Advanced | technique | A reverse reference, referring to the matching group to the other part of the expression itself, for example, in matching HTML tags, we match a <a>, we want to match out a reference out, to find </a>, this time will use a reverse reference.
Syntax
A, reverse reference numbered group, syntax is \number
B, reverse reference named Group, syntax \k <name>
examples
A, matching pairs of HTML tags
@ "(? <tag> [^\s>]+) [^>]*>.*</\k <tag> >
B, matching two two overlapping occurrences of characters
public static void Main ()
{
string s = "AABBC11ASD";
Regex reg = new Regex (@ "(\w) \1");
matchcollection matches = Reg. Matches (s);
foreach (Match m in matches)
Console.WriteLine (M.value);
Console.ReadLine ();
}
return result is AA BB 11
Auxiliary matching group
The following group structures, the pattern in parentheses is not saved as part of the matching result
1, positive statement (? =)
Meaning: The pattern in parentheses must appear on the right side of the declaration, but not as part of the
public static void Main ()
{
string s = "c#.net,vb.net,php,java,jscript.net";
Regex reg = new Regex (@ "[\w\#]+ (? =\.net)", regexoptions.compiled);
MatchCollection mc = Reg. Matches (s);
foreach (Match m in MC)
Console.WriteLine (M.value);
Console.ReadLine ();
//Output C # VB JScript
}
can see that the matching engine requires matching. NET, but does not put. NET in matching results
2, negative statement (?!)
meaning: The pattern in parentheses must not appear on the right side of the declaration
The following example shows how to get the entire contents of a <a> label pair, even if it contains another HTML tag.
public static void Main ()
{
string newscontent = @ "url:<a href=" "1.html" ">test<span style=" "color:red;" >regex </span> </a>. ";
Regex regend = new Regex (@) <\s*a[^>]*> ([^<]|< (?!) /a)) * <\s*/a\s*> ", regexoptions.multiline);
Console.WriteLine (Regend.match (Newscontent). Value);
//result: <a Href= "1.html" >test<span style=" color:red; >regex </span> </a>
Console.ReadLine ();
}
3, Reverse positive statement (? <=)
meaning: The pattern in parentheses must appear on the left side of the declaration, but not as part of the match
4, Reverse negative statement (? <!)
meaning: The pattern in parentheses must not appear on the left side of the declaration
non-backtracking matching
syntax: (? >)
meaning: After the group matches, its matching characters cannot be used to match the subsequent expression by backtracking. Oh, see this sentence must not understand, I had to understand this also spent a lot of time, or through an example to explain it:
"Www.csdn.net" can go through @ "\w+\." (. *) \.\w+ "to match, but not through @" \w+\. (? >.*) \.\w+ "to match!" Why, then?
The reason is that a regular match is greedy, and it will match as many results as possible, so, in the example of the two regular formula, the. * Will match the csdn.net, this time, the first expression at the beginning of the match found \.\w+ no characters to match it, so it will backtrack, the so-called backtracking, is to . * The result of the match is pushed back, the characters that are left out are then used to match the \.\w+ until the \.\w+ match succeeds, and the entire expression returns the successful match result. The second expression, because of the use of a non-backtracking match, is not allowed to match the \.\w+ by backtracking, so the entire expression fails to match.
Note that backtracking matching is a very wasteful way to match a resource, so try to avoid your regular formula to successfully match by backtracking, as in the previous example, you can change to @ "\w+\." ([^\.] +\.) +\w+ "+".
