/// <summary>Scan the tables of which lines are inserted and deleted,
/// producing an edit script in forward order.
/// </summary>
/// dynamic array
private List<Difference> CreateDiffs(FileData DataA, FileData DataB)
{
List<Difference> a = new List<Difference>();
Difference aItem;
int StartA, StartB;
int LineA, LineB;
LineA = 0;
LineB = 0;
while (LineA < DataA.LineAmount || LineB < DataB.LineAmount)
{
if ((LineA < DataA.LineAmount) && (!DataA.ModifiedLinesState[LineA]) && (LineB < DataB.LineAmount) && (!DataB.ModifiedLinesState[LineB]))
{
// equal lines
LineA++;
LineB++;
}
else
{
// maybe deleted and/or inserted lines
StartA = LineA;
StartB = LineB;
while (LineA < DataA.LineAmount && (LineB >= DataB.LineAmount || DataA.ModifiedLinesState[LineA]))
{
LineA++;
}
while (LineB < DataB.LineAmount && (LineA >= DataA.LineAmount || DataB.ModifiedLinesState[LineB]))
{
LineB++;
}
if ((StartA < LineA) || (StartB < LineB))
{
// store a new difference-item
aItem = new Difference(StartA, StartB, LineA - StartA, LineB - StartB);
a.Add(aItem);
}
}
}
return a;
}
/// <summary>
/// This is the divide-and-conquer implementation of the longes common-subsequence (LCS) algorithm.
/// </summary>
/// <param name="DataA">sequence A</param>
/// <param name="LowerA">lower bound of the actual range in DataA</param>
/// <param name="UpperA">upper bound of the actual range in DataA (exclusive)</param>
/// <param name="DataB">sequence B</param>
/// <param name="LowerB">lower bound of the actual range in DataB</param>
/// <param name="UpperB">upper bound of the actual range in DataB (exclusive)</param>
/// <param name="DownVector">a vector for the (0,0) to (x,y) search. Passed as a parameter for speed reasons.</param>
/// <param name="UpVector">a vector for the (u,v) to (N,M) search. Passed as a parameter for speed reasons.</param>
private void LCS(FileData DataA, int LowerA, int UpperA, FileData DataB, int LowerB, int UpperB, int[] DownVector, int[] UpVector)
{
// Fast walkthrough equal lines at the start
while (LowerA < UpperA && LowerB < UpperB && DataA.LinesNumbers[LowerA] == DataB.LinesNumbers[LowerB])
{
LowerA++; LowerB++;
}
// Fast walkthrough equal lines at the end
while (LowerA < UpperA && LowerB < UpperB && DataA.LinesNumbers[UpperA - 1] == DataB.LinesNumbers[UpperB - 1])
{
--UpperA; --UpperB;
}
if (LowerA == UpperA)
{
// mark as inserted lines.
while (LowerB < UpperB)
{
DataB.ModifiedLinesState.Insert(LowerB++, true);
}
}
else if (LowerB == UpperB)
{
// mark as deleted lines.
while (LowerA < UpperA)
{
DataA.ModifiedLinesState.Insert(LowerA++, true);
}
}
else
{
// Find the middle snakea and length of an optimal path for A and B
SMSRD smsrd = GenerateSMSCoords(DataA, LowerA, UpperA, DataB, LowerB, UpperB, DownVector, UpVector);
// The path is from LowerX to (x,y) and (x,y) to UpperX
LCS(DataA, LowerA, smsrd.GetX, DataB, LowerB, smsrd.GetY, DownVector, UpVector);
LCS(DataA, smsrd.GetX, UpperA, DataB, smsrd.GetY, UpperB, DownVector, UpVector);
}
}
/// <summary>
/// Method generates SMS coords
/// </summary>
/// <param name="firstFileData"></param>
/// <param name="LowerA"></param>
/// <param name="UpperA"></param>
/// <param name="secondFileData"></param>
/// <param name="LowerB"></param>
/// <param name="UpperB"></param>
/// <param name="DownVector"></param>
/// <param name="UpVector"></param>
/// <returns></returns>
private static SMSRD GenerateSMSCoords(FileData firstFileData, int LowerA, int UpperA, FileData secondFileData, int LowerB, int UpperB, int[] DownVector, int[] UpVector)
{
int linesSummary = firstFileData.LineAmount + secondFileData.LineAmount + 1;
int DownK = LowerA - LowerB; // the k-line to start the forward search
int UpK = UpperA - UpperB; // the k-line to start the reverse search
int Delta = (UpperA - LowerA) - (UpperB - LowerB);
bool oddDelta = (Delta & 1) != 0;
// The vectors in the publication accepts negative indexes. the vectors implemented here are 0-based
// and are access using a specific offset: UpOffset UpVector and DownOffset for DownVektor
int DownOffset = linesSummary - DownK;
int UpOffset = linesSummary - UpK;
int MaxD = ((UpperA - LowerA + UpperB - LowerB) / 2) + 1;
// init vectors
DownVector[DownOffset + DownK + 1] = LowerA;
UpVector[UpOffset + UpK - 1] = UpperA;
for (int D = 0; D <= MaxD; D++)
{
// Extend the forward path.
for (int k = DownK - D; k <= DownK + D; k += 2)
{
// find the only or better starting point
int x, y;
if (k == DownK - D)
{
// down
x = DownVector[DownOffset + k + 1];
}
else
{
// a step to the right
x = DownVector[DownOffset + k - 1] + 1;
if ((k < DownK + D) && (DownVector[DownOffset + k + 1] >= x))
{
// down
x = DownVector[DownOffset + k + 1];
}
}
y = x - k;
// find the end of the furthest reaching forward D-path in diagonal k.
while ((x < UpperA) && (y < UpperB) && (firstFileData.LinesNumbers[x] == secondFileData.LinesNumbers[y]))
{
x++; y++;
}
DownVector[DownOffset + k] = x;
// overlap
if (oddDelta && (UpK - D < k) && (k < UpK + D))
{
if (UpVector[UpOffset + k] <= DownVector[DownOffset + k])
{
return new SMSRD(DownVector[DownOffset + k], DownVector[DownOffset + k] - k);
}
}
}
// Extend the reverse path.
for (int k = UpK - D; k <= UpK + D; k += 2)
{
// find the only or better starting point
int x, y;
if (k == UpK + D)
{
// up
x = UpVector[UpOffset + k - 1];
}
else
{
// left
x = UpVector[UpOffset + k + 1] - 1;
if ((k > UpK - D) && (UpVector[UpOffset + k - 1] < x))
{
// up
x = UpVector[UpOffset + k - 1];
}
}
y = x - k;
while ((x > LowerA) && (y > LowerB) && (firstFileData.LinesNumbers[x - 1] == secondFileData.LinesNumbers[y - 1]))
{
// diagonal
x--; y--;
}
UpVector[UpOffset + k] = x;
// overlap
if (!oddDelta && (DownK - D <= k) && (k <= DownK + D))
{
if (UpVector[UpOffset + k] <= DownVector[DownOffset + k])
{
return new SMSRD(DownVector[DownOffset + k], DownVector[DownOffset + k] - k);
}
}
}
}
return null;
}
/// <summary>
/// Find the difference in 2 arrays of integers.
/// </summary>
/// <param name="ArrayA">A-version of the numbers (usualy the old one)</param>
/// <param name="ArrayB">B-version of the numbers (usualy the new one)</param>
/// <returns>Returns a array of Items that describe the differences.</returns>
private List<Difference> DiffInt(List<int> ArrayA, List<int> ArrayB)
{
// The A-Version of the data (original data) to be compared.
FileData DataA = new FileData(ArrayA);
// The B-Version of the data (modified data) to be compared.
FileData DataB = new FileData(ArrayB);
int MAX = DataA.LineAmount + DataB.LineAmount + 1;
/// vector for the (0,0) to (x,y) search
int[] DownVector = new int[2 * MAX + 2];
/// vector for the (u,v) to (N,M) search
int[] UpVector = new int[2 * MAX + 2];
LCS(DataA, 0, DataA.LineAmount, DataB, 0, DataB.LineAmount, DownVector, UpVector);
return CreateDiffs(DataA, DataB);
}
/// <summary>
/// If a sequence of modified lines starts with a line that contains the same content
/// as the line that appends the changes, the difference sequence is modified so that the
/// appended line and not the starting line is marked as modified.
/// This leads to more readable diff sequences when comparing text files.
/// </summary>
/// <param name="Data">A Diff data buffer containing the identified changes.</param>
private void Optimize(FileData Data)
{
int StartPos, EndPos;
StartPos = 0;
while (StartPos < Data.LineAmount)
{
while ((StartPos < Data.LineAmount) && (Data.ModifiedLinesState[StartPos] == false))
{
StartPos++;
}
EndPos = StartPos;
while ((EndPos < Data.LineAmount) && (Data.ModifiedLinesState[EndPos] == true))
{
EndPos++;
}
if ((EndPos < Data.LineAmount) && (Data.LinesNumbers[StartPos] == Data.LinesNumbers[EndPos]))
{
Data.ModifiedLinesState[StartPos] = false;
Data.ModifiedLinesState[EndPos] = true;
}
else
{
StartPos = EndPos;
}
}
}
/// <summary>
/// Find the difference in 2 text documents, comparing by textlines.
/// The algorithm itself is comparing 2 arrays of numbers so when comparing 2 text documents
/// each line is converted into a (hash) number. This hash-value is computed by storing all
/// textlines into a common hashtable so i can find dublicates in there, and generating a
/// new number each time a new textline is inserted.
/// </summary>
/// <param name="TextA">A-version of the text (usualy the old one)</param>
/// <param name="TextB">B-version of the text (usualy the new one)</param>
/// <param name="trimSpace">When set to true, all leading and trailing whitespace characters are stripped out before the comparation is done.</param>
/// <param name="ignoreSpace">When set to true, all whitespace characters are converted to a single space character before the comparation is done.</param>
/// <param name="ignoreCase">When set to true, all characters are converted to their lowercase equivivalence before the comparation is done.</param>
/// <returns>Returns a array of Differences that describe the differences.</returns>
private List<Difference> DiffText(List<string> TextA, List<string> TextB, bool ignoreCase)
{
// prepare the input-text and convert to comparable numbers.
Hashtable h = new Hashtable(GetFileLength(TextA) + GetFileLength(TextB));
// The A-Version of the data (original data) to be compared.
FileData DataA = new FileData(DiffCodes(TextA, h, ignoreCase));
// The B-Version of the data (modified data) to be compared.
FileData DataB = new FileData(DiffCodes(TextB, h, ignoreCase));
h = null; // free up hashtable memory (maybe)
int MAX = DataA.LineAmount + DataB.LineAmount + 1;
/// vector for the (0,0) to (x,y) search
int[] DownVector = new int[2 * MAX + 2];
/// vector for the (u,v) to (N,M) search
int[] UpVector = new int[2 * MAX + 2];
LCS(DataA, 0, DataA.LineAmount, DataB, 0, DataB.LineAmount, DownVector, UpVector);
Optimize(DataA);
Optimize(DataB);
return CreateDiffs(DataA, DataB);
}
