/** @file
  Unicode Collation Protocol (English)

Copyright (c) 2006 - 2008, Intel Corporation. <BR>
All rights reserved. This program and the accompanying materials
are licensed and made available under the terms and conditions of the BSD License
which accompanies this distribution.  The full text of the license may be found at
http://opensource.org/licenses/bsd-license.php

THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,
WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.

**/


#include "UnicodeCollationEng.h"

CHAR8 mEngUpperMap[0x100];
CHAR8 mEngLowerMap[0x100];
CHAR8 mEngInfoMap[0x100];

CHAR8 mOtherChars[] = {
  '0',
  '1',
  '2',
  '3',
  '4',
  '5',
  '6',
  '7',
  '8',
  '9',
  '\\',
  '.',
  '_',
  '^',
  '$',
  '~',
  '!',
  '#',
  '%',
  '&',
  '-',
  '{',
  '}',
  '(',
  ')',
  '@',
  '`',
  '\'',
  '\0'
};

STATIC EFI_HANDLE  mHandle = NULL;

//
// EFI Unicode Collation Protocol supporting ISO 639-2 language code
//
GLOBAL_REMOVE_IF_UNREFERENCED EFI_UNICODE_COLLATION_PROTOCOL  UnicodeEng = {
  EngStriColl,
  EngMetaiMatch,
  EngStrLwr,
  EngStrUpr,
  EngFatToStr,
  EngStrToFat,
  "eng"
};

//
// EFI Unicode Collation2 Protocol supporting RFC 3066 language code
//
GLOBAL_REMOVE_IF_UNREFERENCED EFI_UNICODE_COLLATION_PROTOCOL  Unicode2Eng = {
  EngStriColl,
  EngMetaiMatch,
  EngStrLwr,
  EngStrUpr,
  EngFatToStr,
  EngStrToFat,
  "en"
};

//
// Driver entry point.
//
EFI_STATUS
EFIAPI
InitializeUnicodeCollationEng (
  IN EFI_HANDLE       ImageHandle,
  IN EFI_SYSTEM_TABLE *SystemTable
  )
/*++

Routine Description:

  Initializes the Unicode Collation Driver

Arguments:

  ImageHandle -

  SystemTable -

Returns:

  EFI_SUCCESS
  EFI_OUT_OF_RESOURCES

--*/
{
  EFI_STATUS  Status;
  UINTN       Index;
  UINTN       Index2;

  //
  // Initialize mapping tables for the supported languages
  //
  for (Index = 0; Index < 0x100; Index++) {
    mEngUpperMap[Index] = (CHAR8) Index;
    mEngLowerMap[Index] = (CHAR8) Index;
    mEngInfoMap[Index]  = 0;

    if ((Index >= 'a' && Index <= 'z') || (Index >= 0xe0 && Index <= 0xf6) || (Index >= 0xf8 && Index <= 0xfe)) {

      Index2                = Index - 0x20;
      mEngUpperMap[Index]   = (CHAR8) Index2;
      mEngLowerMap[Index2]  = (CHAR8) Index;

      mEngInfoMap[Index] |= CHAR_FAT_VALID;
      mEngInfoMap[Index2] |= CHAR_FAT_VALID;
    }
  }

  for (Index = 0; mOtherChars[Index]; Index++) {
    Index2 = mOtherChars[Index];
    mEngInfoMap[Index2] |= CHAR_FAT_VALID;
  }

  if (FeaturePcdGet (PcdUnicodeCollation2Support)) {
    if (FeaturePcdGet (PcdUnicodeCollationSupport)) {
      Status = gBS->InstallMultipleProtocolInterfaces (
                      &mHandle,
                      &gEfiUnicodeCollationProtocolGuid,
                      &UnicodeEng,
                      &gEfiUnicodeCollation2ProtocolGuid,
                      &Unicode2Eng,
                      NULL
                      );
      ASSERT_EFI_ERROR (Status);
    } else {
      Status = gBS->InstallMultipleProtocolInterfaces (
                      &mHandle,
                      &gEfiUnicodeCollation2ProtocolGuid,
                      &Unicode2Eng,
                      NULL
                      );
      ASSERT_EFI_ERROR (Status);
    }
  } else {
    if (FeaturePcdGet (PcdUnicodeCollationSupport)) {
      Status = gBS->InstallMultipleProtocolInterfaces (
                      &mHandle,
                      &gEfiUnicodeCollationProtocolGuid,
                      &UnicodeEng,
                      NULL
                      );
      ASSERT_EFI_ERROR (Status);
    } else {
      //
      // This module must support to produce at least one of Unicode Collation Protocol
      // and Unicode Collation 2 Protocol.
      //
      ASSERT (FALSE);
      Status = EFI_UNSUPPORTED;
    }
  }

  return Status;
}

INTN
EFIAPI
EngStriColl (
  IN EFI_UNICODE_COLLATION_PROTOCOL   *This,
  IN CHAR16                           *s1,
  IN CHAR16                           *s2
  )
/*++

Routine Description:

  Performs a case-insensitive comparison of two Null-terminated Unicode strings.

Arguments:

  This
  s1
  s2

Returns:

--*/
{
  while (*s1) {
    if (ToUpper (*s1) != ToUpper (*s2)) {
      break;
    }

    s1 += 1;
    s2 += 1;
  }

  return ToUpper (*s1) - ToUpper (*s2);
}

VOID
EFIAPI
EngStrLwr (
  IN EFI_UNICODE_COLLATION_PROTOCOL   *This,
  IN OUT CHAR16                       *Str
  )
/*++

Routine Description:

  Converts all the Unicode characters in a Null-terminated Unicode string
  to lower case Unicode characters.

Arguments:

  This - A pointer to the EFI_UNICODE_COLLATION_PROTOCOL instance.
  Str1  - A pointer to a Null-terminated Unicode string.
  Str2  - A pointer to a Null-terminated Unicode string.

Returns:

  0   - s1 is equivalent to s2.
  > 0 - s1 is lexically greater than s2.
  < 0 - s1 is lexically less than s2.

--*/
{
  while (*Str) {
    *Str = ToLower (*Str);
    Str += 1;
  }
}

VOID
EFIAPI
EngStrUpr (
  IN EFI_UNICODE_COLLATION_PROTOCOL   *This,
  IN OUT CHAR16                       *Str
  )
/*++

Routine Description:

  Converts all the Unicode characters in a Null-terminated
  Unicode string to upper case Unicode characters.

Arguments:
  This
  Str

Returns:
  None

--*/
{
  while (*Str) {
    *Str = ToUpper (*Str);
    Str += 1;
  }
}

BOOLEAN
EFIAPI
EngMetaiMatch (
  IN EFI_UNICODE_COLLATION_PROTOCOL   *This,
  IN CHAR16                           *String,
  IN CHAR16                           *Pattern
  )
/*++

Routine Description:

  Performs a case-insensitive comparison between a Null-terminated
  Unicode pattern string and a Null-terminated Unicode string.

  The pattern string can use the '?' wildcard to match any character,
  and the '*' wildcard to match any sub-string.

Arguments:

  This     - A pointer to the EFI_UNICODE_COLLATION_PROTOCOL instance.
  String   - A pointer to a Null-terminated Unicode string.
  Pattern  - A pointer to a Null-terminated Unicode pattern string.

Returns:

  TRUE  - Pattern was found in String.
  FALSE - Pattern was not found in String.

--*/
{
  CHAR16  CharC;
  CHAR16  CharP;
  CHAR16  Index3;

  for (;;) {
    CharP = *Pattern;
    Pattern += 1;

    switch (CharP) {
    case 0:
      //
      // End of pattern.  If end of string, TRUE match
      //
      if (*String) {
        return FALSE;
      } else {
        return TRUE;
      }

    case '*':
      //
      // Match zero or more chars
      //
      while (*String) {
        if (EngMetaiMatch (This, String, Pattern)) {
          return TRUE;
        }

        String += 1;
      }

      return EngMetaiMatch (This, String, Pattern);

    case '?':
      //
      // Match any one char
      //
      if (!*String) {
        return FALSE;
      }

      String += 1;
      break;

    case '[':
      //
      // Match char set
      //
      CharC = *String;
      if (!CharC) {
        //
        // syntax problem
        //
        return FALSE;
      }

      Index3  = 0;
      CharP   = *Pattern++;
      while (CharP) {
        if (CharP == ']') {
          return FALSE;
        }

        if (CharP == '-') {
          //
          // if range of chars, get high range
          //
          CharP = *Pattern;
          if (CharP == 0 || CharP == ']') {
            //
            // syntax problem
            //
            return FALSE;
          }

          if (ToUpper (CharC) >= ToUpper (Index3) && ToUpper (CharC) <= ToUpper (CharP)) {
            //
            // if in range, it's a match
            //
            break;
          }
        }

        Index3 = CharP;
        if (ToUpper (CharC) == ToUpper (CharP)) {
          //
          // if char matches
          //
          break;
        }

        CharP = *Pattern++;
      }
      //
      // skip to end of match char set
      //
      while (CharP && CharP != ']') {
        CharP = *Pattern;
        Pattern += 1;
      }

      String += 1;
      break;

    default:
      CharC = *String;
      if (ToUpper (CharC) != ToUpper (CharP)) {
        return FALSE;
      }

      String += 1;
      break;
    }
  }
}

VOID
EFIAPI
EngFatToStr (
  IN EFI_UNICODE_COLLATION_PROTOCOL   *This,
  IN UINTN                            FatSize,
  IN CHAR8                            *Fat,
  OUT CHAR16                          *String
  )
/*++

Routine Description:

  Converts an 8.3 FAT file name using an OEM character set
  to a Null-terminated Unicode string.

  BUGBUG: Function has to expand DBCS FAT chars, currently not.

Arguments:
  This
  FatSize
  Fat
  String

Returns:

--*/
{
  //
  // No DBCS issues, just expand and add null terminate to end of string
  //
  while (*Fat && FatSize) {
    *String = *Fat;
    String += 1;
    Fat += 1;
    FatSize -= 1;
  }

  *String = 0;
}

BOOLEAN
EFIAPI
EngStrToFat (
  IN EFI_UNICODE_COLLATION_PROTOCOL   *This,
  IN CHAR16                           *String,
  IN UINTN                            FatSize,
  OUT CHAR8                           *Fat
  )
/*++

Routine Description:

  Converts a Null-terminated Unicode string to legal characters
  in a FAT filename using an OEM character set.

  Functions has to crunch string to a fat string. Replacing
  any chars that can't be represented in the fat name.

Arguments:
  This
  String
  FatSize
  Fat

Returns:
  TRUE
  FALSE
--*/
{
  BOOLEAN SpecialCharExist;

  SpecialCharExist = FALSE;
  while (*String && FatSize) {
    //
    // Skip '.' or ' ' when making a fat name
    //
    if (*String != '.' && *String != ' ') {
      //
      // If this is a valid fat char, move it.
      // Otherwise, move a '_' and flag the fact that the name needs an Lfn
      //
      if (*String < 0x100 && (mEngInfoMap[*String] & CHAR_FAT_VALID)) {
        *Fat = mEngUpperMap[*String];
      } else {
        *Fat              = '_';
        SpecialCharExist  = TRUE;
      }

      Fat += 1;
      FatSize -= 1;
    }

    String += 1;
  }
  //
  // Do not terminate that fat string
  //
  return SpecialCharExist;
}