ktx/Utils/FileEncodingDetector.cs

/*
 * Copyright(C) Dreamgate co.,ltd. All rights reserved.
 */
using System;
using System.IO;
using System.Text;

namespace ktox.Utils
{
        /// <summary>
        /// ファイルのエンコーディングを確認します。
        /// </summary>
        public static class FileEncodingDetector
        {
                public static Encoding GetEncoding(string path)
                {
                        using (var fs = new FileStream(path, FileMode.Open, FileAccess.Read, FileShare.ReadWrite)) {
                                var buffer = new byte[1024];
                                var len = fs.Read(buffer, 0, 4);
                                Encoding result = GetEncodingFromBOM(buffer, len);
                                if (result != null) {
                                        return result;
                                }
                                fs.Seek(0, SeekOrigin.Begin);
                                for (var count = 0; count < 10; count++) {
                                        len = fs.Read(buffer, 0, buffer.Length);
                                        if (IsUTF8(buffer, len))
                                                return Encoding.UTF8;
                                        if (IsSJIS(buffer, len))
                                                return Encoding.GetEncoding(932);
                                }
                        }
                        return null;
                }
                
                private static Encoding GetEncodingFromBOM(byte[] buffer, int length)
                {
                        var bytes = new byte[length];
                        Array.Copy(buffer, bytes, length);
                        if (EqualsBytes(bytes, 0xff, 0xfe, 0x00, 0x00))
                                return Encoding.UTF32;
                        if (EqualsBytes(bytes, 0xfe, 0xff))
                                return Encoding.BigEndianUnicode;
                        if (EqualsBytes(bytes, 0xff, 0xfe))
                                return Encoding.Unicode;
                        if (EqualsBytes(bytes, 0xef, 0xbb, 0xbf))
                                return Encoding.UTF8;
                        return null;
                }
                
                private static bool EqualsBytes(byte[] a, params byte[] b)
                {
                        if (a.Length >= b.Length) {
                                for (int i = 0; i < b.Length; i++) {
                                        if (a[i] != b[i])
                                                return false;
                                }
                                return true;
                        }
                        return false;
                }
                
                private static bool IsUTF8(byte[] buffer, int length)
                {
                        bool multibyte = false;
                        for (int i = 0; i < length; i++) {
                                var b = buffer[i];
                                int charLen = 0;
                                if ((b & 0x80) == 0) {
                                        /* 1 byte */
                                        continue;
                                }
                                if ((b & 0xe0) == 0xc0) {
                                        /* 2 bytes */
                                        charLen = 2;
                                } else if ((b & 0xf0) == 0xe0) {
                                        /* 3 bytes */
                                        charLen = 3;
                                } else if ((b & 0xf8) == 0xf0) {
                                        /* 4 bytes */
                                        charLen = 4;
                                }
                                if (charLen == 0)
                                        return false;
                                for (int j = 1; j < charLen && i + j < length; j++) {
                                        if (!Is2ndByteOfUTF8(buffer[i + j]))
                                                return false;
                                }
                                multibyte = true;
                                i += charLen - 1;
                        }
                        return multibyte;
                }
                
                private static bool IsSJIS(byte[] buffer, int length)
                {
                        bool multibyte = false;
                        for (int i = 0; i < length; i++) {
                                var b = buffer[i];
                                if ((b & 0x80) == 0) {
                                        /* 1 byte */
                                        continue;
                                }
                                if (!IsLeadByteOfSJIS(b))
                                        return false;
                                if (i + 1 < length && !IsTrailByteOfSJIS(buffer[i + 1]))
                                        return false;
                                multibyte = true;
                                i++;
                        }
                        return multibyte;
                }
                
                private static bool Is2ndByteOfUTF8(byte b)
                {
                        return (b & 0xc0) == 0x80;
                }
                
                private static bool IsLeadByteOfSJIS(byte b)
                {
                        return (b >= 0x81 && b <= 0x9f) || (b >= 0xe0 && b <= 0xfc);
                }
                
                private static bool IsTrailByteOfSJIS(byte b)
                {
                        return (b >= 0x40 && b <= 0x7e) || (b >= 0x80 && b <= 0xfc);
                }
        }
}