What is the purpose of HttpHeaders.TryAddWithoutValidation?

Question

In the System.Net.Http.Headers namespace, what is the difference between HttpHeaders.TryAddWithoutValidation and HttpHeaders.Add?

Specifically, what validation is occurring when calling the Add method? The documentation for Add() simply states:

"The header value will be parsed and validated."

Answer 1

TryAddWithoutValidation does not try to parse the value it is given to see if it is valid (e.g. valid "type" of value or multiple values for a single-value header). Add first parses the value to perform the previously mentioned checks before trying the add the header value

Answer 2

Using Reflector, this is what the TryAddWithoutValidation method does internally:

if (!this.TryCheckHeaderName(name))
{
    return false;
}
if (value == null)
{
    value = string.Empty;
}
AddValue(this.GetOrCreateHeaderInfo(name, false), value, StoreLocation.Raw);
return true;

The work happens inside the TryCheckHeaderName() function.

It boils down to checking whether the name is not null and whether it matches the RFC for the HTTP protocol (i.e. it contains no invalid characters etc.) as well as checking the header against a set of not-allowed headers.

Here's the source-code:

bool TryCheckHeaderName(string name)
{
   if (string.IsNullOrEmpty(name))
   {
       return false;
   }
   if (HttpRuleParser.GetTokenLength(name, 0) != name.Length)
   {
       return false;
   }
   if ((this.invalidHeaders != null) && this.invalidHeaders.Contains(name))
   {
       return false;
   }
   return true;
}

In contrast the Add method has similar behavior with the exception(pun intended) that it throws an exception in case the header name fails any one of the conditions in the TryCheckHeaderName function.