// 3.3.1.P1 - Split the text into separate paragraphs.
// A paragraph separator is kept with the previous paragraph.
// Within each paragraph, apply all the other rules of this algorithm.
public static Paragraph[] SplitStringToParagraphs(string logicalString)
{
ArrayList ret = new ArrayList();
int i;
StringBuilder sb = new StringBuilder();
for (i = 0; i < logicalString.Length; ++i)
{
char c = logicalString[i];
BidiCharacterType cType = UnicodeCharacterDataResolver.GetBidiCharacterType(c);
if (cType == BidiCharacterType.B)
{
Paragraph p = new Paragraph(sb.ToString());
p.ParagraphSeparator = c;
ret.Add(p);
sb.Length = 0;
}
else
{
sb.Append(c);
}
}
if (sb.Length > 0) // string ended without a paragraph separator
{
ret.Add(new Paragraph(sb.ToString()));
}
return((Paragraph[])ret.ToArray(typeof(Paragraph)));
}
private char GetPairwiseComposition(char first, char second)
{
if (first < 0 || first > 0xFFFF || second < 0 || second > 0xFFFF)
{
return(BidiChars.NotAChar);
}
return(UnicodeCharacterDataResolver.Compose(first.ToString() + second.ToString()));
}
private void GetRecursiveDecomposition(bool canonical, char ch, StringBuilder builder)
{
string decomp = UnicodeCharacterDataResolver.GetUnicodeDecompositionMapping(ch);
if (decomp != null && !(canonical && UnicodeCharacterDataResolver.GetUnicodeDecompositionType(ch) != UnicodeDecompositionType.None))
{
for (int i = 0; i < decomp.Length; ++i)
{
GetRecursiveDecomposition(canonical, decomp[i], builder);
}
}
else // if no decomp, append
{
builder.Append(ch);
}
}
private StringBuilder InternalDecompose(ArrayList char_lengths)
{
StringBuilder target = new StringBuilder();
StringBuilder buffer = new StringBuilder();
_hasArabic = false;
_hasNSMs = false;
for (int i = 0; i < _text.Length; ++i)
{
BidiCharacterType ct = UnicodeCharacterDataResolver.GetBidiCharacterType(_text[i]);
_hasArabic |= ((ct == BidiCharacterType.AL) || (ct == BidiCharacterType.AN));
_hasNSMs |= (ct == BidiCharacterType.NSM);
buffer.Length = 0;
GetRecursiveDecomposition(false, _text[i], buffer);
char_lengths.Add(1 - buffer.Length);
// add all of the characters in the decomposition.
// (may be just the original character, if there was
// no decomposition mapping)
char ch;
for (int j = 0; j < buffer.Length; ++j)
{
ch = buffer[j];
UnicodeCanonicalClass chClass = UnicodeCharacterDataResolver.GetUnicodeCanonicalClass(ch);
int k = target.Length; // insertion point
if (chClass != UnicodeCanonicalClass.NR)
{
// bubble-sort combining marks as necessary
char ch2;
for (; k > 0; --k)
{
ch2 = target[k - 1];
if (UnicodeCharacterDataResolver.GetUnicodeCanonicalClass(ch2) <= chClass)
{
break;
}
}
}
target.Insert(k, ch);
}
}
return(target);
}
// 3.3.1 The Paragraph Level
// P2 - In each paragraph, find the first character of type L, AL, or R.
// P3 - If a character is found in P2 and it is of type AL or R, then
// set the paragraph embedding level to one; otherwise, set it to zero.
public void RecalculateParagraphEmbeddingLevel()
{
embedding_level = 1;
foreach (char c in _text)
{
BidiCharacterType cType = UnicodeCharacterDataResolver.GetBidiCharacterType(c);
if (cType == BidiCharacterType.R || cType == BidiCharacterType.AL)
{
embedding_level = 1;
break;
}
else if (cType == BidiCharacterType.L)
{
break;
}
}
}
private void InternalCompose(StringBuilder target, ArrayList char_lengths)
{
if (target.Length == 0)
{
return;
}
int starterPos = 0;
int compPos = 1;
int text_idx = 0;
char starterCh = target[0];
char_lengths[starterPos] = (int)char_lengths[starterPos] + 1;
UnicodeCanonicalClass lastClass = UnicodeCharacterDataResolver.GetUnicodeCanonicalClass(starterCh);
if (lastClass != UnicodeCanonicalClass.NR)
{
lastClass = (UnicodeCanonicalClass)256; // fix for strings staring with a combining mark
}
int oldLen = target.Length;
// Loop on the decomposed characters, combining where possible
char ch;
for (int decompPos = compPos; decompPos < target.Length; ++decompPos)
{
ch = target[decompPos];
UnicodeCanonicalClass chClass = UnicodeCharacterDataResolver.GetUnicodeCanonicalClass(ch);
char composite = GetPairwiseComposition(starterCh, ch);
UnicodeDecompositionType composeType = UnicodeCharacterDataResolver.GetUnicodeDecompositionType(composite);
if (composeType == UnicodeDecompositionType.None &&
composite != BidiChars.NotAChar &&
(lastClass < chClass || lastClass == UnicodeCanonicalClass.NR))
{
target[starterPos] = composite;
char_lengths[starterPos] = (int)char_lengths[starterPos] + 1;
// we know that we will only be replacing non-supplementaries by non-supplementaries
// so we don't have to adjust the decompPos
starterCh = composite;
}
else
{
if (chClass == UnicodeCanonicalClass.NR)
{
starterPos = compPos;
starterCh = ch;
text_idx++;
}
lastClass = chClass;
target[compPos] = ch;
//char_lengths[compPos] = (int)char_lengths[compPos] + 1;
int chkPos = compPos;
if ((int)char_lengths[chkPos] < 0)
{
while ((int)char_lengths[chkPos] < 0)
{
char_lengths[chkPos] = (int)char_lengths[chkPos] + 1;
char_lengths.Insert(compPos, 0);
chkPos++;
}
}
else
{
char_lengths[chkPos] = (int)char_lengths[chkPos] + 1;
}
if (target.Length != oldLen) // MAY HAVE TO ADJUST!
{
decompPos += target.Length - oldLen;
oldLen = target.Length;
}
++compPos;
}
}
target.Length = compPos;
char_lengths.RemoveRange(compPos, char_lengths.Count - compPos);
}
// 3.3.2 Explicit Levels and Directions
public void RecalculateCharactersEmbeddingLevels()
{
// This method is implemented in such a way it handles the string in logical order,
// rather than visual order, so it is easier to handle complex layouts. That is why
// it is placed BEFORE ReorderString rather than AFTER it, as its number suggests.
if (_hasArabic)
{
_text = PerformArabicShaping(_text);
}
_text_data = new CharData[_text.Length];
#region rules X1 - X9
// X1
byte embeddingLevel = EmbeddingLevel;
DirectionalOverrideStatus dos = DirectionalOverrideStatus.Neutral;
Stack dosStack = new Stack();
Stack elStack = new Stack();
int idx = 0;
for (int i = 0; i < _text.Length; ++i)
{
bool x9Char = false;
char c = _text[i];
_text_data[i]._ct = UnicodeCharacterDataResolver.GetBidiCharacterType(c);
_text_data[i]._char = c;
_text_data[i]._idx = idx;
idx += _char_lengths[i];
#region rules X2 - X5
// X2. With each RLE, compute the least greater odd embedding level.
// X4. With each RLO, compute the least greater odd embedding level.
if (c == BidiChars.RLE || c == BidiChars.RLO)
{
x9Char = true;
if (embeddingLevel < 60)
{
elStack.Push(embeddingLevel);
dosStack.Push(dos);
++embeddingLevel;
embeddingLevel |= 1;
if (c == BidiChars.RLE)
{
dos = DirectionalOverrideStatus.Neutral;
}
else
{
dos = DirectionalOverrideStatus.RTL;
}
}
}
// X3. With each LRE, compute the least greater even embedding level.
// X5. With each LRO, compute the least greater even embedding level.
else if (c == BidiChars.LRE || c == BidiChars.LRO)
{
x9Char = true;
if (embeddingLevel < 59)
{
elStack.Push(embeddingLevel);
dosStack.Push(dos);
embeddingLevel |= 1;
++embeddingLevel;
if (c == BidiChars.LRE)
{
dos = DirectionalOverrideStatus.Neutral;
}
else
{
dos = DirectionalOverrideStatus.LTR;
}
}
}
#endregion
#region rule X6
// X6. For all types besides RLE, LRE, RLO, LRO, and PDF: (...)
else if (c != BidiChars.PDF)
{
// a. Set the level of the current character to the current embedding level.
_text_data[i]._el = embeddingLevel;
//b. Whenever the directional override status is not neutral,
//reset the current character type to the directional override status.
if (dos == DirectionalOverrideStatus.LTR)
{
_text_data[i]._ct = BidiCharacterType.L;
}
else if (dos == DirectionalOverrideStatus.RTL)
{
_text_data[i]._ct = BidiCharacterType.R;
}
}
#endregion
#region rule X7
//Terminating Embeddings and Overrides
// X7. With each PDF, determine the matching embedding or override code.
// If there was a valid matching code, restore (pop) the last remembered (pushed)
// embedding level and directional override.
else if (c == BidiChars.PDF)
{
x9Char = true;
if (elStack.Count > 0)
{
embeddingLevel = (byte)(elStack.Pop());
dos = (DirectionalOverrideStatus)(dosStack.Pop());
}
}
#endregion
// X8. All explicit directional embeddings and overrides are completely
// terminated at the end of each paragraph. Paragraph separators are not
// included in the embedding.
if (x9Char || _text_data[i]._ct == BidiCharacterType.BN)
{
_text_data[i]._el = embeddingLevel;
}
}
#endregion
// X10. The remaining rules are applied to each run of characters at the same level.
int prevLevel = EmbeddingLevel;
int start = 0;
while (start < _text.Length)
{
#region rule X10 - run level setup
byte level = _text_data[start]._el;
BidiCharacterType sor = TypeForLevel(Math.Max(prevLevel, level));
int limit = start + 1;
while (limit < _text.Length && _text_data[limit]._el == level)
{
++limit;
}
byte nextLevel = limit < _text.Length ? _text_data[limit]._el : EmbeddingLevel;
BidiCharacterType eor = TypeForLevel(Math.Max(nextLevel, level));
#endregion
ResolveWeakTypes(start, limit, sor, eor);
ResolveNeutralTypes(start, limit, sor, eor, level);
ResolveImplicitTypes(start, limit, level);
prevLevel = level;
start = limit;
}
// Wrap lines
ReorderString(0, _text.Length);
FixMirroredCharacters();
ArrayList indexes = new ArrayList();
ArrayList lengths = new ArrayList();
StringBuilder sb = new StringBuilder();
foreach (CharData cd in _text_data)
{
sb.Append(cd._char);
indexes.Add(cd._idx);
lengths.Add(1);
}
_bidi_text = sb.ToString();
_bidi_indexes = (int[])indexes.ToArray(typeof(int));
}
public static ArabicShapeJoiningType GetArabicShapeJoiningType(char c)
{
if (c >= '\u0600' && c <= '\u0603')
{
return(ArabicShapeJoiningType.U);
}
if (c == '\u0608')
{
return(ArabicShapeJoiningType.U);
}
if (c == '\u060B')
{
return(ArabicShapeJoiningType.U);
}
if (c == '\u0621')
{
return(ArabicShapeJoiningType.U);
}
if (c >= '\u0622' && c <= '\u0625')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u0626')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0627')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u0628')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0629')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u062A' && c <= '\u062E')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u062F' && c <= '\u0632')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u0633' && c <= '\u063F')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0640')
{
return(ArabicShapeJoiningType.C);
}
if (c >= '\u0641' && c <= '\u0647')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0648')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u0649' && c <= '\u064A')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u066E' && c <= '\u066F')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u0671' && c <= '\u0673')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u0674')
{
return(ArabicShapeJoiningType.U);
}
if (c >= '\u0675' && c <= '\u0677')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u0678' && c <= '\u0687')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u0688' && c <= '\u0699')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u069A' && c <= '\u06BF')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u06C0')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u06C1' && c <= '\u06C2')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u06C3' && c <= '\u06CB')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u06CC')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u06CD')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u06CE')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u06CF')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u06D0' && c <= '\u06D1')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u06D2' && c <= '\u06D3')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u06D5')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u06DD')
{
return(ArabicShapeJoiningType.U);
}
if (c >= '\u06EE' && c <= '\u06EF')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u06FA' && c <= '\u06FC')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u06FF')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0710')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u0712' && c <= '\u0714')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u0715' && c <= '\u0719')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u071A' && c <= '\u071D')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u071E')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u071F' && c <= '\u0727')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0728')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u0729')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u072A')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u072B')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u072C')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u072D' && c <= '\u072E')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u072F')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u074D')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u074E' && c <= '\u0758')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u0759' && c <= '\u075B')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u075C' && c <= '\u076A')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u076B' && c <= '\u076C')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u076D' && c <= '\u0770')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u0771')
{
return(ArabicShapeJoiningType.R);
}
if (c == '\u0772')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u0773' && c <= '\u0774')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u0775' && c <= '\u0777')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u0778' && c <= '\u0779')
{
return(ArabicShapeJoiningType.R);
}
if (c >= '\u077A' && c <= '\u077F')
{
return(ArabicShapeJoiningType.D);
}
if (c >= '\u07CA' && c <= '\u07EA')
{
return(ArabicShapeJoiningType.D);
}
if (c == '\u07FA')
{
return(ArabicShapeJoiningType.C);
}
if (c == '\u200D')
{
return(ArabicShapeJoiningType.C);
}
UnicodeGeneralCategory ugc = UnicodeCharacterDataResolver.GetUnicodeGeneralCategory(c);
if (ugc == UnicodeGeneralCategory.Mn ||
ugc == UnicodeGeneralCategory.Me ||
ugc == UnicodeGeneralCategory.Cf)
{
return(ArabicShapeJoiningType.T);
}
return(ArabicShapeJoiningType.U);
}
