正则表达式 - 匹配但排除？

Question

I have a very simple task of which I am trying to find and replace special characters within a string. My regex is working but sometimes there are italics tags within the string which I do not want to replace however; it is a requirement for me to replace independent "<" and ">" characters which is causing the italics tags to be morphed. Is there a way for me match the special characters but exclude the italics pattern? Here is my code:

string sampleText = "<i>This should be in italics</i> but this ¶ character needs to be removed"; 
string sPattern = "[―&<>♫♪–‧₢₳-⅓⅟□¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶•¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕ×ØÙÚÛÜÝÞßàáãäåæçèéêëìíîïðñòóôö÷øùüýþÿŒœŠšŸŽžƒ˜-‰›¢€°]";
string replacePattern = "";

string text = System.Text.RegularExpressions.Regex.Replace(sampleText, sPattern, replacePattern, System.Text.RegularExpressions.RegexOptions.IgnoreCase);

When my program executes I get this back:

iThis should be in italics/i but this character needs to be removed

So is it possible to for me to match my special characters but then exclude the italics tags? If this is not possible the only solution I can think of is removing the italics tags with some string processing and then validate the result with my regex then put the italics tags back in..

Any ideas?

Answer 1

Here's an easy way:

string sampleText = "<i>This should be in italics</i> but this ¶ character needs to be removed"; 
string sPattern = "(</?i>)|[―&<>♫♪–‧₢₳-⅓⅟□¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶•¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕ×ØÙÚÛÜÝÞßàáãäåæçèéêëìíîïðñòóôö÷øùüýþÿŒœŠšŸŽžƒ˜-‰›¢€°]";
string replacePattern = "$1";

string text = Regex.Replace(sampleText, sPattern, replacePattern, RegexOptions.IgnoreCase);

Console.WriteLine(text); 
// <i>This should be in italics</i> but this  character needs to be removed

But this will only work with <i> and </i> tags. You can expand that to other tags pretty easily (eg "(</?\\w+>)|..." for any simple tag without attributes) but if you get much more complicated than that, I'd recommend parsing the input as XML first, and only applying the pattern to the text of the nodes you're interested in.

Answer 2

You can use this:

string sPattern = @"(?i)[^<>a-z0-9\s\p{P}]+|<(?!/?i>)|(?<!</?i)>";
string replacePattern = "";

(You can replace \\p{P} by the punctuation you want to preserve)