[tcgc] Adopt the Microsoft emitter name format when pass the emitter name to get the language

[chunyu3]

There are a set of Microsoft emitter whose name convention is @typespec/{protocol}-{client|server}-{language}-{framework} and {framework} part is optional. for example, the csharp emitter name will be @typespec/http-client-csharp. (see loop typespec's codegen ecosystem

TCGC now only adopt "@azure-tools/typespec-{language}" format to parse emitter name( see following). Need to adopt the new emitter name format.

export function parseEmitterName(
  program: Program,
  emitterName?: string
): [string, readonly Diagnostic[]] {
  const diagnostics = createDiagnosticCollector();
  if (!emitterName) {
    diagnostics.add(
      createDiagnostic({
        code: "no-emitter-name",
        format: {},
        target: program.getGlobalNamespaceType(),
      })
    );
    return diagnostics.wrap("none");
  }
  const regex = /(?:cadl|typespec)-([^\\/]*)/;
  const match = emitterName.match(regex);
  if (!match || match.length < 2) return diagnostics.wrap("none");
  const language = match[1];
  if (["typescript", "ts"].includes(language)) return diagnostics.wrap("javascript");
  return diagnostics.wrap(language);
}

Expected:
TCGC adopt the new emitter name format and parse out the language from emitter name.