public RegexCharClass(RegexBuffer buffer)
{
int startLoc = buffer.Offset;
buffer.MoveNext();
Regex regex;
Match match;
regex = new Regex(@"(?<Negated>\^?)(?<Class>.+?)\]");
match = regex.Match(buffer.String);
if (match.Success)
{
if (match.Groups["Negated"].ToString() == "^")
{
description = String.Format("Any character not in \"{0}\"",
match.Groups["Class"]);
}
else
{
description = String.Format("Any character in \"{0}\"",
match.Groups["Class"]);
}
buffer.Offset += match.Groups[0].Length;
}
else
{
description = "missing ']' in character class";
}
buffer.AddLookup(this, startLoc, buffer.Offset - 1);
}
void CheckClosingParen(RegexBuffer buffer)
{
// check for closing ")"
char current = ' ';
try
{
current = buffer.Current;
}
// no closing brace. Set highlight for this capture...
catch (Exception e)
{
buffer.ErrorLocation = startLocation;
buffer.ErrorLength = 1;
throw new Exception(
String.Format("Missing closing ')' in capture"), e);
}
if (current != ')')
{
throw new Exception(
String.Format("Unterminated closure at offset {0}",
buffer.Offset));
}
buffer.Offset++; // eat closing parenthesis
}
bool CheckBalancedGroup(RegexBuffer buffer)
{
// look for ?<Name1-Name2> or ?'Name1-Name2' syntax...
// look for ?<Name> or ?'Name' syntax...
Regex regex = new Regex(@"
^ # anchor to start of string
\?[\<|'] # ?< or ?'
(?<Name1>[a-zA-Z]+?) # Capture name1
-
(?<Name2>[a-zA-Z]+?) # Capture name2
[\>|'] # ?> or ?'
(?<Rest>.+) # The rest of the expression
" ,
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Balancing Group <{0}>-<{1}>", match.Groups["Name1"], match.Groups["Name2"]);
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return(true);
}
return(false);
}
string Interpret(string regex)
{
RegexBuffer buffer = new RegexBuffer(regex);
RegexExpression expression = new RegexExpression(buffer);
string output = expression.ToString(0);
return output;
}
public RegexCharClass(RegexBuffer buffer)
{
int startLoc = buffer.Offset;
buffer.MoveNext();
Regex regex;
Match match;
regex = new Regex(@"(?<Negated>\^?)(?<Class>.+?)\]");
match = regex.Match(buffer.String);
if (match.Success)
{
if (match.Groups["Negated"].ToString() == "^")
{
description = String.Format("Any character not in \"{0}\"",
match.Groups["Class"]);
}
else
{
description = String.Format("Any character in \"{0}\"",
match.Groups["Class"]);
}
buffer.Offset += match.Groups[0].Length;
}
else
{
description = "missing ']' in character class";
}
buffer.AddLookup(this, startLoc, buffer.Offset - 1);
}
// eat the whole comment until the end of line...
void EatComment(RegexBuffer buffer)
{
while (buffer.Current != '\r')
{
buffer.MoveNext();
}
}
bool CheckNamed(RegexBuffer buffer)
{
Regex regex;
Match match;
// look for ?<Name> or ?'Name' syntax...
regex = new Regex(@"
^ # anchor to start of string
\?(\<|') # ?< or ?'
(?<Name>[a-zA-Z0-9]+?) # Capture name
(\>|') # ?> or ?'
(?<Rest>.+) # The rest of the string
" ,
RegexOptions.IgnorePatternWhitespace);
match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Capture to <{0}>", match.Groups["Name"]);
// advance buffer to the rest of the expression
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return(true);
}
return(false);
}
void Parse(RegexBuffer buffer)
{
while (!buffer.AtEnd)
{
// if this regex ignores whitespace, we need to ignore these
if (buffer.IgnorePatternWhitespace &&
((buffer.Current == ' ') ||
(buffer.Current == '\r') ||
(buffer.Current == '\n') ||
(buffer.Current == '\t')))
{
buffer.MoveNext();
}
else
{
switch (buffer.Current)
{
case '(':
items.Add(new RegexCapture(buffer));
break;
case ')':
// end of closure; just return.
return;
case '[':
items.Add(new RegexCharClass(buffer));
break;
case '{':
items.Add(new RegexQuantifier(buffer));
break;
case '|':
items.Add(new RegexAlternate(buffer));
break;
case '\\':
items.Add(new RegexCharacter(buffer));
break;
case '#':
if (buffer.IgnorePatternWhitespace)
{
EatComment(buffer);
}
else
{
items.Add(new RegexCharacter(buffer));
}
break;
default:
items.Add(new RegexCharacter(buffer));
break;
}
}
}
}
string Interpret(string regex)
{
RegexBuffer buffer = new RegexBuffer(regex);
RegexExpression expression = new RegexExpression(buffer);
string output = expression.ToString(0);
return(output);
}
public RegexCapture(RegexBuffer buffer)
{
startLocation = buffer.Offset;
buffer.MoveNext();
// we're not in a series of normal characters, so clear
buffer.ClearInSeries();
// if the first character of the capture is a '?',
// we need to decode what comes after it.
if (buffer.Current == '?')
{
bool decoded = CheckNamed(buffer);
if (!decoded)
{
decoded = CheckBalancedGroup(buffer);
}
if (!decoded)
{
decoded = CheckNonCapturing(buffer);
}
if (!decoded)
{
decoded = CheckOptions(buffer);
}
if (!decoded)
{
decoded = CheckLookahead(buffer);
}
if (!decoded)
{
decoded = CheckNonBacktracking(buffer);
}
if (!decoded)
{
decoded = CheckConditional(buffer);
}
}
else
// plain old capture...
{
if (!HandlePlainOldCapture(buffer))
{
throw new Exception(
String.Format("Unrecognized capture: {0}", buffer.String));
}
}
buffer.AddLookup(this, startLocation, buffer.Offset - 1);
}
void DecodeEscape(RegexBuffer buffer)
{
buffer.MoveNext();
character = (string)escaped[buffer.Current];
if (character == null)
{
bool decoded = false;
decoded = CheckBackReference(buffer);
if (!decoded)
{
// TODO: Handle other items below:
switch (buffer.Current)
{
case 'u':
buffer.MoveNext();
string unicode = buffer.String.Substring(0, 4);
character = "Unicode " + unicode;
buffer.Offset += 4;
break;
case ' ':
character = "' ' (space)";
special = false;
buffer.MoveNext();
break;
case 'c':
buffer.MoveNext();
character = "CTRL-" + buffer.Current;
buffer.MoveNext();
break;
case 'x':
buffer.MoveNext();
string number = buffer.String.Substring(0, 2);
character = "Hex " + number;
buffer.Offset += 2;
break;
default:
character = new String(buffer.Current, 1);
special = false;
buffer.MoveNext();
break;
}
}
}
else
{
special = true;
buffer.MoveNext();
}
}
public RegexCapture(RegexBuffer buffer)
{
startLocation = buffer.Offset;
buffer.MoveNext();
// we're not in a series of normal characters, so clear
buffer.ClearInSeries();
// if the first character of the capture is a '?',
// we need to decode what comes after it.
if (buffer.Current == '?')
{
bool decoded = CheckNamed(buffer);
if (!decoded)
{
decoded = CheckBalancedGroup(buffer);
}
if (!decoded)
{
decoded = CheckNonCapturing(buffer);
}
if (!decoded)
{
decoded = CheckOptions(buffer);
}
if (!decoded)
{
decoded = CheckLookahead(buffer);
}
if (!decoded)
{
decoded = CheckNonBacktracking(buffer);
}
if (!decoded)
{
decoded = CheckConditional(buffer);
}
}
else
// plain old capture...
{
if (!HandlePlainOldCapture(buffer))
{
throw new Exception(
String.Format("Unrecognized capture: {0}", buffer.String));
}
}
buffer.AddLookup(this, startLocation, buffer.Offset - 1);
}
// Handle (?(expression)yes|no)
// when we get called, we're pointing to the first character of the expression
public RegexConditional(RegexBuffer buffer)
{
startLocation = buffer.Offset;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
yesNo = new RegexExpression(buffer);
CheckClosingParen(buffer);
buffer.AddLookup(this, startLocation, buffer.Offset - 1);
}
// Handle (?(expression)yes|no)
// when we get called, we're pointing to the first character of the expression
public RegexConditional(RegexBuffer buffer)
{
startLocation = buffer.Offset;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
yesNo = new RegexExpression(buffer);
CheckClosingParen(buffer);
buffer.AddLookup(this, startLocation, buffer.Offset - 1);
}
bool HandlePlainOldCapture(RegexBuffer buffer)
{
// we're already at the expression. Just create a new
// expression, and make sure that we're at a ")" when
// we're done
if (buffer.ExplicitCapture)
{
description = String.Format("Non-capturing Group");
}
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return(true);
}
public RegexQuantifier(RegexBuffer buffer)
{
int startLoc = buffer.Offset;
buffer.MoveNext();
Regex regex;
Match match;
// look for "n}", "n,}", or "n,m}"
regex = new Regex(@"(?<n>\d+)(?<Comma>,?)(?<m>\d*)\}");
match = regex.Match(buffer.String);
if (match.Success)
{
if (match.Groups["m"].Length != 0)
{
description = String.Format("At least {0}, but not more than {1} times",
match.Groups["n"], match.Groups["m"]);
}
else if (match.Groups["Comma"].Length != 0)
{
description = String.Format("At least {0} times",
match.Groups["n"]);
}
else
{
description = String.Format("Exactly {0} times",
match.Groups["n"]);
}
buffer.Offset += match.Groups[0].Length;
if (!buffer.AtEnd && buffer.Current == '?')
{
description += " (non-greedy)";
buffer.MoveNext();
}
}
else
{
description = "missing '}' in quantifier";
}
buffer.AddLookup(this, startLoc, buffer.Offset - 1);
}
bool CheckBackReference(RegexBuffer buffer)
{
// look for \k<name>
Regex regex = new Regex(@"
k\<(?<Name>.+?)\>
" ,
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
special = true;
this.character = String.Format("Backreference to match: {0}", match.Groups["Name"]);
buffer.Offset += match.Groups[0].Length;
return(true);
}
return(false);
}
public RegexQuantifier(RegexBuffer buffer)
{
int startLoc = buffer.Offset;
buffer.MoveNext();
Regex regex;
Match match;
// look for "n}", "n,}", or "n,m}"
regex = new Regex(@"(?<n>\d+)(?<Comma>,?)(?<m>\d*)\}");
match = regex.Match(buffer.String);
if (match.Success)
{
if (match.Groups["m"].Length != 0)
{
description = String.Format("At least {0}, but not more than {1} times",
match.Groups["n"], match.Groups["m"]);
}
else if (match.Groups["Comma"].Length != 0)
{
description = String.Format("At least {0} times",
match.Groups["n"]);
}
else
{
description = String.Format("Exactly {0} times",
match.Groups["n"]);
}
buffer.Offset += match.Groups[0].Length;
if (!buffer.AtEnd && buffer.Current == '?')
{
description += " (non-greedy)";
buffer.MoveNext();
}
}
else
{
description = "missing '}' in quantifier";
}
buffer.AddLookup(this, startLoc, buffer.Offset - 1);
}
bool CheckLookahead(RegexBuffer buffer)
{
Regex regex = new Regex(@"
^ # anchor to start of string
\?
(?<Assertion><=|<!|=|!) # assertion char
(?<Rest>.+) # The rest of the expression
" ,
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
switch (match.Groups["Assertion"].Value)
{
case "=":
description = "zero-width positive lookahead";
break;
case "!":
description = "zero-width negative lookahead";
break;
case "<=":
description = "zero-width positive lookbehind";
break;
case "<!":
description = "zero-width negative lookbehind";
break;
}
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return(true);
}
return(false);
}
bool CheckOptions(RegexBuffer buffer)
{
// look for ?imnsx-imnsx:
Regex regex = new Regex(@"
^ # anchor to start of string
\?(?<Options>[imnsx-]+):
" ,
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
string option = match.Groups["Options"].Value;
description = String.Format("Set options to {0}",
optionNames[option]);
expression = null;
buffer.Offset += match.Groups[0].Length;
return(true);
}
return(false);
}
bool CheckNonBacktracking(RegexBuffer buffer)
{
// Look for non-backtracking sub-expression ?>
Regex regex = new Regex(@"
^ # anchor to start of string
\?\>
(?<Rest>.+) # The rest of the expression
" ,
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Non-backtracking subexpressio");
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
this.CheckClosingParen(buffer);
return(true);
}
return(false);
}
void CheckClosingParen(RegexBuffer buffer)
{
// check for closing ")"
char current = ' ';
try
{
current = buffer.Current;
}
// no closing brace. Set highlight for this capture...
catch (Exception e)
{
buffer.ErrorLocation = startLocation;
buffer.ErrorLength = 1;
throw new Exception(
String.Format("Missing closing ')' in capture"), e);
}
if (current != ')')
{
throw new Exception(
String.Format("Unterminated closure at offset {0}",
buffer.Offset));
}
buffer.Offset++; // eat closing parenthesis
}
bool CheckConditional(RegexBuffer buffer)
{
// Look for conditional (?(name)yesmatch|nomatch)
// (name can also be an expression)
Regex regex = new Regex(@"
^ # anchor to start of string
\?\(
(?<Rest>.+) # The rest of the expression
" ,
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Conditional Subexpression");
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexConditional(buffer);
return(true);
}
return(false);
}
public RegexCharacter(RegexBuffer buffer)
{
int startLoc = buffer.Offset;
bool quantifier = false;
switch (buffer.Current)
{
case '.':
character = ". (any character)";
buffer.MoveNext();
special = true;
break;
case '+':
character = "+ (one or more times)";
buffer.MoveNext();
special = true;
quantifier = true;
break;
case '*':
character = "* (zero or more times)";
buffer.MoveNext();
special = true;
quantifier = true;
break;
case '?':
character = "? (zero or one time)";
buffer.MoveNext();
special = true;
quantifier = true;
break;
case '^':
character = "^ (anchor to start of string)";
buffer.MoveNext();
break;
case '$':
character = "$ (anchor to end of string)";
buffer.MoveNext();
break;
case ' ':
character = "' ' (space)";
buffer.MoveNext();
break;
case '\\':
DecodeEscape(buffer);
break;
default:
character = buffer.Current.ToString();
buffer.MoveNext();
special = false;
break;
}
if (quantifier)
{
if (!buffer.AtEnd && buffer.Current == '?')
{
character += " (non-greedy)";
buffer.MoveNext();
}
}
buffer.AddLookup(this, startLoc, buffer.Offset - 1, (character.Length == 1));
}
bool CheckBackReference(RegexBuffer buffer)
{
// look for \k<name>
Regex regex = new Regex(@"
k\<(?<Name>.+?)\>
",
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
special = true;
this.character = String.Format("Backreference to match: {0}", match.Groups["Name"]);
buffer.Offset += match.Groups[0].Length;
return true;
}
return false;
}
void DecodeEscape(RegexBuffer buffer)
{
buffer.MoveNext();
character = (string) escaped[buffer.Current];
if (character == null)
{
bool decoded = false;
decoded = CheckBackReference(buffer);
if (!decoded)
{
// TODO: Handle other items below:
switch (buffer.Current)
{
case 'u':
buffer.MoveNext();
string unicode = buffer.String.Substring(0, 4);
character = "Unicode " + unicode;
buffer.Offset += 4;
break;
case ' ':
character = "' ' (space)";
special = false;
buffer.MoveNext();
break;
case 'c':
buffer.MoveNext();
character = "CTRL-" + buffer.Current;
buffer.MoveNext();
break;
case 'x':
buffer.MoveNext();
string number = buffer.String.Substring(0, 2);
character = "Hex " + number;
buffer.Offset += 2;
break;
default:
character = new String(buffer.Current, 1);
special = false;
buffer.MoveNext();
break;
}
}
}
else
{
special = true;
buffer.MoveNext();
}
}
public RegexAlternate(RegexBuffer buffer)
{
buffer.AddLookup(this, buffer.Offset, buffer.Offset);
buffer.MoveNext(); // skip "|"
}
bool CheckConditional(RegexBuffer buffer)
{
// Look for conditional (?(name)yesmatch|nomatch)
// (name can also be an expression)
Regex regex = new Regex(@"
^ # anchor to start of string
\?\(
(?<Rest>.+) # The rest of the expression
",
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Conditional Subexpression");
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexConditional(buffer);
return true;
}
return false;
}
bool CheckOptions(RegexBuffer buffer)
{
// look for ?imnsx-imnsx:
Regex regex = new Regex(@"
^ # anchor to start of string
\?(?<Options>[imnsx-]+):
",
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
string option = match.Groups["Options"].Value;
description = String.Format("Set options to {0}",
optionNames[option]);
expression = null;
buffer.Offset += match.Groups[0].Length;
return true;
}
return false;
}
bool CheckLookahead(RegexBuffer buffer)
{
Regex regex = new Regex(@"
^ # anchor to start of string
\?
(?<Assertion><=|<!|=|!) # assertion char
(?<Rest>.+) # The rest of the expression
",
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
switch (match.Groups["Assertion"].Value)
{
case "=":
description = "zero-width positive lookahead";
break;
case "!":
description = "zero-width negative lookahead";
break;
case "<=":
description = "zero-width positive lookbehind";
break;
case "<!":
description = "zero-width negative lookbehind";
break;
}
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return true;
}
return false;
}
bool CheckNamed(RegexBuffer buffer)
{
Regex regex;
Match match;
// look for ?<Name> or ?'Name' syntax...
regex = new Regex(@"
^ # anchor to start of string
\?(\<|') # ?< or ?'
(?<Name>[a-zA-Z0-9]+?) # Capture name
(\>|') # ?> or ?'
(?<Rest>.+) # The rest of the string
",
RegexOptions.IgnorePatternWhitespace);
match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Capture to <{0}>", match.Groups["Name"]);
// advance buffer to the rest of the expression
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return true;
}
return false;
}
void Parse(RegexBuffer buffer)
{
while (!buffer.AtEnd)
{
// if this regex ignores whitespace, we need to ignore these
if (buffer.IgnorePatternWhitespace &&
((buffer.Current == ' ') ||
(buffer.Current == '\r') ||
(buffer.Current == '\n') ||
(buffer.Current == '\t')))
{
buffer.MoveNext();
}
else
{
switch (buffer.Current)
{
case '(':
items.Add(new RegexCapture(buffer));
break;
case ')':
// end of closure; just return.
return;
case '[':
items.Add(new RegexCharClass(buffer));
break;
case '{':
items.Add(new RegexQuantifier(buffer));
break;
case '|':
items.Add(new RegexAlternate(buffer));
break;
case '\\':
items.Add(new RegexCharacter(buffer));
break;
case '#':
if (buffer.IgnorePatternWhitespace)
{
EatComment(buffer);
}
else
{
items.Add(new RegexCharacter(buffer));
}
break;
default:
items.Add(new RegexCharacter(buffer));
break;
}
}
}
}
bool HandlePlainOldCapture(RegexBuffer buffer)
{
// we're already at the expression. Just create a new
// expression, and make sure that we're at a ")" when
// we're done
if (buffer.ExplicitCapture)
{
description = String.Format("Non-capturing Group");
}
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return true;
}
public RegexExpression(RegexBuffer buffer)
{
Parse(buffer);
}
bool CheckBalancedGroup(RegexBuffer buffer)
{
// look for ?<Name1-Name2> or ?'Name1-Name2' syntax...
// look for ?<Name> or ?'Name' syntax...
Regex regex = new Regex(@"
^ # anchor to start of string
\?[\<|'] # ?< or ?'
(?<Name1>[a-zA-Z]+?) # Capture name1
-
(?<Name2>[a-zA-Z]+?) # Capture name2
[\>|'] # ?> or ?'
(?<Rest>.+) # The rest of the expression
",
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Balancing Group <{0}>-<{1}>", match.Groups["Name1"], match.Groups["Name2"]);
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
CheckClosingParen(buffer);
return true;
}
return false;
}
// eat the whole comment until the end of line...
void EatComment(RegexBuffer buffer)
{
while (buffer.Current != '\r')
{
buffer.MoveNext();
}
}
public RegexExpression(RegexBuffer buffer)
{
Parse(buffer);
}
bool CheckNonCapturing(RegexBuffer buffer)
{
// Look for non-capturing ?:
Regex regex = new Regex(@"
^ # anchor to start of string
\?:
(?<Rest>.+) # The rest of the expression
",
RegexOptions.IgnorePatternWhitespace);
Match match = regex.Match(buffer.String);
if (match.Success)
{
description = String.Format("Non-capturing Group");
buffer.Offset += match.Groups["Rest"].Index;
expression = new RegexExpression(buffer);
this.CheckClosingParen(buffer);
return true;
}
return false;
}
public RegexCharacter(RegexBuffer buffer)
{
int startLoc = buffer.Offset;
bool quantifier = false;
switch (buffer.Current)
{
case '.':
character = ". (any character)";
buffer.MoveNext();
special = true;
break;
case '+':
character = "+ (one or more times)";
buffer.MoveNext();
special = true;
quantifier = true;
break;
case '*':
character = "* (zero or more times)";
buffer.MoveNext();
special = true;
quantifier = true;
break;
case '?':
character = "? (zero or one time)";
buffer.MoveNext();
special = true;
quantifier = true;
break;
case '^':
character = "^ (anchor to start of string)";
buffer.MoveNext();
break;
case '$':
character = "$ (anchor to end of string)";
buffer.MoveNext();
break;
case ' ':
character = "' ' (space)";
buffer.MoveNext();
break;
case '\\':
DecodeEscape(buffer);
break;
default:
character = buffer.Current.ToString();
buffer.MoveNext();
special = false;
break;
}
if (quantifier)
{
if (!buffer.AtEnd && buffer.Current == '?')
{
character += " (non-greedy)";
buffer.MoveNext();
}
}
buffer.AddLookup(this, startLoc, buffer.Offset - 1, (character.Length == 1));
}
public RegexAlternate(RegexBuffer buffer)
{
buffer.AddLookup(this, buffer.Offset, buffer.Offset);
buffer.MoveNext(); // skip "|"
}
