Simple helper class for doing text scrubbing, cleaning, and formatting. Generally Regex's behind the scenes, with a few other dictionary mappings to help things move along. Access to a few of the Regex's special features such as maximum execution time and compiled cache size are controllable as well.
- Strip strings from other strings
- Replace by list of regexs
- Replace words by other words
- Translate characters from one set to another
- Pre-Defined list of useful Regex's (runtime expandable)
- Source on Github
// Map any character to any other character. The matchCarArray MUST be only
// have unique characters. The replaceChar array will have the matching translated char.
// The example below of accent chars, and their non-accented equiv
// Both strings must be 1 to 1 mapping and size of strings. This was done as strings
// to make it easier to deal with lots of characters. Can also add directly to the CharTransDict
// if you want instead of a set of strings.
string matchChar = "ŠŒŽšœžŸ¥µÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖØÙÚÛÜÝßàáâãäåæçèéêëìíîïðñòóôõöøùúûüýÿ¡¿";
string replaceChar = "SOZsozYYuAAAAAAACEEEEIIIIDNOOOOOOUUUUYsaaaaaaaceeeeiiiionoooooouuuuyy ";
// Set up a dictionary, if ignore case, set the dict up with a new comparer
// These words are mapped to any instances of other words. See comments
// on how this works vs regx, basically each word from a sentence is passed
// to the dictionary for translation. Current or past changes are not candidates
// for any further changes
StringComparer comparer = StringComparer.OrdinalIgnoreCase; // default is just Ordinal
Dictionary<string, string> wordDictionary = new Dictionary<string, string>(comparer)
{
{"chevrolet", "Ford"},
{"mAzDa", "BMW"},
{"and and", "and"} // will never match
};
// NOTE : Need `System.ValueTuple` package to do this style of init on v4.6 and below.
// Regx list each item is executed in order of the list.
// First element is the Regx match string (C# style) and the second
// is the replacement string if the pattern matches. Matches can affect the entire
// string, and each subsequent match can as well.
List<(string, string)> regxList = new List<(string, string)>
{ // Match, Replace
("BMW", "Fiat"), // swaps 'BMW' (case dependent) with 'Fiat'
(@"\s+", " "), // multi whitespace to 1 space
(@"^\s*|\s*$", "") // trims leading/ending spaces
};
// Test sentence with odd characters, spaces and other things needing scrubbing
string sentence = "¿¡Señor, the Chevrolet guys don't like Dodge guys, and and no one like MaZdA, Ola Senor?! ";
// Dump the orig string
Console.WriteLine("The Sentence : >{0}<", sentence);
Scrub st = new Scrub(sentence);
// Set dictionary up, case insensitive match
st.SetStringTranslator(wordDictionary, true);
// set up character translators
st.SetCharTranslator(matchChar, replaceChar);
// set up list of regx replaces
st.SetRegxTranslator(regxList);
// add a string translation after the fact
st.StringTransDict.Add("dodge", "Mercedes");
// add a Regx translation after the fact
st.RegxTuples.Add(("Senor", "Mr.Magoo"));
// add a chracter Translation after the fact
st.CharTransDict.Add('\'', '#');
// so all sorts of stuff!
string translated = st.Strip("[,]").MapChars().MapWords().RegxTranslate().Strip(@"Mr\.").ToString();
// Should be something like the string below -
// Magoo the Ford guys don#t like Mercedes guys and and no one like Fiat Ola Magoo?!
Console.WriteLine("Translated : >{0}<", translated);
// ** Test Pre-Defined Regex Patterns **
// reset the string with some emails
st.Set("Hank@kimball.com is sending an email to haystack@calhoon.com");
translated = st.RegxDefined("Email", "**Email Removed**").ToString();
Console.WriteLine("Masked : >{0}<", translated);
st.Set(" 前に来た時は北側からで、当時の光景はいまでも思い出せる。 Even now I remember the scene I saw approaching the city from the north. 青竜山脈から流れる川が湖へと流れこむ様、湖の**には純白のホ");
translated = st.RegxDefined("NonAscii", string.Empty).ToString();
Console.WriteLine("To all ASCII : >{0}<", translated);
// reset the string with some emails
st.Set(@"<h1>Title</h1><script>var a=1; \\comment</script> Not In Script Tags");
translated = st.RegxDefined("ScriptTags", string.Empty).RegxDefined("TagsSimple", string.Empty).ToString();
Console.WriteLine("Strip Script and Tags : >{0}<", translated);
// reset and set up a predefined match pattern and set regx case sensitivity
st.Set("wtf does RemoveWTF do? Is WtF Case SeNsItIvE?");
st.RegxMatchesDefined.Add("RemoveWTF", @"(wtf)|(what the)\s+(hell|$hit)");
translated = st.RegxIgnoreCase().RegxDefined("RemoveWTF", "XXX").ToString();
Console.WriteLine("New Pre-defined Match : >{0}<", translated);
More useful functionality, still basically a wrapper around regex stuff Add constant regex patterns for things like space removal, trim, etc. Currently Core 2.0 build.
Check out the Examples project directory on Github to see a general example of how it can be used.
The project has unit and integration tests. Also look at the tests for some additional use patterns.
Help with some ideas, code fixes are welcome. Use Github for opening request, bugs, etc.
MIT