/// <summary>
/// Gets a string representation of the charset entry
/// </summary>
/// <returns>The string representation of this charset entry</returns>
public override string ToString()
{
if (1 == Last - First)
{
return(string.Concat(RegexExpression.EscapeRangeChar(First), RegexExpression.EscapeRangeChar(Last)));
}
if (2 == Last - First)
{
return(string.Concat(RegexExpression.EscapeRangeChar(First), RegexExpression.EscapeRangeChar((char)(First + 1)), RegexExpression.EscapeRangeChar(Last)));
}
return(string.Concat(RegexExpression.EscapeRangeChar(First), "-", RegexExpression.EscapeRangeChar(Last)));
}
/// <summary>
/// Creates a new expression with the specified left and right hand sides
/// </summary>
/// <param name="left">The left expression</param>
/// <param name="right">The right expressions</param>
public RegexConcatExpression(RegexExpression left, params RegexExpression[] right)
{
Left = left;
for (int i = 0; i < right.Length; i++)
{
var r = right[i];
if (null == Right)
{
Right = r;
}
if (i != right.Length - 1)
{
var c = new RegexConcatExpression();
c.Left = Left;
c.Right = Right;
Right = null;
Left = c;
}
}
}
/// <summary>
/// Gets a string representation of the charset entry
/// </summary>
/// <returns>The string representation of this charset entry</returns>
public override string ToString()
{
return(string.Concat(RegexExpression.EscapeRangeChar(First), "-", RegexExpression.EscapeRangeChar(Last)));
}
/// <summary>
/// Gets a string representation of the charset entry
/// </summary>
/// <returns>The string representation of this charset entry</returns>
public override string ToString()
{
return(RegexExpression.EscapeRangeChar(Value));
}
/// <summary>
/// Creates a repeat expression with the specifed target expression, and minimum and maximum occurances
/// </summary>
/// <param name="expression">The target expression</param>
/// <param name="minOccurs">The minimum number of times the target expression can occur or -1</param>
/// <param name="maxOccurs">The maximum number of times the target expression can occur or -1</param>
public RegexRepeatExpression(RegexExpression expression, int minOccurs = -1, int maxOccurs = -1)
{
Expression = expression;
MinOccurs = minOccurs;
MaxOccurs = maxOccurs;
}
/// <summary>
/// Creates an optional expression using the specified target expression
/// </summary>
/// <param name="expression">The target expression to make optional</param>
public RegexOptionalExpression(RegexExpression expression)
{
Expression = expression;
}
static RegexExpression _ParseModifier(RegexExpression expr, ParseContext pc)
{
var line = pc.Line;
var column = pc.Column;
var position = pc.Position;
switch (pc.Current)
{
case '*':
expr = new RegexRepeatExpression(expr);
expr.SetLocation(line, column, position);
pc.Advance();
break;
case '+':
expr = new RegexRepeatExpression(expr, 1);
expr.SetLocation(line, column, position);
pc.Advance();
break;
case '?':
expr = new RegexOptionalExpression(expr);
expr.SetLocation(line, column, position);
pc.Advance();
break;
case '{':
pc.Advance();
pc.TrySkipWhiteSpace();
pc.Expecting('0', '1', '2', '3', '4', '5', '6', '7', '8', '9', ',', '}');
var min = -1;
var max = -1;
if (',' != pc.Current && '}' != pc.Current)
{
var l = pc.CaptureBuffer.Length;
pc.TryReadDigits();
min = int.Parse(pc.GetCapture(l));
pc.TrySkipWhiteSpace();
}
if (',' == pc.Current)
{
pc.Advance();
pc.TrySkipWhiteSpace();
pc.Expecting('0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '}');
if ('}' != pc.Current)
{
var l = pc.CaptureBuffer.Length;
pc.TryReadDigits();
max = int.Parse(pc.GetCapture(l));
pc.TrySkipWhiteSpace();
}
}
else
{
max = min;
}
pc.Expecting('}');
pc.Advance();
expr = new RegexRepeatExpression(expr, min, max);
expr.SetLocation(line, column, position);
break;
}
return(expr);
}
/// <summary>
/// Parses a regular expression from the specified <see cref="ParseContext"/>
/// </summary>
/// <param name="pc">The parse context to use</param>
/// <returns>A new abstract syntax tree representing the expression</returns>
public static RegexExpression Parse(ParseContext pc)
{
RegexExpression result = null, next = null;
int ich;
pc.EnsureStarted();
var line = pc.Line;
var column = pc.Column;
var position = pc.Position;
while (true)
{
switch (pc.Current)
{
case -1:
return(result);
case '.':
var nset = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetRangeEntry(char.MinValue, char.MaxValue) }, false);
nset.SetLocation(line, column, position);
if (null == result)
{
result = nset;
}
else
{
result = new RegexConcatExpression(result, nset);
result.SetLocation(line, column, position);
}
pc.Advance();
result = _ParseModifier(result, pc);
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case '\\':
pc.Advance();
pc.Expecting();
switch (pc.Current)
{
case 'd':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("digit") });
pc.Advance();
break;
case 'D':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("digit") }, true);
pc.Advance();
break;
case 'h':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("blank") });
pc.Advance();
break;
case 'l':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("lower") });
pc.Advance();
break;
case 's':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("space") });
pc.Advance();
break;
case 'S':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("space") }, true);
pc.Advance();
break;
case 'u':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("upper") });
pc.Advance();
break;
case 'w':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("word") });
pc.Advance();
break;
case 'W':
next = new RegexCharsetExpression(new RegexCharsetEntry[] { new RegexCharsetClassEntry("word") }, true);
pc.Advance();
break;
default:
if (-1 != (ich = _ParseEscapePart(pc)))
{
next = new RegexLiteralExpression((char)ich);
}
else
{
pc.Expecting(); // throw an error
return(null); // doesn't execute
}
break;
}
next.SetLocation(line, column, position);
next = _ParseModifier(next, pc);
if (null != result)
{
result = new RegexConcatExpression(result, next);
result.SetLocation(line, column, position);
}
else
{
result = next;
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case ')':
return(result);
case '(':
pc.Advance();
pc.Expecting();
next = Parse(pc);
pc.Expecting(')');
pc.Advance();
next = _ParseModifier(next, pc);
if (null == result)
{
result = next;
}
else
{
result = new RegexConcatExpression(result, next);
result.SetLocation(line, column, position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case '|':
if (-1 != pc.Advance())
{
next = Parse(pc);
result = new RegexOrExpression(result, next);
result.SetLocation(line, column, position);
}
else
{
result = new RegexOrExpression(result, null);
result.SetLocation(line, column, position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
case '[':
pc.ClearCapture();
pc.Advance();
pc.Expecting();
bool not = false;
if ('^' == pc.Current)
{
not = true;
pc.Advance();
pc.Expecting();
}
var ranges = _ParseRanges(pc);
if (ranges.Count == 0)
{
System.Diagnostics.Debugger.Break();
}
pc.Expecting(']');
pc.Advance();
next = new RegexCharsetExpression(ranges, not);
next.SetLocation(line, column, position);
next = _ParseModifier(next, pc);
if (null == result)
{
result = next;
}
else
{
result = new RegexConcatExpression(result, next);
result.SetLocation(pc.Line, pc.Column, pc.Position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
default:
ich = pc.Current;
next = new RegexLiteralExpression((char)ich);
next.SetLocation(line, column, position);
pc.Advance();
next = _ParseModifier(next, pc);
if (null == result)
{
result = next;
}
else
{
result = new RegexConcatExpression(result, next);
result.SetLocation(line, column, position);
}
line = pc.Line;
column = pc.Column;
position = pc.Position;
break;
}
}
}
static RegexExpression _FromFA <TAccept>(CharFA <TAccept> fa, HashSet <CharFA <TAccept> > visited)
{
if (!visited.Add(fa))
{
return(null);
}
var trgs = fa.FillInputTransitionRangesGroupedByState();
bool isAccepting = fa.IsAccepting;
RegexExpression expr = null;
foreach (var trg in trgs)
{
if (1 == trg.Value.Count && 1 == trg.Value[0].Length)
{
RegexExpression le = new RegexLiteralExpression(trg.Value[0][0]);
var next = _FromFA(trg.Key, visited);
if (null != next)
{
le = new RegexConcatExpression(le, next);
}
if (null == expr)
{
expr = le;
}
else
{
expr = new RegexOrExpression(expr, le);
}
}
else
{
var csel = new List <RegexCharsetEntry>();
foreach (var rng in trg.Value)
{
if (rng.First == rng.Last)
{
csel.Add(new RegexCharsetCharEntry(rng.First));
}
else
{
csel.Add(new RegexCharsetRangeEntry(rng.First, rng.Last));
}
}
RegexExpression cse = new RegexCharsetExpression(csel);
var next = _FromFA(trg.Key, visited);
if (null != next)
{
cse = new RegexConcatExpression(cse, next);
}
if (null == expr)
{
expr = cse;
}
else
{
expr = new RegexOrExpression(expr, cse);
}
}
}
var isLoop = false;
foreach (var val in fa.Descendants)
{
if (val == fa)
{
isLoop = true;
break;
}
}
if (isAccepting && !fa.IsFinal && !isLoop)
{
expr = new RegexOptionalExpression(expr);
}
return(expr);
}