) + \
+ self._UndoToken()
+ #'\n\tGot Token: \"%s\" from File %s\n' % (self._Token, FileLineTuple[0]) + \
# At this point, the closest parent would be the included file itself
Profile = GetParentAtLine(X.OriginalLineNumber)
if Profile is not None:
@@ -1386,7 +1336,7 @@ class FdfParser:
else:
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
X.Message += ' near line %d, column %d: %s' \
- % (FileLineTuple[1], self.CurrentOffsetWithinLine + 1, self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :].rstrip('\n').rstrip('\r'))
+ % (FileLineTuple[1], self.CurrentOffsetWithinLine + 1, self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:].rstrip(TAB_LINE_BREAK).rstrip(T_CHAR_CR))
raise
## SectionParser() method
@@ -1403,7 +1353,7 @@ class FdfParser:
and not S.startswith("[VTF.") and not S.startswith("[RULE.") and not S.startswith("[OPTIONROM.") and not S.startswith('[FMPPAYLOAD.'):
raise Warning("Unknown section or section appear sequence error (The correct sequence should be [DEFINES], [FD.], [FV.], [Capsule.], [VTF.], [Rule.], [OptionRom.], [FMPPAYLOAD.])", self.FileName, self.CurrentLineNumber)
- ## __GetDefines() method
+ ## _GetDefines() method
#
# Get Defines section contents and store its data into AllMacrosList
#
@@ -1411,56 +1361,55 @@ class FdfParser:
# @retval True Successfully find a Defines
# @retval False Not able to find a Defines
#
- def __GetDefines(self):
-
- if not self.__GetNextToken():
+ def _GetDefines(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[DEFINES"):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[DEFINES"):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[DEFINES", True):
+ self._UndoToken()
+ if not self._IsToken("[DEFINES", True):
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
#print 'Parsing String: %s in File %s, At line: %d, Offset Within Line: %d' \
- # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
+ # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
raise Warning("expected [DEFINES", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- while self.__GetNextWord():
+ while self._GetNextWord():
# handle the SET statement
- if self.__Token == 'SET':
- self.__UndoToken()
- self.__GetSetStatement(None)
+ if self._Token == 'SET':
+ self._UndoToken()
+ self._GetSetStatement(None)
continue
- Macro = self.__Token
+ Macro = self._Token
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken() or self.__Token.startswith('['):
+ if not self._GetNextToken() or self._Token.startswith(TAB_SECTION_START):
raise Warning("expected MACRO value", self.FileName, self.CurrentLineNumber)
- Value = self.__Token
+ Value = self._Token
return False
- ##__GetError() method
- def __GetError(self):
+ ##_GetError() method
+ def _GetError(self):
#save the Current information
CurrentLine = self.CurrentLineNumber
CurrentOffset = self.CurrentOffsetWithinLine
- while self.__GetNextToken():
- if self.__Token == TAB_ERROR:
- EdkLogger.error('FdfParser', ERROR_STATEMENT, self.__CurrentLine().replace(TAB_ERROR, '', 1), File=self.FileName, Line=self.CurrentLineNumber)
+ while self._GetNextToken():
+ if self._Token == TAB_ERROR:
+ EdkLogger.error('FdfParser', ERROR_STATEMENT, self._CurrentLine().replace(TAB_ERROR, '', 1), File=self.FileName, Line=self.CurrentLineNumber)
self.CurrentLineNumber = CurrentLine
self.CurrentOffsetWithinLine = CurrentOffset
- ## __GetFd() method
+ ## _GetFd() method
#
# Get FD section contents and store its data into FD dictionary of self.Profile
#
@@ -1468,27 +1417,26 @@ class FdfParser:
# @retval True Successfully find a FD
# @retval False Not able to find a FD
#
- def __GetFd(self):
-
- if not self.__GetNextToken():
+ def _GetFd(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[FD."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[FD."):
if not S.startswith("[FV.") and not S.startswith('[FMPPAYLOAD.') and not S.startswith("[CAPSULE.") \
and not S.startswith("[VTF.") and not S.startswith("[RULE.") and not S.startswith("[OPTIONROM."):
raise Warning("Unknown section", self.FileName, self.CurrentLineNumber)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[FD.", True):
+ self._UndoToken()
+ if not self._IsToken("[FD.", True):
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
#print 'Parsing String: %s in File %s, At line: %d, Offset Within Line: %d' \
- # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
+ # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
raise Warning("expected [FD.]", self.FileName, self.CurrentLineNumber)
- FdName = self.__GetUiName()
+ FdName = self._GetUiName()
if FdName == "":
if len (self.Profile.FdDict) == 0:
FdName = GenFdsGlobalVariable.PlatformName
@@ -1502,56 +1450,56 @@ class FdfParser:
if self.CurrentFdName in self.Profile.FdDict:
raise Warning("Unexpected the same FD name", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- FdObj = Fd.FD()
+ FdObj = FD()
FdObj.FdUiName = self.CurrentFdName
self.Profile.FdDict[self.CurrentFdName] = FdObj
if len (self.Profile.FdDict) > 1 and self.Profile.FdNameNotSet:
raise Warning("expected all FDs have their name", self.FileName, self.CurrentLineNumber)
- Status = self.__GetCreateFile(FdObj)
+ Status = self._GetCreateFile(FdObj)
if not Status:
raise Warning("FD name error", self.FileName, self.CurrentLineNumber)
- while self.__GetTokenStatements(FdObj):
+ while self._GetTokenStatements(FdObj):
pass
for Attr in ("BaseAddress", "Size", "ErasePolarity"):
if getattr(FdObj, Attr) is None:
- self.__GetNextToken()
+ self._GetNextToken()
raise Warning("Keyword %s missing" % Attr, self.FileName, self.CurrentLineNumber)
if not FdObj.BlockSizeList:
FdObj.BlockSizeList.append((1, FdObj.Size, None))
- self.__GetDefineStatements(FdObj)
+ self._GetDefineStatements(FdObj)
- self.__GetSetStatements(FdObj)
+ self._GetSetStatements(FdObj)
- if not self.__GetRegionLayout(FdObj):
+ if not self._GetRegionLayout(FdObj):
raise Warning("expected region layout", self.FileName, self.CurrentLineNumber)
- while self.__GetRegionLayout(FdObj):
+ while self._GetRegionLayout(FdObj):
pass
return True
- ## __GetUiName() method
+ ## _GetUiName() method
#
# Return the UI name of a section
#
# @param self The object pointer
# @retval FdName UI name
#
- def __GetUiName(self):
+ def _GetUiName(self):
Name = ""
- if self.__GetNextWord():
- Name = self.__Token
+ if self._GetNextWord():
+ Name = self._Token
return Name
- ## __GetCreateFile() method
+ ## _GetCreateFile() method
#
# Return the output file name of object
#
@@ -1559,16 +1507,15 @@ class FdfParser:
# @param Obj object whose data will be stored in file
# @retval FdName UI name
#
- def __GetCreateFile(self, Obj):
-
- if self.__IsKeyword( "CREATE_FILE"):
- if not self.__IsToken( "="):
+ def _GetCreateFile(self, Obj):
+ if self._IsKeyword("CREATE_FILE"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected file name", self.FileName, self.CurrentLineNumber)
- FileName = self.__Token
+ FileName = self._Token
Obj.CreateFileName = FileName
return True
@@ -1576,25 +1523,25 @@ class FdfParser:
def SetPcdLocalation(self,pcdpair):
self.Profile.PcdLocalDict[pcdpair] = (self.Profile.FileName,self.CurrentLineNumber)
- ## __GetTokenStatements() method
+ ## _GetTokenStatements() method
#
# Get token statements
#
# @param self The object pointer
# @param Obj for whom token statement is got
#
- def __GetTokenStatements(self, Obj):
- if self.__IsKeyword( "BaseAddress"):
- if not self.__IsToken( "="):
+ def _GetTokenStatements(self, Obj):
+ if self._IsKeyword("BaseAddress"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex base address", self.FileName, self.CurrentLineNumber)
- Obj.BaseAddress = self.__Token
+ Obj.BaseAddress = self._Token
- if self.__IsToken( "|"):
- pcdPair = self.__GetNextPcdSettings()
+ if self._IsToken(TAB_VALUE_SPLIT):
+ pcdPair = self._GetNextPcdSettings()
Obj.BaseAddressPcd = pcdPair
self.Profile.PcdDict[pcdPair] = Obj.BaseAddress
self.SetPcdLocalation(pcdPair)
@@ -1602,16 +1549,16 @@ class FdfParser:
self.Profile.PcdFileLineDict[pcdPair] = FileLineTuple
return True
- if self.__IsKeyword( "Size"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("Size"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex size", self.FileName, self.CurrentLineNumber)
- Size = self.__Token
- if self.__IsToken( "|"):
- pcdPair = self.__GetNextPcdSettings()
+ Size = self._Token
+ if self._IsToken(TAB_VALUE_SPLIT):
+ pcdPair = self._GetNextPcdSettings()
Obj.SizePcd = pcdPair
self.Profile.PcdDict[pcdPair] = Size
self.SetPcdLocalation(pcdPair)
@@ -1620,22 +1567,22 @@ class FdfParser:
Obj.Size = long(Size, 0)
return True
- if self.__IsKeyword( "ErasePolarity"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("ErasePolarity"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Erase Polarity", self.FileName, self.CurrentLineNumber)
- if self.__Token != "1" and self.__Token != "0":
+ if self._Token != "1" and self._Token != "0":
raise Warning("expected 1 or 0 Erase Polarity", self.FileName, self.CurrentLineNumber)
- Obj.ErasePolarity = self.__Token
+ Obj.ErasePolarity = self._Token
return True
- return self.__GetBlockStatements(Obj)
+ return self._GetBlockStatements(Obj)
- ## __GetAddressStatements() method
+ ## _GetAddressStatements() method
#
# Get address statements
#
@@ -1644,38 +1591,37 @@ class FdfParser:
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetAddressStatements(self, Obj):
-
- if self.__IsKeyword("BsBaseAddress"):
- if not self.__IsToken( "="):
+ def _GetAddressStatements(self, Obj):
+ if self._IsKeyword("BsBaseAddress"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextDecimalNumber() and not self.__GetNextHexNumber():
+ if not self._GetNextDecimalNumber() and not self._GetNextHexNumber():
raise Warning("expected address", self.FileName, self.CurrentLineNumber)
- BsAddress = long(self.__Token, 0)
+ BsAddress = long(self._Token, 0)
Obj.BsBaseAddress = BsAddress
- if self.__IsKeyword("RtBaseAddress"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("RtBaseAddress"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextDecimalNumber() and not self.__GetNextHexNumber():
+ if not self._GetNextDecimalNumber() and not self._GetNextHexNumber():
raise Warning("expected address", self.FileName, self.CurrentLineNumber)
- RtAddress = long(self.__Token, 0)
+ RtAddress = long(self._Token, 0)
Obj.RtBaseAddress = RtAddress
- ## __GetBlockStatements() method
+ ## _GetBlockStatements() method
#
# Get block statements
#
# @param self The object pointer
# @param Obj for whom block statement is got
#
- def __GetBlockStatements(self, Obj):
+ def _GetBlockStatements(self, Obj):
IsBlock = False
- while self.__GetBlockStatement(Obj):
+ while self._GetBlockStatement(Obj):
IsBlock = True
Item = Obj.BlockSizeList[-1]
@@ -1683,7 +1629,7 @@ class FdfParser:
raise Warning("expected block statement", self.FileName, self.CurrentLineNumber)
return IsBlock
- ## __GetBlockStatement() method
+ ## _GetBlockStatement() method
#
# Get block statement
#
@@ -1692,20 +1638,20 @@ class FdfParser:
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetBlockStatement(self, Obj):
- if not self.__IsKeyword( "BlockSize"):
+ def _GetBlockStatement(self, Obj):
+ if not self._IsKeyword("BlockSize"):
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber() and not self.__GetNextDecimalNumber():
+ if not self._GetNextHexNumber() and not self._GetNextDecimalNumber():
raise Warning("expected Hex or Integer block size", self.FileName, self.CurrentLineNumber)
- BlockSize = self.__Token
+ BlockSize = self._Token
BlockSizePcd = None
- if self.__IsToken( "|"):
- PcdPair = self.__GetNextPcdSettings()
+ if self._IsToken(TAB_VALUE_SPLIT):
+ PcdPair = self._GetNextPcdSettings()
BlockSizePcd = PcdPair
self.Profile.PcdDict[PcdPair] = BlockSize
self.SetPcdLocalation(PcdPair)
@@ -1714,19 +1660,19 @@ class FdfParser:
BlockSize = long(BlockSize, 0)
BlockNumber = None
- if self.__IsKeyword( "NumBlocks"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("NumBlocks"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextDecimalNumber() and not self.__GetNextHexNumber():
+ if not self._GetNextDecimalNumber() and not self._GetNextHexNumber():
raise Warning("expected block numbers", self.FileName, self.CurrentLineNumber)
- BlockNumber = long(self.__Token, 0)
+ BlockNumber = long(self._Token, 0)
Obj.BlockSizeList.append((BlockSize, BlockNumber, BlockSizePcd))
return True
- ## __GetDefineStatements() method
+ ## _GetDefineStatements() method
#
# Get define statements
#
@@ -1735,11 +1681,11 @@ class FdfParser:
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetDefineStatements(self, Obj):
- while self.__GetDefineStatement( Obj):
+ def _GetDefineStatements(self, Obj):
+ while self._GetDefineStatement(Obj):
pass
- ## __GetDefineStatement() method
+ ## _GetDefineStatement() method
#
# Get define statement
#
@@ -1748,24 +1694,24 @@ class FdfParser:
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetDefineStatement(self, Obj):
- if self.__IsKeyword("DEFINE"):
- self.__GetNextToken()
- Macro = self.__Token
- if not self.__IsToken( "="):
+ def _GetDefineStatement(self, Obj):
+ if self._IsKeyword(TAB_DEFINE):
+ self._GetNextToken()
+ Macro = self._Token
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected value", self.FileName, self.CurrentLineNumber)
- Value = self.__Token
+ Value = self._Token
Macro = '$(' + Macro + ')'
Obj.DefineVarDict[Macro] = Value
return True
return False
- ## __GetSetStatements() method
+ ## _GetSetStatements() method
#
# Get set statements
#
@@ -1774,11 +1720,11 @@ class FdfParser:
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetSetStatements(self, Obj):
- while self.__GetSetStatement(Obj):
+ def _GetSetStatements(self, Obj):
+ while self._GetSetStatement(Obj):
pass
- ## __GetSetStatement() method
+ ## _GetSetStatement() method
#
# Get set statement
#
@@ -1787,15 +1733,15 @@ class FdfParser:
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetSetStatement(self, Obj):
- if self.__IsKeyword("SET"):
- PcdPair = self.__GetNextPcdSettings()
+ def _GetSetStatement(self, Obj):
+ if self._IsKeyword("SET"):
+ PcdPair = self._GetNextPcdSettings()
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- Value = self.__GetExpression()
- Value = self.__EvaluateConditional(Value, self.CurrentLineNumber, 'eval', True)
+ Value = self._GetExpression()
+ Value = self._EvaluateConditional(Value, self.CurrentLineNumber, 'eval', True)
if Obj:
Obj.SetVarDict[PcdPair] = Value
@@ -1807,19 +1753,19 @@ class FdfParser:
return False
- ## __CalcRegionExpr(self)
+ ## _CalcRegionExpr(self)
#
# Calculate expression for offset or size of a region
#
# @return: None if invalid expression
# Calculated number if successfully
#
- def __CalcRegionExpr(self):
+ def _CalcRegionExpr(self):
StartPos = self.GetFileBufferPos()
Expr = ''
PairCount = 0
- while not self.__EndOfFile():
- CurCh = self.__CurrentChar()
+ while not self._EndOfFile():
+ CurCh = self._CurrentChar()
if CurCh == '(':
PairCount += 1
elif CurCh == ')':
@@ -1828,106 +1774,106 @@ class FdfParser:
if CurCh in '|\r\n' and PairCount == 0:
break
Expr += CurCh
- self.__GetOneChar()
+ self._GetOneChar()
try:
return long(
ValueExpression(Expr,
- self.__CollectMacroPcd()
+ self._CollectMacroPcd()
)(True), 0)
except Exception:
self.SetFileBufferPos(StartPos)
return None
- ## __GetRegionLayout() method
+ ## _GetRegionLayout() method
#
# Get region layout for FD
#
# @param self The object pointer
- # @param Fd for whom region is got
+ # @param theFd for whom region is got
# @retval True Successfully find
# @retval False Not able to find
#
- def __GetRegionLayout(self, Fd):
- Offset = self.__CalcRegionExpr()
+ def _GetRegionLayout(self, theFd):
+ Offset = self._CalcRegionExpr()
if Offset is None:
return False
- RegionObj = Region.Region()
+ RegionObj = Region()
RegionObj.Offset = Offset
- Fd.RegionList.append(RegionObj)
+ theFd.RegionList.append(RegionObj)
- if not self.__IsToken( "|"):
+ if not self._IsToken(TAB_VALUE_SPLIT):
raise Warning("expected '|'", self.FileName, self.CurrentLineNumber)
- Size = self.__CalcRegionExpr()
+ Size = self._CalcRegionExpr()
if Size is None:
raise Warning("expected Region Size", self.FileName, self.CurrentLineNumber)
RegionObj.Size = Size
- if not self.__GetNextWord():
+ if not self._GetNextWord():
return True
- if not self.__Token in ("SET", BINARY_FILE_TYPE_FV, "FILE", "DATA", "CAPSULE", "INF"):
+ if not self._Token in ("SET", BINARY_FILE_TYPE_FV, "FILE", "DATA", "CAPSULE", "INF"):
#
# If next token is a word which is not a valid FV type, it might be part of [PcdOffset[|PcdSize]]
# Or it might be next region's offset described by an expression which starts with a PCD.
# PcdOffset[|PcdSize] or OffsetPcdExpression|Size
#
- self.__UndoToken()
- IsRegionPcd = (RegionSizeGuidPattern.match(self.__CurrentLine()[self.CurrentOffsetWithinLine:]) or
- RegionOffsetPcdPattern.match(self.__CurrentLine()[self.CurrentOffsetWithinLine:]))
+ self._UndoToken()
+ IsRegionPcd = (RegionSizeGuidPattern.match(self._CurrentLine()[self.CurrentOffsetWithinLine:]) or
+ RegionOffsetPcdPattern.match(self._CurrentLine()[self.CurrentOffsetWithinLine:]))
if IsRegionPcd:
- RegionObj.PcdOffset = self.__GetNextPcdSettings()
- self.Profile.PcdDict[RegionObj.PcdOffset] = "0x%08X" % (RegionObj.Offset + long(Fd.BaseAddress, 0))
+ RegionObj.PcdOffset = self._GetNextPcdSettings()
+ self.Profile.PcdDict[RegionObj.PcdOffset] = "0x%08X" % (RegionObj.Offset + long(theFd.BaseAddress, 0))
self.SetPcdLocalation(RegionObj.PcdOffset)
- self.__PcdDict['%s.%s' % (RegionObj.PcdOffset[1], RegionObj.PcdOffset[0])] = "0x%x" % RegionObj.Offset
+ self._PcdDict['%s.%s' % (RegionObj.PcdOffset[1], RegionObj.PcdOffset[0])] = "0x%x" % RegionObj.Offset
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
self.Profile.PcdFileLineDict[RegionObj.PcdOffset] = FileLineTuple
- if self.__IsToken( "|"):
- RegionObj.PcdSize = self.__GetNextPcdSettings()
+ if self._IsToken(TAB_VALUE_SPLIT):
+ RegionObj.PcdSize = self._GetNextPcdSettings()
self.Profile.PcdDict[RegionObj.PcdSize] = "0x%08X" % RegionObj.Size
self.SetPcdLocalation(RegionObj.PcdSize)
- self.__PcdDict['%s.%s' % (RegionObj.PcdSize[1], RegionObj.PcdSize[0])] = "0x%x" % RegionObj.Size
+ self._PcdDict['%s.%s' % (RegionObj.PcdSize[1], RegionObj.PcdSize[0])] = "0x%x" % RegionObj.Size
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
self.Profile.PcdFileLineDict[RegionObj.PcdSize] = FileLineTuple
- if not self.__GetNextWord():
+ if not self._GetNextWord():
return True
- if self.__Token == "SET":
- self.__UndoToken()
- self.__GetSetStatements( RegionObj)
- if not self.__GetNextWord():
+ if self._Token == "SET":
+ self._UndoToken()
+ self._GetSetStatements(RegionObj)
+ if not self._GetNextWord():
return True
- elif self.__Token == BINARY_FILE_TYPE_FV:
- self.__UndoToken()
- self.__GetRegionFvType( RegionObj)
+ elif self._Token == BINARY_FILE_TYPE_FV:
+ self._UndoToken()
+ self._GetRegionFvType(RegionObj)
- elif self.__Token == "CAPSULE":
- self.__UndoToken()
- self.__GetRegionCapType( RegionObj)
+ elif self._Token == "CAPSULE":
+ self._UndoToken()
+ self._GetRegionCapType(RegionObj)
- elif self.__Token == "FILE":
- self.__UndoToken()
- self.__GetRegionFileType(RegionObj)
+ elif self._Token == "FILE":
+ self._UndoToken()
+ self._GetRegionFileType(RegionObj)
- elif self.__Token == "INF":
- self.__UndoToken()
+ elif self._Token == "INF":
+ self._UndoToken()
RegionObj.RegionType = "INF"
- while self.__IsKeyword("INF"):
- self.__UndoToken()
- ffsInf = self.__ParseInfStatement()
+ while self._IsKeyword("INF"):
+ self._UndoToken()
+ ffsInf = self._ParseInfStatement()
if not ffsInf:
break
RegionObj.RegionDataList.append(ffsInf)
- elif self.__Token == "DATA":
- self.__UndoToken()
- self.__GetRegionDataType(RegionObj)
+ elif self._Token == "DATA":
+ self._UndoToken()
+ self._GetRegionDataType(RegionObj)
else:
- self.__UndoToken()
- if self.__GetRegionLayout(Fd):
+ self._UndoToken()
+ if self._GetRegionLayout(theFd):
return True
raise Warning("A valid region type was not found. "
"Valid types are [SET, FV, CAPSULE, FILE, DATA, INF]. This error occurred",
@@ -1935,189 +1881,185 @@ class FdfParser:
return True
- ## __GetRegionFvType() method
+ ## _GetRegionFvType() method
#
# Get region fv data for region
#
# @param self The object pointer
# @param RegionObj for whom region data is got
#
- def __GetRegionFvType(self, RegionObj):
-
- if not self.__IsKeyword( BINARY_FILE_TYPE_FV):
+ def _GetRegionFvType(self, RegionObj):
+ if not self._IsKeyword(BINARY_FILE_TYPE_FV):
raise Warning("expected Keyword BINARY_FILE_TYPE_FV", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV name", self.FileName, self.CurrentLineNumber)
RegionObj.RegionType = BINARY_FILE_TYPE_FV
- RegionObj.RegionDataList.append((self.__Token).upper())
+ RegionObj.RegionDataList.append((self._Token).upper())
- while self.__IsKeyword( BINARY_FILE_TYPE_FV):
+ while self._IsKeyword(BINARY_FILE_TYPE_FV):
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV name", self.FileName, self.CurrentLineNumber)
- RegionObj.RegionDataList.append((self.__Token).upper())
+ RegionObj.RegionDataList.append((self._Token).upper())
- ## __GetRegionCapType() method
+ ## _GetRegionCapType() method
#
# Get region capsule data for region
#
# @param self The object pointer
# @param RegionObj for whom region data is got
#
- def __GetRegionCapType(self, RegionObj):
-
- if not self.__IsKeyword("CAPSULE"):
+ def _GetRegionCapType(self, RegionObj):
+ if not self._IsKeyword("CAPSULE"):
raise Warning("expected Keyword 'CAPSULE'", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected CAPSULE name", self.FileName, self.CurrentLineNumber)
RegionObj.RegionType = "CAPSULE"
- RegionObj.RegionDataList.append(self.__Token)
+ RegionObj.RegionDataList.append(self._Token)
- while self.__IsKeyword("CAPSULE"):
+ while self._IsKeyword("CAPSULE"):
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected CAPSULE name", self.FileName, self.CurrentLineNumber)
- RegionObj.RegionDataList.append(self.__Token)
+ RegionObj.RegionDataList.append(self._Token)
- ## __GetRegionFileType() method
+ ## _GetRegionFileType() method
#
# Get region file data for region
#
# @param self The object pointer
# @param RegionObj for whom region data is got
#
- def __GetRegionFileType(self, RegionObj):
-
- if not self.__IsKeyword( "FILE"):
+ def _GetRegionFileType(self, RegionObj):
+ if not self._IsKeyword("FILE"):
raise Warning("expected Keyword 'FILE'", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected File name", self.FileName, self.CurrentLineNumber)
RegionObj.RegionType = "FILE"
- RegionObj.RegionDataList.append( self.__Token)
+ RegionObj.RegionDataList.append(self._Token)
- while self.__IsKeyword( "FILE"):
+ while self._IsKeyword("FILE"):
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FILE name", self.FileName, self.CurrentLineNumber)
- RegionObj.RegionDataList.append(self.__Token)
+ RegionObj.RegionDataList.append(self._Token)
- ## __GetRegionDataType() method
+ ## _GetRegionDataType() method
#
# Get region array data for region
#
# @param self The object pointer
# @param RegionObj for whom region data is got
#
- def __GetRegionDataType(self, RegionObj):
-
- if not self.__IsKeyword( "DATA"):
+ def _GetRegionDataType(self, RegionObj):
+ if not self._IsKeyword("DATA"):
raise Warning("expected Region Data type", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex byte", self.FileName, self.CurrentLineNumber)
- if len(self.__Token) > 18:
+ if len(self._Token) > 18:
raise Warning("Hex string can't be converted to a valid UINT64 value", self.FileName, self.CurrentLineNumber)
# convert hex string value to byte hex string array
- AllString = self.__Token
+ AllString = self._Token
AllStrLen = len (AllString)
DataString = ""
while AllStrLen > 4:
- DataString = DataString + "0x" + AllString[AllStrLen - 2: AllStrLen] + ","
+ DataString = DataString + "0x" + AllString[AllStrLen - 2: AllStrLen] + TAB_COMMA_SPLIT
AllStrLen = AllStrLen - 2
- DataString = DataString + AllString[:AllStrLen] + ","
+ DataString = DataString + AllString[:AllStrLen] + TAB_COMMA_SPLIT
# byte value array
- if len (self.__Token) <= 4:
- while self.__IsToken(","):
- if not self.__GetNextHexNumber():
+ if len (self._Token) <= 4:
+ while self._IsToken(TAB_COMMA_SPLIT):
+ if not self._GetNextHexNumber():
raise Warning("Invalid Hex number", self.FileName, self.CurrentLineNumber)
- if len(self.__Token) > 4:
+ if len(self._Token) > 4:
raise Warning("Hex byte(must be 2 digits) too long", self.FileName, self.CurrentLineNumber)
- DataString += self.__Token
- DataString += ","
+ DataString += self._Token
+ DataString += TAB_COMMA_SPLIT
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- DataString = DataString.rstrip(",")
+ DataString = DataString.rstrip(TAB_COMMA_SPLIT)
RegionObj.RegionType = "DATA"
- RegionObj.RegionDataList.append( DataString)
+ RegionObj.RegionDataList.append(DataString)
- while self.__IsKeyword( "DATA"):
+ while self._IsKeyword("DATA"):
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex byte", self.FileName, self.CurrentLineNumber)
- if len(self.__Token) > 18:
+ if len(self._Token) > 18:
raise Warning("Hex string can't be converted to a valid UINT64 value", self.FileName, self.CurrentLineNumber)
# convert hex string value to byte hex string array
- AllString = self.__Token
+ AllString = self._Token
AllStrLen = len (AllString)
DataString = ""
while AllStrLen > 4:
- DataString = DataString + "0x" + AllString[AllStrLen - 2: AllStrLen] + ","
+ DataString = DataString + "0x" + AllString[AllStrLen - 2: AllStrLen] + TAB_COMMA_SPLIT
AllStrLen = AllStrLen - 2
- DataString = DataString + AllString[:AllStrLen] + ","
+ DataString = DataString + AllString[:AllStrLen] + TAB_COMMA_SPLIT
# byte value array
- if len (self.__Token) <= 4:
- while self.__IsToken(","):
- if not self.__GetNextHexNumber():
+ if len (self._Token) <= 4:
+ while self._IsToken(TAB_COMMA_SPLIT):
+ if not self._GetNextHexNumber():
raise Warning("Invalid Hex number", self.FileName, self.CurrentLineNumber)
- if len(self.__Token) > 4:
+ if len(self._Token) > 4:
raise Warning("Hex byte(must be 2 digits) too long", self.FileName, self.CurrentLineNumber)
- DataString += self.__Token
- DataString += ","
+ DataString += self._Token
+ DataString += TAB_COMMA_SPLIT
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- DataString = DataString.rstrip(",")
- RegionObj.RegionDataList.append( DataString)
+ DataString = DataString.rstrip(TAB_COMMA_SPLIT)
+ RegionObj.RegionDataList.append(DataString)
- ## __GetFv() method
+ ## _GetFv() method
#
# Get FV section contents and store its data into FV dictionary of self.Profile
#
@@ -2125,68 +2067,68 @@ class FdfParser:
# @retval True Successfully find a FV
# @retval False Not able to find a FV
#
- def __GetFv(self):
- if not self.__GetNextToken():
+ def _GetFv(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[FV."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[FV."):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[FV.", True):
+ self._UndoToken()
+ if not self._IsToken("[FV.", True):
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
#print 'Parsing String: %s in File %s, At line: %d, Offset Within Line: %d' \
- # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
- raise Warning("Unknown Keyword '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
+ raise Warning("Unknown Keyword '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
- FvName = self.__GetUiName()
+ FvName = self._GetUiName()
self.CurrentFvName = FvName.upper()
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- FvObj = Fv.FV()
+ FvObj = FV()
FvObj.UiFvName = self.CurrentFvName
self.Profile.FvDict[self.CurrentFvName] = FvObj
- Status = self.__GetCreateFile(FvObj)
+ Status = self._GetCreateFile(FvObj)
if not Status:
raise Warning("FV name error", self.FileName, self.CurrentLineNumber)
- self.__GetDefineStatements(FvObj)
+ self._GetDefineStatements(FvObj)
- self.__GetAddressStatements(FvObj)
+ self._GetAddressStatements(FvObj)
FvObj.FvExtEntryTypeValue = []
FvObj.FvExtEntryType = []
FvObj.FvExtEntryData = []
while True:
- self.__GetSetStatements(FvObj)
+ self._GetSetStatements(FvObj)
- if not (self.__GetBlockStatement(FvObj) or self.__GetFvBaseAddress(FvObj) or
- self.__GetFvForceRebase(FvObj) or self.__GetFvAlignment(FvObj) or
- self.__GetFvAttributes(FvObj) or self.__GetFvNameGuid(FvObj) or
- self.__GetFvExtEntryStatement(FvObj) or self.__GetFvNameString(FvObj)):
+ if not (self._GetBlockStatement(FvObj) or self._GetFvBaseAddress(FvObj) or
+ self._GetFvForceRebase(FvObj) or self._GetFvAlignment(FvObj) or
+ self._GetFvAttributes(FvObj) or self._GetFvNameGuid(FvObj) or
+ self._GetFvExtEntryStatement(FvObj) or self._GetFvNameString(FvObj)):
break
if FvObj.FvNameString == 'TRUE' and not FvObj.FvNameGuid:
raise Warning("FvNameString found but FvNameGuid was not found", self.FileName, self.CurrentLineNumber)
- self.__GetAprioriSection(FvObj, FvObj.DefineVarDict.copy())
- self.__GetAprioriSection(FvObj, FvObj.DefineVarDict.copy())
+ self._GetAprioriSection(FvObj, FvObj.DefineVarDict.copy())
+ self._GetAprioriSection(FvObj, FvObj.DefineVarDict.copy())
while True:
- isInf = self.__GetInfStatement(FvObj)
- isFile = self.__GetFileStatement(FvObj, MacroDict = FvObj.DefineVarDict.copy())
+ isInf = self._GetInfStatement(FvObj)
+ isFile = self._GetFileStatement(FvObj, MacroDict = FvObj.DefineVarDict.copy())
if not isInf and not isFile:
break
return True
- ## __GetFvAlignment() method
+ ## _GetFvAlignment() method
#
# Get alignment for FV
#
@@ -2195,26 +2137,25 @@ class FdfParser:
# @retval True Successfully find a alignment statement
# @retval False Not able to find a alignment statement
#
- def __GetFvAlignment(self, Obj):
-
- if not self.__IsKeyword( "FvAlignment"):
+ def _GetFvAlignment(self, Obj):
+ if not self._IsKeyword("FvAlignment"):
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected alignment value", self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() not in ("1", "2", "4", "8", "16", "32", "64", "128", "256", "512", \
+ if self._Token.upper() not in ("1", "2", "4", "8", "16", "32", "64", "128", "256", "512", \
"1K", "2K", "4K", "8K", "16K", "32K", "64K", "128K", "256K", "512K", \
"1M", "2M", "4M", "8M", "16M", "32M", "64M", "128M", "256M", "512M", \
"1G", "2G"):
- raise Warning("Unknown alignment value '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- Obj.FvAlignment = self.__Token
+ raise Warning("Unknown alignment value '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ Obj.FvAlignment = self._Token
return True
- ## __GetFvBaseAddress() method
+ ## _GetFvBaseAddress() method
#
# Get BaseAddress for FV
#
@@ -2223,23 +2164,22 @@ class FdfParser:
# @retval True Successfully find a FvBaseAddress statement
# @retval False Not able to find a FvBaseAddress statement
#
- def __GetFvBaseAddress(self, Obj):
-
- if not self.__IsKeyword("FvBaseAddress"):
+ def _GetFvBaseAddress(self, Obj):
+ if not self._IsKeyword("FvBaseAddress"):
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV base address value", self.FileName, self.CurrentLineNumber)
- if not BaseAddrValuePattern.match(self.__Token.upper()):
- raise Warning("Unknown FV base address value '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- Obj.FvBaseAddress = self.__Token
+ if not BaseAddrValuePattern.match(self._Token.upper()):
+ raise Warning("Unknown FV base address value '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ Obj.FvBaseAddress = self._Token
return True
- ## __GetFvForceRebase() method
+ ## _GetFvForceRebase() method
#
# Get FvForceRebase for FV
#
@@ -2248,23 +2188,22 @@ class FdfParser:
# @retval True Successfully find a FvForceRebase statement
# @retval False Not able to find a FvForceRebase statement
#
- def __GetFvForceRebase(self, Obj):
-
- if not self.__IsKeyword("FvForceRebase"):
+ def _GetFvForceRebase(self, Obj):
+ if not self._IsKeyword("FvForceRebase"):
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FvForceRebase value", self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() not in ["TRUE", "FALSE", "0", "0X0", "0X00", "1", "0X1", "0X01"]:
- raise Warning("Unknown FvForceRebase value '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ if self._Token.upper() not in ["TRUE", "FALSE", "0", "0X0", "0X00", "1", "0X1", "0X01"]:
+ raise Warning("Unknown FvForceRebase value '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() in ["TRUE", "1", "0X1", "0X01"]:
+ if self._Token.upper() in ["TRUE", "1", "0X1", "0X01"]:
Obj.FvForceRebase = True
- elif self.__Token.upper() in ["FALSE", "0", "0X0", "0X00"]:
+ elif self._Token.upper() in ["FALSE", "0", "0X0", "0X00"]:
Obj.FvForceRebase = False
else:
Obj.FvForceRebase = None
@@ -2272,7 +2211,7 @@ class FdfParser:
return True
- ## __GetFvAttributes() method
+ ## _GetFvAttributes() method
#
# Get attributes for FV
#
@@ -2280,31 +2219,31 @@ class FdfParser:
# @param Obj for whom attribute is got
# @retval None
#
- def __GetFvAttributes(self, FvObj):
+ def _GetFvAttributes(self, FvObj):
IsWordToken = False
- while self.__GetNextWord():
+ while self._GetNextWord():
IsWordToken = True
- name = self.__Token
+ name = self._Token
if name not in ("ERASE_POLARITY", "MEMORY_MAPPED", \
"STICKY_WRITE", "LOCK_CAP", "LOCK_STATUS", "WRITE_ENABLED_CAP", \
"WRITE_DISABLED_CAP", "WRITE_STATUS", "READ_ENABLED_CAP", \
"READ_DISABLED_CAP", "READ_STATUS", "READ_LOCK_CAP", \
"READ_LOCK_STATUS", "WRITE_LOCK_CAP", "WRITE_LOCK_STATUS", \
"WRITE_POLICY_RELIABLE", "WEAK_ALIGNMENT", "FvUsedSizeEnable"):
- self.__UndoToken()
+ self._UndoToken()
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken() or self.__Token.upper() not in ("TRUE", "FALSE", "1", "0"):
+ if not self._GetNextToken() or self._Token.upper() not in ("TRUE", "FALSE", "1", "0"):
raise Warning("expected TRUE/FALSE (1/0)", self.FileName, self.CurrentLineNumber)
- FvObj.FvAttributeDict[name] = self.__Token
+ FvObj.FvAttributeDict[name] = self._Token
return IsWordToken
- ## __GetFvNameGuid() method
+ ## _GetFvNameGuid() method
#
# Get FV GUID for FV
#
@@ -2312,110 +2251,107 @@ class FdfParser:
# @param Obj for whom GUID is got
# @retval None
#
- def __GetFvNameGuid(self, FvObj):
-
- if not self.__IsKeyword( "FvNameGuid"):
+ def _GetFvNameGuid(self, FvObj):
+ if not self._IsKeyword("FvNameGuid"):
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextGuid():
+ if not self._GetNextGuid():
raise Warning("expected FV GUID value", self.FileName, self.CurrentLineNumber)
- FvObj.FvNameGuid = self.__Token
+ FvObj.FvNameGuid = self._Token
return True
- def __GetFvNameString(self, FvObj):
-
- if not self.__IsKeyword( "FvNameString"):
+ def _GetFvNameString(self, FvObj):
+ if not self._IsKeyword("FvNameString"):
return False
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken() or self.__Token not in ('TRUE', 'FALSE'):
+ if not self._GetNextToken() or self._Token not in ('TRUE', 'FALSE'):
raise Warning("expected TRUE or FALSE for FvNameString", self.FileName, self.CurrentLineNumber)
- FvObj.FvNameString = self.__Token
+ FvObj.FvNameString = self._Token
return True
- def __GetFvExtEntryStatement(self, FvObj):
-
- if not (self.__IsKeyword( "FV_EXT_ENTRY") or self.__IsKeyword( "FV_EXT_ENTRY_TYPE")):
+ def _GetFvExtEntryStatement(self, FvObj):
+ if not (self._IsKeyword("FV_EXT_ENTRY") or self._IsKeyword("FV_EXT_ENTRY_TYPE")):
return False
- if not self.__IsKeyword ("TYPE"):
+ if not self._IsKeyword ("TYPE"):
raise Warning("expected 'TYPE'", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber() and not self.__GetNextDecimalNumber():
+ if not self._GetNextHexNumber() and not self._GetNextDecimalNumber():
raise Warning("expected Hex FV extension entry type value At Line ", self.FileName, self.CurrentLineNumber)
- FvObj.FvExtEntryTypeValue.append(self.__Token)
+ FvObj.FvExtEntryTypeValue.append(self._Token)
- if not self.__IsToken( "{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- if not self.__IsKeyword ("FILE") and not self.__IsKeyword ("DATA"):
+ if not self._IsKeyword ("FILE") and not self._IsKeyword ("DATA"):
raise Warning("expected 'FILE' or 'DATA'", self.FileName, self.CurrentLineNumber)
- FvObj.FvExtEntryType.append(self.__Token)
+ FvObj.FvExtEntryType.append(self._Token)
- if self.__Token == 'DATA':
+ if self._Token == 'DATA':
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex byte", self.FileName, self.CurrentLineNumber)
- if len(self.__Token) > 4:
+ if len(self._Token) > 4:
raise Warning("Hex byte(must be 2 digits) too long", self.FileName, self.CurrentLineNumber)
- DataString = self.__Token
- DataString += ","
+ DataString = self._Token
+ DataString += TAB_COMMA_SPLIT
- while self.__IsToken(","):
- if not self.__GetNextHexNumber():
+ while self._IsToken(TAB_COMMA_SPLIT):
+ if not self._GetNextHexNumber():
raise Warning("Invalid Hex number", self.FileName, self.CurrentLineNumber)
- if len(self.__Token) > 4:
+ if len(self._Token) > 4:
raise Warning("Hex byte(must be 2 digits) too long", self.FileName, self.CurrentLineNumber)
- DataString += self.__Token
- DataString += ","
+ DataString += self._Token
+ DataString += TAB_COMMA_SPLIT
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- DataString = DataString.rstrip(",")
+ DataString = DataString.rstrip(TAB_COMMA_SPLIT)
FvObj.FvExtEntryData.append(DataString)
- if self.__Token == 'FILE':
+ if self._Token == 'FILE':
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV Extension Entry file path At Line ", self.FileName, self.CurrentLineNumber)
- FvObj.FvExtEntryData.append(self.__Token)
+ FvObj.FvExtEntryData.append(self._Token)
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
return True
- ## __GetAprioriSection() method
+ ## _GetAprioriSection() method
#
# Get token statements
#
@@ -2425,56 +2361,55 @@ class FdfParser:
# @retval True Successfully find apriori statement
# @retval False Not able to find apriori statement
#
- def __GetAprioriSection(self, FvObj, MacroDict = {}):
-
- if not self.__IsKeyword( "APRIORI"):
+ def _GetAprioriSection(self, FvObj, MacroDict = {}):
+ if not self._IsKeyword("APRIORI"):
return False
- if not self.__IsKeyword("PEI") and not self.__IsKeyword("DXE"):
+ if not self._IsKeyword("PEI") and not self._IsKeyword("DXE"):
raise Warning("expected Apriori file type", self.FileName, self.CurrentLineNumber)
- AprType = self.__Token
+ AprType = self._Token
- if not self.__IsToken( "{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- AprSectionObj = AprioriSection.AprioriSection()
+ AprSectionObj = AprioriSection()
AprSectionObj.AprioriType = AprType
- self.__GetDefineStatements(AprSectionObj)
+ self._GetDefineStatements(AprSectionObj)
MacroDict.update(AprSectionObj.DefineVarDict)
while True:
- IsInf = self.__GetInfStatement(AprSectionObj)
- IsFile = self.__GetFileStatement( AprSectionObj)
+ IsInf = self._GetInfStatement(AprSectionObj)
+ IsFile = self._GetFileStatement(AprSectionObj)
if not IsInf and not IsFile:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
FvObj.AprioriSectionList.append(AprSectionObj)
return True
- def __ParseInfStatement(self):
- if not self.__IsKeyword("INF"):
+ def _ParseInfStatement(self):
+ if not self._IsKeyword("INF"):
return None
- ffsInf = FfsInfStatement.FfsInfStatement()
- self.__GetInfOptions(ffsInf)
+ ffsInf = FfsInfStatement()
+ self._GetInfOptions(ffsInf)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected INF file path", self.FileName, self.CurrentLineNumber)
- ffsInf.InfFileName = self.__Token
+ ffsInf.InfFileName = self._Token
if not ffsInf.InfFileName.endswith('.inf'):
raise Warning("expected .inf file path", self.FileName, self.CurrentLineNumber)
ffsInf.CurrentLineNum = self.CurrentLineNumber
- ffsInf.CurrentLineContent = self.__CurrentLine()
+ ffsInf.CurrentLineContent = self._CurrentLine()
#Replace $(SAPCE) with real space
ffsInf.InfFileName = ffsInf.InfFileName.replace('$(SPACE)', ' ')
- if ffsInf.InfFileName.replace('$(WORKSPACE)', '').find('$') == -1:
+ if ffsInf.InfFileName.replace(TAB_WORKSPACE, '').find('$') == -1:
#do case sensitive check for file path
ErrorCode, ErrorInfo = PathClass(NormPath(ffsInf.InfFileName), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
@@ -2492,16 +2427,16 @@ class FdfParser:
else:
self.Profile.InfDict['ArchTBD'].append(ffsInf.InfFileName)
- if self.__IsToken('|'):
- if self.__IsKeyword('RELOCS_STRIPPED'):
+ if self._IsToken(TAB_VALUE_SPLIT):
+ if self._IsKeyword('RELOCS_STRIPPED'):
ffsInf.KeepReloc = False
- elif self.__IsKeyword('RELOCS_RETAINED'):
+ elif self._IsKeyword('RELOCS_RETAINED'):
ffsInf.KeepReloc = True
else:
- raise Warning("Unknown reloc strip flag '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ raise Warning("Unknown reloc strip flag '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
return ffsInf
- ## __GetInfStatement() method
+ ## _GetInfStatement() method
#
# Get INF statements
#
@@ -2510,86 +2445,86 @@ class FdfParser:
# @retval True Successfully find inf statement
# @retval False Not able to find inf statement
#
- def __GetInfStatement(self, Obj, ForCapsule=False):
- ffsInf = self.__ParseInfStatement()
+ def _GetInfStatement(self, Obj, ForCapsule=False):
+ ffsInf = self._ParseInfStatement()
if not ffsInf:
return False
if ForCapsule:
- capsuleFfs = CapsuleData.CapsuleFfs()
+ capsuleFfs = CapsuleFfs()
capsuleFfs.Ffs = ffsInf
Obj.CapsuleDataList.append(capsuleFfs)
else:
Obj.FfsList.append(ffsInf)
return True
- ## __GetInfOptions() method
+ ## _GetInfOptions() method
#
# Get options for INF
#
# @param self The object pointer
# @param FfsInfObj for whom option is got
#
- def __GetInfOptions(self, FfsInfObj):
- if self.__IsKeyword("FILE_GUID"):
- if not self.__IsToken("="):
+ def _GetInfOptions(self, FfsInfObj):
+ if self._IsKeyword("FILE_GUID"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextGuid():
+ if not self._GetNextGuid():
raise Warning("expected GUID value", self.FileName, self.CurrentLineNumber)
- FfsInfObj.OverrideGuid = self.__Token
+ FfsInfObj.OverrideGuid = self._Token
- if self.__IsKeyword( "RuleOverride"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("RuleOverride"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Rule name", self.FileName, self.CurrentLineNumber)
- FfsInfObj.Rule = self.__Token
+ FfsInfObj.Rule = self._Token
- if self.__IsKeyword( "VERSION"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("VERSION"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Version", self.FileName, self.CurrentLineNumber)
- if self.__GetStringData():
- FfsInfObj.Version = self.__Token
+ if self._GetStringData():
+ FfsInfObj.Version = self._Token
- if self.__IsKeyword( BINARY_FILE_TYPE_UI):
- if not self.__IsToken( "="):
+ if self._IsKeyword(BINARY_FILE_TYPE_UI):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected UI name", self.FileName, self.CurrentLineNumber)
- if self.__GetStringData():
- FfsInfObj.Ui = self.__Token
+ if self._GetStringData():
+ FfsInfObj.Ui = self._Token
- if self.__IsKeyword( "USE"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("USE"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected ARCH name", self.FileName, self.CurrentLineNumber)
- FfsInfObj.UseArch = self.__Token
+ FfsInfObj.UseArch = self._Token
- if self.__GetNextToken():
- p = re.compile(r'([a-zA-Z0-9\-]+|\$\(TARGET\)|\*)_([a-zA-Z0-9\-]+|\$\(TOOL_CHAIN_TAG\)|\*)_([a-zA-Z0-9\-]+|\$\(ARCH\))')
- if p.match(self.__Token) and p.match(self.__Token).span()[1] == len(self.__Token):
- FfsInfObj.KeyStringList.append(self.__Token)
- if not self.__IsToken(","):
+ if self._GetNextToken():
+ p = compile(r'([a-zA-Z0-9\-]+|\$\(TARGET\)|\*)_([a-zA-Z0-9\-]+|\$\(TOOL_CHAIN_TAG\)|\*)_([a-zA-Z0-9\-]+|\$\(ARCH\))')
+ if p.match(self._Token) and p.match(self._Token).span()[1] == len(self._Token):
+ FfsInfObj.KeyStringList.append(self._Token)
+ if not self._IsToken(TAB_COMMA_SPLIT):
return
else:
- self.__UndoToken()
+ self._UndoToken()
return
- while self.__GetNextToken():
- if not p.match(self.__Token):
+ while self._GetNextToken():
+ if not p.match(self._Token):
raise Warning("expected KeyString \"Target_Tag_Arch\"", self.FileName, self.CurrentLineNumber)
- FfsInfObj.KeyStringList.append(self.__Token)
+ FfsInfObj.KeyStringList.append(self._Token)
- if not self.__IsToken(","):
+ if not self._IsToken(TAB_COMMA_SPLIT):
break
- ## __GetFileStatement() method
+ ## _GetFileStatement() method
#
# Get FILE statements
#
@@ -2599,42 +2534,41 @@ class FdfParser:
# @retval True Successfully find FILE statement
# @retval False Not able to find FILE statement
#
- def __GetFileStatement(self, Obj, ForCapsule = False, MacroDict = {}):
-
- if not self.__IsKeyword( "FILE"):
+ def _GetFileStatement(self, Obj, ForCapsule = False, MacroDict = {}):
+ if not self._IsKeyword("FILE"):
return False
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected FFS type", self.FileName, self.CurrentLineNumber)
- if ForCapsule and self.__Token == 'DATA':
- self.__UndoToken()
- self.__UndoToken()
+ if ForCapsule and self._Token == 'DATA':
+ self._UndoToken()
+ self._UndoToken()
return False
- FfsFileObj = FfsFileStatement.FileStatement()
- FfsFileObj.FvFileType = self.__Token
+ FfsFileObj = FileStatement()
+ FfsFileObj.FvFileType = self._Token
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextGuid():
- if not self.__GetNextWord():
+ if not self._GetNextGuid():
+ if not self._GetNextWord():
raise Warning("expected File GUID", self.FileName, self.CurrentLineNumber)
- if self.__Token == 'PCD':
- if not self.__IsToken( "("):
+ if self._Token == 'PCD':
+ if not self._IsToken("("):
raise Warning("expected '('", self.FileName, self.CurrentLineNumber)
- PcdPair = self.__GetNextPcdSettings()
- if not self.__IsToken( ")"):
+ PcdPair = self._GetNextPcdSettings()
+ if not self._IsToken(")"):
raise Warning("expected ')'", self.FileName, self.CurrentLineNumber)
- self.__Token = 'PCD('+PcdPair[1]+'.'+PcdPair[0]+')'
+ self._Token = 'PCD('+PcdPair[1]+TAB_SPLIT+PcdPair[0]+')'
- FfsFileObj.NameGuid = self.__Token
+ FfsFileObj.NameGuid = self._Token
- self.__GetFilePart( FfsFileObj, MacroDict.copy())
+ self._GetFilePart(FfsFileObj, MacroDict.copy())
if ForCapsule:
- capsuleFfs = CapsuleData.CapsuleFfs()
+ capsuleFfs = CapsuleFfs()
capsuleFfs.Ffs = FfsFileObj
Obj.CapsuleDataList.append(capsuleFfs)
else:
@@ -2642,7 +2576,7 @@ class FdfParser:
return True
- ## __FileCouldHaveRelocFlag() method
+ ## _FileCouldHaveRelocFlag() method
#
# Check whether reloc strip flag can be set for a file type.
#
@@ -2651,13 +2585,13 @@ class FdfParser:
# @retval False No way to have it
#
@staticmethod
- def __FileCouldHaveRelocFlag (FileType):
+ def _FileCouldHaveRelocFlag (FileType):
if FileType in (SUP_MODULE_SEC, SUP_MODULE_PEI_CORE, SUP_MODULE_PEIM, 'PEI_DXE_COMBO'):
return True
else:
return False
- ## __SectionCouldHaveRelocFlag() method
+ ## _SectionCouldHaveRelocFlag() method
#
# Check whether reloc strip flag can be set for a section type.
#
@@ -2666,13 +2600,13 @@ class FdfParser:
# @retval False No way to have it
#
@staticmethod
- def __SectionCouldHaveRelocFlag (SectionType):
+ def _SectionCouldHaveRelocFlag (SectionType):
if SectionType in (BINARY_FILE_TYPE_TE, BINARY_FILE_TYPE_PE32):
return True
else:
return False
- ## __GetFilePart() method
+ ## _GetFilePart() method
#
# Get components for FILE statement
#
@@ -2680,58 +2614,57 @@ class FdfParser:
# @param FfsFileObj for whom component is got
# @param MacroDict dictionary used to replace macro
#
- def __GetFilePart(self, FfsFileObj, MacroDict = {}):
+ def _GetFilePart(self, FfsFileObj, MacroDict = {}):
+ self._GetFileOpts(FfsFileObj)
- self.__GetFileOpts( FfsFileObj)
-
- if not self.__IsToken("{"):
- if self.__IsKeyword('RELOCS_STRIPPED') or self.__IsKeyword('RELOCS_RETAINED'):
- if self.__FileCouldHaveRelocFlag(FfsFileObj.FvFileType):
- if self.__Token == 'RELOCS_STRIPPED':
+ if not self._IsToken("{"):
+ if self._IsKeyword('RELOCS_STRIPPED') or self._IsKeyword('RELOCS_RETAINED'):
+ if self._FileCouldHaveRelocFlag(FfsFileObj.FvFileType):
+ if self._Token == 'RELOCS_STRIPPED':
FfsFileObj.KeepReloc = False
else:
FfsFileObj.KeepReloc = True
else:
raise Warning("File type %s could not have reloc strip flag%d" % (FfsFileObj.FvFileType, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected File name or section data", self.FileName, self.CurrentLineNumber)
- if self.__Token == BINARY_FILE_TYPE_FV:
- if not self.__IsToken( "="):
+ if self._Token == BINARY_FILE_TYPE_FV:
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV name", self.FileName, self.CurrentLineNumber)
- FfsFileObj.FvName = self.__Token
+ FfsFileObj.FvName = self._Token
- elif self.__Token == "FD":
- if not self.__IsToken( "="):
+ elif self._Token == "FD":
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FD name", self.FileName, self.CurrentLineNumber)
- FfsFileObj.FdName = self.__Token
+ FfsFileObj.FdName = self._Token
- elif self.__Token in ("DEFINE", "APRIORI", "SECTION"):
- self.__UndoToken()
- self.__GetSectionData( FfsFileObj, MacroDict)
+ elif self._Token in (TAB_DEFINE, "APRIORI", "SECTION"):
+ self._UndoToken()
+ self._GetSectionData(FfsFileObj, MacroDict)
elif hasattr(FfsFileObj, 'FvFileType') and FfsFileObj.FvFileType == 'RAW':
- self.__UndoToken()
- self.__GetRAWData(FfsFileObj, MacroDict)
+ self._UndoToken()
+ self._GetRAWData(FfsFileObj, MacroDict)
else:
FfsFileObj.CurrentLineNum = self.CurrentLineNumber
- FfsFileObj.CurrentLineContent = self.__CurrentLine()
- FfsFileObj.FileName = self.__Token.replace('$(SPACE)', ' ')
- self.__VerifyFile(FfsFileObj.FileName)
+ FfsFileObj.CurrentLineContent = self._CurrentLine()
+ FfsFileObj.FileName = self._Token.replace('$(SPACE)', ' ')
+ self._VerifyFile(FfsFileObj.FileName)
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- ## __GetRAWData() method
+ ## _GetRAWData() method
#
# Get RAW data for FILE statement
#
@@ -2739,33 +2672,32 @@ class FdfParser:
# @param FfsFileObj for whom section is got
# @param MacroDict dictionary used to replace macro
#
- def __GetRAWData(self, FfsFileObj, MacroDict = {}):
+ def _GetRAWData(self, FfsFileObj, MacroDict = {}):
FfsFileObj.FileName = []
FfsFileObj.SubAlignment = []
while True:
AlignValue = None
- if self.__GetAlignment():
- if self.__Token not in ("Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENTS:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
#For FFS, Auto is default option same to ""
- if not self.__Token == "Auto":
- AlignValue = self.__Token
- if not self.__GetNextToken():
+ if not self._Token == "Auto":
+ AlignValue = self._Token
+ if not self._GetNextToken():
raise Warning("expected Filename value", self.FileName, self.CurrentLineNumber)
- FileName = self.__Token.replace('$(SPACE)', ' ')
+ FileName = self._Token.replace('$(SPACE)', ' ')
if FileName == '}':
- self.__UndoToken()
+ self._UndoToken()
raise Warning("expected Filename value", self.FileName, self.CurrentLineNumber)
- self.__VerifyFile(FileName)
+ self._VerifyFile(FileName)
File = PathClass(NormPath(FileName), GenFdsGlobalVariable.WorkSpaceDir)
FfsFileObj.FileName.append(File.Path)
FfsFileObj.SubAlignment.append(AlignValue)
- if self.__IsToken( "}"):
- self.__UndoToken()
+ if self._IsToken("}"):
+ self._UndoToken()
break
if len(FfsFileObj.SubAlignment) == 1:
@@ -2773,45 +2705,43 @@ class FdfParser:
if len(FfsFileObj.FileName) == 1:
FfsFileObj.FileName = FfsFileObj.FileName[0]
- ## __GetFileOpts() method
+ ## _GetFileOpts() method
#
# Get options for FILE statement
#
# @param self The object pointer
# @param FfsFileObj for whom options is got
#
- def __GetFileOpts(self, FfsFileObj):
-
- if self.__GetNextToken():
- if TokenFindPattern.match(self.__Token):
- FfsFileObj.KeyStringList.append(self.__Token)
- if self.__IsToken(","):
- while self.__GetNextToken():
- if not TokenFindPattern.match(self.__Token):
+ def _GetFileOpts(self, FfsFileObj):
+ if self._GetNextToken():
+ if TokenFindPattern.match(self._Token):
+ FfsFileObj.KeyStringList.append(self._Token)
+ if self._IsToken(TAB_COMMA_SPLIT):
+ while self._GetNextToken():
+ if not TokenFindPattern.match(self._Token):
raise Warning("expected KeyString \"Target_Tag_Arch\"", self.FileName, self.CurrentLineNumber)
- FfsFileObj.KeyStringList.append(self.__Token)
+ FfsFileObj.KeyStringList.append(self._Token)
- if not self.__IsToken(","):
+ if not self._IsToken(TAB_COMMA_SPLIT):
break
else:
- self.__UndoToken()
+ self._UndoToken()
- if self.__IsKeyword( "FIXED", True):
+ if self._IsKeyword("FIXED", True):
FfsFileObj.Fixed = True
- if self.__IsKeyword( "CHECKSUM", True):
+ if self._IsKeyword("CHECKSUM", True):
FfsFileObj.CheckSum = True
- if self.__GetAlignment():
- if self.__Token not in ("Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENTS:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
#For FFS, Auto is default option same to ""
- if not self.__Token == "Auto":
- FfsFileObj.Alignment = self.__Token
+ if not self._Token == "Auto":
+ FfsFileObj.Alignment = self._Token
- ## __GetAlignment() method
+ ## _GetAlignment() method
#
# Return the alignment value
#
@@ -2819,18 +2749,18 @@ class FdfParser:
# @retval True Successfully find alignment
# @retval False Not able to find alignment
#
- def __GetAlignment(self):
- if self.__IsKeyword( "Align", True):
- if not self.__IsToken( "="):
+ def _GetAlignment(self):
+ if self._IsKeyword("Align", True):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected alignment value", self.FileName, self.CurrentLineNumber)
return True
return False
- ## __GetFilePart() method
+ ## _GetFilePart() method
#
# Get section data for FILE statement
#
@@ -2838,23 +2768,23 @@ class FdfParser:
# @param FfsFileObj for whom section is got
# @param MacroDict dictionary used to replace macro
#
- def __GetSectionData(self, FfsFileObj, MacroDict = {}):
+ def _GetSectionData(self, FfsFileObj, MacroDict = {}):
Dict = {}
Dict.update(MacroDict)
- self.__GetDefineStatements(FfsFileObj)
+ self._GetDefineStatements(FfsFileObj)
Dict.update(FfsFileObj.DefineVarDict)
- self.__GetAprioriSection(FfsFileObj, Dict.copy())
- self.__GetAprioriSection(FfsFileObj, Dict.copy())
+ self._GetAprioriSection(FfsFileObj, Dict.copy())
+ self._GetAprioriSection(FfsFileObj, Dict.copy())
while True:
- IsLeafSection = self.__GetLeafSection(FfsFileObj, Dict)
- IsEncapSection = self.__GetEncapsulationSec(FfsFileObj)
+ IsLeafSection = self._GetLeafSection(FfsFileObj, Dict)
+ IsEncapSection = self._GetEncapsulationSec(FfsFileObj)
if not IsLeafSection and not IsEncapSection:
break
- ## __GetLeafSection() method
+ ## _GetLeafSection() method
#
# Get leaf section for Obj
#
@@ -2864,97 +2794,95 @@ class FdfParser:
# @retval True Successfully find section statement
# @retval False Not able to find section statement
#
- def __GetLeafSection(self, Obj, MacroDict = {}):
-
+ def _GetLeafSection(self, Obj, MacroDict = {}):
OldPos = self.GetFileBufferPos()
- if not self.__IsKeyword( "SECTION"):
+ if not self._IsKeyword("SECTION"):
if len(Obj.SectionList) == 0:
raise Warning("expected SECTION", self.FileName, self.CurrentLineNumber)
else:
return False
AlignValue = None
- if self.__GetAlignment():
- if self.__Token not in ("Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- AlignValue = self.__Token
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENTS:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ AlignValue = self._Token
BuildNum = None
- if self.__IsKeyword( "BUILD_NUM"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("BUILD_NUM"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Build number value", self.FileName, self.CurrentLineNumber)
- BuildNum = self.__Token
+ BuildNum = self._Token
- if self.__IsKeyword( "VERSION"):
+ if self._IsKeyword("VERSION"):
if AlignValue == 'Auto':
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected version", self.FileName, self.CurrentLineNumber)
- VerSectionObj = VerSection.VerSection()
+ VerSectionObj = VerSection()
VerSectionObj.Alignment = AlignValue
VerSectionObj.BuildNum = BuildNum
- if self.__GetStringData():
- VerSectionObj.StringData = self.__Token
+ if self._GetStringData():
+ VerSectionObj.StringData = self._Token
else:
- VerSectionObj.FileName = self.__Token
+ VerSectionObj.FileName = self._Token
Obj.SectionList.append(VerSectionObj)
- elif self.__IsKeyword( BINARY_FILE_TYPE_UI):
+ elif self._IsKeyword(BINARY_FILE_TYPE_UI):
if AlignValue == 'Auto':
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected UI", self.FileName, self.CurrentLineNumber)
- UiSectionObj = UiSection.UiSection()
+ UiSectionObj = UiSection()
UiSectionObj.Alignment = AlignValue
- if self.__GetStringData():
- UiSectionObj.StringData = self.__Token
+ if self._GetStringData():
+ UiSectionObj.StringData = self._Token
else:
- UiSectionObj.FileName = self.__Token
+ UiSectionObj.FileName = self._Token
Obj.SectionList.append(UiSectionObj)
- elif self.__IsKeyword( "FV_IMAGE"):
+ elif self._IsKeyword("FV_IMAGE"):
if AlignValue == 'Auto':
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV name or FV file path", self.FileName, self.CurrentLineNumber)
- FvName = self.__Token
+ FvName = self._Token
FvObj = None
- if self.__IsToken( "{"):
- FvObj = Fv.FV()
+ if self._IsToken("{"):
+ FvObj = FV()
FvObj.UiFvName = FvName.upper()
- self.__GetDefineStatements(FvObj)
+ self._GetDefineStatements(FvObj)
MacroDict.update(FvObj.DefineVarDict)
- self.__GetBlockStatement(FvObj)
- self.__GetSetStatements(FvObj)
- self.__GetFvAlignment(FvObj)
- self.__GetFvAttributes(FvObj)
- self.__GetAprioriSection(FvObj, MacroDict.copy())
- self.__GetAprioriSection(FvObj, MacroDict.copy())
+ self._GetBlockStatement(FvObj)
+ self._GetSetStatements(FvObj)
+ self._GetFvAlignment(FvObj)
+ self._GetFvAttributes(FvObj)
+ self._GetAprioriSection(FvObj, MacroDict.copy())
+ self._GetAprioriSection(FvObj, MacroDict.copy())
while True:
- IsInf = self.__GetInfStatement(FvObj)
- IsFile = self.__GetFileStatement(FvObj, MacroDict.copy())
+ IsInf = self._GetInfStatement(FvObj)
+ IsFile = self._GetFileStatement(FvObj, MacroDict.copy())
if not IsInf and not IsFile:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- FvImageSectionObj = FvImageSection.FvImageSection()
+ FvImageSectionObj = FvImageSection()
FvImageSectionObj.Alignment = AlignValue
if FvObj is not None:
FvImageSectionObj.Fv = FvObj
@@ -2965,81 +2893,81 @@ class FdfParser:
Obj.SectionList.append(FvImageSectionObj)
- elif self.__IsKeyword("PEI_DEPEX_EXP") or self.__IsKeyword("DXE_DEPEX_EXP") or self.__IsKeyword("SMM_DEPEX_EXP"):
+ elif self._IsKeyword("PEI_DEPEX_EXP") or self._IsKeyword("DXE_DEPEX_EXP") or self._IsKeyword("SMM_DEPEX_EXP"):
if AlignValue == 'Auto':
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
- DepexSectionObj = DepexSection.DepexSection()
+ DepexSectionObj = DepexSection()
DepexSectionObj.Alignment = AlignValue
- DepexSectionObj.DepexType = self.__Token
+ DepexSectionObj.DepexType = self._Token
- if not self.__IsToken( "="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken( "{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- if not self.__SkipToToken( "}"):
+ if not self._SkipToToken("}"):
raise Warning("expected Depex expression ending '}'", self.FileName, self.CurrentLineNumber)
- DepexSectionObj.Expression = self.__SkippedChars.rstrip('}')
+ DepexSectionObj.Expression = self._SkippedChars.rstrip('}')
Obj.SectionList.append(DepexSectionObj)
else:
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected section type", self.FileName, self.CurrentLineNumber)
# Encapsulation section appear, UndoToken and return
- if self.__Token == "COMPRESS" or self.__Token == "GUIDED":
+ if self._Token == "COMPRESS" or self._Token == "GUIDED":
self.SetFileBufferPos(OldPos)
return False
- if self.__Token not in ("COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
+ if self._Token not in ("COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
BINARY_FILE_TYPE_UI, "VERSION", BINARY_FILE_TYPE_PEI_DEPEX, "SUBTYPE_GUID", BINARY_FILE_TYPE_SMM_DEPEX):
- raise Warning("Unknown section type '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- if AlignValue == 'Auto'and (not self.__Token == BINARY_FILE_TYPE_PE32) and (not self.__Token == BINARY_FILE_TYPE_TE):
+ raise Warning("Unknown section type '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ if AlignValue == 'Auto'and (not self._Token == BINARY_FILE_TYPE_PE32) and (not self._Token == BINARY_FILE_TYPE_TE):
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
# DataSection
- DataSectionObj = DataSection.DataSection()
+ DataSectionObj = DataSection()
DataSectionObj.Alignment = AlignValue
- DataSectionObj.SecType = self.__Token
+ DataSectionObj.SecType = self._Token
- if self.__IsKeyword('RELOCS_STRIPPED') or self.__IsKeyword('RELOCS_RETAINED'):
- if self.__FileCouldHaveRelocFlag(Obj.FvFileType) and self.__SectionCouldHaveRelocFlag(DataSectionObj.SecType):
- if self.__Token == 'RELOCS_STRIPPED':
+ if self._IsKeyword('RELOCS_STRIPPED') or self._IsKeyword('RELOCS_RETAINED'):
+ if self._FileCouldHaveRelocFlag(Obj.FvFileType) and self._SectionCouldHaveRelocFlag(DataSectionObj.SecType):
+ if self._Token == 'RELOCS_STRIPPED':
DataSectionObj.KeepReloc = False
else:
DataSectionObj.KeepReloc = True
else:
raise Warning("File type %s, section type %s, could not have reloc strip flag%d" % (Obj.FvFileType, DataSectionObj.SecType, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
- if self.__IsToken("="):
- if not self.__GetNextToken():
+ if self._IsToken(TAB_EQUAL_SPLIT):
+ if not self._GetNextToken():
raise Warning("expected section file path", self.FileName, self.CurrentLineNumber)
- DataSectionObj.SectFileName = self.__Token
- self.__VerifyFile(DataSectionObj.SectFileName)
+ DataSectionObj.SectFileName = self._Token
+ self._VerifyFile(DataSectionObj.SectFileName)
else:
- if not self.__GetCglSection(DataSectionObj):
+ if not self._GetCglSection(DataSectionObj):
return False
Obj.SectionList.append(DataSectionObj)
return True
- ## __VerifyFile
+ ## _VerifyFile
#
# Check if file exists or not:
# If current phase if GenFds, the file must exist;
# If current phase is AutoGen and the file is not in $(OUTPUT_DIRECTORY), the file must exist
# @param FileName: File path to be verified.
#
- def __VerifyFile(self, FileName):
- if FileName.replace('$(WORKSPACE)', '').find('$') != -1:
+ def _VerifyFile(self, FileName):
+ if FileName.replace(TAB_WORKSPACE, '').find('$') != -1:
return
- if not GlobalData.gAutoGenPhase or not self.__GetMacroValue("OUTPUT_DIRECTORY") in FileName:
+ if not GlobalData.gAutoGenPhase or not self._GetMacroValue(TAB_DSC_DEFINES_OUTPUT_DIRECTORY) in FileName:
ErrorCode, ErrorInfo = PathClass(NormPath(FileName), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
EdkLogger.error("GenFds", ErrorCode, ExtraData=ErrorInfo)
- ## __GetCglSection() method
+ ## _GetCglSection() method
#
# Get compressed or GUIDed section for Obj
#
@@ -3049,28 +2977,28 @@ class FdfParser:
# @retval True Successfully find section statement
# @retval False Not able to find section statement
#
- def __GetCglSection(self, Obj, AlignValue = None):
+ def _GetCglSection(self, Obj, AlignValue = None):
- if self.__IsKeyword( "COMPRESS"):
+ if self._IsKeyword("COMPRESS"):
type = "PI_STD"
- if self.__IsKeyword("PI_STD") or self.__IsKeyword("PI_NONE"):
- type = self.__Token
+ if self._IsKeyword("PI_STD") or self._IsKeyword("PI_NONE"):
+ type = self._Token
- if not self.__IsToken("{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- CompressSectionObj = CompressSection.CompressSection()
+ CompressSectionObj = CompressSection()
CompressSectionObj.Alignment = AlignValue
CompressSectionObj.CompType = type
# Recursive sections...
while True:
- IsLeafSection = self.__GetLeafSection(CompressSectionObj)
- IsEncapSection = self.__GetEncapsulationSec(CompressSectionObj)
+ IsLeafSection = self._GetLeafSection(CompressSectionObj)
+ IsEncapSection = self._GetEncapsulationSec(CompressSectionObj)
if not IsLeafSection and not IsEncapSection:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
Obj.SectionList.append(CompressSectionObj)
@@ -3079,15 +3007,15 @@ class FdfParser:
return True
- elif self.__IsKeyword( "GUIDED"):
+ elif self._IsKeyword("GUIDED"):
GuidValue = None
- if self.__GetNextGuid():
- GuidValue = self.__Token
+ if self._GetNextGuid():
+ GuidValue = self._Token
- AttribDict = self.__GetGuidAttrib()
- if not self.__IsToken("{"):
+ AttribDict = self._GetGuidAttrib()
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- GuidSectionObj = GuidSection.GuidSection()
+ GuidSectionObj = GuidSection()
GuidSectionObj.Alignment = AlignValue
GuidSectionObj.NameGuid = GuidValue
GuidSectionObj.SectionType = "GUIDED"
@@ -3096,12 +3024,12 @@ class FdfParser:
GuidSectionObj.ExtraHeaderSize = AttribDict["EXTRA_HEADER_SIZE"]
# Recursive sections...
while True:
- IsLeafSection = self.__GetLeafSection(GuidSectionObj)
- IsEncapSection = self.__GetEncapsulationSec(GuidSectionObj)
+ IsLeafSection = self._GetLeafSection(GuidSectionObj)
+ IsEncapSection = self._GetEncapsulationSec(GuidSectionObj)
if not IsLeafSection and not IsEncapSection:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
Obj.SectionList.append(GuidSectionObj)
@@ -3109,44 +3037,43 @@ class FdfParser:
return False
- ## __GetGuidAttri() method
+ ## _GetGuidAttri() method
#
# Get attributes for GUID section
#
# @param self The object pointer
# @retval AttribDict Dictionary of key-value pair of section attributes
#
- def __GetGuidAttrib(self):
-
+ def _GetGuidAttrib(self):
AttribDict = {}
AttribDict["PROCESSING_REQUIRED"] = "NONE"
AttribDict["AUTH_STATUS_VALID"] = "NONE"
AttribDict["EXTRA_HEADER_SIZE"] = -1
- while self.__IsKeyword("PROCESSING_REQUIRED") or self.__IsKeyword("AUTH_STATUS_VALID") \
- or self.__IsKeyword("EXTRA_HEADER_SIZE"):
- AttribKey = self.__Token
+ while self._IsKeyword("PROCESSING_REQUIRED") or self._IsKeyword("AUTH_STATUS_VALID") \
+ or self._IsKeyword("EXTRA_HEADER_SIZE"):
+ AttribKey = self._Token
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected TRUE(1)/FALSE(0)/Number", self.FileName, self.CurrentLineNumber)
elif AttribKey == "EXTRA_HEADER_SIZE":
Base = 10
- if self.__Token[0:2].upper() == "0X":
+ if self._Token[0:2].upper() == "0X":
Base = 16
try:
- AttribDict[AttribKey] = int(self.__Token, Base)
+ AttribDict[AttribKey] = int(self._Token, Base)
continue
except ValueError:
raise Warning("expected Number", self.FileName, self.CurrentLineNumber)
- elif self.__Token.upper() not in ("TRUE", "FALSE", "1", "0"):
+ elif self._Token.upper() not in ("TRUE", "FALSE", "1", "0"):
raise Warning("expected TRUE/FALSE (1/0)", self.FileName, self.CurrentLineNumber)
- AttribDict[AttribKey] = self.__Token
+ AttribDict[AttribKey] = self._Token
return AttribDict
- ## __GetEncapsulationSec() method
+ ## _GetEncapsulationSec() method
#
# Get encapsulation section for FILE
#
@@ -3155,91 +3082,89 @@ class FdfParser:
# @retval True Successfully find section statement
# @retval False Not able to find section statement
#
- def __GetEncapsulationSec(self, FfsFileObj):
-
+ def _GetEncapsulationSec(self, FfsFileObj):
OldPos = self.GetFileBufferPos()
- if not self.__IsKeyword( "SECTION"):
+ if not self._IsKeyword("SECTION"):
if len(FfsFileObj.SectionList) == 0:
raise Warning("expected SECTION", self.FileName, self.CurrentLineNumber)
else:
return False
AlignValue = None
- if self.__GetAlignment():
- if self.__Token not in ("8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- AlignValue = self.__Token
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENT_NOAUTO:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ AlignValue = self._Token
- if not self.__GetCglSection(FfsFileObj, AlignValue):
+ if not self._GetCglSection(FfsFileObj, AlignValue):
self.SetFileBufferPos(OldPos)
return False
else:
return True
- def __GetFmp(self):
- if not self.__GetNextToken():
+ def _GetFmp(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[FMPPAYLOAD."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[FMPPAYLOAD."):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- self.__SkipToToken("[FMPPAYLOAD.", True)
- FmpUiName = self.__GetUiName().upper()
+ self._UndoToken()
+ self._SkipToToken("[FMPPAYLOAD.", True)
+ FmpUiName = self._GetUiName().upper()
if FmpUiName in self.Profile.FmpPayloadDict:
raise Warning("Duplicated FMP UI name found: %s" % FmpUiName, self.FileName, self.CurrentLineNumber)
- FmpData = CapsuleData.CapsulePayload()
+ FmpData = CapsulePayload()
FmpData.UiName = FmpUiName
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("The FMP payload section is empty!", self.FileName, self.CurrentLineNumber)
FmpKeyList = ['IMAGE_HEADER_INIT_VERSION', 'IMAGE_TYPE_ID', 'IMAGE_INDEX', 'HARDWARE_INSTANCE', 'CERTIFICATE_GUID', 'MONOTONIC_COUNT']
- while self.__Token in FmpKeyList:
- Name = self.__Token
+ while self._Token in FmpKeyList:
+ Name = self._Token
FmpKeyList.remove(Name)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
if Name == 'IMAGE_TYPE_ID':
- if not self.__GetNextGuid():
+ if not self._GetNextGuid():
raise Warning("expected GUID value for IMAGE_TYPE_ID.", self.FileName, self.CurrentLineNumber)
- FmpData.ImageTypeId = self.__Token
+ FmpData.ImageTypeId = self._Token
elif Name == 'CERTIFICATE_GUID':
- if not self.__GetNextGuid():
+ if not self._GetNextGuid():
raise Warning("expected GUID value for CERTIFICATE_GUID.", self.FileName, self.CurrentLineNumber)
- FmpData.Certificate_Guid = self.__Token
- if uuid.UUID(FmpData.Certificate_Guid) != EFI_CERT_TYPE_RSA2048_SHA256_GUID and uuid.UUID(FmpData.Certificate_Guid) != EFI_CERT_TYPE_PKCS7_GUID:
+ FmpData.Certificate_Guid = self._Token
+ if UUID(FmpData.Certificate_Guid) != EFI_CERT_TYPE_RSA2048_SHA256_GUID and UUID(FmpData.Certificate_Guid) != EFI_CERT_TYPE_PKCS7_GUID:
raise Warning("Only support EFI_CERT_TYPE_RSA2048_SHA256_GUID or EFI_CERT_TYPE_PKCS7_GUID for CERTIFICATE_GUID.", self.FileName, self.CurrentLineNumber)
else:
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected value of %s" % Name, self.FileName, self.CurrentLineNumber)
- Value = self.__Token
+ Value = self._Token
if Name == 'IMAGE_HEADER_INIT_VERSION':
- if FdfParser.__Verify(Name, Value, 'UINT8'):
+ if FdfParser._Verify(Name, Value, 'UINT8'):
FmpData.Version = Value
elif Name == 'IMAGE_INDEX':
- if FdfParser.__Verify(Name, Value, 'UINT8'):
+ if FdfParser._Verify(Name, Value, 'UINT8'):
FmpData.ImageIndex = Value
elif Name == 'HARDWARE_INSTANCE':
- if FdfParser.__Verify(Name, Value, 'UINT8'):
+ if FdfParser._Verify(Name, Value, 'UINT8'):
FmpData.HardwareInstance = Value
elif Name == 'MONOTONIC_COUNT':
- if FdfParser.__Verify(Name, Value, 'UINT64'):
+ if FdfParser._Verify(Name, Value, 'UINT64'):
FmpData.MonotonicCount = Value
if FmpData.MonotonicCount.upper().startswith('0X'):
FmpData.MonotonicCount = (long)(FmpData.MonotonicCount, 16)
else:
FmpData.MonotonicCount = (long)(FmpData.MonotonicCount)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
break
else:
- self.__UndoToken()
+ self._UndoToken()
if (FmpData.MonotonicCount and not FmpData.Certificate_Guid) or (not FmpData.MonotonicCount and FmpData.Certificate_Guid):
EdkLogger.error("FdfParser", FORMAT_INVALID, "CERTIFICATE_GUID and MONOTONIC_COUNT must be work as a pair.")
@@ -3248,7 +3173,7 @@ class FdfParser:
if FmpKeyList and 'IMAGE_TYPE_ID' in FmpKeyList:
raise Warning("Missing keywords IMAGE_TYPE_ID in FMP payload section.", self.FileName, self.CurrentLineNumber)
# get the Image file and Vendor code file
- self.__GetFMPCapsuleData(FmpData)
+ self._GetFMPCapsuleData(FmpData)
if not FmpData.ImageFile:
raise Warning("Missing image file in FMP payload section.", self.FileName, self.CurrentLineNumber)
# check whether more than one Vendor code file
@@ -3257,7 +3182,7 @@ class FdfParser:
self.Profile.FmpPayloadDict[FmpUiName] = FmpData
return True
- ## __GetCapsule() method
+ ## _GetCapsule() method
#
# Get capsule section contents and store its data into capsule list of self.Profile
#
@@ -3265,90 +3190,89 @@ class FdfParser:
# @retval True Successfully find a capsule
# @retval False Not able to find a capsule
#
- def __GetCapsule(self):
-
- if not self.__GetNextToken():
+ def _GetCapsule(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[CAPSULE."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[CAPSULE."):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[CAPSULE.", True):
+ self._UndoToken()
+ if not self._IsToken("[CAPSULE.", True):
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
#print 'Parsing String: %s in File %s, At line: %d, Offset Within Line: %d' \
- # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
+ # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
raise Warning("expected [Capsule.]", self.FileName, self.CurrentLineNumber)
- CapsuleObj = Capsule.Capsule()
+ CapsuleObj = Capsule()
- CapsuleName = self.__GetUiName()
+ CapsuleName = self._GetUiName()
if not CapsuleName:
raise Warning("expected capsule name", self.FileName, self.CurrentLineNumber)
CapsuleObj.UiCapsuleName = CapsuleName.upper()
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- if self.__IsKeyword("CREATE_FILE"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("CREATE_FILE"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected file name", self.FileName, self.CurrentLineNumber)
- CapsuleObj.CreateFile = self.__Token
+ CapsuleObj.CreateFile = self._Token
- self.__GetCapsuleStatements(CapsuleObj)
+ self._GetCapsuleStatements(CapsuleObj)
self.Profile.CapsuleDict[CapsuleObj.UiCapsuleName] = CapsuleObj
return True
- ## __GetCapsuleStatements() method
+ ## _GetCapsuleStatements() method
#
# Get statements for capsule
#
# @param self The object pointer
# @param Obj for whom statements are got
#
- def __GetCapsuleStatements(self, Obj):
- self.__GetCapsuleTokens(Obj)
- self.__GetDefineStatements(Obj)
- self.__GetSetStatements(Obj)
- self.__GetCapsuleData(Obj)
+ def _GetCapsuleStatements(self, Obj):
+ self._GetCapsuleTokens(Obj)
+ self._GetDefineStatements(Obj)
+ self._GetSetStatements(Obj)
+ self._GetCapsuleData(Obj)
- ## __GetCapsuleTokens() method
+ ## _GetCapsuleTokens() method
#
# Get token statements for capsule
#
# @param self The object pointer
# @param Obj for whom token statements are got
#
- def __GetCapsuleTokens(self, Obj):
- if not self.__GetNextToken():
+ def _GetCapsuleTokens(self, Obj):
+ if not self._GetNextToken():
return False
- while self.__Token in ("CAPSULE_GUID", "CAPSULE_HEADER_SIZE", "CAPSULE_FLAGS", "OEM_CAPSULE_FLAGS", "CAPSULE_HEADER_INIT_VERSION"):
- Name = self.__Token.strip()
- if not self.__IsToken("="):
+ while self._Token in ("CAPSULE_GUID", "CAPSULE_HEADER_SIZE", "CAPSULE_FLAGS", "OEM_CAPSULE_FLAGS", "CAPSULE_HEADER_INIT_VERSION"):
+ Name = self._Token.strip()
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected value", self.FileName, self.CurrentLineNumber)
if Name == 'CAPSULE_FLAGS':
- if not self.__Token in ("PersistAcrossReset", "PopulateSystemTable", "InitiateReset"):
+ if not self._Token in ("PersistAcrossReset", "PopulateSystemTable", "InitiateReset"):
raise Warning("expected PersistAcrossReset, PopulateSystemTable, or InitiateReset", self.FileName, self.CurrentLineNumber)
- Value = self.__Token.strip()
- while self.__IsToken(","):
- Value += ','
- if not self.__GetNextToken():
+ Value = self._Token.strip()
+ while self._IsToken(TAB_COMMA_SPLIT):
+ Value += TAB_COMMA_SPLIT
+ if not self._GetNextToken():
raise Warning("expected value", self.FileName, self.CurrentLineNumber)
- if not self.__Token in ("PersistAcrossReset", "PopulateSystemTable", "InitiateReset"):
+ if not self._Token in ("PersistAcrossReset", "PopulateSystemTable", "InitiateReset"):
raise Warning("expected PersistAcrossReset, PopulateSystemTable, or InitiateReset", self.FileName, self.CurrentLineNumber)
- Value += self.__Token.strip()
+ Value += self._Token.strip()
elif Name == 'OEM_CAPSULE_FLAGS':
- Value = self.__Token.strip()
+ Value = self._Token.strip()
if not Value.upper().startswith('0X'):
raise Warning("expected hex value between 0x0000 and 0xFFFF", self.FileName, self.CurrentLineNumber)
try:
@@ -3357,51 +3281,49 @@ class FdfParser:
raise Warning("expected hex value between 0x0000 and 0xFFFF", self.FileName, self.CurrentLineNumber)
if not 0x0000 <= Value <= 0xFFFF:
raise Warning("expected hex value between 0x0000 and 0xFFFF", self.FileName, self.CurrentLineNumber)
- Value = self.__Token.strip()
+ Value = self._Token.strip()
else:
- Value = self.__Token.strip()
+ Value = self._Token.strip()
Obj.TokensDict[Name] = Value
- if not self.__GetNextToken():
+ if not self._GetNextToken():
return False
- self.__UndoToken()
+ self._UndoToken()
- ## __GetCapsuleData() method
+ ## _GetCapsuleData() method
#
# Get capsule data for capsule
#
# @param self The object pointer
# @param Obj for whom capsule data are got
#
- def __GetCapsuleData(self, Obj):
-
+ def _GetCapsuleData(self, Obj):
while True:
- IsInf = self.__GetInfStatement(Obj, True)
- IsFile = self.__GetFileStatement(Obj, True)
- IsFv = self.__GetFvStatement(Obj)
- IsFd = self.__GetFdStatement(Obj)
- IsAnyFile = self.__GetAnyFileStatement(Obj)
- IsAfile = self.__GetAfileStatement(Obj)
- IsFmp = self.__GetFmpStatement(Obj)
+ IsInf = self._GetInfStatement(Obj, True)
+ IsFile = self._GetFileStatement(Obj, True)
+ IsFv = self._GetFvStatement(Obj)
+ IsFd = self._GetFdStatement(Obj)
+ IsAnyFile = self._GetAnyFileStatement(Obj)
+ IsAfile = self._GetAfileStatement(Obj)
+ IsFmp = self._GetFmpStatement(Obj)
if not (IsInf or IsFile or IsFv or IsFd or IsAnyFile or IsAfile or IsFmp):
break
- ## __GetFMPCapsuleData() method
+ ## _GetFMPCapsuleData() method
#
# Get capsule data for FMP capsule
#
# @param self The object pointer
# @param Obj for whom capsule data are got
#
- def __GetFMPCapsuleData(self, Obj):
-
+ def _GetFMPCapsuleData(self, Obj):
while True:
- IsFv = self.__GetFvStatement(Obj, True)
- IsFd = self.__GetFdStatement(Obj, True)
- IsAnyFile = self.__GetAnyFileStatement(Obj, True)
+ IsFv = self._GetFvStatement(Obj, True)
+ IsFd = self._GetFdStatement(Obj, True)
+ IsAnyFile = self._GetAnyFileStatement(Obj, True)
if not (IsFv or IsFd or IsAnyFile):
break
- ## __GetFvStatement() method
+ ## _GetFvStatement() method
#
# Get FV for capsule
#
@@ -3410,32 +3332,31 @@ class FdfParser:
# @retval True Successfully find a FV statement
# @retval False Not able to find a FV statement
#
- def __GetFvStatement(self, CapsuleObj, FMPCapsule = False):
-
- if not self.__IsKeyword(BINARY_FILE_TYPE_FV):
+ def _GetFvStatement(self, CapsuleObj, FMPCapsule = False):
+ if not self._IsKeyword(BINARY_FILE_TYPE_FV):
return False
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FV name", self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() not in self.Profile.FvDict:
+ if self._Token.upper() not in self.Profile.FvDict:
raise Warning("FV name does not exist", self.FileName, self.CurrentLineNumber)
- CapsuleFv = CapsuleData.CapsuleFv()
- CapsuleFv.FvName = self.__Token
+ myCapsuleFv = CapsuleFv()
+ myCapsuleFv.FvName = self._Token
if FMPCapsule:
if not CapsuleObj.ImageFile:
- CapsuleObj.ImageFile.append(CapsuleFv)
+ CapsuleObj.ImageFile.append(myCapsuleFv)
else:
- CapsuleObj.VendorCodeFile.append(CapsuleFv)
+ CapsuleObj.VendorCodeFile.append(myCapsuleFv)
else:
- CapsuleObj.CapsuleDataList.append(CapsuleFv)
+ CapsuleObj.CapsuleDataList.append(myCapsuleFv)
return True
- ## __GetFdStatement() method
+ ## _GetFdStatement() method
#
# Get FD for capsule
#
@@ -3444,22 +3365,21 @@ class FdfParser:
# @retval True Successfully find a FD statement
# @retval False Not able to find a FD statement
#
- def __GetFdStatement(self, CapsuleObj, FMPCapsule = False):
-
- if not self.__IsKeyword("FD"):
+ def _GetFdStatement(self, CapsuleObj, FMPCapsule = False):
+ if not self._IsKeyword("FD"):
return False
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected FD name", self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() not in self.Profile.FdDict:
+ if self._Token.upper() not in self.Profile.FdDict:
raise Warning("FD name does not exist", self.FileName, self.CurrentLineNumber)
- CapsuleFd = CapsuleData.CapsuleFd()
- CapsuleFd.FdName = self.__Token
+ CapsuleFd = CapsuleFd()
+ CapsuleFd.FdName = self._Token
if FMPCapsule:
if not CapsuleObj.ImageFile:
CapsuleObj.ImageFile.append(CapsuleFd)
@@ -3469,49 +3389,49 @@ class FdfParser:
CapsuleObj.CapsuleDataList.append(CapsuleFd)
return True
- def __GetFmpStatement(self, CapsuleObj):
- if not self.__IsKeyword("FMP_PAYLOAD"):
- if not self.__IsKeyword("FMP"):
+ def _GetFmpStatement(self, CapsuleObj):
+ if not self._IsKeyword("FMP_PAYLOAD"):
+ if not self._IsKeyword("FMP"):
return False
- if not self.__IsKeyword("PAYLOAD"):
- self.__UndoToken()
+ if not self._IsKeyword("PAYLOAD"):
+ self._UndoToken()
return False
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected payload name after FMP_PAYLOAD =", self.FileName, self.CurrentLineNumber)
- Payload = self.__Token.upper()
+ Payload = self._Token.upper()
if Payload not in self.Profile.FmpPayloadDict:
- raise Warning("This FMP Payload does not exist: %s" % self.__Token, self.FileName, self.CurrentLineNumber)
+ raise Warning("This FMP Payload does not exist: %s" % self._Token, self.FileName, self.CurrentLineNumber)
CapsuleObj.FmpPayloadList.append(self.Profile.FmpPayloadDict[Payload])
return True
- def __ParseRawFileStatement(self):
- if not self.__IsKeyword("FILE"):
+ def _ParseRawFileStatement(self):
+ if not self._IsKeyword("FILE"):
return None
- if not self.__IsKeyword("DATA"):
- self.__UndoToken()
+ if not self._IsKeyword("DATA"):
+ self._UndoToken()
return None
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected File name", self.FileName, self.CurrentLineNumber)
- AnyFileName = self.__Token
- self.__VerifyFile(AnyFileName)
+ AnyFileName = self._Token
+ self._VerifyFile(AnyFileName)
if not os.path.isabs(AnyFileName):
AnyFileName = mws.join(GenFdsGlobalVariable.WorkSpaceDir, AnyFileName)
return AnyFileName
- ## __GetAnyFileStatement() method
+ ## _GetAnyFileStatement() method
#
# Get AnyFile for capsule
#
@@ -3520,12 +3440,12 @@ class FdfParser:
# @retval True Successfully find a Anyfile statement
# @retval False Not able to find a AnyFile statement
#
- def __GetAnyFileStatement(self, CapsuleObj, FMPCapsule = False):
- AnyFileName = self.__ParseRawFileStatement()
+ def _GetAnyFileStatement(self, CapsuleObj, FMPCapsule = False):
+ AnyFileName = self._ParseRawFileStatement()
if not AnyFileName:
return False
- CapsuleAnyFile = CapsuleData.CapsuleAnyFile()
+ CapsuleAnyFile = CapsuleAnyFile()
CapsuleAnyFile.FileName = AnyFileName
if FMPCapsule:
if not CapsuleObj.ImageFile:
@@ -3536,7 +3456,7 @@ class FdfParser:
CapsuleObj.CapsuleDataList.append(CapsuleAnyFile)
return True
- ## __GetAfileStatement() method
+ ## _GetAfileStatement() method
#
# Get Afile for capsule
#
@@ -3545,18 +3465,17 @@ class FdfParser:
# @retval True Successfully find a Afile statement
# @retval False Not able to find a Afile statement
#
- def __GetAfileStatement(self, CapsuleObj):
-
- if not self.__IsKeyword("APPEND"):
+ def _GetAfileStatement(self, CapsuleObj):
+ if not self._IsKeyword("APPEND"):
return False
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Afile name", self.FileName, self.CurrentLineNumber)
- AfileName = self.__Token
+ AfileName = self._Token
AfileBaseName = os.path.basename(AfileName)
if os.path.splitext(AfileBaseName)[1] not in [".bin", ".BIN", ".Bin", ".dat", ".DAT", ".Dat", ".data", ".DATA", ".Data"]:
@@ -3565,19 +3484,19 @@ class FdfParser:
if not os.path.isabs(AfileName):
AfileName = GenFdsGlobalVariable.ReplaceWorkspaceMacro(AfileName)
- self.__VerifyFile(AfileName)
+ self._VerifyFile(AfileName)
else:
if not os.path.exists(AfileName):
raise Warning('%s does not exist' % AfileName, self.FileName, self.CurrentLineNumber)
else:
pass
- CapsuleAfile = CapsuleData.CapsuleAfile()
+ CapsuleAfile = CapsuleAfile()
CapsuleAfile.FileName = AfileName
CapsuleObj.CapsuleDataList.append(CapsuleAfile)
return True
- ## __GetRule() method
+ ## _GetRule() method
#
# Get Rule section contents and store its data into rule list of self.Profile
#
@@ -3585,143 +3504,140 @@ class FdfParser:
# @retval True Successfully find a Rule
# @retval False Not able to find a Rule
#
- def __GetRule(self):
-
- if not self.__GetNextToken():
+ def _GetRule(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[RULE."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[RULE."):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[Rule.", True):
+ self._UndoToken()
+ if not self._IsToken("[Rule.", True):
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
#print 'Parsing String: %s in File %s, At line: %d, Offset Within Line: %d' \
- # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
+ # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
raise Warning("expected [Rule.]", self.FileName, self.CurrentLineNumber)
- if not self.__SkipToToken("."):
+ if not self._SkipToToken(TAB_SPLIT):
raise Warning("expected '.'", self.FileName, self.CurrentLineNumber)
- Arch = self.__SkippedChars.rstrip(".")
+ Arch = self._SkippedChars.rstrip(TAB_SPLIT)
if Arch.upper() not in ARCH_SET_FULL:
raise Warning("Unknown Arch '%s'" % Arch, self.FileName, self.CurrentLineNumber)
- ModuleType = self.__GetModuleType()
+ ModuleType = self._GetModuleType()
TemplateName = ""
- if self.__IsToken("."):
- if not self.__GetNextWord():
+ if self._IsToken(TAB_SPLIT):
+ if not self._GetNextWord():
raise Warning("expected template name", self.FileName, self.CurrentLineNumber)
- TemplateName = self.__Token
+ TemplateName = self._Token
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- RuleObj = self.__GetRuleFileStatements()
+ RuleObj = self._GetRuleFileStatements()
RuleObj.Arch = Arch.upper()
RuleObj.ModuleType = ModuleType
RuleObj.TemplateName = TemplateName
- if TemplateName == '' :
+ if TemplateName == '':
self.Profile.RuleDict['RULE' + \
- '.' + \
+ TAB_SPLIT + \
Arch.upper() + \
- '.' + \
+ TAB_SPLIT + \
ModuleType.upper() ] = RuleObj
- else :
+ else:
self.Profile.RuleDict['RULE' + \
- '.' + \
+ TAB_SPLIT + \
Arch.upper() + \
- '.' + \
+ TAB_SPLIT + \
ModuleType.upper() + \
- '.' + \
+ TAB_SPLIT + \
TemplateName.upper() ] = RuleObj
return True
- ## __GetModuleType() method
+ ## _GetModuleType() method
#
# Return the module type
#
# @param self The object pointer
# @retval string module type
#
- def __GetModuleType(self):
-
- if not self.__GetNextWord():
+ def _GetModuleType(self):
+ if not self._GetNextWord():
raise Warning("expected Module type", self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() not in (SUP_MODULE_SEC, SUP_MODULE_PEI_CORE, SUP_MODULE_PEIM, SUP_MODULE_DXE_CORE, \
+ if self._Token.upper() not in (SUP_MODULE_SEC, SUP_MODULE_PEI_CORE, SUP_MODULE_PEIM, SUP_MODULE_DXE_CORE, \
SUP_MODULE_DXE_DRIVER, SUP_MODULE_DXE_SAL_DRIVER, \
SUP_MODULE_DXE_SMM_DRIVER, SUP_MODULE_DXE_RUNTIME_DRIVER, \
- SUP_MODULE_UEFI_DRIVER, SUP_MODULE_UEFI_APPLICATION, SUP_MODULE_USER_DEFINED, "DEFAULT", SUP_MODULE_BASE, \
+ SUP_MODULE_UEFI_DRIVER, SUP_MODULE_UEFI_APPLICATION, SUP_MODULE_USER_DEFINED, TAB_DEFAULT, SUP_MODULE_BASE, \
EDK_COMPONENT_TYPE_SECURITY_CORE, EDK_COMPONENT_TYPE_COMBINED_PEIM_DRIVER, EDK_COMPONENT_TYPE_PIC_PEIM, EDK_COMPONENT_TYPE_RELOCATABLE_PEIM, \
"PE32_PEIM", EDK_COMPONENT_TYPE_BS_DRIVER, EDK_COMPONENT_TYPE_RT_DRIVER, EDK_COMPONENT_TYPE_SAL_RT_DRIVER, EDK_COMPONENT_TYPE_APPLICATION, "ACPITABLE", SUP_MODULE_SMM_CORE, SUP_MODULE_MM_STANDALONE, SUP_MODULE_MM_CORE_STANDALONE):
- raise Warning("Unknown Module type '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- return self.__Token
+ raise Warning("Unknown Module type '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ return self._Token
- ## __GetFileExtension() method
+ ## _GetFileExtension() method
#
# Return the file extension
#
# @param self The object pointer
# @retval string file name extension
#
- def __GetFileExtension(self):
- if not self.__IsToken("."):
+ def _GetFileExtension(self):
+ if not self._IsToken(TAB_SPLIT):
raise Warning("expected '.'", self.FileName, self.CurrentLineNumber)
Ext = ""
- if self.__GetNextToken():
- if FileExtensionPattern.match(self.__Token):
- Ext = self.__Token
- return '.' + Ext
+ if self._GetNextToken():
+ if FileExtensionPattern.match(self._Token):
+ Ext = self._Token
+ return TAB_SPLIT + Ext
else:
- raise Warning("Unknown file extension '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ raise Warning("Unknown file extension '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
else:
raise Warning("expected file extension", self.FileName, self.CurrentLineNumber)
- ## __GetRuleFileStatement() method
+ ## _GetRuleFileStatement() method
#
# Get rule contents
#
# @param self The object pointer
# @retval Rule Rule object
#
- def __GetRuleFileStatements(self):
-
- if not self.__IsKeyword("FILE"):
+ def _GetRuleFileStatements(self):
+ if not self._IsKeyword("FILE"):
raise Warning("expected FILE", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected FFS type", self.FileName, self.CurrentLineNumber)
- Type = self.__Token.strip().upper()
+ Type = self._Token.strip().upper()
if Type not in ("RAW", "FREEFORM", SUP_MODULE_SEC, SUP_MODULE_PEI_CORE, SUP_MODULE_PEIM,\
"PEI_DXE_COMBO", "DRIVER", SUP_MODULE_DXE_CORE, EDK_COMPONENT_TYPE_APPLICATION, "FV_IMAGE", "SMM", SUP_MODULE_SMM_CORE, SUP_MODULE_MM_STANDALONE, SUP_MODULE_MM_CORE_STANDALONE):
- raise Warning("Unknown FV type '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ raise Warning("Unknown FV type '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__IsKeyword("$(NAMED_GUID)"):
- if not self.__GetNextWord():
+ if not self._IsKeyword("$(NAMED_GUID)"):
+ if not self._GetNextWord():
raise Warning("expected $(NAMED_GUID)", self.FileName, self.CurrentLineNumber)
- if self.__Token == 'PCD':
- if not self.__IsToken( "("):
+ if self._Token == 'PCD':
+ if not self._IsToken("("):
raise Warning("expected '('", self.FileName, self.CurrentLineNumber)
- PcdPair = self.__GetNextPcdSettings()
- if not self.__IsToken( ")"):
+ PcdPair = self._GetNextPcdSettings()
+ if not self._IsToken(")"):
raise Warning("expected ')'", self.FileName, self.CurrentLineNumber)
- self.__Token = 'PCD('+PcdPair[1]+'.'+PcdPair[0]+')'
+ self._Token = 'PCD('+PcdPair[1]+TAB_SPLIT+PcdPair[0]+')'
- NameGuid = self.__Token
+ NameGuid = self._Token
KeepReloc = None
- if self.__IsKeyword('RELOCS_STRIPPED') or self.__IsKeyword('RELOCS_RETAINED'):
- if self.__FileCouldHaveRelocFlag(Type):
- if self.__Token == 'RELOCS_STRIPPED':
+ if self._IsKeyword('RELOCS_STRIPPED') or self._IsKeyword('RELOCS_RETAINED'):
+ if self._FileCouldHaveRelocFlag(Type):
+ if self._Token == 'RELOCS_STRIPPED':
KeepReloc = False
else:
KeepReloc = True
@@ -3729,111 +3645,109 @@ class FdfParser:
raise Warning("File type %s could not have reloc strip flag%d" % (Type, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
KeyStringList = []
- if self.__GetNextToken():
- if TokenFindPattern.match(self.__Token):
- KeyStringList.append(self.__Token)
- if self.__IsToken(","):
- while self.__GetNextToken():
- if not TokenFindPattern.match(self.__Token):
+ if self._GetNextToken():
+ if TokenFindPattern.match(self._Token):
+ KeyStringList.append(self._Token)
+ if self._IsToken(TAB_COMMA_SPLIT):
+ while self._GetNextToken():
+ if not TokenFindPattern.match(self._Token):
raise Warning("expected KeyString \"Target_Tag_Arch\"", self.FileName, self.CurrentLineNumber)
- KeyStringList.append(self.__Token)
+ KeyStringList.append(self._Token)
- if not self.__IsToken(","):
+ if not self._IsToken(TAB_COMMA_SPLIT):
break
else:
- self.__UndoToken()
+ self._UndoToken()
Fixed = False
- if self.__IsKeyword("Fixed", True):
+ if self._IsKeyword("Fixed", True):
Fixed = True
CheckSum = False
- if self.__IsKeyword("CheckSum", True):
+ if self._IsKeyword("CheckSum", True):
CheckSum = True
AlignValue = ""
- if self.__GetAlignment():
- if self.__Token not in ("Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENTS:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
#For FFS, Auto is default option same to ""
- if not self.__Token == "Auto":
- AlignValue = self.__Token
+ if not self._Token == "Auto":
+ AlignValue = self._Token
- if self.__IsToken("{"):
+ if self._IsToken("{"):
# Complex file rule expected
- Rule = RuleComplexFile.RuleComplexFile()
- Rule.FvFileType = Type
- Rule.NameGuid = NameGuid
- Rule.Alignment = AlignValue
- Rule.CheckSum = CheckSum
- Rule.Fixed = Fixed
- Rule.KeyStringList = KeyStringList
+ NewRule = RuleComplexFile()
+ NewRule.FvFileType = Type
+ NewRule.NameGuid = NameGuid
+ NewRule.Alignment = AlignValue
+ NewRule.CheckSum = CheckSum
+ NewRule.Fixed = Fixed
+ NewRule.KeyStringList = KeyStringList
if KeepReloc is not None:
- Rule.KeepReloc = KeepReloc
+ NewRule.KeepReloc = KeepReloc
while True:
- IsEncapsulate = self.__GetRuleEncapsulationSection(Rule)
- IsLeaf = self.__GetEfiSection(Rule)
+ IsEncapsulate = self._GetRuleEncapsulationSection(NewRule)
+ IsLeaf = self._GetEfiSection(NewRule)
if not IsEncapsulate and not IsLeaf:
break
- if not self.__IsToken("}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- return Rule
+ return NewRule
else:
# Simple file rule expected
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected leaf section type", self.FileName, self.CurrentLineNumber)
- SectionName = self.__Token
+ SectionName = self._Token
if SectionName not in ("COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
BINARY_FILE_TYPE_UI, BINARY_FILE_TYPE_PEI_DEPEX, "VERSION", "SUBTYPE_GUID", BINARY_FILE_TYPE_SMM_DEPEX):
raise Warning("Unknown leaf section name '%s'" % SectionName, self.FileName, self.CurrentLineNumber)
- if self.__IsKeyword("Fixed", True):
+ if self._IsKeyword("Fixed", True):
Fixed = True
- if self.__IsKeyword("CheckSum", True):
+ if self._IsKeyword("CheckSum", True):
CheckSum = True
SectAlignment = ""
- if self.__GetAlignment():
- if self.__Token not in ("Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- if self.__Token == 'Auto' and (not SectionName == BINARY_FILE_TYPE_PE32) and (not SectionName == BINARY_FILE_TYPE_TE):
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENTS:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ if self._Token == 'Auto' and (not SectionName == BINARY_FILE_TYPE_PE32) and (not SectionName == BINARY_FILE_TYPE_TE):
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
- SectAlignment = self.__Token
+ SectAlignment = self._Token
Ext = None
- if self.__IsToken('|'):
- Ext = self.__GetFileExtension()
- elif not self.__GetNextToken():
+ if self._IsToken(TAB_VALUE_SPLIT):
+ Ext = self._GetFileExtension()
+ elif not self._GetNextToken():
raise Warning("expected File name", self.FileName, self.CurrentLineNumber)
- Rule = RuleSimpleFile.RuleSimpleFile()
- Rule.SectionType = SectionName
- Rule.FvFileType = Type
- Rule.NameGuid = NameGuid
- Rule.Alignment = AlignValue
- Rule.SectAlignment = SectAlignment
- Rule.CheckSum = CheckSum
- Rule.Fixed = Fixed
- Rule.KeyStringList = KeyStringList
+ NewRule = RuleSimpleFile()
+ NewRule.SectionType = SectionName
+ NewRule.FvFileType = Type
+ NewRule.NameGuid = NameGuid
+ NewRule.Alignment = AlignValue
+ NewRule.SectAlignment = SectAlignment
+ NewRule.CheckSum = CheckSum
+ NewRule.Fixed = Fixed
+ NewRule.KeyStringList = KeyStringList
if KeepReloc is not None:
- Rule.KeepReloc = KeepReloc
- Rule.FileExtension = Ext
- Rule.FileName = self.__Token
- return Rule
+ NewRule.KeepReloc = KeepReloc
+ NewRule.FileExtension = Ext
+ NewRule.FileName = self._Token
+ return NewRule
- ## __GetEfiSection() method
+ ## _GetEfiSection() method
#
# Get section list for Rule
#
@@ -3842,127 +3756,124 @@ class FdfParser:
# @retval True Successfully find section statement
# @retval False Not able to find section statement
#
- def __GetEfiSection(self, Obj):
-
+ def _GetEfiSection(self, Obj):
OldPos = self.GetFileBufferPos()
- if not self.__GetNextWord():
+ if not self._GetNextWord():
return False
- SectionName = self.__Token
+ SectionName = self._Token
if SectionName not in ("COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
BINARY_FILE_TYPE_UI, "VERSION", BINARY_FILE_TYPE_PEI_DEPEX, BINARY_FILE_TYPE_GUID, BINARY_FILE_TYPE_SMM_DEPEX):
- self.__UndoToken()
+ self._UndoToken()
return False
if SectionName == "FV_IMAGE":
- FvImageSectionObj = FvImageSection.FvImageSection()
- if self.__IsKeyword("FV_IMAGE"):
+ FvImageSectionObj = FvImageSection()
+ if self._IsKeyword("FV_IMAGE"):
pass
- if self.__IsToken( "{"):
- FvObj = Fv.FV()
- self.__GetDefineStatements(FvObj)
- self.__GetBlockStatement(FvObj)
- self.__GetSetStatements(FvObj)
- self.__GetFvAlignment(FvObj)
- self.__GetFvAttributes(FvObj)
- self.__GetAprioriSection(FvObj)
- self.__GetAprioriSection(FvObj)
+ if self._IsToken("{"):
+ FvObj = FV()
+ self._GetDefineStatements(FvObj)
+ self._GetBlockStatement(FvObj)
+ self._GetSetStatements(FvObj)
+ self._GetFvAlignment(FvObj)
+ self._GetFvAttributes(FvObj)
+ self._GetAprioriSection(FvObj)
+ self._GetAprioriSection(FvObj)
while True:
- IsInf = self.__GetInfStatement(FvObj)
- IsFile = self.__GetFileStatement(FvObj)
+ IsInf = self._GetInfStatement(FvObj)
+ IsFile = self._GetFileStatement(FvObj)
if not IsInf and not IsFile:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
FvImageSectionObj.Fv = FvObj
FvImageSectionObj.FvName = None
else:
- if not self.__IsKeyword(BINARY_FILE_TYPE_FV):
- raise Warning("expected BINARY_FILE_TYPE_FV", self.FileName, self.CurrentLineNumber)
- FvImageSectionObj.FvFileType = self.__Token
+ if not self._IsKeyword(BINARY_FILE_TYPE_FV):
+ raise Warning("expected 'FV'", self.FileName, self.CurrentLineNumber)
+ FvImageSectionObj.FvFileType = self._Token
- if self.__GetAlignment():
- if self.__Token not in ("8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- FvImageSectionObj.Alignment = self.__Token
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENT_NOAUTO:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ FvImageSectionObj.Alignment = self._Token
- if self.__IsToken('|'):
- FvImageSectionObj.FvFileExtension = self.__GetFileExtension()
- elif self.__GetNextToken():
- if self.__Token not in ("}", "COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
+ if self._IsToken(TAB_VALUE_SPLIT):
+ FvImageSectionObj.FvFileExtension = self._GetFileExtension()
+ elif self._GetNextToken():
+ if self._Token not in ("}", "COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
BINARY_FILE_TYPE_UI, "VERSION", BINARY_FILE_TYPE_PEI_DEPEX, BINARY_FILE_TYPE_GUID, BINARY_FILE_TYPE_SMM_DEPEX):
- FvImageSectionObj.FvFileName = self.__Token
+ FvImageSectionObj.FvFileName = self._Token
else:
- self.__UndoToken()
+ self._UndoToken()
else:
raise Warning("expected FV file name", self.FileName, self.CurrentLineNumber)
Obj.SectionList.append(FvImageSectionObj)
return True
- EfiSectionObj = EfiSection.EfiSection()
+ EfiSectionObj = EfiSection()
EfiSectionObj.SectionType = SectionName
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected file type", self.FileName, self.CurrentLineNumber)
- if self.__Token == "STRING":
- if not self.__RuleSectionCouldHaveString(EfiSectionObj.SectionType):
+ if self._Token == "STRING":
+ if not self._RuleSectionCouldHaveString(EfiSectionObj.SectionType):
raise Warning("%s section could NOT have string data%d" % (EfiSectionObj.SectionType, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
- if not self.__IsToken('='):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Quoted String", self.FileName, self.CurrentLineNumber)
- if self.__GetStringData():
- EfiSectionObj.StringData = self.__Token
+ if self._GetStringData():
+ EfiSectionObj.StringData = self._Token
- if self.__IsKeyword("BUILD_NUM"):
- if not self.__RuleSectionCouldHaveBuildNum(EfiSectionObj.SectionType):
+ if self._IsKeyword("BUILD_NUM"):
+ if not self._RuleSectionCouldHaveBuildNum(EfiSectionObj.SectionType):
raise Warning("%s section could NOT have BUILD_NUM%d" % (EfiSectionObj.SectionType, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Build number", self.FileName, self.CurrentLineNumber)
- EfiSectionObj.BuildNum = self.__Token
+ EfiSectionObj.BuildNum = self._Token
else:
- EfiSectionObj.FileType = self.__Token
- self.__CheckRuleSectionFileType(EfiSectionObj.SectionType, EfiSectionObj.FileType)
+ EfiSectionObj.FileType = self._Token
+ self._CheckRuleSectionFileType(EfiSectionObj.SectionType, EfiSectionObj.FileType)
- if self.__IsKeyword("Optional"):
- if not self.__RuleSectionCouldBeOptional(EfiSectionObj.SectionType):
+ if self._IsKeyword("Optional"):
+ if not self._RuleSectionCouldBeOptional(EfiSectionObj.SectionType):
raise Warning("%s section could NOT be optional%d" % (EfiSectionObj.SectionType, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
EfiSectionObj.Optional = True
- if self.__IsKeyword("BUILD_NUM"):
- if not self.__RuleSectionCouldHaveBuildNum(EfiSectionObj.SectionType):
+ if self._IsKeyword("BUILD_NUM"):
+ if not self._RuleSectionCouldHaveBuildNum(EfiSectionObj.SectionType):
raise Warning("%s section could NOT have BUILD_NUM%d" % (EfiSectionObj.SectionType, self.CurrentLineNumber), self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Build number", self.FileName, self.CurrentLineNumber)
- EfiSectionObj.BuildNum = self.__Token
+ EfiSectionObj.BuildNum = self._Token
- if self.__GetAlignment():
- if self.__Token not in ("Auto", "8", "16", "32", "64", "128", "512", "1K", "4K", "32K", "64K", "128K",
- "256K", "512K", "1M", "2M", "4M", "8M", "16M"):
- raise Warning("Incorrect alignment '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- if self.__Token == 'Auto' and (not SectionName == BINARY_FILE_TYPE_PE32) and (not SectionName == BINARY_FILE_TYPE_TE):
+ if self._GetAlignment():
+ if self._Token not in ALIGNMENTS:
+ raise Warning("Incorrect alignment '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ if self._Token == 'Auto' and (not SectionName == BINARY_FILE_TYPE_PE32) and (not SectionName == BINARY_FILE_TYPE_TE):
raise Warning("Auto alignment can only be used in PE32 or TE section ", self.FileName, self.CurrentLineNumber)
- EfiSectionObj.Alignment = self.__Token
+ EfiSectionObj.Alignment = self._Token
- if self.__IsKeyword('RELOCS_STRIPPED') or self.__IsKeyword('RELOCS_RETAINED'):
- if self.__SectionCouldHaveRelocFlag(EfiSectionObj.SectionType):
- if self.__Token == 'RELOCS_STRIPPED':
+ if self._IsKeyword('RELOCS_STRIPPED') or self._IsKeyword('RELOCS_RETAINED'):
+ if self._SectionCouldHaveRelocFlag(EfiSectionObj.SectionType):
+ if self._Token == 'RELOCS_STRIPPED':
EfiSectionObj.KeepReloc = False
else:
EfiSectionObj.KeepReloc = True
@@ -3972,35 +3883,35 @@ class FdfParser:
raise Warning("Section type %s could not have reloc strip flag" % EfiSectionObj.SectionType, self.FileName, self.CurrentLineNumber)
- if self.__IsToken('|'):
- EfiSectionObj.FileExtension = self.__GetFileExtension()
- elif self.__GetNextToken():
- if self.__Token not in ("}", "COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
+ if self._IsToken(TAB_VALUE_SPLIT):
+ EfiSectionObj.FileExtension = self._GetFileExtension()
+ elif self._GetNextToken():
+ if self._Token not in ("}", "COMPAT16", BINARY_FILE_TYPE_PE32, BINARY_FILE_TYPE_PIC, BINARY_FILE_TYPE_TE, "FV_IMAGE", "RAW", BINARY_FILE_TYPE_DXE_DEPEX,\
BINARY_FILE_TYPE_UI, "VERSION", BINARY_FILE_TYPE_PEI_DEPEX, BINARY_FILE_TYPE_GUID, BINARY_FILE_TYPE_SMM_DEPEX):
- if self.__Token.startswith('PCD'):
- self.__UndoToken()
- self.__GetNextWord()
+ if self._Token.startswith('PCD'):
+ self._UndoToken()
+ self._GetNextWord()
- if self.__Token == 'PCD':
- if not self.__IsToken( "("):
+ if self._Token == 'PCD':
+ if not self._IsToken("("):
raise Warning("expected '('", self.FileName, self.CurrentLineNumber)
- PcdPair = self.__GetNextPcdSettings()
- if not self.__IsToken( ")"):
+ PcdPair = self._GetNextPcdSettings()
+ if not self._IsToken(")"):
raise Warning("expected ')'", self.FileName, self.CurrentLineNumber)
- self.__Token = 'PCD('+PcdPair[1]+'.'+PcdPair[0]+')'
+ self._Token = 'PCD('+PcdPair[1]+TAB_SPLIT+PcdPair[0]+')'
- EfiSectionObj.FileName = self.__Token
+ EfiSectionObj.FileName = self._Token
else:
- self.__UndoToken()
+ self._UndoToken()
else:
raise Warning("expected section file name", self.FileName, self.CurrentLineNumber)
Obj.SectionList.append(EfiSectionObj)
return True
- ## __RuleSectionCouldBeOptional() method
+ ## _RuleSectionCouldBeOptional() method
#
# Get whether a section could be optional
#
@@ -4009,13 +3920,13 @@ class FdfParser:
# @retval False section never optional
#
@staticmethod
- def __RuleSectionCouldBeOptional(SectionType):
+ def _RuleSectionCouldBeOptional(SectionType):
if SectionType in (BINARY_FILE_TYPE_DXE_DEPEX, BINARY_FILE_TYPE_UI, "VERSION", BINARY_FILE_TYPE_PEI_DEPEX, "RAW", BINARY_FILE_TYPE_SMM_DEPEX):
return True
else:
return False
- ## __RuleSectionCouldHaveBuildNum() method
+ ## _RuleSectionCouldHaveBuildNum() method
#
# Get whether a section could have build number information
#
@@ -4024,13 +3935,13 @@ class FdfParser:
# @retval False section never have build number information
#
@staticmethod
- def __RuleSectionCouldHaveBuildNum(SectionType):
+ def _RuleSectionCouldHaveBuildNum(SectionType):
if SectionType in ("VERSION"):
return True
else:
return False
- ## __RuleSectionCouldHaveString() method
+ ## _RuleSectionCouldHaveString() method
#
# Get whether a section could have string
#
@@ -4039,13 +3950,13 @@ class FdfParser:
# @retval False section never have string
#
@staticmethod
- def __RuleSectionCouldHaveString(SectionType):
+ def _RuleSectionCouldHaveString(SectionType):
if SectionType in (BINARY_FILE_TYPE_UI, "VERSION"):
return True
else:
return False
- ## __CheckRuleSectionFileType() method
+ ## _CheckRuleSectionFileType() method
#
# Get whether a section matches a file type
#
@@ -4053,7 +3964,7 @@ class FdfParser:
# @param SectionType The section type to check
# @param FileType The file type to check
#
- def __CheckRuleSectionFileType(self, SectionType, FileType):
+ def _CheckRuleSectionFileType(self, SectionType, FileType):
if SectionType == "COMPAT16":
if FileType not in ("COMPAT16", "SEC_COMPAT16"):
raise Warning("Incorrect section file type '%s'" % FileType, self.FileName, self.CurrentLineNumber)
@@ -4085,53 +3996,53 @@ class FdfParser:
if FileType not in (BINARY_FILE_TYPE_PE32, "SEC_GUID"):
raise Warning("Incorrect section file type '%s'" % FileType, self.FileName, self.CurrentLineNumber)
- ## __GetRuleEncapsulationSection() method
+ ## _GetRuleEncapsulationSection() method
#
# Get encapsulation section for Rule
#
# @param self The object pointer
- # @param Rule for whom section is got
+ # @param theRule for whom section is got
# @retval True Successfully find section statement
# @retval False Not able to find section statement
#
- def __GetRuleEncapsulationSection(self, Rule):
- if self.__IsKeyword( "COMPRESS"):
+ def _GetRuleEncapsulationSection(self, theRule):
+ if self._IsKeyword("COMPRESS"):
Type = "PI_STD"
- if self.__IsKeyword("PI_STD") or self.__IsKeyword("PI_NONE"):
- Type = self.__Token
+ if self._IsKeyword("PI_STD") or self._IsKeyword("PI_NONE"):
+ Type = self._Token
- if not self.__IsToken("{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- CompressSectionObj = CompressSection.CompressSection()
+ CompressSectionObj = CompressSection()
CompressSectionObj.CompType = Type
# Recursive sections...
while True:
- IsEncapsulate = self.__GetRuleEncapsulationSection(CompressSectionObj)
- IsLeaf = self.__GetEfiSection(CompressSectionObj)
+ IsEncapsulate = self._GetRuleEncapsulationSection(CompressSectionObj)
+ IsLeaf = self._GetEfiSection(CompressSectionObj)
if not IsEncapsulate and not IsLeaf:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- Rule.SectionList.append(CompressSectionObj)
+ theRule.SectionList.append(CompressSectionObj)
return True
- elif self.__IsKeyword( "GUIDED"):
+ elif self._IsKeyword("GUIDED"):
GuidValue = None
- if self.__GetNextGuid():
- GuidValue = self.__Token
+ if self._GetNextGuid():
+ GuidValue = self._Token
- if self.__IsKeyword( "$(NAMED_GUID)"):
- GuidValue = self.__Token
+ if self._IsKeyword("$(NAMED_GUID)"):
+ GuidValue = self._Token
- AttribDict = self.__GetGuidAttrib()
+ AttribDict = self._GetGuidAttrib()
- if not self.__IsToken("{"):
+ if not self._IsToken("{"):
raise Warning("expected '{'", self.FileName, self.CurrentLineNumber)
- GuidSectionObj = GuidSection.GuidSection()
+ GuidSectionObj = GuidSection()
GuidSectionObj.NameGuid = GuidValue
GuidSectionObj.SectionType = "GUIDED"
GuidSectionObj.ProcessRequired = AttribDict["PROCESSING_REQUIRED"]
@@ -4140,20 +4051,20 @@ class FdfParser:
# Efi sections...
while True:
- IsEncapsulate = self.__GetRuleEncapsulationSection(GuidSectionObj)
- IsLeaf = self.__GetEfiSection(GuidSectionObj)
+ IsEncapsulate = self._GetRuleEncapsulationSection(GuidSectionObj)
+ IsLeaf = self._GetEfiSection(GuidSectionObj)
if not IsEncapsulate and not IsLeaf:
break
- if not self.__IsToken( "}"):
+ if not self._IsToken("}"):
raise Warning("expected '}'", self.FileName, self.CurrentLineNumber)
- Rule.SectionList.append(GuidSectionObj)
+ theRule.SectionList.append(GuidSectionObj)
return True
return False
- ## __GetVtf() method
+ ## _GetVtf() method
#
# Get VTF section contents and store its data into VTF list of self.Profile
#
@@ -4161,69 +4072,69 @@ class FdfParser:
# @retval True Successfully find a VTF
# @retval False Not able to find a VTF
#
- def __GetVtf(self):
- if not self.__GetNextToken():
+ def _GetVtf(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[VTF."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[VTF."):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[VTF.", True):
+ self._UndoToken()
+ if not self._IsToken("[VTF.", True):
FileLineTuple = GetRealFileLine(self.FileName, self.CurrentLineNumber)
#print 'Parsing String: %s in File %s, At line: %d, Offset Within Line: %d' \
- # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine :], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
+ # % (self.Profile.FileLinesList[self.CurrentLineNumber - 1][self.CurrentOffsetWithinLine:], FileLineTuple[0], FileLineTuple[1], self.CurrentOffsetWithinLine)
raise Warning("expected [VTF.]", self.FileName, self.CurrentLineNumber)
- if not self.__SkipToToken("."):
+ if not self._SkipToToken(TAB_SPLIT):
raise Warning("expected '.'", self.FileName, self.CurrentLineNumber)
- Arch = self.__SkippedChars.rstrip(".").upper()
- if Arch not in ("IA32", "X64", "IPF", "ARM", "AARCH64"):
+ Arch = self._SkippedChars.rstrip(TAB_SPLIT).upper()
+ if Arch not in {TAB_ARCH_IA32, TAB_ARCH_X64, TAB_ARCH_IPF, TAB_ARCH_ARM, TAB_ARCH_AARCH64}:
raise Warning("Unknown Arch '%s'" % Arch, self.FileName, self.CurrentLineNumber)
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected VTF name", self.FileName, self.CurrentLineNumber)
- Name = self.__Token.upper()
+ Name = self._Token.upper()
- VtfObj = Vtf.Vtf()
+ VtfObj = Vtf()
VtfObj.UiName = Name
VtfObj.KeyArch = Arch
- if self.__IsToken(","):
- if not self.__GetNextWord():
+ if self._IsToken(TAB_COMMA_SPLIT):
+ if not self._GetNextWord():
raise Warning("expected Arch list", self.FileName, self.CurrentLineNumber)
- if self.__Token.upper() not in ("IA32", "X64", "IPF", "ARM", "AARCH64"):
- raise Warning("Unknown Arch '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- VtfObj.ArchList = self.__Token.upper()
+ if self._Token.upper() not in (TAB_ARCH_IA32, TAB_ARCH_X64, TAB_ARCH_IPF, TAB_ARCH_ARM, TAB_ARCH_AARCH64):
+ raise Warning("Unknown Arch '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ VtfObj.ArchList = self._Token.upper()
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- if self.__IsKeyword("IA32_RST_BIN"):
- if not self.__IsToken("="):
+ if self._IsKeyword("IA32_RST_BIN"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Reset file", self.FileName, self.CurrentLineNumber)
- VtfObj.ResetBin = self.__Token
- if VtfObj.ResetBin.replace('$(WORKSPACE)', '').find('$') == -1:
+ VtfObj.ResetBin = self._Token
+ if VtfObj.ResetBin.replace(TAB_WORKSPACE, '').find('$') == -1:
#check for file path
ErrorCode, ErrorInfo = PathClass(NormPath(VtfObj.ResetBin), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
EdkLogger.error("GenFds", ErrorCode, ExtraData=ErrorInfo)
- while self.__GetComponentStatement(VtfObj):
+ while self._GetComponentStatement(VtfObj):
pass
self.Profile.VtfList.append(VtfObj)
return True
- ## __GetComponentStatement() method
+ ## _GetComponentStatement() method
#
# Get components in VTF
#
@@ -4232,132 +4143,132 @@ class FdfParser:
# @retval True Successfully find a component
# @retval False Not able to find a component
#
- def __GetComponentStatement(self, VtfObj):
- if not self.__IsKeyword("COMP_NAME"):
+ def _GetComponentStatement(self, VtfObj):
+ if not self._IsKeyword("COMP_NAME"):
return False
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextWord():
+ if not self._GetNextWord():
raise Warning("expected Component Name", self.FileName, self.CurrentLineNumber)
- CompStatementObj = ComponentStatement.ComponentStatement()
- CompStatementObj.CompName = self.__Token
+ CompStatementObj = ComponentStatement()
+ CompStatementObj.CompName = self._Token
- if not self.__IsKeyword("COMP_LOC"):
+ if not self._IsKeyword("COMP_LOC"):
raise Warning("expected COMP_LOC", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
CompStatementObj.CompLoc = ""
- if self.__GetNextWord():
- CompStatementObj.CompLoc = self.__Token
- if self.__IsToken('|'):
- if not self.__GetNextWord():
+ if self._GetNextWord():
+ CompStatementObj.CompLoc = self._Token
+ if self._IsToken(TAB_VALUE_SPLIT):
+ if not self._GetNextWord():
raise Warning("Expected Region Name", self.FileName, self.CurrentLineNumber)
- if self.__Token not in ("F", "N", "S"): #, "H", "L", "PH", "PL"): not support
- raise Warning("Unknown location type '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ if self._Token not in ("F", "N", "S"): #, "H", "L", "PH", "PL"): not support
+ raise Warning("Unknown location type '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
- CompStatementObj.FilePos = self.__Token
+ CompStatementObj.FilePos = self._Token
else:
self.CurrentLineNumber += 1
self.CurrentOffsetWithinLine = 0
- if not self.__IsKeyword("COMP_TYPE"):
+ if not self._IsKeyword("COMP_TYPE"):
raise Warning("expected COMP_TYPE", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Component type", self.FileName, self.CurrentLineNumber)
- if self.__Token not in ("FIT", "PAL_B", "PAL_A", "OEM"):
- if not self.__Token.startswith("0x") or len(self.__Token) < 3 or len(self.__Token) > 4 or \
- not self.__Token[2] in string.hexdigits or not self.__Token[-1] in string.hexdigits:
- raise Warning("Unknown location type '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- CompStatementObj.CompType = self.__Token
+ if self._Token not in ("FIT", "PAL_B", "PAL_A", "OEM"):
+ if not self._Token.startswith("0x") or len(self._Token) < 3 or len(self._Token) > 4 or \
+ not self._Token[2] in hexdigits or not self._Token[-1] in hexdigits:
+ raise Warning("Unknown location type '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ CompStatementObj.CompType = self._Token
- if not self.__IsKeyword("COMP_VER"):
+ if not self._IsKeyword("COMP_VER"):
raise Warning("expected COMP_VER", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Component version", self.FileName, self.CurrentLineNumber)
- Pattern = re.compile('-$|[0-9a-fA-F]{1,2}\.[0-9a-fA-F]{1,2}$', re.DOTALL)
- if Pattern.match(self.__Token) is None:
- raise Warning("Unknown version format '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- CompStatementObj.CompVer = self.__Token
+ Pattern = compile('-$|[0-9a-fA-F]{1,2}\.[0-9a-fA-F]{1,2}$', DOTALL)
+ if Pattern.match(self._Token) is None:
+ raise Warning("Unknown version format '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ CompStatementObj.CompVer = self._Token
- if not self.__IsKeyword("COMP_CS"):
+ if not self._IsKeyword("COMP_CS"):
raise Warning("expected COMP_CS", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Component CS", self.FileName, self.CurrentLineNumber)
- if self.__Token not in ("1", "0"):
- raise Warning("Unknown Component CS '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
- CompStatementObj.CompCs = self.__Token
+ if self._Token not in ("1", "0"):
+ raise Warning("Unknown Component CS '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
+ CompStatementObj.CompCs = self._Token
- if not self.__IsKeyword("COMP_BIN"):
+ if not self._IsKeyword("COMP_BIN"):
raise Warning("expected COMP_BIN", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Component file", self.FileName, self.CurrentLineNumber)
- CompStatementObj.CompBin = self.__Token
- if CompStatementObj.CompBin != '-' and CompStatementObj.CompBin.replace('$(WORKSPACE)', '').find('$') == -1:
+ CompStatementObj.CompBin = self._Token
+ if CompStatementObj.CompBin != '-' and CompStatementObj.CompBin.replace(TAB_WORKSPACE, '').find('$') == -1:
#check for file path
ErrorCode, ErrorInfo = PathClass(NormPath(CompStatementObj.CompBin), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
EdkLogger.error("GenFds", ErrorCode, ExtraData=ErrorInfo)
- if not self.__IsKeyword("COMP_SYM"):
+ if not self._IsKeyword("COMP_SYM"):
raise Warning("expected COMP_SYM", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected Component symbol file", self.FileName, self.CurrentLineNumber)
- CompStatementObj.CompSym = self.__Token
- if CompStatementObj.CompSym != '-' and CompStatementObj.CompSym.replace('$(WORKSPACE)', '').find('$') == -1:
+ CompStatementObj.CompSym = self._Token
+ if CompStatementObj.CompSym != '-' and CompStatementObj.CompSym.replace(TAB_WORKSPACE, '').find('$') == -1:
#check for file path
ErrorCode, ErrorInfo = PathClass(NormPath(CompStatementObj.CompSym), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
EdkLogger.error("GenFds", ErrorCode, ExtraData=ErrorInfo)
- if not self.__IsKeyword("COMP_SIZE"):
+ if not self._IsKeyword("COMP_SIZE"):
raise Warning("expected COMP_SIZE", self.FileName, self.CurrentLineNumber)
- if not self.__IsToken("="):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if self.__IsToken("-"):
- CompStatementObj.CompSize = self.__Token
- elif self.__GetNextDecimalNumber():
- CompStatementObj.CompSize = self.__Token
- elif self.__GetNextHexNumber():
- CompStatementObj.CompSize = self.__Token
+ if self._IsToken("-"):
+ CompStatementObj.CompSize = self._Token
+ elif self._GetNextDecimalNumber():
+ CompStatementObj.CompSize = self._Token
+ elif self._GetNextHexNumber():
+ CompStatementObj.CompSize = self._Token
else:
- raise Warning("Unknown size '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ raise Warning("Unknown size '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
VtfObj.ComponentStatementList.append(CompStatementObj)
return True
- ## __GetOptionRom() method
+ ## _GetOptionRom() method
#
# Get OptionROM section contents and store its data into OptionROM list of self.Profile
#
@@ -4365,38 +4276,37 @@ class FdfParser:
# @retval True Successfully find a OptionROM
# @retval False Not able to find a OptionROM
#
- def __GetOptionRom(self):
- if not self.__GetNextToken():
+ def _GetOptionRom(self):
+ if not self._GetNextToken():
return False
- S = self.__Token.upper()
- if S.startswith("[") and not S.startswith("[OPTIONROM."):
+ S = self._Token.upper()
+ if S.startswith(TAB_SECTION_START) and not S.startswith("[OPTIONROM."):
self.SectionParser(S)
- self.__UndoToken()
+ self._UndoToken()
return False
- self.__UndoToken()
- if not self.__IsToken("[OptionRom.", True):
- raise Warning("Unknown Keyword '%s'" % self.__Token, self.FileName, self.CurrentLineNumber)
+ self._UndoToken()
+ if not self._IsToken("[OptionRom.", True):
+ raise Warning("Unknown Keyword '%s'" % self._Token, self.FileName, self.CurrentLineNumber)
- OptRomName = self.__GetUiName()
+ OptRomName = self._GetUiName()
- if not self.__IsToken( "]"):
+ if not self._IsToken(TAB_SECTION_END):
raise Warning("expected ']'", self.FileName, self.CurrentLineNumber)
- OptRomObj = OptionRom.OPTIONROM()
- OptRomObj.DriverName = OptRomName
+ OptRomObj = OPTIONROM(OptRomName)
self.Profile.OptRomDict[OptRomName] = OptRomObj
while True:
- isInf = self.__GetOptRomInfStatement(OptRomObj)
- isFile = self.__GetOptRomFileStatement(OptRomObj)
+ isInf = self._GetOptRomInfStatement(OptRomObj)
+ isFile = self._GetOptRomFileStatement(OptRomObj)
if not isInf and not isFile:
break
return True
- ## __GetOptRomInfStatement() method
+ ## _GetOptRomInfStatement() method
#
# Get INF statements
#
@@ -4405,17 +4315,17 @@ class FdfParser:
# @retval True Successfully find inf statement
# @retval False Not able to find inf statement
#
- def __GetOptRomInfStatement(self, Obj):
- if not self.__IsKeyword( "INF"):
+ def _GetOptRomInfStatement(self, Obj):
+ if not self._IsKeyword("INF"):
return False
- ffsInf = OptRomInfStatement.OptRomInfStatement()
- self.__GetInfOptions( ffsInf)
+ ffsInf = OptRomInfStatement()
+ self._GetInfOptions(ffsInf)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected INF file path", self.FileName, self.CurrentLineNumber)
- ffsInf.InfFileName = self.__Token
- if ffsInf.InfFileName.replace('$(WORKSPACE)', '').find('$') == -1:
+ ffsInf.InfFileName = self._Token
+ if ffsInf.InfFileName.replace(TAB_WORKSPACE, '').find('$') == -1:
#check for file path
ErrorCode, ErrorInfo = PathClass(NormPath(ffsInf.InfFileName), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
@@ -4434,71 +4344,71 @@ class FdfParser:
self.Profile.InfDict['ArchTBD'].append(ffsInf.InfFileName)
- self.__GetOptRomOverrides (ffsInf)
+ self._GetOptRomOverrides (ffsInf)
Obj.FfsList.append(ffsInf)
return True
- ## __GetOptRomOverrides() method
+ ## _GetOptRomOverrides() method
#
# Get overrides for OptROM INF & FILE
#
# @param self The object pointer
# @param FfsInfObj for whom overrides is got
#
- def __GetOptRomOverrides(self, Obj):
- if self.__IsToken('{'):
- Overrides = OptRomInfStatement.OverrideAttribs()
+ def _GetOptRomOverrides(self, Obj):
+ if self._IsToken('{'):
+ Overrides = OverrideAttribs()
while True:
- if self.__IsKeyword( "PCI_VENDOR_ID"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("PCI_VENDOR_ID"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex vendor id", self.FileName, self.CurrentLineNumber)
- Overrides.PciVendorId = self.__Token
+ Overrides.PciVendorId = self._Token
continue
- if self.__IsKeyword( "PCI_CLASS_CODE"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("PCI_CLASS_CODE"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex class code", self.FileName, self.CurrentLineNumber)
- Overrides.PciClassCode = self.__Token
+ Overrides.PciClassCode = self._Token
continue
- if self.__IsKeyword( "PCI_DEVICE_ID"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("PCI_DEVICE_ID"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex device id", self.FileName, self.CurrentLineNumber)
- Overrides.PciDeviceId = self.__Token
+ Overrides.PciDeviceId = self._Token
continue
- if self.__IsKeyword( "PCI_REVISION"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("PCI_REVISION"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextHexNumber():
+ if not self._GetNextHexNumber():
raise Warning("expected Hex revision", self.FileName, self.CurrentLineNumber)
- Overrides.PciRevision = self.__Token
+ Overrides.PciRevision = self._Token
continue
- if self.__IsKeyword( "PCI_COMPRESS"):
- if not self.__IsToken( "="):
+ if self._IsKeyword("PCI_COMPRESS"):
+ if not self._IsToken(TAB_EQUAL_SPLIT):
raise Warning("expected '='", self.FileName, self.CurrentLineNumber)
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected TRUE/FALSE for compress", self.FileName, self.CurrentLineNumber)
- Overrides.NeedCompress = self.__Token.upper() == 'TRUE'
+ Overrides.NeedCompress = self._Token.upper() == 'TRUE'
continue
- if self.__IsToken( "}"):
+ if self._IsToken("}"):
break
else:
EdkLogger.error("FdfParser", FORMAT_INVALID, File=self.FileName, Line=self.CurrentLineNumber)
Obj.OverrideAttribs = Overrides
- ## __GetOptRomFileStatement() method
+ ## _GetOptRomFileStatement() method
#
# Get FILE statements
#
@@ -4507,33 +4417,33 @@ class FdfParser:
# @retval True Successfully find FILE statement
# @retval False Not able to find FILE statement
#
- def __GetOptRomFileStatement(self, Obj):
- if not self.__IsKeyword( "FILE"):
+ def _GetOptRomFileStatement(self, Obj):
+ if not self._IsKeyword("FILE"):
return False
- FfsFileObj = OptRomFileStatement.OptRomFileStatement()
+ FfsFileObj = OptRomFileStatement()
- if not self.__IsKeyword("EFI") and not self.__IsKeyword(BINARY_FILE_TYPE_BIN):
+ if not self._IsKeyword("EFI") and not self._IsKeyword(BINARY_FILE_TYPE_BIN):
raise Warning("expected Binary type (EFI/BIN)", self.FileName, self.CurrentLineNumber)
- FfsFileObj.FileType = self.__Token
+ FfsFileObj.FileType = self._Token
- if not self.__GetNextToken():
+ if not self._GetNextToken():
raise Warning("expected File path", self.FileName, self.CurrentLineNumber)
- FfsFileObj.FileName = self.__Token
- if FfsFileObj.FileName.replace('$(WORKSPACE)', '').find('$') == -1:
+ FfsFileObj.FileName = self._Token
+ if FfsFileObj.FileName.replace(TAB_WORKSPACE, '').find('$') == -1:
#check for file path
ErrorCode, ErrorInfo = PathClass(NormPath(FfsFileObj.FileName), GenFdsGlobalVariable.WorkSpaceDir).Validate()
if ErrorCode != 0:
EdkLogger.error("GenFds", ErrorCode, ExtraData=ErrorInfo)
if FfsFileObj.FileType == 'EFI':
- self.__GetOptRomOverrides(FfsFileObj)
+ self._GetOptRomOverrides(FfsFileObj)
Obj.FfsList.append(FfsFileObj)
return True
- ## __GetCapInFd() method
+ ## _GetCapInFd() method
#
# Get Cap list contained in FD
#
@@ -4541,7 +4451,7 @@ class FdfParser:
# @param FdName FD name
# @retval CapList List of Capsule in FD
#
- def __GetCapInFd (self, FdName):
+ def _GetCapInFd (self, FdName):
CapList = []
if FdName.upper() in self.Profile.FdDict:
FdObj = self.Profile.FdDict[FdName.upper()]
@@ -4554,7 +4464,7 @@ class FdfParser:
CapList.append(elementRegionData.upper())
return CapList
- ## __GetReferencedFdCapTuple() method
+ ## _GetReferencedFdCapTuple() method
#
# Get FV and FD list referenced by a capsule image
#
@@ -4563,22 +4473,22 @@ class FdfParser:
# @param RefFdList referenced FD by section
# @param RefFvList referenced FV by section
#
- def __GetReferencedFdCapTuple(self, CapObj, RefFdList = [], RefFvList = []):
- for CapsuleDataObj in CapObj.CapsuleDataList :
+ def _GetReferencedFdCapTuple(self, CapObj, RefFdList = [], RefFvList = []):
+ for CapsuleDataObj in CapObj.CapsuleDataList:
if hasattr(CapsuleDataObj, 'FvName') and CapsuleDataObj.FvName is not None and CapsuleDataObj.FvName.upper() not in RefFvList:
RefFvList.append (CapsuleDataObj.FvName.upper())
elif hasattr(CapsuleDataObj, 'FdName') and CapsuleDataObj.FdName is not None and CapsuleDataObj.FdName.upper() not in RefFdList:
RefFdList.append (CapsuleDataObj.FdName.upper())
elif CapsuleDataObj.Ffs is not None:
- if isinstance(CapsuleDataObj.Ffs, FfsFileStatement.FileStatement):
+ if isinstance(CapsuleDataObj.Ffs, FileStatement):
if CapsuleDataObj.Ffs.FvName is not None and CapsuleDataObj.Ffs.FvName.upper() not in RefFvList:
RefFvList.append(CapsuleDataObj.Ffs.FvName.upper())
elif CapsuleDataObj.Ffs.FdName is not None and CapsuleDataObj.Ffs.FdName.upper() not in RefFdList:
RefFdList.append(CapsuleDataObj.Ffs.FdName.upper())
else:
- self.__GetReferencedFdFvTupleFromSection(CapsuleDataObj.Ffs, RefFdList, RefFvList)
+ self._GetReferencedFdFvTupleFromSection(CapsuleDataObj.Ffs, RefFdList, RefFvList)
- ## __GetFvInFd() method
+ ## _GetFvInFd() method
#
# Get FV list contained in FD
#
@@ -4586,7 +4496,7 @@ class FdfParser:
# @param FdName FD name
# @retval FvList list of FV in FD
#
- def __GetFvInFd (self, FdName):
+ def _GetFvInFd (self, FdName):
FvList = []
if FdName.upper() in self.Profile.FdDict:
FdObj = self.Profile.FdDict[FdName.upper()]
@@ -4599,7 +4509,7 @@ class FdfParser:
FvList.append(elementRegionData.upper())
return FvList
- ## __GetReferencedFdFvTuple() method
+ ## _GetReferencedFdFvTuple() method
#
# Get FD and FV list referenced by a FFS file
#
@@ -4608,17 +4518,17 @@ class FdfParser:
# @param RefFdList referenced FD by section
# @param RefFvList referenced FV by section
#
- def __GetReferencedFdFvTuple(self, FvObj, RefFdList = [], RefFvList = []):
+ def _GetReferencedFdFvTuple(self, FvObj, RefFdList = [], RefFvList = []):
for FfsObj in FvObj.FfsList:
- if isinstance(FfsObj, FfsFileStatement.FileStatement):
+ if isinstance(FfsObj, FileStatement):
if FfsObj.FvName is not None and FfsObj.FvName.upper() not in RefFvList:
RefFvList.append(FfsObj.FvName.upper())
elif FfsObj.FdName is not None and FfsObj.FdName.upper() not in RefFdList:
RefFdList.append(FfsObj.FdName.upper())
else:
- self.__GetReferencedFdFvTupleFromSection(FfsObj, RefFdList, RefFvList)
+ self._GetReferencedFdFvTupleFromSection(FfsObj, RefFdList, RefFvList)
- ## __GetReferencedFdFvTupleFromSection() method
+ ## _GetReferencedFdFvTupleFromSection() method
#
# Get FD and FV list referenced by a FFS section
#
@@ -4627,19 +4537,19 @@ class FdfParser:
# @param FdList referenced FD by section
# @param FvList referenced FV by section
#
- def __GetReferencedFdFvTupleFromSection(self, FfsFile, FdList = [], FvList = []):
+ def _GetReferencedFdFvTupleFromSection(self, FfsFile, FdList = [], FvList = []):
SectionStack = []
SectionStack.extend(FfsFile.SectionList)
while SectionStack != []:
SectionObj = SectionStack.pop()
- if isinstance(SectionObj, FvImageSection.FvImageSection):
+ if isinstance(SectionObj, FvImageSection):
if SectionObj.FvName is not None and SectionObj.FvName.upper() not in FvList:
FvList.append(SectionObj.FvName.upper())
if SectionObj.Fv is not None and SectionObj.Fv.UiFvName is not None and SectionObj.Fv.UiFvName.upper() not in FvList:
FvList.append(SectionObj.Fv.UiFvName.upper())
- self.__GetReferencedFdFvTuple(SectionObj.Fv, FdList, FvList)
+ self._GetReferencedFdFvTuple(SectionObj.Fv, FdList, FvList)
- if isinstance(SectionObj, CompressSection.CompressSection) or isinstance(SectionObj, GuidSection.GuidSection):
+ if isinstance(SectionObj, CompressSection) or isinstance(SectionObj, GuidSection):
SectionStack.extend(SectionObj.SectionList)
## CycleReferenceCheck() method
@@ -4672,14 +4582,14 @@ class FdfParser:
RefFdList = []
RefFvList = []
- self.__GetReferencedFdFvTuple(FvObj, RefFdList, RefFvList)
+ self._GetReferencedFdFvTuple(FvObj, RefFdList, RefFvList)
for RefFdName in RefFdList:
if RefFdName in FdAnalyzedList:
continue
LogStr += "FV %s contains FD %s\n" % (FvNameFromStack, RefFdName)
- FvInFdList = self.__GetFvInFd(RefFdName)
+ FvInFdList = self._GetFvInFd(RefFdName)
if FvInFdList != []:
for FvNameInFd in FvInFdList:
LogStr += "FD %s contains FV %s\n" % (RefFdName, FvNameInFd)
@@ -4725,7 +4635,7 @@ class FdfParser:
RefFvList = []
RefFdList = []
- self.__GetReferencedFdCapTuple(CapObj, RefFdList, RefFvList)
+ self._GetReferencedFdCapTuple(CapObj, RefFdList, RefFvList)
FvListLength = 0
FdListLength = 0
@@ -4735,7 +4645,7 @@ class FdfParser:
continue
LogStr += "Capsule %s contains FD %s\n" % (CapNameFromStack, RefFdName)
- CapInFdList = self.__GetCapInFd(RefFdName)
+ CapInFdList = self._GetCapInFd(RefFdName)
if CapInFdList != []:
for CapNameInFd in CapInFdList:
LogStr += "FD %s contains Capsule %s\n" % (RefFdName, CapNameInFd)
@@ -4746,7 +4656,7 @@ class FdfParser:
EdkLogger.info(LogStr)
return True
- FvInFdList = self.__GetFvInFd(RefFdName)
+ FvInFdList = self._GetFvInFd(RefFdName)
if FvInFdList != []:
for FvNameInFd in FvInFdList:
LogStr += "FD %s contains FV %s\n" % (RefFdName, FvNameInFd)
@@ -4767,7 +4677,7 @@ class FdfParser:
FvObj = self.Profile.FvDict[RefFvName.upper()]
else:
continue
- self.__GetReferencedFdFvTuple(FvObj, RefFdList, RefFvList)
+ self._GetReferencedFdFvTuple(FvObj, RefFdList, RefFvList)
FvAnalyzedList.append(RefFvName)
return False
diff --git a/BaseTools/Source/Python/GenFds/Ffs.py b/BaseTools/Source/Python/GenFds/Ffs.py
index 598b2d023124..4c1b43572964 100644
--- a/BaseTools/Source/Python/GenFds/Ffs.py
+++ b/BaseTools/Source/Python/GenFds/Ffs.py
@@ -15,53 +15,41 @@
##
# Import Modules
#
-from CommonDataClass.FdfClass import FDClassObject
from Common.DataType import *
-## generate FFS
-#
-#
-class Ffs(FDClassObject):
- # mapping between FILE type in FDF and file type for GenFfs
- FdfFvFileTypeToFileType = {
- SUP_MODULE_SEC : 'EFI_FV_FILETYPE_SECURITY_CORE',
- SUP_MODULE_PEI_CORE : 'EFI_FV_FILETYPE_PEI_CORE',
- SUP_MODULE_PEIM : 'EFI_FV_FILETYPE_PEIM',
- SUP_MODULE_DXE_CORE : 'EFI_FV_FILETYPE_DXE_CORE',
- 'FREEFORM' : 'EFI_FV_FILETYPE_FREEFORM',
- 'DRIVER' : 'EFI_FV_FILETYPE_DRIVER',
- 'APPLICATION' : 'EFI_FV_FILETYPE_APPLICATION',
- 'FV_IMAGE' : 'EFI_FV_FILETYPE_FIRMWARE_VOLUME_IMAGE',
- 'RAW' : 'EFI_FV_FILETYPE_RAW',
- 'PEI_DXE_COMBO' : 'EFI_FV_FILETYPE_COMBINED_PEIM_DRIVER',
- 'SMM' : 'EFI_FV_FILETYPE_SMM',
- SUP_MODULE_SMM_CORE : 'EFI_FV_FILETYPE_SMM_CORE',
- SUP_MODULE_MM_STANDALONE : 'EFI_FV_FILETYPE_MM_STANDALONE',
- SUP_MODULE_MM_CORE_STANDALONE : 'EFI_FV_FILETYPE_MM_CORE_STANDALONE'
- }
+# mapping between FILE type in FDF and file type for GenFfs
+FdfFvFileTypeToFileType = {
+ SUP_MODULE_SEC : 'EFI_FV_FILETYPE_SECURITY_CORE',
+ SUP_MODULE_PEI_CORE : 'EFI_FV_FILETYPE_PEI_CORE',
+ SUP_MODULE_PEIM : 'EFI_FV_FILETYPE_PEIM',
+ SUP_MODULE_DXE_CORE : 'EFI_FV_FILETYPE_DXE_CORE',
+ 'FREEFORM' : 'EFI_FV_FILETYPE_FREEFORM',
+ 'DRIVER' : 'EFI_FV_FILETYPE_DRIVER',
+ 'APPLICATION' : 'EFI_FV_FILETYPE_APPLICATION',
+ 'FV_IMAGE' : 'EFI_FV_FILETYPE_FIRMWARE_VOLUME_IMAGE',
+ 'RAW' : 'EFI_FV_FILETYPE_RAW',
+ 'PEI_DXE_COMBO' : 'EFI_FV_FILETYPE_COMBINED_PEIM_DRIVER',
+ 'SMM' : 'EFI_FV_FILETYPE_SMM',
+ SUP_MODULE_SMM_CORE : 'EFI_FV_FILETYPE_SMM_CORE',
+ SUP_MODULE_MM_STANDALONE : 'EFI_FV_FILETYPE_MM_STANDALONE',
+ SUP_MODULE_MM_CORE_STANDALONE : 'EFI_FV_FILETYPE_MM_CORE_STANDALONE'
+}
- # mapping between section type in FDF and file suffix
- SectionSuffix = {
- BINARY_FILE_TYPE_PE32 : '.pe32',
- BINARY_FILE_TYPE_PIC : '.pic',
- BINARY_FILE_TYPE_TE : '.te',
- BINARY_FILE_TYPE_DXE_DEPEX : '.dpx',
- 'VERSION' : '.ver',
- BINARY_FILE_TYPE_UI : '.ui',
- 'COMPAT16' : '.com16',
- 'RAW' : '.raw',
- 'FREEFORM_SUBTYPE_GUID': '.guid',
- 'SUBTYPE_GUID' : '.guid',
- 'FV_IMAGE' : 'fv.sec',
- 'COMPRESS' : '.com',
- 'GUIDED' : '.guided',
- BINARY_FILE_TYPE_PEI_DEPEX : '.dpx',
- BINARY_FILE_TYPE_SMM_DEPEX : '.dpx'
- }
-
- ## The constructor
- #
- # @param self The object pointer
- #
- def __init__(self):
- FfsClassObject.__init__(self)
+# mapping between section type in FDF and file suffix
+SectionSuffix = {
+ BINARY_FILE_TYPE_PE32 : '.pe32',
+ BINARY_FILE_TYPE_PIC : '.pic',
+ BINARY_FILE_TYPE_TE : '.te',
+ BINARY_FILE_TYPE_DXE_DEPEX : '.dpx',
+ 'VERSION' : '.ver',
+ BINARY_FILE_TYPE_UI : '.ui',
+ 'COMPAT16' : '.com16',
+ 'RAW' : '.raw',
+ 'FREEFORM_SUBTYPE_GUID': '.guid',
+ 'SUBTYPE_GUID' : '.guid',
+ 'FV_IMAGE' : 'fv.sec',
+ 'COMPRESS' : '.com',
+ 'GUIDED' : '.guided',
+ BINARY_FILE_TYPE_PEI_DEPEX : '.dpx',
+ BINARY_FILE_TYPE_SMM_DEPEX : '.dpx'
+}
diff --git a/BaseTools/Source/Python/GenFds/FfsFileStatement.py b/BaseTools/Source/Python/GenFds/FfsFileStatement.py
index 4c35aac76a00..7479efff04c7 100644
--- a/BaseTools/Source/Python/GenFds/FfsFileStatement.py
+++ b/BaseTools/Source/Python/GenFds/FfsFileStatement.py
@@ -16,26 +16,22 @@
# Import Modules
#
from __future__ import absolute_import
-from . import Ffs
-from . import Rule
-import Common.LongFilePathOs as os
from io import BytesIO
-import subprocess
-
-from .GenFdsGlobalVariable import GenFdsGlobalVariable
+from struct import pack
from CommonDataClass.FdfClass import FileStatementClassObject
from Common import EdkLogger
-from Common.BuildToolError import *
-from Common.Misc import GuidStructureByteArrayToGuidString
+from Common.BuildToolError import GENFDS_ERROR
+from Common.Misc import GuidStructureByteArrayToGuidString, SaveFileOnChange
+import Common.LongFilePathOs as os
from .GuidSection import GuidSection
from .FvImageSection import FvImageSection
-from Common.Misc import SaveFileOnChange
-from struct import *
+from .Ffs import FdfFvFileTypeToFileType
+from .GenFdsGlobalVariable import GenFdsGlobalVariable
## generate FFS from FILE
#
#
-class FileStatement (FileStatementClassObject) :
+class FileStatement (FileStatementClassObject):
## The constructor
#
# @param self The object pointer
@@ -60,7 +56,7 @@ class FileStatement (FileStatementClassObject) :
#
def GenFfs(self, Dict = {}, FvChildAddr=[], FvParentAddr=None, IsMakefile=False, FvName=None):
- if self.NameGuid is not None and self.NameGuid.startswith('PCD('):
+ if self.NameGuid and self.NameGuid.startswith('PCD('):
PcdValue = GenFdsGlobalVariable.GetPcdValue(self.NameGuid)
if len(PcdValue) == 0:
EdkLogger.error("GenFds", GENFDS_ERROR, '%s NOT defined.' \
@@ -82,7 +78,7 @@ class FileStatement (FileStatementClassObject) :
Dict.update(self.DefineVarDict)
SectionAlignments = None
- if self.FvName is not None :
+ if self.FvName:
Buffer = BytesIO('')
if self.FvName.upper() not in GenFdsGlobalVariable.FdfParser.Profile.FvDict:
EdkLogger.error("GenFds", GENFDS_ERROR, "FV (%s) is NOT described in FDF file!" % (self.FvName))
@@ -90,14 +86,14 @@ class FileStatement (FileStatementClassObject) :
FileName = Fv.AddToBuffer(Buffer)
SectionFiles = [FileName]
- elif self.FdName is not None:
+ elif self.FdName:
if self.FdName.upper() not in GenFdsGlobalVariable.FdfParser.Profile.FdDict:
EdkLogger.error("GenFds", GENFDS_ERROR, "FD (%s) is NOT described in FDF file!" % (self.FdName))
Fd = GenFdsGlobalVariable.FdfParser.Profile.FdDict.get(self.FdName.upper())
FileName = Fd.GenFd()
SectionFiles = [FileName]
- elif self.FileName is not None:
+ elif self.FileName:
if hasattr(self, 'FvFileType') and self.FvFileType == 'RAW':
if isinstance(self.FileName, list) and isinstance(self.SubAlignment, list) and len(self.FileName) == len(self.SubAlignment):
FileContent = ''
@@ -111,7 +107,7 @@ class FileStatement (FileStatementClassObject) :
Content = f.read()
f.close()
AlignValue = 1
- if self.SubAlignment[Index] is not None:
+ if self.SubAlignment[Index]:
AlignValue = GenFdsGlobalVariable.GetAlignment(self.SubAlignment[Index])
if AlignValue > MaxAlignValue:
MaxAlignIndex = Index
@@ -143,7 +139,7 @@ class FileStatement (FileStatementClassObject) :
SectionFiles = []
Index = 0
SectionAlignments = []
- for section in self.SectionList :
+ for section in self.SectionList:
Index = Index + 1
SecIndex = '%d' %Index
# process the inside FvImage from FvSection or GuidSection
@@ -152,7 +148,7 @@ class FileStatement (FileStatementClassObject) :
section.FvAddr = FvChildAddr.pop(0)
elif isinstance(section, GuidSection):
section.FvAddr = FvChildAddr
- if FvParentAddr is not None and isinstance(section, GuidSection):
+ if FvParentAddr and isinstance(section, GuidSection):
section.FvParentAddr = FvParentAddr
if self.KeepReloc == False:
@@ -168,7 +164,7 @@ class FileStatement (FileStatementClassObject) :
#
FfsFileOutput = os.path.join(OutputDir, self.NameGuid + '.ffs')
GenFdsGlobalVariable.GenerateFfs(FfsFileOutput, SectionFiles,
- Ffs.Ffs.FdfFvFileTypeToFileType.get(self.FvFileType),
+ FdfFvFileTypeToFileType.get(self.FvFileType),
self.NameGuid,
Fixed=self.Fixed,
CheckSum=self.CheckSum,
@@ -177,6 +173,3 @@ class FileStatement (FileStatementClassObject) :
)
return FfsFileOutput
-
-
-
diff --git a/BaseTools/Source/Python/GenFds/FfsInfStatement.py b/BaseTools/Source/Python/GenFds/FfsInfStatement.py
index 6149bc81b4ef..c1073c96e9aa 100644
--- a/BaseTools/Source/Python/GenFds/FfsInfStatement.py
+++ b/BaseTools/Source/Python/GenFds/FfsInfStatement.py
@@ -22,7 +22,7 @@ import Common.LongFilePathOs as os
from io import BytesIO
from struct import *
from .GenFdsGlobalVariable import GenFdsGlobalVariable
-from . import Ffs
+from .Ffs import SectionSuffix,FdfFvFileTypeToFileType
import subprocess
import sys
from . import Section
@@ -760,7 +760,7 @@ class FfsInfStatement(FfsInfStatementClassObject):
SecNum = '%d' %Index
GenSecOutputFile= self.__ExtendMacro__(Rule.NameGuid) + \
- Ffs.Ffs.SectionSuffix[SectionType] + SUP_MODULE_SEC + SecNum
+ SectionSuffix[SectionType] + SUP_MODULE_SEC + SecNum
Index = Index + 1
OutputFile = os.path.join(self.OutputPath, GenSecOutputFile)
File = GenFdsGlobalVariable.MacroExtend(File, Dict, self.CurrentArch)
@@ -803,7 +803,7 @@ class FfsInfStatement(FfsInfStatementClassObject):
else:
SecNum = '%d' %Index
GenSecOutputFile= self.__ExtendMacro__(Rule.NameGuid) + \
- Ffs.Ffs.SectionSuffix[SectionType] + SUP_MODULE_SEC + SecNum
+ SectionSuffix[SectionType] + SUP_MODULE_SEC + SecNum
OutputFile = os.path.join(self.OutputPath, GenSecOutputFile)
GenSecInputFile = GenFdsGlobalVariable.MacroExtend(GenSecInputFile, Dict, self.CurrentArch)
@@ -882,7 +882,7 @@ class FfsInfStatement(FfsInfStatementClassObject):
self.ModuleGuid = RegistryGuidStr
GenFdsGlobalVariable.GenerateFfs(FfsOutput, InputSection,
- Ffs.Ffs.FdfFvFileTypeToFileType[Rule.FvFileType],
+ FdfFvFileTypeToFileType[Rule.FvFileType],
self.ModuleGuid, Fixed=Rule.Fixed,
CheckSum=Rule.CheckSum, Align=Rule.Alignment,
SectionAlign=SectionAlignments,
@@ -1055,7 +1055,7 @@ class FfsInfStatement(FfsInfStatementClassObject):
FfsOutput = os.path.join( self.OutputPath, self.ModuleGuid + '.ffs')
GenFdsGlobalVariable.GenerateFfs(FfsOutput, InputFile,
- Ffs.Ffs.FdfFvFileTypeToFileType[Rule.FvFileType],
+ FdfFvFileTypeToFileType[Rule.FvFileType],
self.ModuleGuid, Fixed=Rule.Fixed,
CheckSum=Rule.CheckSum, Align=Rule.Alignment,
SectionAlign=Alignments,
diff --git a/BaseTools/Source/Python/GenFds/Fv.py b/BaseTools/Source/Python/GenFds/Fv.py
index 0d005ebf5bb6..acb133573d72 100644
--- a/BaseTools/Source/Python/GenFds/Fv.py
+++ b/BaseTools/Source/Python/GenFds/Fv.py
@@ -1,4 +1,3 @@
-from __future__ import absolute_import
## @file
# process FV generation
#
@@ -16,16 +15,13 @@ from __future__ import absolute_import
##
# Import Modules
#
+from __future__ import absolute_import
import Common.LongFilePathOs as os
import subprocess
from io import BytesIO
from struct import *
-
-from . import Ffs
-from . import AprioriSection
from . import FfsFileStatement
from .GenFdsGlobalVariable import GenFdsGlobalVariable
-from CommonDataClass.FdfClass import FvClassObject
from Common.Misc import SaveFileOnChange, PackGUID
from Common.LongFilePathSupport import CopyLongFilePath
from Common.LongFilePathSupport import OpenLongFilePath as open
@@ -36,13 +32,25 @@ FV_UI_EXT_ENTY_GUID = 'A67DF1FA-8DE8-4E98-AF09-4BDF2EFFBC7C'
## generate FV
#
#
-class FV (FvClassObject):
+class FV (object):
## The constructor
#
# @param self The object pointer
#
def __init__(self):
- FvClassObject.__init__(self)
+ self.UiFvName = None
+ self.CreateFileName = None
+ self.BlockSizeList = []
+ self.DefineVarDict = {}
+ self.SetVarDict = {}
+ self.FvAlignment = None
+ self.FvAttributeDict = {}
+ self.FvNameGuid = None
+ self.FvNameString = None
+ self.AprioriSectionList = []
+ self.FfsList = []
+ self.BsBaseAddress = None
+ self.RtBaseAddress = None
self.FvInfFile = None
self.FvAddressFile = None
self.BaseAddress = None
@@ -68,7 +76,7 @@ class FV (FvClassObject):
# @param MacroDict macro value pair
# @retval string Generated FV file path
#
- def AddToBuffer (self, Buffer, BaseAddress=None, BlockSize= None, BlockNum=None, ErasePloarity='1', VtfDict=None, MacroDict = {}, Flag=False) :
+ def AddToBuffer (self, Buffer, BaseAddress=None, BlockSize= None, BlockNum=None, ErasePloarity='1', VtfDict=None, MacroDict = {}, Flag=False):
if BaseAddress is None and self.UiFvName.upper() + 'fv' in GenFdsGlobalVariable.ImageBinDict:
return GenFdsGlobalVariable.ImageBinDict[self.UiFvName.upper() + 'fv']
@@ -96,7 +104,7 @@ class FV (FvClassObject):
if self.FvBaseAddress is not None:
BaseAddress = self.FvBaseAddress
if not Flag:
- self.__InitializeInf__(BaseAddress, BlockSize, BlockNum, ErasePloarity, VtfDict)
+ self._InitializeInf(BaseAddress, BlockSize, BlockNum, ErasePloarity, VtfDict)
#
# First Process the Apriori section
#
@@ -114,7 +122,7 @@ class FV (FvClassObject):
TAB_LINE_BREAK)
# Process Modules in FfsList
- for FfsFile in self.FfsList :
+ for FfsFile in self.FfsList:
if Flag:
if isinstance(FfsFile, FfsFileStatement.FileStatement):
continue
@@ -177,7 +185,7 @@ class FV (FvClassObject):
if FvChildAddr != []:
# Update Ffs again
- for FfsFile in self.FfsList :
+ for FfsFile in self.FfsList:
FileName = FfsFile.GenFfs(MacroDict, FvChildAddr, BaseAddress, IsMakefile=Flag, FvName=self.UiFvName)
if GenFdsGlobalVariable.LargeFileInFvFlags[-1]:
@@ -248,7 +256,7 @@ class FV (FvClassObject):
return True
return False
- ## __InitializeInf__()
+ ## _InitializeInf()
#
# Initilize the inf file to create FV
#
@@ -259,7 +267,7 @@ class FV (FvClassObject):
# @param ErasePolarity Flash erase polarity
# @param VtfDict VTF objects
#
- def __InitializeInf__ (self, BaseAddress = None, BlockSize= None, BlockNum = None, ErasePloarity='1', VtfDict=None) :
+ def _InitializeInf (self, BaseAddress = None, BlockSize= None, BlockNum = None, ErasePloarity='1', VtfDict=None):
#
# Create FV inf file
#
@@ -271,7 +279,7 @@ class FV (FvClassObject):
# Add [Options]
#
self.FvInfFile.writelines("[options]" + TAB_LINE_BREAK)
- if BaseAddress is not None :
+ if BaseAddress is not None:
self.FvInfFile.writelines("EFI_BASE_ADDRESS = " + \
BaseAddress + \
TAB_LINE_BREAK)
@@ -290,7 +298,7 @@ class FV (FvClassObject):
#set default block size is 1
self.FvInfFile.writelines("EFI_BLOCK_SIZE = 0x1" + TAB_LINE_BREAK)
- for BlockSize in self.BlockSizeList :
+ for BlockSize in self.BlockSizeList:
if BlockSize[0] is not None:
self.FvInfFile.writelines("EFI_BLOCK_SIZE = " + \
'0x%X' %BlockSize[0] + \
@@ -316,9 +324,9 @@ class FV (FvClassObject):
' %s' %ErasePloarity + \
TAB_LINE_BREAK)
if not (self.FvAttributeDict is None):
- for FvAttribute in self.FvAttributeDict.keys() :
+ for FvAttribute in self.FvAttributeDict.keys():
if FvAttribute == "FvUsedSizeEnable":
- if self.FvAttributeDict[FvAttribute].upper() in ('TRUE', '1') :
+ if self.FvAttributeDict[FvAttribute].upper() in ('TRUE', '1'):
self.UsedSizeEnable = True
continue
self.FvInfFile.writelines("EFI_" + \
@@ -361,8 +369,8 @@ class FV (FvClassObject):
Guid = FV_UI_EXT_ENTY_GUID.split('-')
#
# Layout:
- # EFI_FIRMWARE_VOLUME_EXT_ENTRY : size 4
- # GUID : size 16
+ # EFI_FIRMWARE_VOLUME_EXT_ENTRY: size 4
+ # GUID: size 16
# FV UI name
#
Buffer += (pack('HH', (FvUiLen + 16 + 4), 0x0002)
@@ -418,7 +426,6 @@ class FV (FvClassObject):
FvExtHeaderFileName + \
TAB_LINE_BREAK)
-
#
# Add [Files]
#
diff --git a/BaseTools/Source/Python/GenFds/FvImageSection.py b/BaseTools/Source/Python/GenFds/FvImageSection.py
index 04556fc87099..d6e1f3315b7e 100644
--- a/BaseTools/Source/Python/GenFds/FvImageSection.py
+++ b/BaseTools/Source/Python/GenFds/FvImageSection.py
@@ -18,7 +18,7 @@
from __future__ import absolute_import
from . import Section
from io import BytesIO
-from .Ffs import Ffs
+from .Ffs import SectionSuffix
import subprocess
from .GenFdsGlobalVariable import GenFdsGlobalVariable
import Common.LongFilePathOs as os
@@ -76,7 +76,7 @@ class FvImageSection(FvImageSectionClassObject):
if FvAlignmentValue > MaxFvAlignment:
MaxFvAlignment = FvAlignmentValue
- OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + Num + Ffs.SectionSuffix.get("FV_IMAGE"))
+ OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + Num + SectionSuffix.get("FV_IMAGE"))
GenFdsGlobalVariable.GenerateSection(OutputFile, [FvFileName], 'EFI_SECTION_FIRMWARE_VOLUME_IMAGE', IsMakefile=IsMakefile)
OutputFileList.append(OutputFile)
@@ -146,7 +146,7 @@ class FvImageSection(FvImageSectionClassObject):
#
# Prepare the parameter of GenSection
#
- OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + SecNum + Ffs.SectionSuffix.get("FV_IMAGE"))
+ OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + SecNum + SectionSuffix.get("FV_IMAGE"))
GenFdsGlobalVariable.GenerateSection(OutputFile, [FvFileName], 'EFI_SECTION_FIRMWARE_VOLUME_IMAGE', IsMakefile=IsMakefile)
OutputFileList.append(OutputFile)
diff --git a/BaseTools/Source/Python/GenFds/GenFds.py b/BaseTools/Source/Python/GenFds/GenFds.py
index 9dec9c5eb57d..0c8091b79841 100644
--- a/BaseTools/Source/Python/GenFds/GenFds.py
+++ b/BaseTools/Source/Python/GenFds/GenFds.py
@@ -17,34 +17,30 @@
#
from __future__ import print_function
from __future__ import absolute_import
+from re import compile
from optparse import OptionParser
-import sys
-import Common.LongFilePathOs as os
-import linecache
-from . import FdfParser
-import Common.BuildToolError as BuildToolError
-from .GenFdsGlobalVariable import GenFdsGlobalVariable
-from Workspace.WorkspaceDatabase import WorkspaceDatabase
-from Workspace.BuildClassObject import PcdClassObject
-from . import RuleComplexFile
-from .EfiSection import EfiSection
+from sys import exit
+from glob import glob
+from struct import unpack
+from linecache import getlines
from io import BytesIO
-import Common.TargetTxtClassObject as TargetTxtClassObject
-import Common.ToolDefClassObject as ToolDefClassObject
+
+import Common.LongFilePathOs as os
+from Common.TargetTxtClassObject import TargetTxtClassObject
from Common.DataType import *
import Common.GlobalData as GlobalData
from Common import EdkLogger
-from Common.StringUtils import *
-from Common.Misc import DirCache, PathClass
-from Common.Misc import SaveFileOnChange
-from Common.Misc import ClearDuplicatedInf
-from Common.Misc import GuidStructureStringToGuidString
+from Common.StringUtils import NormPath
+from Common.Misc import DirCache, PathClass, GuidStructureStringToGuidString
+from Common.Misc import SaveFileOnChange, ClearDuplicatedInf
from Common.BuildVersion import gBUILD_VERSION
from Common.MultipleWorkspace import MultipleWorkspace as mws
-from . import FfsFileStatement
-import glob
-from struct import unpack
-from Common.GlobalData import gGuidPattern
+from Common.BuildToolError import FatalError, GENFDS_ERROR, CODE_ERROR, FORMAT_INVALID, RESOURCE_NOT_AVAILABLE, FILE_NOT_FOUND, OPTION_MISSING, FORMAT_NOT_SUPPORTED,OPTION_VALUE_INVALID
+from Workspace.WorkspaceDatabase import WorkspaceDatabase
+
+from .FdfParser import FdfParser, Warning
+from .GenFdsGlobalVariable import GenFdsGlobalVariable
+from .FfsFileStatement import FileStatement
## Version and Copyright
versionNumber = "1.0" + ' ' + gBUILD_VERSION
@@ -71,22 +67,22 @@ def main():
EdkLogger.Initialize()
try:
- if Options.verbose is not None:
+ if Options.verbose:
EdkLogger.SetLevel(EdkLogger.VERBOSE)
GenFdsGlobalVariable.VerboseMode = True
- if Options.FixedAddress is not None:
+ if Options.FixedAddress:
GenFdsGlobalVariable.FixedLoadAddress = True
- if Options.quiet is not None:
+ if Options.quiet:
EdkLogger.SetLevel(EdkLogger.QUIET)
- if Options.debug is not None:
+ if Options.debug:
EdkLogger.SetLevel(Options.debug + 1)
GenFdsGlobalVariable.DebugLevel = Options.debug
else:
EdkLogger.SetLevel(EdkLogger.INFO)
- if (Options.Workspace is None):
+ if not Options.Workspace:
EdkLogger.error("GenFds", OPTION_MISSING, "WORKSPACE not defined",
ExtraData="Please use '-w' switch to pass it or set the WORKSPACE environment variable.")
elif not os.path.exists(Options.Workspace):
@@ -97,7 +93,7 @@ def main():
GenFdsGlobalVariable.WorkSpaceDir = Workspace
if 'EDK_SOURCE' in os.environ:
GenFdsGlobalVariable.EdkSourceDir = os.path.normcase(os.environ['EDK_SOURCE'])
- if (Options.debug):
+ if Options.debug:
GenFdsGlobalVariable.VerboseLogger("Using Workspace:" + Workspace)
if Options.GenfdsMultiThread:
GenFdsGlobalVariable.EnableGenfdsMultiThread = True
@@ -107,7 +103,7 @@ def main():
PackagesPath = os.getenv("PACKAGES_PATH")
mws.setWs(GenFdsGlobalVariable.WorkSpaceDir, PackagesPath)
- if (Options.filename):
+ if Options.filename:
FdfFilename = Options.filename
FdfFilename = GenFdsGlobalVariable.ReplaceWorkspaceMacro(FdfFilename)
@@ -123,13 +119,13 @@ def main():
else:
EdkLogger.error("GenFds", OPTION_MISSING, "Missing FDF filename")
- if (Options.BuildTarget):
+ if Options.BuildTarget:
GenFdsGlobalVariable.TargetName = Options.BuildTarget
- if (Options.ToolChain):
+ if Options.ToolChain:
GenFdsGlobalVariable.ToolChainTag = Options.ToolChain
- if (Options.activePlatform):
+ if Options.activePlatform:
ActivePlatform = Options.activePlatform
ActivePlatform = GenFdsGlobalVariable.ReplaceWorkspaceMacro(ActivePlatform)
@@ -139,15 +135,15 @@ def main():
if not os.path.isabs (ActivePlatform):
ActivePlatform = mws.join(GenFdsGlobalVariable.WorkSpaceDir, ActivePlatform)
- if not os.path.exists(ActivePlatform) :
+ if not os.path.exists(ActivePlatform):
EdkLogger.error("GenFds", FILE_NOT_FOUND, "ActivePlatform doesn't exist!")
else:
EdkLogger.error("GenFds", OPTION_MISSING, "Missing active platform")
- GlobalData.BuildOptionPcd = Options.OptionPcd if Options.OptionPcd else {}
+ GlobalData.BuildOptionPcd = Options.OptionPcd if Options.OptionPcd else {}
GenFdsGlobalVariable.ActivePlatform = PathClass(NormPath(ActivePlatform))
- if (Options.ConfDirectory):
+ if Options.ConfDirectory:
# Get alternate Conf location, if it is absolute, then just use the absolute directory name
ConfDirectoryPath = os.path.normpath(Options.ConfDirectory)
if ConfDirectoryPath.startswith('"'):
@@ -169,7 +165,7 @@ def main():
GlobalData.gConfDirectory = GenFdsGlobalVariable.ConfDir
BuildConfigurationFile = os.path.normpath(os.path.join(ConfDirectoryPath, "target.txt"))
if os.path.isfile(BuildConfigurationFile) == True:
- TargetTxt = TargetTxtClassObject.TargetTxtClassObject()
+ TargetTxt = TargetTxtClassObject()
TargetTxt.LoadTargetTxtFile(BuildConfigurationFile)
# if no build target given in command line, get it from target.txt
if not GenFdsGlobalVariable.TargetName:
@@ -237,10 +233,9 @@ def main():
GlobalData.gAllFiles = DirCache(Workspace)
GlobalData.gWorkspace = Workspace
- if (Options.archList) :
+ if Options.archList:
ArchList = Options.archList.split(',')
else:
-# EdkLogger.error("GenFds", OPTION_MISSING, "Missing build ARCH")
ArchList = BuildWorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, TAB_COMMON, Options.BuildTarget, Options.ToolChain].SupArchList
TargetArchList = set(BuildWorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, TAB_COMMON, Options.BuildTarget, Options.ToolChain].SupArchList) & set(ArchList)
@@ -249,9 +244,11 @@ def main():
for Arch in ArchList:
GenFdsGlobalVariable.OutputDirFromDscDict[Arch] = NormPath(BuildWorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, Options.BuildTarget, Options.ToolChain].OutputDirectory)
- GenFdsGlobalVariable.PlatformName = BuildWorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, Options.BuildTarget, Options.ToolChain].PlatformName
- if (Options.outputDir):
+ # assign platform name based on last entry in ArchList
+ GenFdsGlobalVariable.PlatformName = BuildWorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, ArchList[-1], Options.BuildTarget, Options.ToolChain].PlatformName
+
+ if Options.outputDir:
OutputDirFromCommandLine = GenFdsGlobalVariable.ReplaceWorkspaceMacro(Options.outputDir)
if not os.path.isabs (OutputDirFromCommandLine):
OutputDirFromCommandLine = os.path.join(GenFdsGlobalVariable.WorkSpaceDir, OutputDirFromCommandLine)
@@ -274,27 +271,27 @@ def main():
GenFdsGlobalVariable.OutputDirDict[Key] = OutputDir
""" Parse Fdf file, has to place after build Workspace as FDF may contain macros from DSC file """
- FdfParserObj = FdfParser.FdfParser(FdfFilename)
+ FdfParserObj = FdfParser(FdfFilename)
FdfParserObj.ParseFile()
if FdfParserObj.CycleReferenceCheck():
EdkLogger.error("GenFds", FORMAT_NOT_SUPPORTED, "Cycle Reference Detected in FDF file")
- if (Options.uiFdName) :
+ if Options.uiFdName:
if Options.uiFdName.upper() in FdfParserObj.Profile.FdDict:
GenFds.OnlyGenerateThisFd = Options.uiFdName
else:
EdkLogger.error("GenFds", OPTION_VALUE_INVALID,
"No such an FD in FDF file: %s" % Options.uiFdName)
- if (Options.uiFvName) :
+ if Options.uiFvName:
if Options.uiFvName.upper() in FdfParserObj.Profile.FvDict:
GenFds.OnlyGenerateThisFv = Options.uiFvName
else:
EdkLogger.error("GenFds", OPTION_VALUE_INVALID,
"No such an FV in FDF file: %s" % Options.uiFvName)
- if (Options.uiCapName) :
+ if Options.uiCapName:
if Options.uiCapName.upper() in FdfParserObj.Profile.CapsuleDict:
GenFds.OnlyGenerateThisCap = Options.uiCapName
else:
@@ -302,7 +299,7 @@ def main():
"No such a Capsule in FDF file: %s" % Options.uiCapName)
GenFdsGlobalVariable.WorkSpace = BuildWorkSpace
- if ArchList is not None:
+ if ArchList:
GenFdsGlobalVariable.ArchList = ArchList
# Dsc Build Data will handle Pcd Settings from CommandLine.
@@ -313,10 +310,8 @@ def main():
# Record the FV Region info that may specific in the FD
if FdfParserObj.Profile.FvDict and FdfParserObj.Profile.FdDict:
- for Fv in FdfParserObj.Profile.FvDict:
- FvObj = FdfParserObj.Profile.FvDict[Fv]
- for Fd in FdfParserObj.Profile.FdDict:
- FdObj = FdfParserObj.Profile.FdDict[Fd]
+ for FvObj in FdfParserObj.Profile.FvDict.values():
+ for FdObj in FdfParserObj.Profile.FdDict.values():
for RegionObj in FdObj.RegionList:
if RegionObj.RegionType != BINARY_FILE_TYPE_FV:
continue
@@ -338,7 +333,7 @@ def main():
"""Display FV space info."""
GenFds.DisplayFvSpaceInfo(FdfParserObj)
- except FdfParser.Warning as X:
+ except Warning as X:
EdkLogger.error(X.ToolName, FORMAT_INVALID, File=X.FileName, Line=X.LineNumber, ExtraData=X.Message, RaiseError=False)
ReturnCode = FORMAT_INVALID
except FatalError as X:
@@ -374,7 +369,6 @@ def SingleCheckCallback(option, opt_str, value, parser):
# Using standard Python module optparse to parse command line option of this tool.
#
# @retval Opt A optparse.Values object containing the parsed options
-# @retval Args Target of build command
#
def myOptionParser():
usage = "%prog [options] -f input_file -a arch_list -b build_target -p active_platform -t tool_chain_tag -D \"MacroName [= MacroValue]\""
@@ -404,7 +398,7 @@ def myOptionParser():
Parser.add_option("--pcd", action="append", dest="OptionPcd", help="Set PCD value by command line. Format: \"PcdName=Value\" ")
Parser.add_option("--genfds-multi-thread", action="store_true", dest="GenfdsMultiThread", default=False, help="Enable GenFds multi thread to generate ffs file.")
- (Options, args) = Parser.parse_args()
+ Options, _ = Parser.parse_args()
return Options
## The class implementing the EDK2 flash image generation process
@@ -414,7 +408,7 @@ def myOptionParser():
# 2. Call methods of Fd class to generate FD
# 3. Call methods of Fv class to generate FV that not belong to FD
#
-class GenFds :
+class GenFds(object):
FdfParsef = None
OnlyGenerateThisFd = None
OnlyGenerateThisFv = None
@@ -423,12 +417,13 @@ class GenFds :
## GenFd()
#
# @param OutputDir Output directory
- # @param FdfParser FDF contents parser
+ # @param FdfParserObject FDF contents parser
# @param Workspace The directory of workspace
# @param ArchList The Arch list of platform
#
- def GenFd (OutputDir, FdfParser, WorkSpace, ArchList):
- GenFdsGlobalVariable.SetDir ('', FdfParser, WorkSpace, ArchList)
+ @staticmethod
+ def GenFd (OutputDir, FdfParserObject, WorkSpace, ArchList):
+ GenFdsGlobalVariable.SetDir ('', FdfParserObject, WorkSpace, ArchList)
GenFdsGlobalVariable.VerboseLogger(" Generate all Fd images and their required FV and Capsule images!")
if GenFds.OnlyGenerateThisCap is not None and GenFds.OnlyGenerateThisCap.upper() in GenFdsGlobalVariable.FdfParser.Profile.CapsuleDict:
@@ -470,9 +465,10 @@ class GenFds :
GenFdsGlobalVariable.VerboseLogger("\n Generate all Option ROM!")
for OptRomObj in GenFdsGlobalVariable.FdfParser.Profile.OptRomDict.values():
OptRomObj.AddToBuffer(None)
+
@staticmethod
- def GenFfsMakefile(OutputDir, FdfParser, WorkSpace, ArchList, GlobalData):
- GenFdsGlobalVariable.SetEnv(FdfParser, WorkSpace, ArchList, GlobalData)
+ def GenFfsMakefile(OutputDir, FdfParserObject, WorkSpace, ArchList, GlobalData):
+ GenFdsGlobalVariable.SetEnv(FdfParserObject, WorkSpace, ArchList, GlobalData)
for FdObj in GenFdsGlobalVariable.FdfParser.Profile.FdDict.values():
FdObj.GenFd(Flag=True)
@@ -490,6 +486,7 @@ class GenFds :
# @param FvObj Whose block size to get
# @retval int Block size value
#
+ @staticmethod
def GetFvBlockSize(FvObj):
DefaultBlockSize = 0x1
FdObj = None
@@ -524,16 +521,17 @@ class GenFds :
# @param FvObj Whose block size to get
# @retval None
#
- def DisplayFvSpaceInfo(FdfParser):
+ @staticmethod
+ def DisplayFvSpaceInfo(FdfParserObject):
FvSpaceInfoList = []
MaxFvNameLength = 0
- for FvName in FdfParser.Profile.FvDict:
+ for FvName in FdfParserObject.Profile.FvDict:
if len(FvName) > MaxFvNameLength:
MaxFvNameLength = len(FvName)
FvSpaceInfoFileName = os.path.join(GenFdsGlobalVariable.FvDir, FvName.upper() + '.Fv.map')
if os.path.exists(FvSpaceInfoFileName):
- FileLinesList = linecache.getlines(FvSpaceInfoFileName)
+ FileLinesList = getlines(FvSpaceInfoFileName)
TotalFound = False
Total = ''
UsedFound = False
@@ -575,6 +573,7 @@ class GenFds :
# @param DscFile modules from dsc file will be preprocessed
# @retval None
#
+ @staticmethod
def PreprocessImage(BuildDb, DscFile):
PcdDict = BuildDb.BuildObject[DscFile, TAB_COMMON, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag].Pcds
PcdValue = ''
@@ -600,6 +599,7 @@ class GenFds :
ModuleObj = BuildDb.BuildObject[Key, TAB_COMMON, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag]
print(ModuleObj.BaseName + ' ' + ModuleObj.ModuleType)
+ @staticmethod
def GenerateGuidXRefFile(BuildDb, ArchList, FdfParserObj):
GuidXRefFileName = os.path.join(GenFdsGlobalVariable.FvDir, "Guid.xref")
GuidXRefFile = BytesIO('')
@@ -607,7 +607,6 @@ class GenFds :
GuidDict = {}
ModuleList = []
FileGuidList = []
- GuidPattern = gGuidPattern
for Arch in ArchList:
PlatformDataBase = BuildDb.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag]
PkgList = GenFdsGlobalVariable.WorkSpace.GetPackageList(GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag)
@@ -626,20 +625,16 @@ class GenFds :
continue
else:
ModuleList.append(Module)
- GuidMatch = GuidPattern.match(ModuleFile.BaseName)
- if GuidMatch is not None:
+ if GlobalData.gGuidPattern.match(ModuleFile.BaseName):
GuidXRefFile.write("%s %s\n" % (ModuleFile.BaseName, Module.BaseName))
else:
GuidXRefFile.write("%s %s\n" % (Module.Guid, Module.BaseName))
- for key, item in Module.Protocols.items():
- GuidDict[key] = item
- for key, item in Module.Guids.items():
- GuidDict[key] = item
- for key, item in Module.Ppis.items():
- GuidDict[key] = item
+ GuidDict.update(Module.Protocols)
+ GuidDict.update(Module.Guids)
+ GuidDict.update(Module.Ppis)
for FvName in FdfParserObj.Profile.FvDict:
for FfsObj in FdfParserObj.Profile.FvDict[FvName].FfsList:
- if not isinstance(FfsObj, FfsFileStatement.FileStatement):
+ if not isinstance(FfsObj, FileStatement):
InfPath = PathClass(NormPath(mws.join(GenFdsGlobalVariable.WorkSpaceDir, FfsObj.InfFileName)))
FdfModule = BuildDb.BuildObject[InfPath, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag]
if FdfModule in ModuleList:
@@ -647,12 +642,9 @@ class GenFds :
else:
ModuleList.append(FdfModule)
GuidXRefFile.write("%s %s\n" % (FdfModule.Guid, FdfModule.BaseName))
- for key, item in FdfModule.Protocols.items():
- GuidDict[key] = item
- for key, item in FdfModule.Guids.items():
- GuidDict[key] = item
- for key, item in FdfModule.Ppis.items():
- GuidDict[key] = item
+ GuidDict.update(FdfModule.Protocols)
+ GuidDict.update(FdfModule.Guids)
+ GuidDict.update(FdfModule.Ppis)
else:
FileStatementGuid = FfsObj.NameGuid
if FileStatementGuid in FileGuidList:
@@ -661,13 +653,13 @@ class GenFds :
FileGuidList.append(FileStatementGuid)
Name = []
FfsPath = os.path.join(GenFdsGlobalVariable.FvDir, 'Ffs')
- FfsPath = glob.glob(os.path.join(FfsPath, FileStatementGuid) + '*')
+ FfsPath = glob(os.path.join(FfsPath, FileStatementGuid) + '*')
if not FfsPath:
continue
if not os.path.exists(FfsPath[0]):
continue
MatchDict = {}
- ReFileEnds = re.compile('\S+(.ui)$|\S+(fv.sec.txt)$|\S+(.pe32.txt)$|\S+(.te.txt)$|\S+(.pic.txt)$|\S+(.raw.txt)$|\S+(.ffs.txt)$')
+ ReFileEnds = compile('\S+(.ui)$|\S+(fv.sec.txt)$|\S+(.pe32.txt)$|\S+(.te.txt)$|\S+(.pic.txt)$|\S+(.raw.txt)$|\S+(.ffs.txt)$')
FileList = os.listdir(FfsPath[0])
for File in FileList:
Match = ReFileEnds.search(File)
@@ -724,16 +716,10 @@ class GenFds :
os.remove(GuidXRefFileName)
GuidXRefFile.close()
- ##Define GenFd as static function
- GenFd = staticmethod(GenFd)
- GetFvBlockSize = staticmethod(GetFvBlockSize)
- DisplayFvSpaceInfo = staticmethod(DisplayFvSpaceInfo)
- PreprocessImage = staticmethod(PreprocessImage)
- GenerateGuidXRefFile = staticmethod(GenerateGuidXRefFile)
-
if __name__ == '__main__':
r = main()
## 0-127 is a safe return range, and 1 is a standard default error
- if r < 0 or r > 127: r = 1
- sys.exit(r)
+ if r < 0 or r > 127:
+ r = 1
+ exit(r)
diff --git a/BaseTools/Source/Python/GenFds/GenFdsGlobalVariable.py b/BaseTools/Source/Python/GenFds/GenFdsGlobalVariable.py
index 14578a92a9b1..e867693d7c88 100644
--- a/BaseTools/Source/Python/GenFds/GenFdsGlobalVariable.py
+++ b/BaseTools/Source/Python/GenFds/GenFdsGlobalVariable.py
@@ -16,13 +16,15 @@
# Import Modules
#
from __future__ import print_function
+from __future__ import absolute_import
+
import Common.LongFilePathOs as os
-import sys
-import subprocess
-import struct
-import array
+from sys import stdout
+from subprocess import PIPE,Popen
+from struct import Struct
+from array import array
-from Common.BuildToolError import *
+from Common.BuildToolError import COMMAND_FAILURE,GENFDS_ERROR
from Common import EdkLogger
from Common.Misc import SaveFileOnChange
@@ -89,7 +91,7 @@ class GenFdsGlobalVariable:
EFI_FIRMWARE_FILE_SYSTEM3_GUID = '5473C07A-3DCB-4dca-BD6F-1E9689E7349A'
LARGE_FILE_SIZE = 0x1000000
- SectionHeader = struct.Struct("3B 1B")
+ SectionHeader = Struct("3B 1B")
# FvName, FdName, CapName in FDF, Image file name
ImageBinDict = {}
@@ -97,7 +99,7 @@ class GenFdsGlobalVariable:
## LoadBuildRule
#
@staticmethod
- def __LoadBuildRule():
+ def _LoadBuildRule():
if GenFdsGlobalVariable.__BuildRuleDatabase:
return GenFdsGlobalVariable.__BuildRuleDatabase
BuildConfigurationFile = os.path.normpath(os.path.join(GenFdsGlobalVariable.ConfDir, "target.txt"))
@@ -139,44 +141,43 @@ class GenFdsGlobalVariable:
if not Arch in GenFdsGlobalVariable.OutputDirDict:
return {}
- BuildRuleDatabase = GenFdsGlobalVariable.__LoadBuildRule()
+ BuildRuleDatabase = GenFdsGlobalVariable._LoadBuildRule()
if not BuildRuleDatabase:
return {}
PathClassObj = PathClass(Inf.MetaFile.File,
GenFdsGlobalVariable.WorkSpaceDir)
- Macro = {}
- Macro["WORKSPACE" ] = GenFdsGlobalVariable.WorkSpaceDir
- Macro["MODULE_NAME" ] = Inf.BaseName
- Macro["MODULE_GUID" ] = Inf.Guid
- Macro["MODULE_VERSION" ] = Inf.Version
- Macro["MODULE_TYPE" ] = Inf.ModuleType
- Macro["MODULE_FILE" ] = str(PathClassObj)
- Macro["MODULE_FILE_BASE_NAME" ] = PathClassObj.BaseName
- Macro["MODULE_RELATIVE_DIR" ] = PathClassObj.SubDir
- Macro["MODULE_DIR" ] = PathClassObj.SubDir
-
- Macro["BASE_NAME" ] = Inf.BaseName
-
- Macro["ARCH" ] = Arch
- Macro["TOOLCHAIN" ] = GenFdsGlobalVariable.ToolChainTag
- Macro["TOOLCHAIN_TAG" ] = GenFdsGlobalVariable.ToolChainTag
- Macro["TOOL_CHAIN_TAG" ] = GenFdsGlobalVariable.ToolChainTag
- Macro["TARGET" ] = GenFdsGlobalVariable.TargetName
-
- Macro["BUILD_DIR" ] = GenFdsGlobalVariable.OutputDirDict[Arch]
- Macro["BIN_DIR" ] = os.path.join(GenFdsGlobalVariable.OutputDirDict[Arch], Arch)
- Macro["LIB_DIR" ] = os.path.join(GenFdsGlobalVariable.OutputDirDict[Arch], Arch)
BuildDir = os.path.join(
GenFdsGlobalVariable.OutputDirDict[Arch],
Arch,
PathClassObj.SubDir,
PathClassObj.BaseName
)
- Macro["MODULE_BUILD_DIR" ] = BuildDir
- Macro["OUTPUT_DIR" ] = os.path.join(BuildDir, "OUTPUT")
- Macro["DEBUG_DIR" ] = os.path.join(BuildDir, "DEBUG")
-
+ BinDir = os.path.join(GenFdsGlobalVariable.OutputDirDict[Arch], Arch)
+ Macro = {
+ "WORKSPACE":GenFdsGlobalVariable.WorkSpaceDir,
+ "MODULE_NAME":Inf.BaseName,
+ "MODULE_GUID":Inf.Guid,
+ "MODULE_VERSION":Inf.Version,
+ "MODULE_TYPE":Inf.ModuleType,
+ "MODULE_FILE":str(PathClassObj),
+ "MODULE_FILE_BASE_NAME":PathClassObj.BaseName,
+ "MODULE_RELATIVE_DIR":PathClassObj.SubDir,
+ "MODULE_DIR":PathClassObj.SubDir,
+ "BASE_NAME":Inf.BaseName,
+ "ARCH":Arch,
+ "TOOLCHAIN":GenFdsGlobalVariable.ToolChainTag,
+ "TOOLCHAIN_TAG":GenFdsGlobalVariable.ToolChainTag,
+ "TOOL_CHAIN_TAG":GenFdsGlobalVariable.ToolChainTag,
+ "TARGET":GenFdsGlobalVariable.TargetName,
+ "BUILD_DIR":GenFdsGlobalVariable.OutputDirDict[Arch],
+ "BIN_DIR":BinDir,
+ "LIB_DIR":BinDir,
+ "MODULE_BUILD_DIR":BuildDir,
+ "OUTPUT_DIR":os.path.join(BuildDir, "OUTPUT"),
+ "DEBUG_DIR":os.path.join(BuildDir, "DEBUG")
+ }
+
BuildRules = {}
for Type in BuildRuleDatabase.FileTypeList:
#first try getting build rule by BuildRuleFamily
@@ -216,12 +217,12 @@ class GenFdsGlobalVariable:
if not Inf.IsBinaryModule:
for File in Inf.Sources:
- if File.TagName in ("", "*", GenFdsGlobalVariable.ToolChainTag) and \
- File.ToolChainFamily in ("", "*", GenFdsGlobalVariable.ToolChainFamily):
+ if File.TagName in {"", "*", GenFdsGlobalVariable.ToolChainTag} and \
+ File.ToolChainFamily in {"", "*", GenFdsGlobalVariable.ToolChainFamily}:
FileList.append((File, DataType.TAB_UNKNOWN_FILE))
for File in Inf.Binaries:
- if File.Target in [DataType.TAB_COMMON, '*', GenFdsGlobalVariable.TargetName]:
+ if File.Target in {DataType.TAB_COMMON, '*', GenFdsGlobalVariable.TargetName}:
FileList.append((File, File.Type))
for File, FileType in FileList:
@@ -233,7 +234,7 @@ class GenFdsGlobalVariable:
Source = SourceList[Index]
Index = Index + 1
- if File.IsBinary and File == Source and Inf.Binaries is not None and File in Inf.Binaries:
+ if File.IsBinary and File == Source and Inf.Binaries and File in Inf.Binaries:
# Skip all files that are not binary libraries
if not Inf.LibraryClass:
continue
@@ -287,19 +288,18 @@ class GenFdsGlobalVariable:
# @param Workspace The directory of workspace
# @param ArchList The Arch list of platform
#
+ @staticmethod
def SetDir (OutputDir, FdfParser, WorkSpace, ArchList):
- GenFdsGlobalVariable.VerboseLogger("GenFdsGlobalVariable.OutputDir :%s" % OutputDir)
-# GenFdsGlobalVariable.OutputDirDict = OutputDir
+ GenFdsGlobalVariable.VerboseLogger("GenFdsGlobalVariable.OutputDir:%s" % OutputDir)
GenFdsGlobalVariable.FdfParser = FdfParser
GenFdsGlobalVariable.WorkSpace = WorkSpace
GenFdsGlobalVariable.FvDir = os.path.join(GenFdsGlobalVariable.OutputDirDict[ArchList[0]], DataType.TAB_FV_DIRECTORY)
- if not os.path.exists(GenFdsGlobalVariable.FvDir) :
+ if not os.path.exists(GenFdsGlobalVariable.FvDir):
os.makedirs(GenFdsGlobalVariable.FvDir)
GenFdsGlobalVariable.FfsDir = os.path.join(GenFdsGlobalVariable.FvDir, 'Ffs')
- if not os.path.exists(GenFdsGlobalVariable.FfsDir) :
+ if not os.path.exists(GenFdsGlobalVariable.FfsDir):
os.makedirs(GenFdsGlobalVariable.FfsDir)
- T_CHAR_LF = '\n'
#
# Create FV Address inf file
#
@@ -308,7 +308,7 @@ class GenFdsGlobalVariable:
#
# Add [Options]
#
- FvAddressFile.writelines("[options]" + T_CHAR_LF)
+ FvAddressFile.writelines("[options]" + DataType.TAB_LINE_BREAK)
BsAddress = '0'
for Arch in ArchList:
if GenFdsGlobalVariable.WorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag].BsBaseAddress:
@@ -317,19 +317,22 @@ class GenFdsGlobalVariable:
FvAddressFile.writelines("EFI_BOOT_DRIVER_BASE_ADDRESS = " + \
BsAddress + \
- T_CHAR_LF)
+ DataType.TAB_LINE_BREAK)
RtAddress = '0'
- for Arch in ArchList:
- if GenFdsGlobalVariable.WorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag].RtBaseAddress:
- RtAddress = GenFdsGlobalVariable.WorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag].RtBaseAddress
+ for Arch in reversed(ArchList):
+ temp = GenFdsGlobalVariable.WorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag].RtBaseAddress
+ if temp:
+ RtAddress = temp
+ break
FvAddressFile.writelines("EFI_RUNTIME_DRIVER_BASE_ADDRESS = " + \
RtAddress + \
- T_CHAR_LF)
+ DataType.TAB_LINE_BREAK)
FvAddressFile.close()
+ @staticmethod
def SetEnv(FdfParser, WorkSpace, ArchList, GlobalData):
GenFdsGlobalVariable.ModuleFile = WorkSpace.ModuleFile
GenFdsGlobalVariable.FdfParser = FdfParser
@@ -360,7 +363,6 @@ class GenFdsGlobalVariable:
if not os.path.exists(GenFdsGlobalVariable.FfsDir):
os.makedirs(GenFdsGlobalVariable.FfsDir)
- T_CHAR_LF = '\n'
#
# Create FV Address inf file
#
@@ -369,7 +371,7 @@ class GenFdsGlobalVariable:
#
# Add [Options]
#
- FvAddressFile.writelines("[options]" + T_CHAR_LF)
+ FvAddressFile.writelines("[options]" + DataType.TAB_LINE_BREAK)
BsAddress = '0'
for Arch in ArchList:
BsAddress = GenFdsGlobalVariable.WorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch,
@@ -380,20 +382,20 @@ class GenFdsGlobalVariable:
FvAddressFile.writelines("EFI_BOOT_DRIVER_BASE_ADDRESS = " + \
BsAddress + \
- T_CHAR_LF)
+ DataType.TAB_LINE_BREAK)
RtAddress = '0'
- for Arch in ArchList:
- if GenFdsGlobalVariable.WorkSpace.BuildObject[
+ for Arch in reversed(ArchList):
+ temp = GenFdsGlobalVariable.WorkSpace.BuildObject[
GenFdsGlobalVariable.ActivePlatform, Arch, GlobalData.gGlobalDefines['TARGET'],
- GlobalData.gGlobalDefines["TOOL_CHAIN_TAG"]].RtBaseAddress:
- RtAddress = GenFdsGlobalVariable.WorkSpace.BuildObject[
- GenFdsGlobalVariable.ActivePlatform, Arch, GlobalData.gGlobalDefines['TARGET'],
- GlobalData.gGlobalDefines["TOOL_CHAIN_TAG"]].RtBaseAddress
+ GlobalData.gGlobalDefines["TOOL_CHAIN_TAG"]].RtBaseAddress
+ if temp:
+ RtAddress = temp
+ break
FvAddressFile.writelines("EFI_RUNTIME_DRIVER_BASE_ADDRESS = " + \
RtAddress + \
- T_CHAR_LF)
+ DataType.TAB_LINE_BREAK)
FvAddressFile.close()
@@ -401,6 +403,7 @@ class GenFdsGlobalVariable:
#
# @param String String that may contain macro
#
+ @staticmethod
def ReplaceWorkspaceMacro(String):
String = mws.handleWsMacro(String)
Str = String.replace('$(WORKSPACE)', GenFdsGlobalVariable.WorkSpaceDir)
@@ -424,7 +427,7 @@ class GenFdsGlobalVariable:
if not os.path.exists(Output):
return True
# always update "Output" if no "Input" given
- if Input is None or len(Input) == 0:
+ if not Input:
return True
# if fdf file is changed after the 'Output" is generated, update the 'Output'
@@ -449,9 +452,9 @@ class GenFdsGlobalVariable:
Cmd += ("-s", Type)
if CompressionType:
Cmd += ("-c", CompressionType)
- if Guid is not None:
+ if Guid:
Cmd += ("-g", Guid)
- if DummyFile is not None:
+ if DummyFile:
Cmd += ("--dummy", DummyFile)
if GuidHdrLen:
Cmd += ("-l", GuidHdrLen)
@@ -473,7 +476,7 @@ class GenFdsGlobalVariable:
if ' '.join(Cmd).strip() not in GenFdsGlobalVariable.SecCmdList:
GenFdsGlobalVariable.SecCmdList.append(' '.join(Cmd).strip())
else:
- SectionData = array.array('B', [0, 0, 0, 0])
+ SectionData = array('B', [0, 0, 0, 0])
SectionData.fromstring(Ui.encode("utf_16_le"))
SectionData.append(0)
SectionData.append(0)
@@ -571,9 +574,9 @@ class GenFdsGlobalVariable:
if BaseAddress:
Cmd += ("-r", BaseAddress)
- if ForceRebase == False:
+ if not ForceRebase:
Cmd += ("-F", "FALSE")
- elif ForceRebase == True:
+ else:
Cmd += ("-F", "TRUE")
if Capsule:
@@ -634,7 +637,7 @@ class GenFdsGlobalVariable:
Revision=None, DeviceId=None, VendorId=None, IsMakefile=False):
InputList = []
Cmd = ["EfiRom"]
- if len(EfiInput) > 0:
+ if EfiInput:
if Compress:
Cmd.append("-ec")
@@ -645,7 +648,7 @@ class GenFdsGlobalVariable:
Cmd.append(EfiFile)
InputList.append (EfiFile)
- if len(BinaryInput) > 0:
+ if BinaryInput:
Cmd.append("-b")
for BinFile in BinaryInput:
Cmd.append(BinFile)
@@ -656,13 +659,13 @@ class GenFdsGlobalVariable:
return
GenFdsGlobalVariable.DebugLogger(EdkLogger.DEBUG_5, "%s needs update because of newer %s" % (Output, InputList))
- if ClassCode is not None:
+ if ClassCode:
Cmd += ("-l", ClassCode)
- if Revision is not None:
+ if Revision:
Cmd += ("-r", Revision)
- if DeviceId is not None:
+ if DeviceId:
Cmd += ("-i", DeviceId)
- if VendorId is not None:
+ if VendorId:
Cmd += ("-f", VendorId)
Cmd += ("-o", Output)
@@ -688,6 +691,7 @@ class GenFdsGlobalVariable:
else:
GenFdsGlobalVariable.CallExternalTool(Cmd, "Failed to call " + ToolPath, returnValue)
+ @staticmethod
def CallExternalTool (cmd, errorMess, returnValue=[]):
if type(cmd) not in (tuple, list):
@@ -701,19 +705,19 @@ class GenFdsGlobalVariable:
cmd += ('-v',)
GenFdsGlobalVariable.InfLogger (cmd)
else:
- sys.stdout.write ('#')
- sys.stdout.flush()
+ stdout.write ('#')
+ stdout.flush()
GenFdsGlobalVariable.SharpCounter = GenFdsGlobalVariable.SharpCounter + 1
if GenFdsGlobalVariable.SharpCounter % GenFdsGlobalVariable.SharpNumberPerLine == 0:
- sys.stdout.write('\n')
+ stdout.write('\n')
try:
- PopenObject = subprocess.Popen(' '.join(cmd), stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
+ PopenObject = Popen(' '.join(cmd), stdout=PIPE, stderr=PIPE, shell=True)
except Exception as X:
EdkLogger.error("GenFds", COMMAND_FAILURE, ExtraData="%s: %s" % (str(X), cmd[0]))
(out, error) = PopenObject.communicate()
- while PopenObject.returncode is None :
+ while PopenObject.returncode is None:
PopenObject.wait()
if returnValue != [] and returnValue[0] != 0:
#get command return value
@@ -727,45 +731,52 @@ class GenFdsGlobalVariable:
print("###", cmd)
EdkLogger.error("GenFds", COMMAND_FAILURE, errorMess)
+ @staticmethod
def VerboseLogger (msg):
EdkLogger.verbose(msg)
+ @staticmethod
def InfLogger (msg):
EdkLogger.info(msg)
+ @staticmethod
def ErrorLogger (msg, File=None, Line=None, ExtraData=None):
EdkLogger.error('GenFds', GENFDS_ERROR, msg, File, Line, ExtraData)
+ @staticmethod
def DebugLogger (Level, msg):
EdkLogger.debug(Level, msg)
- ## ReplaceWorkspaceMacro()
+ ## MacroExtend()
#
# @param Str String that may contain macro
# @param MacroDict Dictionary that contains macro value pair
#
+ @staticmethod
def MacroExtend (Str, MacroDict={}, Arch=DataType.TAB_COMMON):
- if Str is None :
+ if Str is None:
return None
- Dict = {'$(WORKSPACE)' : GenFdsGlobalVariable.WorkSpaceDir,
- '$(EDK_SOURCE)' : GenFdsGlobalVariable.EdkSourceDir,
+ Dict = {'$(WORKSPACE)': GenFdsGlobalVariable.WorkSpaceDir,
+ '$(EDK_SOURCE)': GenFdsGlobalVariable.EdkSourceDir,
# '$(OUTPUT_DIRECTORY)': GenFdsGlobalVariable.OutputDirFromDsc,
- '$(TARGET)' : GenFdsGlobalVariable.TargetName,
- '$(TOOL_CHAIN_TAG)' : GenFdsGlobalVariable.ToolChainTag,
- '$(SPACE)' : ' '
+ '$(TARGET)': GenFdsGlobalVariable.TargetName,
+ '$(TOOL_CHAIN_TAG)': GenFdsGlobalVariable.ToolChainTag,
+ '$(SPACE)': ' '
}
- OutputDir = GenFdsGlobalVariable.OutputDirFromDscDict[GenFdsGlobalVariable.ArchList[0]]
+
if Arch != DataType.TAB_COMMON and Arch in GenFdsGlobalVariable.ArchList:
OutputDir = GenFdsGlobalVariable.OutputDirFromDscDict[Arch]
+ else:
+ OutputDir = GenFdsGlobalVariable.OutputDirFromDscDict[GenFdsGlobalVariable.ArchList[0]]
Dict['$(OUTPUT_DIRECTORY)'] = OutputDir
- if MacroDict is not None and len (MacroDict) != 0:
+ if MacroDict:
Dict.update(MacroDict)
for key in Dict:
- if Str.find(key) >= 0 :
+ if Str.find(key) >= 0:
Str = Str.replace (key, Dict[key])
if Str.find('$(ARCH)') >= 0:
@@ -780,14 +791,14 @@ class GenFdsGlobalVariable:
#
# @param PcdPattern pattern that labels a PCD.
#
+ @staticmethod
def GetPcdValue (PcdPattern):
- if PcdPattern is None :
+ if PcdPattern is None:
return None
PcdPair = PcdPattern.lstrip('PCD(').rstrip(')').strip().split('.')
TokenSpace = PcdPair[0]
TokenCName = PcdPair[1]
- PcdValue = ''
for Arch in GenFdsGlobalVariable.ArchList:
Platform = GenFdsGlobalVariable.WorkSpace.BuildObject[GenFdsGlobalVariable.ActivePlatform, Arch, GenFdsGlobalVariable.TargetName, GenFdsGlobalVariable.ToolChainTag]
PcdDict = Platform.Pcds
@@ -799,8 +810,7 @@ class GenFdsGlobalVariable:
if PcdObj.DatumType != DataType.TAB_VOID:
EdkLogger.error("GenFds", GENFDS_ERROR, "%s is not VOID* datum type." % PcdPattern)
- PcdValue = PcdObj.DefaultValue
- return PcdValue
+ return PcdObj.DefaultValue
for Package in GenFdsGlobalVariable.WorkSpace.GetPackageList(GenFdsGlobalVariable.ActivePlatform,
Arch,
@@ -815,21 +825,9 @@ class GenFdsGlobalVariable:
if PcdObj.DatumType != DataType.TAB_VOID:
EdkLogger.error("GenFds", GENFDS_ERROR, "%s is not VOID* datum type." % PcdPattern)
- PcdValue = PcdObj.DefaultValue
- return PcdValue
+ return PcdObj.DefaultValue
- return PcdValue
-
- SetDir = staticmethod(SetDir)
- SetEnv = staticmethod(SetEnv)
- ReplaceWorkspaceMacro = staticmethod(ReplaceWorkspaceMacro)
- CallExternalTool = staticmethod(CallExternalTool)
- VerboseLogger = staticmethod(VerboseLogger)
- InfLogger = staticmethod(InfLogger)
- ErrorLogger = staticmethod(ErrorLogger)
- DebugLogger = staticmethod(DebugLogger)
- MacroExtend = staticmethod (MacroExtend)
- GetPcdValue = staticmethod(GetPcdValue)
+ return ''
## FindExtendTool()
#
@@ -863,7 +861,7 @@ def FindExtendTool(KeyStringList, CurrentArchList, NameGuid):
ToolOptionKey = None
KeyList = None
for ToolDef in ToolDefinition.items():
- if NameGuid.lower() == ToolDef[1].lower() :
+ if NameGuid.lower() == ToolDef[1].lower():
KeyList = ToolDef[0].split('_')
Key = KeyList[0] + \
'_' + \
diff --git a/BaseTools/Source/Python/GenFds/GuidSection.py b/BaseTools/Source/Python/GenFds/GuidSection.py
index fd660737f421..25609e6f2ae1 100644
--- a/BaseTools/Source/Python/GenFds/GuidSection.py
+++ b/BaseTools/Source/Python/GenFds/GuidSection.py
@@ -19,7 +19,7 @@
from __future__ import absolute_import
from . import Section
import subprocess
-from .Ffs import Ffs
+from .Ffs import SectionSuffix
import Common.LongFilePathOs as os
from .GenFdsGlobalVariable import GenFdsGlobalVariable
from .GenFdsGlobalVariable import FindExtendTool
@@ -126,7 +126,7 @@ class GuidSection(GuidSectionClassObject) :
ModuleName + \
SUP_MODULE_SEC + \
SecNum + \
- Ffs.SectionSuffix['GUIDED']
+ SectionSuffix['GUIDED']
OutputFile = os.path.normpath(OutputFile)
ExternalTool = None
diff --git a/BaseTools/Source/Python/GenFds/OptionRom.py b/BaseTools/Source/Python/GenFds/OptionRom.py
index bb7071fa7fca..9d7ad25ecc43 100644
--- a/BaseTools/Source/Python/GenFds/OptionRom.py
+++ b/BaseTools/Source/Python/GenFds/OptionRom.py
@@ -26,8 +26,6 @@ from Common.Misc import SaveFileOnChange
from Common import EdkLogger
from Common.BuildToolError import *
-T_CHAR_LF = '\n'
-
##
#
#
@@ -36,9 +34,9 @@ class OPTIONROM (OptionRomClassObject):
#
# @param self The object pointer
#
- def __init__(self):
+ def __init__(self, Name = ""):
OptionRomClassObject.__init__(self)
-
+ self.DriverName = Name
## AddToBuffer()
#
diff --git a/BaseTools/Source/Python/GenFds/Region.py b/BaseTools/Source/Python/GenFds/Region.py
index 5242b74c9e70..9aeaa45d4566 100644
--- a/BaseTools/Source/Python/GenFds/Region.py
+++ b/BaseTools/Source/Python/GenFds/Region.py
@@ -20,7 +20,6 @@ from struct import *
from .GenFdsGlobalVariable import GenFdsGlobalVariable
from io import BytesIO
import string
-from CommonDataClass.FdfClass import RegionClassObject
import Common.LongFilePathOs as os
from stat import *
from Common import EdkLogger
@@ -32,15 +31,20 @@ from Common.DataType import BINARY_FILE_TYPE_FV
## generate Region
#
#
-class Region(RegionClassObject):
+class Region(object):
## The constructor
#
# @param self The object pointer
#
def __init__(self):
- RegionClassObject.__init__(self)
-
+ self.Offset = None # The begin position of the Region
+ self.Size = None # The Size of the Region
+ self.PcdOffset = None
+ self.PcdSize = None
+ self.SetVarDict = {}
+ self.RegionType = None
+ self.RegionDataList = []
## PadBuffer()
#
diff --git a/BaseTools/Source/Python/GenFds/UiSection.py b/BaseTools/Source/Python/GenFds/UiSection.py
index 24f2f3ca938d..c1dff188be96 100644
--- a/BaseTools/Source/Python/GenFds/UiSection.py
+++ b/BaseTools/Source/Python/GenFds/UiSection.py
@@ -17,7 +17,7 @@
#
from __future__ import absolute_import
from . import Section
-from .Ffs import Ffs
+from .Ffs import SectionSuffix
import subprocess
import Common.LongFilePathOs as os
from .GenFdsGlobalVariable import GenFdsGlobalVariable
@@ -59,7 +59,7 @@ class UiSection (UiSectionClassObject):
self.StringData = FfsInf.__ExtendMacro__(self.StringData)
self.FileName = FfsInf.__ExtendMacro__(self.FileName)
- OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + SecNum + Ffs.SectionSuffix.get(BINARY_FILE_TYPE_UI))
+ OutputFile = os.path.join(OutputPath, ModuleName + SUP_MODULE_SEC + SecNum + SectionSuffix.get(BINARY_FILE_TYPE_UI))
if self.StringData is not None :
NameString = self.StringData
diff --git a/BaseTools/Source/Python/GenFds/VerSection.py b/BaseTools/Source/Python/GenFds/VerSection.py
index db71fe9653d4..051743f10c21 100644
--- a/BaseTools/Source/Python/GenFds/VerSection.py
+++ b/BaseTools/Source/Python/GenFds/VerSection.py
@@ -1,7 +1,7 @@
## @file
# process Version section generation
#
-# Copyright (c) 2007 - 2017, Intel Corporation. All rights reserved.<BR>
+# Copyright (c) 2007 - 2018, Intel Corporation. All rights reserved.<BR>
#
# This program and the accompanying materials
# are licensed and made available under the terms and conditions of the BSD License
@@ -16,10 +16,8 @@
# Import Modules
#
from __future__ import absolute_import
-from .Ffs import Ffs
-from . import Section
+from .Ffs import SectionSuffix
import Common.LongFilePathOs as os
-import subprocess
from .GenFdsGlobalVariable import GenFdsGlobalVariable
from CommonDataClass.FdfClass import VerSectionClassObject
from Common.LongFilePathSupport import OpenLongFilePath as open
@@ -54,29 +52,27 @@ class VerSection (VerSectionClassObject):
#
# Prepare the parameter of GenSection
#
- if FfsInf is not None:
+ if FfsInf:
self.Alignment = FfsInf.__ExtendMacro__(self.Alignment)
self.BuildNum = FfsInf.__ExtendMacro__(self.BuildNum)
self.StringData = FfsInf.__ExtendMacro__(self.StringData)
self.FileName = FfsInf.__ExtendMacro__(self.FileName)
OutputFile = os.path.join(OutputPath,
- ModuleName + SUP_MODULE_SEC + SecNum + Ffs.SectionSuffix.get('VERSION'))
+ ModuleName + SUP_MODULE_SEC + SecNum + SectionSuffix.get('VERSION'))
OutputFile = os.path.normpath(OutputFile)
# Get String Data
StringData = ''
- if self.StringData is not None:
+ if self.StringData:
StringData = self.StringData
- elif self.FileName is not None:
+ elif self.FileName:
FileNameStr = GenFdsGlobalVariable.ReplaceWorkspaceMacro(self.FileName)
FileNameStr = GenFdsGlobalVariable.MacroExtend(FileNameStr, Dict)
FileObj = open(FileNameStr, 'r')
StringData = FileObj.read()
StringData = '"' + StringData + '"'
FileObj.close()
- else:
- StringData = ''
GenFdsGlobalVariable.GenerateSection(OutputFile, [], 'EFI_SECTION_VERSION',
Ver=StringData, BuildNumber=self.BuildNum, IsMakefile=IsMakefile)
OutputFileList = []
diff --git a/BaseTools/Source/Python/GenFds/Vtf.py b/BaseTools/Source/Python/GenFds/Vtf.py
index 5cb2d4acfb1c..9dcd48b2d812 100644
--- a/BaseTools/Source/Python/GenFds/Vtf.py
+++ b/BaseTools/Source/Python/GenFds/Vtf.py
@@ -18,21 +18,24 @@
from __future__ import absolute_import
from .GenFdsGlobalVariable import GenFdsGlobalVariable
import Common.LongFilePathOs as os
-from CommonDataClass.FdfClass import VtfClassObject
from Common.LongFilePathSupport import OpenLongFilePath as open
-T_CHAR_LF = '\n'
+from Common.DataType import TAB_LINE_BREAK
## generate VTF
#
#
-class Vtf (VtfClassObject):
+class Vtf (object):
## The constructor
#
# @param self The object pointer
#
def __init__(self):
- VtfClassObject.__init__(self)
+ self.KeyArch = None
+ self.ArchList = None
+ self.UiName = None
+ self.ResetBin = None
+ self.ComponentStatementList = []
## GenVtf() method
#
@@ -44,7 +47,6 @@ class Vtf (VtfClassObject):
#
def GenVtf(self, FdAddressDict) :
self.GenBsfInf()
- OutputFile = os.path.join(GenFdsGlobalVariable.FvDir, self.UiName + '.Vtf')
BaseAddArg = self.GetBaseAddressArg(FdAddressDict)
OutputArg, VtfRawDict = self.GenOutputArg()
@@ -69,57 +71,57 @@ class Vtf (VtfClassObject):
FvList = self.GetFvList()
self.BsfInfName = os.path.join(GenFdsGlobalVariable.FvDir, self.UiName + '.inf')
BsfInf = open(self.BsfInfName, 'w+')
- if self.ResetBin is not None:
- BsfInf.writelines ("[OPTIONS]" + T_CHAR_LF)
+ if self.ResetBin:
+ BsfInf.writelines ("[OPTIONS]" + TAB_LINE_BREAK)
BsfInf.writelines ("IA32_RST_BIN" + \
" = " + \
GenFdsGlobalVariable.MacroExtend(GenFdsGlobalVariable.ReplaceWorkspaceMacro(self.ResetBin)) + \
- T_CHAR_LF)
- BsfInf.writelines (T_CHAR_LF)
+ TAB_LINE_BREAK)
+ BsfInf.writelines (TAB_LINE_BREAK)
- BsfInf.writelines ("[COMPONENTS]" + T_CHAR_LF)
+ BsfInf.writelines ("[COMPONENTS]" + TAB_LINE_BREAK)
for ComponentObj in self.ComponentStatementList :
BsfInf.writelines ("COMP_NAME" + \
" = " + \
ComponentObj.CompName + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
if ComponentObj.CompLoc.upper() == 'NONE':
BsfInf.writelines ("COMP_LOC" + \
" = " + \
'N' + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
- elif ComponentObj.FilePos is not None:
+ elif ComponentObj.FilePos:
BsfInf.writelines ("COMP_LOC" + \
" = " + \
ComponentObj.FilePos + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
else:
Index = FvList.index(ComponentObj.CompLoc.upper())
if Index == 0:
BsfInf.writelines ("COMP_LOC" + \
" = " + \
'F' + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
elif Index == 1:
BsfInf.writelines ("COMP_LOC" + \
" = " + \
'S' + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
BsfInf.writelines ("COMP_TYPE" + \
" = " + \
ComponentObj.CompType + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
BsfInf.writelines ("COMP_VER" + \
" = " + \
ComponentObj.CompVer + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
BsfInf.writelines ("COMP_CS" + \
" = " + \
ComponentObj.CompCs + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
BinPath = ComponentObj.CompBin
if BinPath != '-':
@@ -127,7 +129,7 @@ class Vtf (VtfClassObject):
BsfInf.writelines ("COMP_BIN" + \
" = " + \
BinPath + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
SymPath = ComponentObj.CompSym
if SymPath != '-':
@@ -135,12 +137,12 @@ class Vtf (VtfClassObject):
BsfInf.writelines ("COMP_SYM" + \
" = " + \
SymPath + \
- T_CHAR_LF)
+ TAB_LINE_BREAK)
BsfInf.writelines ("COMP_SIZE" + \
" = " + \
ComponentObj.CompSize + \
- T_CHAR_LF)
- BsfInf.writelines (T_CHAR_LF)
+ TAB_LINE_BREAK)
+ BsfInf.writelines (TAB_LINE_BREAK)
BsfInf.close()
diff --git a/BaseTools/Source/Python/build/BuildReport.py b/BaseTools/Source/Python/build/BuildReport.py
index 4fe29f124d84..4ca456f96f24 100644
--- a/BaseTools/Source/Python/build/BuildReport.py
+++ b/BaseTools/Source/Python/build/BuildReport.py
@@ -46,6 +46,7 @@ from Common.StringUtils import NormPath
from Common.DataType import *
import collections
from Common.Expression import *
+from GenFds.AprioriSection import DXE_APRIORI_GUID, PEI_APRIORI_GUID
## Pattern to extract contents in EDK DXS files
gDxsDependencyPattern = re.compile(r"DEPENDENCY_START(.+)DEPENDENCY_END", re.DOTALL)
@@ -1840,8 +1841,8 @@ class FdRegionReport(object):
#
# Add PEI and DXE a priori files GUIDs defined in PI specification.
#
- self._GuidsDb["1B45CC0A-156A-428A-AF62-49864DA0E6E6"] = "PEI Apriori"
- self._GuidsDb["FC510EE7-FFDC-11D4-BD41-0080C73C8881"] = "DXE Apriori"
+ self._GuidsDb[PEI_APRIORI_GUID] = "PEI Apriori"
+ self._GuidsDb[DXE_APRIORI_GUID] = "DXE Apriori"
#
# Add ACPI table storage file
#
--
2.16.2.windows.1
next prev parent reply other threads:[~2018-09-28 21:55 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-28 21:54 [PATCH v2 0/7] BaseTools/GenFds: cleanup GenFds Jaben Carsey
2018-09-28 21:54 ` Jaben Carsey [this message]
2018-09-28 21:54 ` [PATCH v2 2/7] BaseTools/GenFds: change objects to sets Jaben Carsey
2018-09-28 21:54 ` [PATCH v2 3/7] Basetools/GenFds: refactor class FV Jaben Carsey
2018-09-28 21:54 ` [PATCH v2 4/7] BaseTools/GenFds: remove MacroDict parameter Jaben Carsey
2018-09-28 21:54 ` [PATCH v2 5/7] BaseTools/GenFds: refactor FdfParser warnings Jaben Carsey
2018-09-28 21:54 ` [PATCH v2 6/7] BaseTools/GenFds: Remove duplicate function calls Jaben Carsey
2018-09-28 21:54 ` [PATCH v2 7/7] BaseTools/GenFds: create and use new variable in FdfParser Jaben Carsey
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-list from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cec710b033a13b8ddd2865b9548fc19f60898797.1538171557.git.jaben.carsey@intel.com \
--to=devel@edk2.groups.io \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox