/// <summary>
/// Changes fully translated Unicode data, e.g. with Japanaese Kanji characters, into
/// a string of characters that represents the raw, ISO character repertoire representation, i.e.
/// with escape sequences, but encoded as a Unicode string
/// </summary>
/// <param name="unicodeData">Fully translated Unicode data</param>
/// <param name="repertoire">Target repertoire to be transformed into</param>
/// <returns></returns>
private static string Encode(string unicodeData, CharacterSetInfo repertoire)
{
byte[] rawBytes;
Encode(unicodeData, repertoire, out rawBytes);
char[] rawCharacters = IsomorphicEncoding.GetChars(rawBytes);
return(new string(rawCharacters));
}
public Last(string specificCharacterSet, CharacterSetInfo defaultRepertoire,
Dictionary <string, CharacterSetInfo> extensionRepertoires)
{
SpecificCharacterSet = specificCharacterSet;
DefaultRepertoire = defaultRepertoire;
ExtensionRepertoires = extensionRepertoires;
}
public Last(string specificCharacterSet, CharacterSetInfo defaultRepertoire,
Dictionary<string, CharacterSetInfo> extensionRepertoires)
{
SpecificCharacterSet = specificCharacterSet;
DefaultRepertoire = defaultRepertoire;
ExtensionRepertoires = extensionRepertoires;
}
/// <summary>
/// Takes a string that is a representation of the raw sequence of bytes
/// encoded using an ISO repertoire, but current encoded as a Unicode string
/// and gives back a true Unicode string, e.g. containing the Japanese Kanji
/// characters
/// </summary>
/// <param name="rawData">Sequence of bytes formatted in Unicode</param>
/// <param name="repertoire">Original ISO repertoire used in the encoding</param>
/// <returns>True Unicode string</returns>
private static string Decode(string rawData, CharacterSetInfo repertoire)
{
// get it back to byte array form using a character set that includes
// both GR and GL areas (characters up to \xff in binary value)
// and it seems Windows-1252 works better than ISO-8859-1
byte[] rawBytes = IsomorphicEncoding.GetBytes(rawData);
return(Decode(rawBytes, repertoire));
}
/// <summary>
/// Takes a string that is encoded using the ISO repertoires, as a raw sequence
/// of bytes, and then gives back a fully translated Unicode representation, e.g.
/// with the correct Japanese Kanji characters in Unicode
/// </summary>
/// <param name="rawData">Byte sequence encoded using ISO repertoire</param>
/// <param name="repertoire">Repertoire the byte sequence is encoded using</param>
/// <returns>Unicode string</returns>
private static string Decode(byte[] rawData, CharacterSetInfo repertoire)
{
Encoding rawEncoding = Encoding.GetEncoding(repertoire.MicrosoftCodePage);
string rawDataDecoded = new string(rawEncoding.GetChars(rawData));
// get rid of any escape sequences, if they appear in the decoded string,
// like the case of Korean, using code page 20949 for some reason
if ("" != repertoire.G1Sequence)
{
return(rawDataDecoded.Replace(repertoire.G1Sequence, ""));
}
else
{
return(rawDataDecoded);
}
}
/// <summary>
/// Takes a string that is encoded using the ISO repertoires, as a raw sequence
/// of bytes, and then gives back a fully translated Unicode representation, e.g.
/// with the correct Japanese Kanji characters in Unicode
/// </summary>
/// <param name="rawData">Byte sequence encoded using ISO repertoire</param>
/// <param name="repertoire">Repertoire the byte sequence is encoded using</param>
/// <returns>Unicode string</returns>
private static string Decode(byte[] rawData, CharacterSetInfo repertoire)
{
Encoding rawEncoding = Encoding.GetEncoding(repertoire.MicrosoftCodePage);
string rawDataDecoded = new string(rawEncoding.GetChars(rawData));
// get rid of any escape sequences, if they appear in the decoded string,
// like the case of Korean, using code page 20949 for some reason
if ("" != repertoire.G1Sequence)
return rawDataDecoded.Replace(repertoire.G1Sequence, "");
else
return rawDataDecoded;
}
/// <summary>
/// Takes a string that is a representation of the raw sequence of bytes
/// encoded using an ISO repertoire, but current encoded as a Unicode string
/// and gives back a true Unicode string, e.g. containing the Japanese Kanji
/// characters
/// </summary>
/// <param name="rawData">Sequence of bytes formatted in Unicode</param>
/// <param name="repertoire">Original ISO repertoire used in the encoding</param>
/// <returns>True Unicode string</returns>
private static string Decode(string rawData, CharacterSetInfo repertoire)
{
// get it back to byte array form using a character set that includes
// both GR and GL areas (characters up to \xff in binary value)
// and it seems Windows-1252 works better than ISO-8859-1
byte[] rawBytes = Encoding.GetEncoding(IsomorphicCodePage).GetBytes(rawData);
return Decode(rawBytes, repertoire);
}
/// <summary>
/// Changes fully translated Unicode data, e.g. with Japanese Kanji characters, into
/// a raw byte array containing the 8-bit representation in the target repertoire
/// </summary>
/// <param name="unicodeData">Fully translated Unicode data</param>
/// <param name="repertoire">Target repertoire to be transformed into</param>
/// <param name="encoded">Output: byte array to hold the results</param>
private static void Encode(string unicodeData, CharacterSetInfo repertoire, out byte[] encoded)
{
byte[] rawBytes = Encoding.GetEncoding(repertoire.MicrosoftCodePage).GetBytes(unicodeData);
encoded = rawBytes;
}
/// <summary>
/// Changes fully translated Unicode data, e.g. with Japanaese Kanji characters, into
/// a string of characters that represents the raw, ISO character repertoire representation, i.e.
/// with escape sequences, but encoded as a Unicode string
/// </summary>
/// <param name="unicodeData">Fully translated Unicode data</param>
/// <param name="repertoire">Target repertoire to be transformed into</param>
/// <returns></returns>
private static string Encode(string unicodeData, CharacterSetInfo repertoire)
{
byte[] rawBytes;
Encode(unicodeData, repertoire, out rawBytes);
char[] rawCharacters = Encoding.GetEncoding(IsomorphicCodePage).GetChars(rawBytes);
return new string(rawCharacters);
}
private static void GetRepertoires(string specificCharacterSet, out CharacterSetInfo defaultRepertoire, out Dictionary<string, CharacterSetInfo> extensionRepertoires)
{
// TODO:
// Specific Character Set may have up to n values if
// Code Extensions are used. We accomodate for that here
// by parsing out all the different possible defined terms.
// At this point, however, we're not going to handle escaping
// between character sets from different code pages within
// a single string. For example, DICOM implies that you should
// be able to have JIS-encoded Japanese, ISO European characters,
// Thai characters and Korean characters on the same line, using
// Code Extensions (escape sequences). (Chinese is not included
// since the only support for Chinese is through GB18030 and
// UTF-8, both of which do not support Code Extensions.)
string[] specificCharacterSetValues = specificCharacterSet.Split('\\');
defaultRepertoire = null;
// set the default repertoire from Value 1
if (specificCharacterSetValues.GetUpperBound(0) >= 0)
{
if (!CharacterSetDatabase.TryGetValue(specificCharacterSetValues[0], out defaultRepertoire))
// we put in the default repertoire. Technically, it may
// not be ISO 2022 IR 6, but ISO_IR 6, but the information
// we want to use is the same
defaultRepertoire = CharacterSetDatabase["ISO 2022 IR 6"];
}
// Here we are accounting for cases where the same character sets are repeated, so
// we need to select out the unique ones. It should never really happen, but it
// does happen with a particular dataset when querying JDicom.
List<string> uniqueExtensionRepertoireDefinedTerms = new List<string>();
for (int i = 1; i < specificCharacterSetValues.Length; ++i)
{
string value = specificCharacterSetValues[i];
if (value != defaultRepertoire.DefinedTerm && !uniqueExtensionRepertoireDefinedTerms.Contains(value))
uniqueExtensionRepertoireDefinedTerms.Add(value);
}
// parse out the extension repertoires
extensionRepertoires = new Dictionary<string, CharacterSetInfo>();
foreach (string value in uniqueExtensionRepertoireDefinedTerms)
{
if (CharacterSetDatabase.ContainsKey(value) && !extensionRepertoires.ContainsKey(value))
{
// special robustness handling of GB18030 and UTF-8
if ("GB18030" == value || "ISO_IR 192" == value)
{
// these two character sets can't use code extensions, so there should really only be 1
// character set in the repertoire
extensionRepertoires.Clear();
extensionRepertoires.Add(value, CharacterSetDatabase[value]);
break;
}
extensionRepertoires.Add(value, CharacterSetDatabase[value]);
}
else if (!extensionRepertoires.ContainsKey("ISO 2022 IR 6"))
{
// we put in the default repertoire. Technically, it may
// not be ISO 2022 IR 6, but ISO_IR 6, but the information
// we want to use is the same
extensionRepertoires.Add(value, SpecificCharacterSetParser.CharacterSetDatabase["ISO 2022 IR 6"]);
}
}
}
/// <summary>
/// Changes fully translated Unicode data, e.g. with Japanese Kanji characters, into
/// a raw byte array containing the 8-bit representation in the target repertoire
/// </summary>
/// <param name="unicodeData">Fully translated Unicode data</param>
/// <param name="repertoire">Target repertoire to be transformed into</param>
/// <param name="encoded">Output: byte array to hold the results</param>
private static void Encode(string unicodeData, CharacterSetInfo repertoire, out byte[] encoded)
{
byte[] rawBytes = Encoding.GetEncoding(repertoire.MicrosoftCodePage).GetBytes(unicodeData);
encoded = rawBytes;
}
private static void GetRepertoires(string specificCharacterSet, out CharacterSetInfo defaultRepertoire, out Dictionary <string, CharacterSetInfo> extensionRepertoires)
{
//Most of the time, especially on the same thread, the specific character set will be the same.
//This simple check avoids having to figure it out over and over again, which gets expensive.
var last = _last;
if (last != null && specificCharacterSet == last.SpecificCharacterSet)
{
defaultRepertoire = last.DefaultRepertoire;
extensionRepertoires = last.ExtensionRepertoires;
return;
}
// TODO:
// Specific Character Set may have up to n values if
// Code Extensions are used. We accomodate for that here
// by parsing out all the different possible defined terms.
// At this point, however, we're not going to handle escaping
// between character sets from different code pages within
// a single string. For example, DICOM implies that you should
// be able to have JIS-encoded Japanese, ISO European characters,
// Thai characters and Korean characters on the same line, using
// Code Extensions (escape sequences). (Chinese is not included
// since the only support for Chinese is through GB18030 and
// UTF-8, both of which do not support Code Extensions.)
string[] specificCharacterSetValues = specificCharacterSet.Split('\\');
defaultRepertoire = null;
// set the default repertoire from Value 1
if (specificCharacterSetValues.GetUpperBound(0) >= 0)
{
if (!CharacterSetDatabase.TryGetValue(specificCharacterSetValues[0], out defaultRepertoire))
{
// we put in the default repertoire. Technically, it may
// not be ISO 2022 IR 6, but ISO_IR 6, but the information
// we want to use is the same
defaultRepertoire = CharacterSetDatabase["ISO 2022 IR 6"];
}
}
// Here we are accounting for cases where the same character sets are repeated, so
// we need to select out the unique ones. It should never really happen, but it
// does happen with a particular dataset when querying JDicom.
List <string> uniqueExtensionRepertoireDefinedTerms = new List <string>();
for (int i = 1; i < specificCharacterSetValues.Length; ++i)
{
string value = specificCharacterSetValues[i];
if (value != defaultRepertoire.DefinedTerm && !uniqueExtensionRepertoireDefinedTerms.Contains(value))
{
uniqueExtensionRepertoireDefinedTerms.Add(value);
}
}
// parse out the extension repertoires
extensionRepertoires = new Dictionary <string, CharacterSetInfo>();
foreach (string value in uniqueExtensionRepertoireDefinedTerms)
{
if (CharacterSetDatabase.ContainsKey(value) && !extensionRepertoires.ContainsKey(value))
{
// special robustness handling of GB18030 and UTF-8
if ("GB18030" == value || "ISO_IR 192" == value)
{
// these two character sets can't use code extensions, so there should really only be 1
// character set in the repertoire
extensionRepertoires.Clear();
extensionRepertoires.Add(value, CharacterSetDatabase[value]);
break;
}
extensionRepertoires.Add(value, CharacterSetDatabase[value]);
}
else if (!extensionRepertoires.ContainsKey("ISO 2022 IR 6"))
{
// we put in the default repertoire. Technically, it may
// not be ISO 2022 IR 6, but ISO_IR 6, but the information
// we want to use is the same
extensionRepertoires.Add(value, CharacterSetDatabase["ISO 2022 IR 6"]);
}
}
_last = new Last(specificCharacterSet, defaultRepertoire, extensionRepertoires);
}
