mono/netcore/System.Private.CoreLib/shared/System/Text/EncodingNLS.cs

// Licensed to the .NET Foundation under one or more agreements.
// The .NET Foundation licenses this file to you under the MIT license.
// See the LICENSE file in the project root for more information.

using System;
using System.Collections;
using System.Globalization;
using System.Runtime.InteropServices;
using System.Threading;

namespace System.Text
{
    // This class overrides Encoding with the things we need for our NLS Encodings
    //
    // All of the GetBytes/Chars GetByte/CharCount methods are just wrappers for the pointer
    // plus decoder/encoder method that is our real workhorse.  Note that this is an internal
    // class, so our public classes cannot derive from this class.  Because of this, all of the
    // GetBytes/Chars GetByte/CharCount wrapper methods are duplicated in all of our public
    // encodings, which currently include:
    //
    //      EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, & UnicodeEncoding
    //
    // So if you change the wrappers in this class, you must change the wrappers in the other classes
    // as well because they should have the same behavior.

    internal abstract class EncodingNLS : Encoding
    {
        protected EncodingNLS(int codePage) : base(codePage)
        {
        }

        // Returns the number of bytes required to encode a range of characters in
        // a character array.
        // 
        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        // parent method is safe
        public override unsafe int GetByteCount(char[] chars, int index, int count)
        {
            // Validate input parameters
            if (chars == null)
                throw new ArgumentNullException(nameof(chars), SR.ArgumentNull_Array);

            if (index < 0 || count < 0)
                throw new ArgumentOutOfRangeException((index < 0 ? nameof(index) : nameof(count)), SR.ArgumentOutOfRange_NeedNonNegNum);

            if (chars.Length - index < count)
                throw new ArgumentOutOfRangeException(nameof(chars), SR.ArgumentOutOfRange_IndexCountBuffer);

            // If no input, return 0, avoid fixed empty array problem
            if (count == 0)
                return 0;

            // Just call the pointer version
            fixed (char* pChars = chars)
                return GetByteCount(pChars + index, count, null);
        }

        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        // parent method is safe
        public override unsafe int GetByteCount(string s)
        {
            // Validate input
            if (s==null)
                throw new ArgumentNullException(nameof(s));

            fixed (char* pChars = s)
                return GetByteCount(pChars, s.Length, null);
        }       

        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        public override unsafe int GetByteCount(char* chars, int count)
        {
            // Validate Parameters
            if (chars == null)
                throw new ArgumentNullException(nameof(chars), SR.ArgumentNull_Array);

            if (count < 0)
                throw new ArgumentOutOfRangeException(nameof(count), SR.ArgumentOutOfRange_NeedNonNegNum);
            
            // Call it with empty encoder
            return GetByteCount(chars, count, null);
        }

        // Parent method is safe.
        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        public override unsafe int GetBytes(string s, int charIndex, int charCount,
                                              byte[] bytes, int byteIndex)
        {
            if (s == null || bytes == null)
                throw new ArgumentNullException((s == null ? nameof(s) : nameof(bytes)), SR.ArgumentNull_Array);

            if (charIndex < 0 || charCount < 0)
                throw new ArgumentOutOfRangeException((charIndex < 0 ? nameof(charIndex) : nameof(charCount)), SR.ArgumentOutOfRange_NeedNonNegNum);

            if (s.Length - charIndex < charCount)
                throw new ArgumentOutOfRangeException(nameof(s), SR.ArgumentOutOfRange_IndexCount);

            if (byteIndex < 0 || byteIndex > bytes.Length)
                throw new ArgumentOutOfRangeException(nameof(byteIndex), SR.ArgumentOutOfRange_Index);

            int byteCount = bytes.Length - byteIndex;

            fixed (char* pChars = s) fixed (byte* pBytes = &MemoryMarshal.GetReference((Span<byte>)bytes))
                return GetBytes(pChars + charIndex, charCount, pBytes + byteIndex, byteCount, null);
        }

        // Encodes a range of characters in a character array into a range of bytes
        // in a byte array. An exception occurs if the byte array is not large
        // enough to hold the complete encoding of the characters. The
        // GetByteCount method can be used to determine the exact number of
        // bytes that will be produced for a given range of characters.
        // Alternatively, the GetMaxByteCount method can be used to
        // determine the maximum number of bytes that will be produced for a given
        // number of characters, regardless of the actual character values.
        // 
        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        // parent method is safe
        public override unsafe int GetBytes(char[] chars, int charIndex, int charCount,
                                               byte[] bytes, int byteIndex)
        {
            // Validate parameters
            if (chars == null || bytes == null)
                throw new ArgumentNullException((chars == null ? nameof(chars) : nameof(bytes)), SR.ArgumentNull_Array);

            if (charIndex < 0 || charCount < 0)
                throw new ArgumentOutOfRangeException((charIndex < 0 ? nameof(charIndex) : nameof(charCount)), SR.ArgumentOutOfRange_NeedNonNegNum);

            if (chars.Length - charIndex < charCount)
                throw new ArgumentOutOfRangeException(nameof(chars), SR.ArgumentOutOfRange_IndexCountBuffer);

            if (byteIndex < 0 || byteIndex > bytes.Length)
                throw new ArgumentOutOfRangeException(nameof(byteIndex), SR.ArgumentOutOfRange_Index);

            // If nothing to encode return 0, avoid fixed problem
            if (charCount == 0)
                return 0;

            // Just call pointer version
            int byteCount = bytes.Length - byteIndex;

            fixed (char* pChars = chars) fixed (byte* pBytes = &MemoryMarshal.GetReference((Span<byte>)bytes))
                // Remember that byteCount is # to decode, not size of array.
                return GetBytes(pChars + charIndex, charCount, pBytes + byteIndex, byteCount, null);
        }

        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        public override unsafe int GetBytes(char* chars, int charCount, byte* bytes, int byteCount)
        {
            // Validate Parameters
            if (bytes == null || chars == null)
                throw new ArgumentNullException(bytes == null ? nameof(bytes) : nameof(chars), SR.ArgumentNull_Array);

            if (charCount < 0 || byteCount < 0)
                throw new ArgumentOutOfRangeException((charCount < 0 ? nameof(charCount) : nameof(byteCount)),  SR.ArgumentOutOfRange_NeedNonNegNum);

            return GetBytes(chars, charCount, bytes, byteCount, null);
        }

        // Returns the number of characters produced by decoding a range of bytes
        // in a byte array.
        // 
        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        // parent method is safe
        public override unsafe int GetCharCount(byte[] bytes, int index, int count)
        {
            // Validate Parameters
            if (bytes == null)
                throw new ArgumentNullException(nameof(bytes), SR.ArgumentNull_Array);

            if (index < 0 || count < 0)
                throw new ArgumentOutOfRangeException((index < 0 ? nameof(index) : nameof(count)), SR.ArgumentOutOfRange_NeedNonNegNum);
                                   
            if (bytes.Length - index < count)
                throw new ArgumentOutOfRangeException(nameof(bytes), SR.ArgumentOutOfRange_IndexCountBuffer);

            // If no input just return 0, fixed doesn't like 0 length arrays
            if (count == 0)
                return 0;

            // Just call pointer version
            fixed (byte* pBytes = bytes)
                return GetCharCount(pBytes + index, count, null);
        }

        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        public override unsafe int GetCharCount(byte* bytes, int count)
        {
            // Validate Parameters
            if (bytes == null)
                throw new ArgumentNullException(nameof(bytes), SR.ArgumentNull_Array);

            if (count < 0)
                throw new ArgumentOutOfRangeException(nameof(count), SR.ArgumentOutOfRange_NeedNonNegNum);
            
            return GetCharCount(bytes, count, null);
        }

        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        // parent method is safe
        public override unsafe int GetChars(byte[] bytes, int byteIndex, int byteCount,
                                              char[] chars, int charIndex)
        {
            // Validate Parameters
            if (bytes == null || chars == null)
                throw new ArgumentNullException(bytes == null ? nameof(bytes) : nameof(chars), SR.ArgumentNull_Array);

            if (byteIndex < 0 || byteCount < 0)
                throw new ArgumentOutOfRangeException((byteIndex < 0 ? nameof(byteIndex) : nameof(byteCount)), SR.ArgumentOutOfRange_NeedNonNegNum);

            if ( bytes.Length - byteIndex < byteCount)
                throw new ArgumentOutOfRangeException(nameof(bytes), SR.ArgumentOutOfRange_IndexCountBuffer);

            if (charIndex < 0 || charIndex > chars.Length)
                throw new ArgumentOutOfRangeException(nameof(charIndex), SR.ArgumentOutOfRange_Index);

            // If no input, return 0 & avoid fixed problem
            if (byteCount == 0)
                return 0;

            // Just call pointer version
            int charCount = chars.Length - charIndex;

            fixed (byte* pBytes = bytes) fixed (char* pChars = &MemoryMarshal.GetReference((Span<char>)chars))
                // Remember that charCount is # to decode, not size of array
                return GetChars(pBytes + byteIndex, byteCount, pChars + charIndex, charCount, null);
        }

        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        public unsafe override int GetChars(byte* bytes, int byteCount, char* chars, int charCount)
        {
            // Validate Parameters
            if (bytes == null || chars == null)
                throw new ArgumentNullException(bytes == null ? nameof(bytes) : nameof(chars), SR.ArgumentNull_Array);

            if (charCount < 0 || byteCount < 0)
                throw new ArgumentOutOfRangeException((charCount < 0 ? nameof(charCount) : nameof(byteCount)), SR.ArgumentOutOfRange_NeedNonNegNum);
            
            return GetChars(bytes, byteCount, chars, charCount, null);
        }

        // Returns a string containing the decoded representation of a range of
        // bytes in a byte array.
        // 
        // All of our public Encodings that don't use EncodingNLS must have this (including EncodingNLS)
        // So if you fix this, fix the others.  Currently those include:
        // EncodingNLS, UTF7Encoding, UTF8Encoding, UTF32Encoding, ASCIIEncoding, UnicodeEncoding
        // parent method is safe
        public override unsafe string GetString(byte[] bytes, int index, int count)
        {
            // Validate Parameters
            if (bytes == null)
                throw new ArgumentNullException(nameof(bytes), SR.ArgumentNull_Array);

            if (index < 0 || count < 0)
                throw new ArgumentOutOfRangeException((index < 0 ? nameof(index) : nameof(count)), SR.ArgumentOutOfRange_NeedNonNegNum);

            if (bytes.Length - index < count)
                throw new ArgumentOutOfRangeException(nameof(bytes), SR.ArgumentOutOfRange_IndexCountBuffer);

            // Avoid problems with empty input buffer
            if (count == 0) return string.Empty;
            
            fixed (byte* pBytes = bytes)
                return string.CreateStringFromEncoding(
                    pBytes + index, count, this);
        }

        public override Decoder GetDecoder()
        {
            return new DecoderNLS(this);
        }

        public override Encoder GetEncoder()
        {
            return new EncoderNLS(this);
        }
    }
}