[pschweitzer] 34573: - Reimplemented FsRtlIsFatDbcsLegal to match with Windows one (XP SP3 was the test platform) - Fixed bugs in FsRtlDissectDbcs - Fixed the kernl exports for FsRtlIsFatDbcsLegal - Fixed build
Author: pschweitzer Date: Fri Jul 18 04:15:53 2008 New Revision: 34573
URL: http://svn.reactos.org/svn/reactos?rev=34573&view=rev Log: - Reimplemented FsRtlIsFatDbcsLegal to match with Windows one (XP SP3 was the test platform) - Fixed bugs in FsRtlDissectDbcs - Fixed the kernl exports for FsRtlIsFatDbcsLegal - Fixed build
Modified: branches/pierre-fsd/include/ddk/ntifs.h branches/pierre-fsd/ntoskrnl/fsrtl/dbcsname.c branches/pierre-fsd/ntoskrnl/ntoskrnl_arm.def branches/pierre-fsd/ntoskrnl/ntoskrnl_i386.def
Modified: branches/pierre-fsd/include/ddk/ntifs.h URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/include/ddk/ntifs.h?r... ============================================================================== --- branches/pierre-fsd/include/ddk/ntifs.h [iso-8859-1] (original) +++ branches/pierre-fsd/include/ddk/ntifs.h [iso-8859-1] Fri Jul 18 04:15:53 2008 @@ -3287,6 +3287,12 @@ IN POPLOCK Oplock );
+typedef VOID +(NTAPI *PFSRTL_STACK_OVERFLOW_ROUTINE) ( + IN PVOID Context, + IN PKEVENT Event +); + NTKERNELAPI VOID NTAPI @@ -3294,12 +3300,6 @@ IN PVOID Context, IN PKEVENT Event, IN PFSRTL_STACK_OVERFLOW_ROUTINE StackOverflowRoutine -); - -typedef VOID -(NTAPI *PFSRTL_STACK_OVERFLOW_ROUTINE) ( - IN PVOID Context, - IN PKEVENT Event );
NTKERNELAPI
Modified: branches/pierre-fsd/ntoskrnl/fsrtl/dbcsname.c URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/ntoskrnl/fsrtl/dbcsna... ============================================================================== --- branches/pierre-fsd/ntoskrnl/fsrtl/dbcsname.c [iso-8859-1] (original) +++ branches/pierre-fsd/ntoskrnl/fsrtl/dbcsname.c [iso-8859-1] Fri Jul 18 04:15:53 2008 @@ -4,6 +4,7 @@ * FILE: ntoskrnl/fsrtl/name.c * PURPOSE: Provides DBCS parsing and other support routines for FSDs * PROGRAMMERS: Alex Ionescu (alex.ionescu@reactos.org) + * Pierre Schweitzer (heis_spiter@hotmail.com) */
/* INCLUDES ******************************************************************/ @@ -52,11 +53,16 @@ if (!Name.Length) return;
/* Find first backslash */ - FirstPosition = Name.Length / sizeof(CHAR) ; + FirstPosition = Name.Length / sizeof(CHAR); for (i = 0; i < Name.Length / sizeof(CHAR); i++) { + /* First make sure the character it's not the Lead DBCS */ + if (FsRtlIsLeadDbcsCharacter(Name->Buffer[i])) + { + i++; + } /* If we found one... */ - if (Name.Buffer[i] == '\') + else if (Name.Buffer[i] == '\') { /* If it begins string, just notice it and continue */ if (i == 0) @@ -81,7 +87,7 @@ if (FirstPosition < (Name.Length / sizeof(CHAR))) { RemainingPart->Buffer = Name.Buffer + FirstPosition + 1; - RemainingPart->Length = (Name.Length - FirstPosition) * sizeof(CHAR); + RemainingPart->Length = Name.Length - (FirstPosition + 1) * sizeof(CHAR); RemainingPart->MaximumLength = Name.MaximumLength - RemainingPart->Length; } } @@ -185,97 +191,108 @@ IN BOOLEAN PathNamePermissible, IN BOOLEAN LeadingBackslashPermissible) { - BOOLEAN FirstSlash = FALSE, IsIllegal = FALSE; - ULONG i, LastSlash = 0, LastDot = 0; - ANSI_STRING FileName; - + ANSI_STRING FirstPart, RemainingPart, Name; + ULONG i, LastDot = 0; + /* Just quit if the string is empty */ if (!DbcsName.Length) return FALSE;
- /* Check if we have a filename and a pathname - * Continue until we get the last one to extract the filename - */ - for (i = 0; i < DbcsName.Length / sizeof(CHAR); i++) - { - if (DbcsName.Buffer[i] == '\') - { - LastSlash = i; - if (!i) - { - FirstSlash = TRUE; - } - } - } - - /* DbcsName was to be a filename */ - if (!PathNamePermissible && LastSlash) + /* DbcsName wasn't supposed to be started with \ */ + if (!LeadingBackslashPermissible && DbcsName.Buffer[0] == '\') return FALSE; - - /* DbcsName wasn't supposed to be started with \ if it's a filename */ - if (!LeadingBackslashPermissible && !LastSlash && FirstSlash) - return FALSE; - - /* Now, only work on filename */ - if (LastSlash || FirstSlash) - { - FileName.Buffer = DbcsName.Buffer + LastSlash + 1; - FileName.Length = (DbcsName.Length - LastSlash - 1) * sizeof(CHAR); - FileName.MaximumLength = DbcsName.MaximumLength - FileName.Length; - - } - else - { - FileName.Buffer = DbcsName.Buffer; - FileName.Length = DbcsName.Length; - FileName.MaximumLength = DbcsName.MaximumLength; - } - - /* Filename must be 8.3 filename */ - if (FileName.Length < 3 || FileName.Length > 12) - return FALSE; - - if (!WildCardsPermissible && FsRtlDoesDbcsContainWildCards(&FileName)) - return FALSE; - - /* Now, we will parse the filename to find everything bad in - * It mustn't contain: - * 0x00-0x1F, 0x22, 0x2B, 0x2C, 0x2F, 0x3A, 0x3B, 0x3D, 0x5B, 0x5D, 0x7C */ - for (i = 0; i < FileName.Length / sizeof(CHAR); i++) - { - if ((FileName.Buffer[i] < 0x1F) || (FileName.Buffer[i] == 0x22) || - (FileName.Buffer[i] == 0x2B) || (FileName.Buffer[i] == 0x2C) || - (FileName.Buffer[i] == 0x2F) || (FileName.Buffer[i] == 0x3A) || - (FileName.Buffer[i] == 0x3B) || (FileName.Buffer[i] == 0x3D) || - (FileName.Buffer[i] == 0x5B) || (FileName.Buffer[i] == 0x5D) || - (FileName.Buffer[i] == 0x7C)) - { - IsIllegal = TRUE; - break; - } - if (FileName.Buffer[i] == '.') - { - LastDot = i; - if (!i) - { - IsIllegal = TRUE; - break; - } - } - } - if (IsIllegal || LastDot == (FileName.Length / sizeof(CHAR)) - 1) - return FALSE; - - /* We mustn't have spaces before dot or at the end of the filename */ - if ((LastDot && FileName.Buffer[LastDot - 1] == ' ') || - (FileName.Buffer[FileName.Length / sizeof(CHAR) - 1] == ' ')) - return FALSE; - - /* Filename must be 8.3 filename and not 3.8 filename */ - if (LastDot && ((FileName.Length / sizeof(CHAR) - 1) - LastDot > 3)) - return FALSE; - - /* We have a valid filename */ + /* DbcsName was allowed to be started with , but now, remove it */ + else if (LeadingBackslashPermissible && DbcsName.Buffer[0] == '\') + { + DbcsName.Buffer = DbcsName.Buffer + 1; + DbcsName.Length = DbcsName.Length - 1; + DbcsName.MaximumLength = DbcsName.MaximumLength - 1; + } + + /* Zero strings for further use */ + RtlZeroMemory(&FirstPart, sizeof(FirstPart)); + RtlZeroMemory(&RemainingPart, sizeof(RemainingPart)); + + /* Extract first part of the DbcsName to work on */ + FsRtlDissectDbcs(DbcsName, &FirstPart, &RemainingPart); + while (FirstPart.Length > 0) + { + /* Accept special filename if wildcards are allowed */ + if (WildCardsPermissible && (FirstPart.Length == 1 || FirstPart.Length == 2) && FirstPart.Buffer[0] == '.') + { + if (FirstPart.Length == 2) + { + if (FirstPart.Buffer[1] == '.') + { + goto EndLoop; + } + } + else + { + goto EndLoop; + } + } + + /* Filename must be 8.3 filename */ + if (FirstPart.Length < 3 || FirstPart.Length > 12) + return FALSE; + + if (!WildCardsPermissible && FsRtlDoesDbcsContainWildCards(&FirstPart)) + return FALSE; + + /* Now, we will parse the filename to find everything bad in + * It mustn't contain: + * 0x00-0x1F, 0x22, 0x2B, 0x2C, 0x2F, 0x3A, 0x3B, 0x3D, 0x5B, 0x5D, 0x7C */ + for (i = 0; i < FirstPart.Length / sizeof(CHAR); i++) + { + if ((FirstPart.Buffer[i] < 0x1F) || (FirstPart.Buffer[i] == 0x22) || + (FirstPart.Buffer[i] == 0x2B) || (FirstPart.Buffer[i] == 0x2C) || + (FirstPart.Buffer[i] == 0x2F) || (FirstPart.Buffer[i] == 0x3A) || + (FirstPart.Buffer[i] == 0x3B) || (FirstPart.Buffer[i] == 0x3D) || + (FirstPart.Buffer[i] == 0x5B) || (FirstPart.Buffer[i] == 0x5D) || + (FirstPart.Buffer[i] == 0x7C)) + { + return FALSE; + } + /* First make sure the character it's not the Lead DBCS */ + if (FsRtlIsLeadDbcsCharacter(Name->Buffer[i])) + { + i++; + } + else if (FirstPart.Buffer[i] == '.') + { + LastDot = i; + + /* We mustn't have spaces before dot or at the end of the filename + * and no dot at the beginning of the filename */ + if ((i == (FirstPart.Length / sizeof(CHAR)) - 1) || i == 0) + return FALSE; + + if (i > 0) + if (FirstPart.Buffer[i - 1] == ' ') + return FALSE; + } + } + + /* Filename must be 8.3 filename and not 3.8 filename */ + if (LastDot && ((FirstPart.Length / sizeof(CHAR) - 1) - LastDot > 3)) + return FALSE; + + EndLoop: + /* Preparing next loop */ + Name.Buffer = RemainingPart.Buffer; + Name.Length = RemainingPart.Length; + Name.MaximumLength = RemainingPart.MaximumLength; + RtlZeroMemory(&FirstPart, sizeof(FirstPart)); + RtlZeroMemory(&RemainingPart, sizeof(RemainingPart)); + + /* Call once again our dissect function */ + FsRtlDissectDbcs(Name, &FirstPart, &RemainingPart); + + /* We found a pathname, it wasn't allowed */ + if (FirstPart.Length > 0 && !PathNamePermissible) + return FALSE; + } return TRUE; }
Modified: branches/pierre-fsd/ntoskrnl/ntoskrnl_arm.def URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/ntoskrnl/ntoskrnl_arm... ============================================================================== --- branches/pierre-fsd/ntoskrnl/ntoskrnl_arm.def [iso-8859-1] (original) +++ branches/pierre-fsd/ntoskrnl/ntoskrnl_arm.def [iso-8859-1] Fri Jul 18 04:15:53 2008 @@ -216,7 +216,7 @@ FsRtlInsertPerFileObjectContext FsRtlInsertPerStreamContext FsRtlIsDbcsInExpression -FsRtlIsFatDbcsLegal +FsRtlIsFatDbcsLegal@20 FsRtlIsHpfsDbcsLegal FsRtlIsNameInExpression FsRtlIsNtstatusExpected
Modified: branches/pierre-fsd/ntoskrnl/ntoskrnl_i386.def URL: http://svn.reactos.org/svn/reactos/branches/pierre-fsd/ntoskrnl/ntoskrnl_i38... ============================================================================== --- branches/pierre-fsd/ntoskrnl/ntoskrnl_i386.def [iso-8859-1] (original) +++ branches/pierre-fsd/ntoskrnl/ntoskrnl_i386.def [iso-8859-1] Fri Jul 18 04:15:53 2008 @@ -233,7 +233,7 @@ FsRtlInsertPerFileObjectContext FsRtlInsertPerStreamContext FsRtlIsDbcsInExpression -FsRtlIsFatDbcsLegal +FsRtlIsFatDbcsLegal@20 FsRtlIsHpfsDbcsLegal FsRtlIsNameInExpression@16 FsRtlIsNtstatusExpected@4
-
pschweitzer＠svn.reactos.org