public CommandProcessingResult ProcessCommand(CommandContext context)
{
var source = context.CommandSource;
var position = context.ArgumentsStartPosition;
var topFormula = context.Parser.Parse(
TexFormulaParser.ReadElement(source, ref position),
context.Formula.TextStyle,
context.Environment.CreateChildEnvironment());
var bottomFormula = context.Parser.Parse(
TexFormulaParser.ReadElement(source, ref position),
context.Formula.TextStyle,
context.Environment.CreateChildEnvironment());
var start = context.CommandNameStartPosition;
var atomSource = source.Segment(start, position - start);
var topAtom = new List <Atom?> {
topFormula.RootAtom
};
var bottomAtom = new List <Atom?> {
bottomFormula.RootAtom
};
var atoms = new List <List <Atom?> > {
topAtom, bottomAtom
};
var matrixAtom = new MatrixAtom(atomSource, atoms, MatrixCellAlignment.Center);
var left = new SymbolAtom(atomSource, "(", TexAtomType.Opening, true);
var right = new SymbolAtom(atomSource, ")", TexAtomType.Closing, true);
var fencedAtom = new FencedAtom(atomSource, matrixAtom, left, right);
return(new CommandProcessingResult(fencedAtom, position));
}
public CommandProcessingResult ProcessCommand(CommandContext context)
{
var source = context.CommandSource;
var prevFormulaAtom = context.Formula.RootAtom;
var nextLineAtom = context.Parser.Parse(
source.Segment(context.ArgumentsStartPosition),
context.Formula.TextStyle,
context.Environment).RootAtom;
// An optimization: if the new content itself is a matrix with suitable parameters, then we won't
// wrap it into another formula, but will combine it with the content on top.
var newMatrix = nextLineAtom is MatrixAtom m &&
m.MatrixCellAlignment == MatrixCellAlignment.Left &&
m.HorizontalPadding == MatrixAtom.DefaultPadding &&
m.VerticalPadding == MatrixAtom.DefaultPadding
? m
: null;
var topRow = new[] { prevFormulaAtom };
var rows = new List <IEnumerable <Atom?> > {
topRow
};
if (newMatrix != null)
{
rows.AddRange(newMatrix.MatrixCells);
}
else
{
var bottomRow = new[] { nextLineAtom };
rows.Add(bottomRow);
}
// We'll always use source = null for the resulting matrix, because it's a structural element and not a
// useful atom generated from any particular sources.
var atom = new MatrixAtom(null, rows, MatrixCellAlignment.Left);
var position = source.Length; // we always parse the provided source until the end
return(new CommandProcessingResult(atom, position, AtomAppendMode.Replace));
}
public CommandProcessingResult ProcessCommand(CommandContext context)
{
var position = context.ArgumentsStartPosition;
var source = context.CommandSource;
if (position == source.Length)
{
throw new TexParseException("illegal end!");
}
var cellsSource = TexFormulaParser.ReadElement(source, ref position);
var matrixSource = context.CommandSource.Segment(
context.CommandNameStartPosition,
position - context.CommandNameStartPosition);
var cells = ReadMatrixCells(context.Parser, context.Formula, cellsSource, context.Environment);
var matrix = new MatrixAtom(matrixSource, cells, _cellAlignment);
SymbolAtom GetDelimiter(string name) =>
name == null
? null
: TexFormulaParser.GetDelimiterSymbol(name, null) ??
throw new TexParseException($"The delimiter {name} could not be found");
var leftDelimiter = GetDelimiter(_leftDelimiterSymbolName);
var rightDelimiter = GetDelimiter(_rightDelimiterSymbolName);
var atom = leftDelimiter == null && rightDelimiter == null
? (Atom)matrix
: new FencedAtom(
matrixSource,
matrix,
leftDelimiter,
rightDelimiter);
return(new CommandProcessingResult(atom, position));
}
