Develop Reference

format string template with named parameters to literal c#

I have an application that creates string templates with named variables. This is done in accordance to the logging guide for ASP.NET Core
Now I find myself wanting to deliver these strings through the API itself as well, but this time with all the parameters filled in.
Basicly I'd want to use:
var template = "ID {ID} not found";
var para = new object[] {"value"};
String.Format(template, para);
However this gives an invalid input string.
Ofcourse I also cannot guarantee that somebody din't make a string template the 'classic' way with indexes.
var template2 = "ID {0} not found";
Is there a new way of formatting strings that I'm missing or are we supposed to work around this ?
I do not want to rework the existing code base to use numbers or use the $"...{para}" syntax. As this would lose information when it is being logged.
I'm guessing I could do a regex search and see if there's a '{0}' or a named parameter, and replace the named with indexes before formatting. But I wanted to know if there are some easier/cleaner ways of doing this.
Update - regex solution:
Bellow is the current work-around I've made using regex
public static class StringUtils
{
public static string Format(string template, params object[] para)
{
var match = Regex.Match(template, #"\{#?\w+}");
if (!match.Success) return template;
if (int.TryParse(match.Value.Substring(1, match.Value.Length - 2), out int n))
return string.Format(template, para);
else
{
var list = new List<string>();
var nextStartIndex = 0;
var i = 0;
while (match.Success)
{
if (match.Index > nextStartIndex)
list.Add(template.Substring(nextStartIndex , match.Index - nextStartIndex) + $"{{{i}}}");
else
list.Add($"{{{i}}}");
nextStartIndex = match.Index + match.Value.Length;
match = match.NextMatch();
i++;
}
return string.Format(string.Join("",list.ToArray()), para);
}
}
}

How can I find the first strong directionality character of a string in C#?

Assuming I get a string that can have mixed left-to-right and right-to-left content, I want to find the first strong directionality character in it, as defined here.
I think I found a good starting point in this question, but I still can't figure out how the BiDi category is related to the strong directionality characteristic. Is it possible to figure this out in C#?

Instead of relying on the internal implementation I took a slightly different approach that is open for optimizations but gives enough of a basis to answer your question.
I simply download the UnicodeData.txt that is part of the official release of a unicode version. That file contains for each unicodecharacter its number and some semicolon delimited fields. A typical line looks like this:
0041;LATIN CAPITAL LETTER A;Lu;0;L;;;;;N;;;;0061;
the fifth field contains the Bidirectional Class Value.
Armed with this knowledge the naive parser that reads the data and then inspects a demo string with it looks like this:
// hold chars with their Bidi Class Value
var udb = new Dictionary<char, string>();
// download UnicodeData txt file
var cli = new WebClient();
var data = cli.DownloadData("http://www.unicode.org/Public/UNIDATA/UnicodeData.txt");
// parse
using (var ms = new MemoryStream(data))
{
var sr = new StreamReader(ms, Encoding.UTF8);
var line = sr.ReadLine();
while (line != null)
{
var fields = line.Split(';');
int uc = int.Parse(fields[0], NumberStyles.HexNumber);
// above 0xffff we're lost
if (uc > 0xffff) break;
var ch = (char) uc;
var bca = fields[4];
udb.Add(ch, bca);
line = sr.ReadLine();
}
}
// test string
var s = "123A\xfb1d\x0620";
Console.WriteLine(s);
var pos = 0;
foreach(var c in s)
{
var bcv = udb[c]; // for a char get the Bidi Class Value
if (bcv == "L" || bcv == "R" || bcv == "AL")
{
Console.WriteLine(
"{0} - {1} : {2} [{3}]",
c,
pos,
CharUnicodeInfo.GetUnicodeCategory(c),
bcv);
}
pos++;
}
When run, you'll see the characters that are of the Strong Type and at which position they were found.

Converting HTML entities to Unicode Characters in C#

I found similar questions and answers for Python and Javascript, but not for C# or any other WinRT compatible language.
The reason I think I need it, is because I'm displaying text I get from websites in a Windows 8 store app. E.g. é should become é.
Or is there a better way? I'm not displaying websites or rss feeds, but just a list of websites and their titles.

I recommend using System.Net.WebUtility.HtmlDecode and NOT HttpUtility.HtmlDecode.
This is due to the fact that the System.Web reference does not exist in Winforms/WPF/Console applications and you can get the exact same result using this class (which is already added as a reference in all those projects).
Usage:
string s = System.Net.WebUtility.HtmlDecode("é"); // Returns é

Use HttpUtility.HtmlDecode() .Read on msdn here
decodedString = HttpUtility.HtmlDecode(myEncodedString)

This might be useful, replaces all (for as far as my requirements go) entities with their unicode equivalent.
public string EntityToUnicode(string html) {
var replacements = new Dictionary<string, string>();
var regex = new Regex("(&[a-z]{2,5};)");
foreach (Match match in regex.Matches(html)) {
if (!replacements.ContainsKey(match.Value)) {
var unicode = HttpUtility.HtmlDecode(match.Value);
if (unicode.Length == 1) {
replacements.Add(match.Value, string.Concat("&#", Convert.ToInt32(unicode[0]), ";"));
}
}
}
foreach (var replacement in replacements) {
html = html.Replace(replacement.Key, replacement.Value);
}
return html;
}

Different coding/encoding of HTML entities and HTML numbers in Metro App and WP8 App.
With Windows Runtime Metro App
{
string inStr = "ó";
string auxStr = System.Net.WebUtility.HtmlEncode(inStr);
// auxStr == ó
string outStr = System.Net.WebUtility.HtmlDecode(auxStr);
// outStr == ó
string outStr2 = System.Net.WebUtility.HtmlDecode("ó");
// outStr2 == ó
}
With Windows Phone 8.0
{
string inStr = "ó";
string auxStr = System.Net.WebUtility.HtmlEncode(inStr);
// auxStr == ó
string outStr = System.Net.WebUtility.HtmlDecode(auxStr);
// outStr == ó
string outStr2 = System.Net.WebUtility.HtmlDecode("ó");
// outStr2 == ó
}
To solve this, in WP8, I have implemented the table in HTML ISO-8859-1 Reference before calling System.Net.WebUtility.HtmlDecode().

This worked for me, replaces both common and unicode entities.
private static readonly Regex HtmlEntityRegex = new Regex("&(#)?([a-zA-Z0-9]*);");
public static string HtmlDecode(this string html)
{
if (html.IsNullOrEmpty()) return html;
return HtmlEntityRegex.Replace(html, x => x.Groups[1].Value == "#"
? ((char)int.Parse(x.Groups[2].Value)).ToString()
: HttpUtility.HtmlDecode(x.Groups[0].Value));
}
[Test]
[TestCase(null, null)]
[TestCase("", "")]
[TestCase("'fark'", "'fark'")]
[TestCase(""fark"", "\"fark\"")]
public void should_remove_html_entities(string html, string expected)
{
html.HtmlDecode().ShouldEqual(expected);
}

Improved Zumey method (I can`t comment there). Max char size is in the entity: &exclamation; (11). Upper case in the entities are also possible, ex. À (Source from wiki)
public string EntityToUnicode(string html) {
var replacements = new Dictionary<string, string>();
var regex = new Regex("(&[a-zA-Z]{2,11};)");
foreach (Match match in regex.Matches(html)) {
if (!replacements.ContainsKey(match.Value)) {
var unicode = HttpUtility.HtmlDecode(match.Value);
if (unicode.Length == 1) {
replacements.Add(match.Value, string.Concat("&#", Convert.ToInt32(unicode[0]), ";"));
}
}
}
foreach (var replacement in replacements) {
html = html.Replace(replacement.Key, replacement.Value);
}
return html;
}

RegEx - Please help in forming this RegEx

Can you please help me to write regular expression for this.
Name = "Windows Product for .Net"
Type = "Software Programming"
Quantity = "Pack of 3"
I want to do a match like this in c# for which I need RegEx.
If Name.contains(".Net") && (Type.Contains("Programming") || Type.Contains("Hardware")
{
// output will be a Match.
}
else
{
// no match.
}
The approach I want to take here is , specify regular expression for each condition and then apply logical operand && , logical grouping paranthesis and then logical operand ||.
I have come up with all regular expressions for these. How can I provide logical operands for each of them to execute in appropriate order?
string Name = "Windows Product for .Net";
string Type = "Software Programming";
string patternForName = ".*Net";
Regex rgxName = new Regex(patternForName);
Match matchName = rgx.Match(Name);
string patternForType = ".*Programming";
Regex rgxType = new Regex(patternForType);
Match matchType = rgx.Match(Type);
string patternForType1 = ".*Hardware";
Regex rgxType1 = new Regex(patternForType1);
Match matchType1 = rgx.Match(Type);
Please note - We are making it dynamic, in the sense the patterns , operands and regEx are coming from xml file. So that's why I do not want to write one big regEx for above.

First of all you don't need a leading .* in your expression unless you want the whole match (i.e. when working with matches). Just for a simple "is it there" you won't need it as the pattern might match any position.
Just use one regular expression for each field (i.e. one for Name, one for Type, one for Quantity:
string patternForName = "\\.Net"; // escaping the dot so it will match real dots only
string patternForType = "Programming|Hardware"; // | will result in "left side or right side"
string patternForQuantity = ".?"; // will match any string, even empty ones
To check everything:
bool match = rgxName.IsMatch(Name) && rgxType.IsMatch(Type) && rgx.IsMatch(Quantity);

You can make them dynamic without using regex. Using regex won't really save you any time or effort, since the code's going to be about the same size either way. Following your pattern above, you can do something like this:
var names = new[] { "Net", "Programming" };
var types = new[] { "Hardware" };
bool matchFound = true;
foreach (string n in names)
matchFound &= Name.Contains(n);
foreach (string t in types)
matchFound |= Type.Contains(t);
The above code assumes you want to match all of "names" and any of "types", but you can substitute any logic you want.
The real crux of your problem is these boolean combinations; regex won't help you with the logic for those, so you're better off using string.Contains unless the patterns you're looking for become much more variable. Regex is distracting you from your real goal here, in my opinion.

It sounds like you're asking how you should handle the logical part of the problem. If you're pulling it from an xml file, you could structure your file in the way you want to structure your logic.
for example, have And and Or groups:
<And>
<Name Match=".Net"/>
<Or>
<Type Match="Programming"/>
<Type Match="Hardware"/>
</Or>
</And>
Create classes for each of these types. For brevity, I didnt define the classes with properties or create constructors, but you can fill them out however you want:
class YourType
{
string Name;
string Type;
string Quantity;
}
abstract class Test
{
public abstract bool RunTest(YourType o);
}
class AndTest : Test
{
public List<Test> Children;
public bool RunTest(YourType o)
{
foreach (var test in Children)
{
if (!test.RunTest(o)) return false;
}
return true;
}
}
class OrTest : Test
{
public List<Test> Children;
public bool RunTest(YourType o)
{
foreach (var test in Children)
{
if (test.RunTest(o)) return true;
}
return false;
}
}
class NameTest : Test
{
public string Match;
public bool RunTest(YourType o)
{
return o.Name.Contains(Match);
}
}
class TypeTest : Test
{
public string Match;
public bool RunTest(YourType o)
{
return o.Type.Contains(Match);
}
}
Build the class structure from the xml file and just call RunTest from the top level Test. This way you can do any type of logic youd like. I just used Contains instead of a Regex for ease of the example, but you can easily replace the string match with a regex match.

if (rgxName.IsMatch(Name) && (rgxType.IsMatch(Type) || rgxType1.IsMatch(Type))
{
...
}

In .NET, Regex.Match matches anywhere in the string, so you don't need the any-characters (.*) prefix on your pattern. So, to check for ".NET", it would simply be:
Regex regexName = new Regex(#"\.NET", RegexOptions.IgnoreCase);
// IsMatch returns true/false, Match returns a Match object
bool nameMatches = regexName.IsMatch(name);
Your patterns for Programming and Hardware would just be
new Regex(#"Programming", RegexOptions.IgnoreCase) // Or leave out IgnoreCase if you're case-sensitive
new Regex(#"Hardware")
If you have a list of Name patterns and a list of type patterns, you could do something similar to this:
bool nameIsMatch = false;
bool typeIsMatch = false;
foreach (string namePattern in namePatterns)
{
nameIsMatch = nameIsMatch || Regex.IsMatch(nameString, namePattern);
}
foreach (string typePattern in typePatterns)
{
typeIsMatch = typeIsMatch || Regex.IsMatch(typeString, typePattern);
}
if (nameIsMatch && typeIsMatch)
{
// Whatever you want to do
}

patternForName = ".Net"
patternForType = "Programming"
patternForType1 = "Hardware"
You might find The Regex Coach to be useful.

What is the best way of performing a string concatenation in making a dynamic URL?

I am constructing a URL at runtime. So far I have done like
public string BuildURLAndNavigate(CodeType codeType)
{
string vURL = string.Empty;
string mstrNavServer = "http://some.com/nav";
vURL = ConcatenateString(mstrNavServer , "/somepage.asp?app=myapp");
//Build Code Type
switch (codeType)
{
case CodeType.Series:
vURL = ConcatenateString(vURL , "&tools=ser");
break;
case CodeType.DataType:
vURL = ConcatenateString(vURL , "&tools=dt");
break;
}
//build version
string VER_NUM = "5.0";
vURL = ConcatenateString(vURL , ConcatenateString("&vsn=" , VER_NUM));
return vURL;
}
private string ConcatenateString(string expression1, string expression2)
{
return string.Concat(expression1 + expression2);
}
But I am not happy with the one I am doing.
I am sure that there is definitely a best practice / better approach than this.
Kindly help me out in guiding for the same.
Thanks

You could use a StringBuilder:
public string BuildURLAndNavigate(CodeType codeType)
{
StringBuilder vURL = new StringBuilder();
vURL.Append("http://some.com/nav");
vURL.Append("/somepage.asp?app=myapp");
//Build Code Type
switch (codeType)
{
case CodeType.Series:
vURL.Append("&tools=ser");
break;
case CodeType.DataType:
vURL.Append("&tools=dt");
break;
}
//build version
string VER_NUM = "5.0";
vURL.AppendFormat("&vsn={0}", VER_NUM);
return vURL.ToString();
}

Never build urls using strings, string builders, string concatenations.
You could start by defining a custom collection which will take care of properly URL encoding any value being added to it:
public class HttpNameValueCollection : NameValueCollection
{
public override void Add(string name, string value)
{
base.Add(name, HttpUtility.UrlEncode(value));
}
public override string ToString()
{
return string.Join("&", Keys.Cast<string>().Select(
key => string.Format("{0}={1}", key, this[key])));
}
}
And then simply:
public string BuildURLAndNavigate()
{
var uriBuilder = new UriBuilder("http://some.com/nav/somepage.asp");
var values = new HttpNameValueCollection();
values.Add("app", "myapp");
switch (codeType)
{
case CodeType.Series:
values.Add("tools", "ser");
break;
case CodeType.DataType:
values.Add("tools", "dt");
break;
}
// You could even do things like this without breaking your urls
values.Add("p", "http://www.example.com?p1=v1&p2=v2");
string VER_NUM = "5.0";
values.Add("vsn", VER_NUM);
uriBuilder.Query = values.ToString();
return uriBuilder.ToString();
}

Like Saxon Druce said: You could use a StringBuilder, but, depending on CodeType values, you could eliminate the switch too:
public string BuildURLAndNavigate(CodeType codeType)
{
StringBuilder vURL = new StringBuilder();
vURL.Append("http://some.com/nav");
vURL.Append("/somepage.asp?app=myapp");
//Build Code Type
vURL.Append(codeType == CodeType.Series ? "&tools=ser" : "&tools=dt");
//build version
string VER_NUM = "5.0";
vURL.AppendFormat("&vsn={0}", VER_NUM);
return vURL.ToString();
}

Do
return string.Concat(expression1, expression2);
not
return string.Concat(expression1 + expression2);

wouldn't the right way to do that be to use the Uri-class or the UriBuilder class?
for example the Uri ctor overload Uri(Uri, string):
public Uri(
Uri baseUri,
string relativeUri
);
Uri baseUri = new Uri("http://www.contoso.com");
Uri myUri = new Uri(baseUri, "catalog/shownew.htm");
Console.WriteLine(myUri.ToString());
http://msdn.microsoft.com/en-us/library/aa332624(v=VS.71).aspx

What you are doing is fine - it is simple and understandable. Anyone who reads the code can understand what you are doing.
In terms of performance - you are not doing much string manipulation, so unless you are building huge strings or doing this operation thousands of times a minute, you will not gain much by using StringBuilder. Before optimizing this code, test its performance. You will probably find that there are other bigger bottlenecks to work on first.
The only real comment I have is that your ConcatenateString function seems superfluous. It is not really adding anything to the code and all the call to it can simply be replaced by string.Concat. As mentioned in the answer from #abatishchev, you should be using (str1, str2) not (str1 + str2), as that defeats the reason for the call.

Yes, StringBuilder is the best solution here. You can find more information on MSDN page: http://msdn.microsoft.com/en-us/library/system.text.stringbuilder.aspx
StringBuilder contains very usefull methods:
StringBuilder.Append
Appends information to the end of the current StringBuilder.
StringBuilder.AppendFormat
Replaces a format specifier passed in a string with formatted text.
StringBuilder.Insert
Inserts a string or object into the specified index of the current StringBuilder.
StringBuilder.Remove
Removes a specified number of characters from the current StringBuilder.
StringBuilder.Replace
Replaces a specified character at a specified index.

I'd personally be inclined to just string.format something like that:
public string BuildURLAndNavigate(CodeType codeType)
{
string vURL = "http://some.com/nav/somepage.asp?app=myapp&tools={0}&vsn={1}";
string codevalue = "";
//Build Code Type
switch (codeType)
{
case CodeType.Series:
codevalue = "ser";
break;
case CodeType.DataType:
codevalue = "dt";
break;
}
//build version
string version = "5.0";
return string.Format(vURL, codevalue, version);
}
}
Apologies if there are any mistakes in that, I wasn't in VS when writing it so there might be a few typos I didn't notice - you should get the idea though.
The reason I like this method is because you can immediately see what your url's overall form is which can make it a bit easier to understand roughly what the url being returned is.

In order to keep all variables at one place, could we use following solution
public string BuildURLAndNavigate(CodeType codeType)
{
//code here - switch statement to calculate the codeValue
//Anonymous type - To keep all variables at one place
var urlComponents = new {
server = "http://some.com/nav",
pageName="/somepage.asp?app=myapp",
codevalue = "", //Replace with the actual value calculated in switch statement
versionPart="&vsn=",
version = "5.0"
};
StringBuilder vURL = new StringBuilder();
vURL.Append(urlComponents.server);
vURL.Append(urlComponents.pageName);
vURL.Append(urlComponents.codevalue);
vURL.Append(urlComponents.versionPart);
vURL.Append(urlComponents.version);
return vURL.ToString();
}