This question already has answers here:
Is there a "String.Format" that can accept named input parameters instead of index placeholders? [duplicate]
(9 answers)
Closed 4 years ago.
I'm trying to format some string dynamically with available variables in a specific context/scope.
This strings would have parts with things like {{parameter1}}, {{parameter2}} and these variables would exist in the scope where I'll try to reformat the string. The variable names should match.
I looked for something like a dynamically string interpolation approach, or how to use FormattableStringFactory, but I found nothing that really gives me what I need.
var parameter1 = DateTime.Now.ToString();
var parameter2 = "Hello world!";
var retrievedString = "{{parameter2}} Today we're {{parameter1}}";
var result = MagicMethod(retrievedString, parameter1, parameter2);
// or, var result = MagicMethod(retrievedString, new { parameter1, parameter2 });
Is there an existing solution or should I (in MagicMethod) replace these parts in the retrievedString with matching members of the anonymous object given as parameter (using reflection or something like that)?
EDIT:
Finally, I created an extension method to handle this:
internal static string SpecialFormat(this string input, object parameters) {
var type = parameters.GetType();
System.Text.RegularExpressions.Regex regex = new System.Text.RegularExpressions.Regex( "\\{(.*?)\\}" );
var sb = new System.Text.StringBuilder();
var pos = 0;
foreach (System.Text.RegularExpressions.Match toReplace in regex.Matches( input )) {
var capture = toReplace.Groups[ 0 ];
var paramName = toReplace.Groups[ toReplace.Groups.Count - 1 ].Value;
var property = type.GetProperty( paramName );
if (property == null) continue;
sb.Append( input.Substring( pos, capture.Index - pos) );
sb.Append( property.GetValue( parameters, null ) );
pos = capture.Index + capture.Length;
}
if (input.Length > pos + 1) sb.Append( input.Substring( pos ) );
return sb.ToString();
}
and I call it like this:
var parameter1 = DateTime.Now.ToString();
var parameter2 = "Hello world!";
var retrievedString = "{parameter2} Today we're {parameter1}";
var result = retrievedString.SpecialFormat( new { parameter1, parameter2 } );
Now, I don't use double braces anymore.
You can use reflection coupled with an anonymous type to do this:
public string StringFormat(string input, object parameters)
{
var properties = parameters.GetType().GetProperties();
var result = input;
foreach (var property in properties)
{
result = result.Replace(
$"{{{{{property.Name}}}}}", //This is assuming your param names are in format "{{abc}}"
property.GetValue(parameters).ToString());
}
return result;
}
And call it like this:
var result = StringFormat(retrievedString, new { parameter1, parameter2 });
While not understanding what is the dificulty you're having, I'm placing my bet on
Replace( string oldValue, string newValue )
You can replace your "tags" with data you want.
var parameter1 = DateTime.Now.ToString();
var parameter2 = "Hello world!";
var retrievedString = "{{parameter2}} Today we're {{parameter1}}";
var result = retrievedString.Replace("{{parameter2}}", parameter2).Replace({{parameter1}}, parameter1);
EDIT
Author mentioned that he's looking at something that will take parameters and iterate the list. It can be done by something like
public static void Main(string[] args)
{
//your "unmodified" srting
string text = "{{parameter2}} Today we're {{parameter1}}";
//key = tag(explicitly) value = new string
Dictionary<string, string> tagToStringDict = new Dictionary<string,string>();
//add tags and it's respective replacement
tagToStringDict.Add("{{parameter1}}", "Foo");
tagToStringDict.Add("{{parameter2}}", "Bar");
//this returns your "modified string"
changeTagWithText(text, tagToStringDict);
}
public static string changeTagWithText(string text, Dictionary<string, string> dict)
{
foreach (KeyValuePair<string, string> entry in dict)
{
//key is the tag ; value is the replacement
text = text.Replace(entry.Key, entry.Value);
}
return text;
}
The function changeTagWithText will return:
"Bar Today we're Foo"
Using this method you can add all the tags to the Dictionary and it'll replace all automatically.
If you know order of parameters, you can use string.Format() method (msdn). Then, your code will look like:
var parameter1 = DateTime.Now.ToString();
var parameter2 = "Hello world!";
var retrievedString = "{{0}} Today we're {{1}}";
var result = string.Format(retrievedString, parameter2, parameter1);
Related
Let's say I have the following class:
public class TestClass
{
public Dictionary<string, string> Property1 { get; set; }
}
If I do the following:
var dictionary = new Dictionary<string, string>();
dictionary.add("key1", "value1");
dictionary.add("key2", "value2");
var newClass = new TestClass();
newClass.Property1 = dictionary;
I am trying to URL encode this dictionary to the following:
https://baseaddress.com/resource/?Property1[key1]=value1&Property1[key2]=value2
When attempting to URL encode the dictionary via HttpUtility it is returning the ToString() method of the Dictionary which comes out as:
System.Collections.Generic.Dictionary`2[System.String,System.String]
I am trying to pass this dictionary to a .netcore API that binds to a similar Dictionary<string, string>
Edit
I was able to get it working by using the following code:
var builder = new UriBuilder(uri);
var query = HttpUtility.ParseQueryString(string.Empty);
foreach (var propInfo in obj.GetType().GetProperties())
{
var propName = propInfo.Name;
var propValue = propInfo.GetValue(obj);
if (propValue != null)
{
var dict = propValue as IDictionary;
if (dict != null)
{
foreach (var key in dict.Keys)
{
var keyName = key;
var keyValue = dict[key];
query.Add($"{propName}[{keyName}]", keyValue.ToString());
}
}
else
{
query.Add(propName, propValue.ToString());
}
}
}
builder.Query = query.ToString();
return builder.Uri;
I was hoping there was a more efficient way to make this work.
If you want to get the standard format and avoid any problem with your QueryString you can "leverage" .net core approach which indeed is a way larger than the old approach. With that said here is what you can do:
One thing....Notice that they are strings so you can add your brackets :)
Dictionary<String,StringValues>() queryString = QueryHelpers.ParseQuery("?param1=value");
StringValues secondValue=StringValues.Concat(queryString["param2"], "my other value");
parsedQueryString["yourkey"] = secondValue;
//At this point you can start concatenating as many time as needed.
QueryString.Create(parsedQueryString).ToString();
// creates the following string "?param1=value¶m2=my%20other%20value"
A plus :)
// Getting a param value
var param2Value = queryString["param2"];
param2Value.ToString(); // Get the values concatenated together
param2Value.ToArray(); // Gets an array of strings
// Modifying a parameter
queryString["param1"] = "another value";
// NOTE, if there were two values, this overwrites both and leaves a single value
I want to create a single object (possibly Dictionary) with string keys that will have different variable types as the value (string, int, bool, Dictionary<string,string> etc). Is this possible?
*I understand this might just be a fundamental difference of two languages AKA square peg round hole
You can use dynamic as values type, that match better than object to the question and you need no future castings:
var dictionary = new Dictionary<string, dynamic>();
dictionary.Add("1", 10);
dictionary.Add("2", "test");
dictionary.Add("3", true);
foreach ( var item in dictionary )
Console.WriteLine($"{item.Key} is type: {item.Value.GetType().Name} = {item.Value}");
Console.WriteLine();
int v = dictionary["1"] + 10;
Console.WriteLine(v);
string s = dictionary["2"] + " one";
Console.WriteLine(s);
bool b = !dictionary["3"];
Console.WriteLine(b);
Output
1 is type: Int32 = 10
2 is type: String = test
3 is type: Boolean = True
20
test one
False
https://learn.microsoft.com/dotnet/csharp/programming-guide/types/using-type-dynamic
A Dictionary<string, object> is roughly equivalent to an object in JavaScript.
Example:
var dictionary = new Dictionary<string, object>
{
"myString" = "helloWorld",
"myChild" = new Dictionary<string, object>
{
"myName" = "bobby tables"
}
};
var myString = (string)dictionary["myString"];
var myName = (string)((Dictionary<string, object>)dictionary["myChild"])["myName"];
You can also use the dynamic keyword and ExpandoObject.
dynamic obj = new ExpandoObject();
obj.MyString = "helloWorld";
obj.MyChild = new ExpandoObject();
obj.MyChild.MyName = "bobby tables";
string myString = obj.MyString;
string myName = obj.MyChild.MyName;
This is a function to work with lists in string interpolation. It takes a List and an inner Func, and it appends the string result of the inner Func called for each member of the list, with a separator.
So the following builds a valid start of an Insert statement...
static void Main(string[] args)
{
var tableName = "customers";
var cols = new List<dynamic>
{
new { Name = "surname"},
new { Name = "firstname"},
new { Name = "dateOfBirth"}
};
Func<List<dynamic>, Func<dynamic, string>, string, string> ForEach = (list, func, separator) =>
{
var bldr = new StringBuilder();
var first = true;
foreach (var obj in list)
{
if (!first)
bldr.Append(separator);
first = false;
bldr.Append(func(obj));
}
return bldr.ToString();
};
var InsertStatement = $"Insert into { tableName } ( {ForEach(cols, col => col.Name, ", ")} )";
Console.WriteLine(InsertStatement);
Console.ReadLine();
}
Outputs...
Insert into customers ( surname, firstname, dateOfBirth )
It works for dynamic. How do I make it work for any type? The outer Func shouldn't care about the Type in the list, it just passes it through to the inner Func.
The .NET framework already gives you a generic function to achieve what you are trying to do String.Join and you can combine it with a LINQ Select statement, which will allow you to use a lambda on a generic type to select the property that you want to print. You can view the source code of these methods if you are interested as they are open source.
using System;
using System.Collections.Generic;
using System.Linq;
public class MyType
{
public string Name { get; set; }
}
public class Program
{
public static void Main()
{
var tableName = "customers";
var cols = new List<MyType>
{
new MyType { Name = "surname"},
new MyType { Name = "firstname"},
new MyType { Name = "dateOfBirth"}
};
var InsertStatement = $"Insert into { tableName } ( {String.Join(", ", cols.Select(col => col.Name))} )";
Console.WriteLine(InsertStatement);
}
}
Replace dynamic with object, or TValue with a type constraint stipulating it must be a class (where TValue : class), and call obj.ToString() instead of just obj
However, this doesn't guarantee it would "work with any type" - for that you need to know that those types all follow a contract to output the desired column name as their string representation. To get more specificity, require that your accepted types must implement some interface eg IColumnName and put that interface into the type constraint instead
You can create the text easily like this:
var query = $"INSERT INTO {tableName}({string.Join(",", cols.Select(x=>x.Name))})";
However, if for learning purpose you are going to handle the case using a generic method, you can create a generic function like the following and then easily use a for loop and strip additional separator using TrimEnd, or as a better option, like String.Join implementation of .NET Framework get enumerator like this:
string Join<TItem>(
IEnumerable<TItem> items, Func<TItem, string> itemTextSelecor, string separator)
{
var en = items.GetEnumerator();
if (!en.MoveNext())
return String.Empty;
var builder = new StringBuilder();
if (en.Current != null)
builder.Append(itemTextSelecor(en.Current));
while (en.MoveNext())
{
builder.Append(separator);
if (en.Current != null)
builder.Append(itemTextSelecor(en.Current));
}
return builder.ToString();
}
And use it this way:
var tableName = "customers";
var cols = new[]
{
new { Name = "surname"},
new { Name = "firstname"},
new { Name = "dateOfBirth"}
};
var InsertStatement = $"INSERT INTO {tableName} ({Join(cols, col => col.Name, ", ")})"
+ $"VALUES({Join(cols, col => $"#{col.Name}", ", ")})";
I wrote a function to convert LocalDeclaration's to Global Resources. Right now I'm replacing with each definition with a property, but I want to replace it with a property using the new syntax =>
public PropertyDeclarationSyntax ConvertToResourceProperty(string resouceClassIdentifier, string fieldName, string resourceKey, CSharpSyntaxNode field)
{
var stringType = SyntaxFactory.ParseTypeName("string");
var resourceReturnIdentifier = SyntaxFactory.IdentifierName(resouceClassIdentifier + "." + resourceKey);
var returnResourceStatement = SyntaxFactory.ReturnStatement(resourceReturnIdentifier).NormalizeWhitespace();
var getRescourceBlock = SyntaxFactory.Block(returnResourceStatement);
var getAccessor = SyntaxFactory.AccessorDeclaration(SyntaxKind.GetAccessorDeclaration, getRescourceBlock).WithAdditionalAnnotations(Formatter.Annotation, Simplifier.Annotation);
var propertyDeclaration = SyntaxFactory.PropertyDeclaration(stringType, fieldName).AddModifiers(SyntaxFactory.Token(SyntaxKind.PublicKeyword), SyntaxFactory.Token(SyntaxKind.StaticKeyword)).NormalizeWhitespace();
propertyDeclaration = propertyDeclaration.AddAccessorListAccessors(getAccessor).WithAdditionalAnnotations(Formatter.Annotation);
SyntaxTrivia[] leadingTrivia = field.GetLeadingTrivia().ToArray() ?? new[] { SyntaxFactory.Whitespace("\t") };
return propertyDeclaration.WithTrailingTrivia(SyntaxFactory.Whitespace("\r\n"))
.WithLeadingTrivia(leadingTrivia)
.WithAdditionalAnnotations(Simplifier.Annotation);
}
This code create a property like so:
public static string LocalResourceName
{
get{ return Resources.LocalResourceName; }
}
I would like it to make the property like so:
public static string LocalResourceName =>Resources.LocalResourceName;
I'm not too sure what will create an expression bodied property from the syntaxfactory? Can anyone point me to the right method?
After scouring the internet I've found a way to do it. Why is there no documentation for roslyn?
public PropertyDeclarationSyntax ConvertToResourceProperty(string resouceClassIdentifier, string fieldName, string resourceKey, CSharpSyntaxNode field)
{
var stringType = SyntaxFactory.ParseTypeName("string");
var resourceClassName = SyntaxFactory.IdentifierName(resouceClassIdentifier);
var resourceKeyName = SyntaxFactory.IdentifierName(resourceKey);
var memberaccess = SyntaxFactory.MemberAccessExpression(SyntaxKind.SimpleMemberAccessExpression, resourceClassName, resourceKeyName);
var propertyLambda = SyntaxFactory.ArrowExpressionClause(memberaccess);
var propertyDeclaration = SyntaxFactory.PropertyDeclaration(new SyntaxList<AttributeListSyntax>(), new SyntaxTokenList(),
stringType, null, SyntaxFactory.Identifier(fieldName), null,
propertyLambda, null, SyntaxFactory.Token(SyntaxKind.SemicolonToken))
.AddModifiers(SyntaxFactory.Token(SyntaxKind.PublicKeyword),
SyntaxFactory.Token(SyntaxKind.StaticKeyword)).WithAdditionalAnnotations(Formatter.Annotation).NormalizeWhitespace();
return propertyDeclaration.WithTrailingTrivia(SyntaxFactory.ElasticCarriageReturnLineFeed)
.WithLeadingTrivia(field.GetLeadingTrivia().ToArray())
.WithAdditionalAnnotations(Simplifier.Annotation);
}
I have a rather complicated issue. I am trying to get a unique key from a method and its formal and actual parameters. The goal of the method, is to take a method call, and return a unique key based on 1) The name of the class and method and 2) The name and values of the parameters it is called with.
The method looks like this (sorry for all the details, but I can't find a sensible way to make the example smaller yet still explain my problem)
public class MethodKey
{
public static string GetKey<T>(Expression<Func<T>> method, params string[] paramMembers)
{
var keys = new Dictionary<string, string>();
string scope = null;
string prefix = null;
ParameterInfo[] formalParams = null;
object[] actual = null;
var methodCall = method.Body as MethodCallExpression;
if (methodCall != null)
{
scope = methodCall.Method.DeclaringType.FullName;
prefix = methodCall.Method.Name;
IEnumerable<Expression> actualParams = methodCall.Arguments;
actual = actualParams.Select(GetValueOfParameter<T>).ToArray();
formalParams = methodCall.Method.GetParameters();
}
else
{
// TODO: Check if the supplied expression is something that makes sense to evaluate as a method, e.g. MemberExpression (method.Body as MemberExpression)
var objectMember = Expression.Convert(method.Body, typeof (object));
var getterLambda = Expression.Lambda<Func<object>>(objectMember);
var getter = getterLambda.Compile();
var m = getter();
var m2 = ((System.Delegate) m);
var delegateDeclaringType = m2.Method.DeclaringType;
var actualMethodDeclaringType = delegateDeclaringType.DeclaringType;
scope = actualMethodDeclaringType.FullName;
var ar = m2.Target;
formalParams = m2.Method.GetParameters();
//var m = (System.MulticastDelegate)((Expression.Lambda<Func<object>>(Expression.Convert(method.Body, typeof(object)))).Compile()())
//throw new ArgumentException("Caller is not a method", "method");
}
// null list of paramMembers should disregard all parameters when creating key.
if (paramMembers != null)
{
for (var i = 0; i < formalParams.Length; i++)
{
var par = formalParams[i];
// empty list of paramMembers should be treated as using all parameters
if (paramMembers.Length == 0 || paramMembers.Contains(par.Name))
{
var value = actual[i];
keys.Add(par.Name, value.ToString());
}
}
if (paramMembers.Length != 0 && keys.Count != paramMembers.Length)
{
var notFound = paramMembers.Where(x => !keys.ContainsKey(x));
var notFoundString = string.Join(", ", notFound);
throw new ArgumentException("Unable to find the following parameters in supplied method: " + notFoundString, "paramMembers");
}
}
return scope + "¤" + prefix + "¤" + Flatten(keys);
}
private static object GetValueOfParameter<T>(Expression parameter)
{
LambdaExpression lambda = Expression.Lambda(parameter);
var compiledExpression = lambda.Compile();
var value = compiledExpression.DynamicInvoke();
return value;
}
}
Then, I have the following test, which works OK:
[Test]
public void GetKey_From_Expression_Returns_Expected_Scope()
{
const string expectedScope = "MethodNameTests.DummyObject";
var expected = expectedScope + "¤" + "SayHello" + "¤" + MethodKey.Flatten(new Dictionary<string, string>() { { "name", "Jens" } });
var dummy = new DummyObject();
var actual = MethodKey.GetKey(() => dummy.SayHello("Jens"), "name");
Assert.That(actual, Is.Not.Null);
Assert.That(actual, Is.EqualTo(expected));
}
However, if I put the () => dummy.SayHello("Jens") call in a variable, the call fails. Because I then no longer get a MethodCallExpression in my GetKey method, but a FieldExpression (subclass of MemberExpression. The test is:
[Test]
public void GetKey_Works_With_func_variable()
{
const string expectedScope = "MethodNameTests.DummyObject";
var expected = expectedScope + "¤" + "SayHello" + "¤" + MethodKey.Flatten(new Dictionary<string, string>() { { "name", "Jens" } });
var dummy = new DummyObject();
Func<string> indirection = (() => dummy.SayHello("Jens"));
// This fails. I would like to do the following, but the compiler
// doesn't agree :)
// var actual = MethodKey.GetKey(indirection, "name");
var actual = MethodKey.GetKey(() => indirection, "name");
Assert.That(actual, Is.Not.Null);
Assert.That(actual, Is.EqualTo(expected));
}
The Dummy class SayHello method definitions are trivial:
public class DummyObject
{
public string SayHello(string name)
{
return "Hello " + name;
}
public string Meet(string person1, string person2 )
{
return person1 + " met " + person2;
}
}
I have two questions:
Is there any way to send the variable indirection to MethodKey.GetKey, and get it as a MethodCallExpression type?
If not, how can I get the name and value of the method supplied if I get a MemberExpression instead? I have tried a few bits in the "else" part of the code, but haven't succeeded.
Any help is appreciated.
Thanks in advance, and sorry for the long post.
The problem is you are putting it into the wrong type of variable. Your method expects Expression<Func<T>> and you are using a variable of type Func<string> to store it. The following should fix your problem:
Expression<Func<string>> foo = () => dummy.SayHello("Jens");
var actual = MethodKey.GetKey<string>(foo, "name");
converting a .net Func<T> to a .net Expression<Func<T>> discusses the differences between a Func and an Expression<Func> and converting between the two and at a glance it says don't. The compiler makes them into totally different things. So make it the right thing at compile time and it should work fine.
If this isn't an option then possibly an overload that takes a Func instead of an Expression might work for you.
Note that in both cases I would pass the variable directly rather than trying to make it into a new expression in your call.