Let's teach Sierra Chart how to talk

August 14th, 2012, 10:31 AM

The title of this thread is a tribute to the thread " Let's Teach [AUTOLINK]NinjaTrader[/AUTOLINK] how to talk" thanks to which I have learnt that Windows has built-in speech generation capability, and which gives some code example for NinjaTrader.

Objective: have Sierra Chart speak "New Candle" at each new candle
Of course, this is not very useful.

But the idea is that the code may be adapted for other uses, as sophisticated alerts.

Important disclaimer. I am a total newbie in C++, .NET, Windows programming, etc. I have solved all my Windows-related problems by Googling. And I am not sure to have understood what I have done. What follows is experimental, may be dirty and non-optimal. If another nexusfi.com (formerly BMT) member wants to improve this solution, she/he shall not hesitate!

1. Open a new solution in Visual C++ with the 40-line code below.
Do not forget to adapt the first line to reflect the exact directory of your sierrachart.h
Do not forget to update the output .DLL directory (as usual)

2. Try to build the solution.
You should obtain the following error message: fatal error C1190: managed targeted code requires a '/clr' option
Go to Properties of the solution >> Configuration Properties >> General >> Common Language Runtime Support
Choose "Common Language Runtime Support (/clr)"
Reference : managed targeted code requires a /clr option

3. Try to build the solution.
You should obtain the following error message: Command line error D8016 : '/MT' and '/clr' command-line options are incompatible
Go to Properties of the solution >> Configuration Properties >> C++ >> Code Generation >> Runtime Library
Choose "Multi-threaded DLL (/MD)"
References : '/MT' and '/clr' command-line options are incompatible and /MD, /MT, /LD (Use Run-Time Library)

4. Try to build the solution.
You should obtain the following error message: fatal error C1107: could not find assembly 'System.Speech.dll': please specify the assembly search path using /AI or by setting the LIBPATH environment variable
Go to Properties of the solution >> Configuration Properties >> C++ >> General >> Resolve #using References
Add the directory of System.Speech.dll
On my computer, this directory is:
C:\Program Files\Reference Assemblies\Microsoft\Framework\v3.0\
Reference: /AI (Specify Metadata Directories) (C++)

5. Try to build the solution: should be OK!

Then add the study "Voice" to your chart, as usual.
You should hear "Hello" once or twice.
And "New candle" at each new candle.

6. Some links for my own reference:
On #pragma managed: managed type or function cannot be used in an unmanaged function - .NET Framework and Mixing Managed and Unmanaged code - CodeProject
On managed global variable: compile error: may not declare a global or static variable, or a member of a native type that refers to objects in the gc heap
On gcnew and ^ : C3673: class does not have a copy-constructor
"A Very Easy Introduction to Microsoft .NET Speech Synthesis (VB, C#, C++)": A Very Easy Introduction to Microsoft .NET Speech Synthesis (VB, C#, C++) - CodeProject
Reference of Microsoft Speech API: Microsoft Speech API (SAPI) 5.3 and System.Speech Namespaces ()

Once more, the above is experimental. It may exist cleaner solutions.

Nicolas

Code

#include "C:\Sierrachart-IQFeed\ACS_Source\sierrachart.h"

#using <System.dll>
#using <System.Speech.dll>
using namespace System::Speech::Synthesis;

SCDLLName("Voice")

#pragma managed
public ref class Globals abstract sealed {
public:
   static SpeechSynthesizer^ speaker = gcnew SpeechSynthesizer();
};

#pragma managed
SCSFExport scsf_Voice(SCStudyGraphRef sc)
{

	// Persistent variable:
	int& LastScIndex = sc.PersistVars->i1;

	if ( sc.SetDefaults )
	{	
		sc.GraphName = "Voice";
		sc.StudyDescription = "Voice";
		sc.AutoLoop = 1;
		sc.FreeDLL = 1;

		Globals::speaker->SpeakAsync("Hello");

		return;
	}

	if ( sc.Index == 0 ) LastScIndex = -1;

	// If new candle...
	if ( sc.Index != LastScIndex ) {
		LastScIndex = sc.Index;
		// If in real-time, let's talk!
		if ( sc.Index == sc.ArraySize-1 ) {
			Globals::speaker->SpeakAsync("New candle");
		}
	}
}

August 14th, 2012, 10:34 AM

Dude, you are on a roll!

Mike

August 18th, 2012, 03:21 PM

A code a little bit cleaner. I have put in bold all the specific parts related to the voice.

Code

#include "C:\Sierrachart-IQFeed\ACS_Source\sierrachart.h"

#using <System.dll>
#using <System.Speech.dll>

using namespace System;
using namespace System::Speech::Synthesis;

#pragma managed
public ref class Globals abstract sealed {
public:
   static SpeechSynthesizer^ speaker = gcnew SpeechSynthesizer();
};

SCDLLName("Voice")

SCSFExport scsf_Voice(SCStudyGraphRef sc)
{

	// Persistent variable:
	int& LastScIndex = sc.PersistVars->i1;

	if ( sc.SetDefaults ) {	
		sc.GraphName = "Voice";
		sc.StudyDescription = "Voice";
		sc.AutoLoop = 1;
		sc.FreeDLL = 1;

		return;
	}

	if ( sc.LastCallToFunction ) {
		delete Globals::speaker;
	}

	if ( sc.Index == 0 ) LastScIndex = -1;

	// If new candle...
	if ( sc.Index != LastScIndex ) {
		LastScIndex = sc.Index;
		// If in real-time, let's talk!
		if ( sc.Index == sc.ArraySize-1 ) {
			Globals::speaker->SpeakAsync("New candle");
		}
	}
}

Let's teach Sierra Chart how to talk

Discussion in Sierra Chart

Let's teach Sierra Chart how to talk