In order for AI to work with string T v3.. we need to split it up to maybe 3 or 4 classes and then maybe anthropic can fit it into projects… does it need to be this big a class and all in one module.. monster power house class that is perfect for document management and file name replacement and date searches ect.
processing go pro data , ect..fit files…
then we can bind to it..
I expect the AI limitations will be lifted in the not too distant future. So I’m probably not going to make sweeping architectural changes based on that.
Personally I prefer monolithic classes that “do everything”. They are easier to ue because I don’t have to keep remembering which class which function is in. Plus I’m not limited with regard to which methods can use which other methods. Or which methods are available in this class, but not that class and so on.
with the Linux binding service machine software running inside clarion win32 now or your clarion win 32 DLL loaded as an Add on as a full client bindable service what can AI do with ST V3 - M…M stand now for Modules as we have split STV3… well AI now thinks it can built the following and has blurted out the clarion classes to do it plus create a bindable for Window and Linux to load it solution to git hub.. to big to show here but here is it summary …
With StringTheory added to the project, the Clarion WinSock module becomes a fully capable REST API and OAuth 2 client platform. The combination provides:
The StringTheory + QuickSocket combination creates a powerful, enterprise-grade REST API client that can compete with any modern HTTP client library while maintaining native Clarion integration.
AI thinks it can take Clarions WinSock module, ST V3 Modules, The Linux Binding Service Software , SQL API ++, and create a runtime able script powered REST Server… well if it runs good on AI and String Theory… the Linux Binding Service handles the runtime scripting but String Theory lets Clarions Win Sock play bigger game… well if it works well done AI and String T V3.
Probably doesnt work but we will see what AI can do with Claron.
Test code …
MEMBER
MAP
END
INCLUDE(‘RestAPIServer.inc’),ONCE
INCLUDE(‘QuickSocket.inc’),ONCE
INCLUDE(‘StringTh.inc’),ONCE
!===================================================================
! HTTPRequestParser Implementation
!===================================================================
HTTPRequestParser.CONSTRUCT PROCEDURE()
CODE
SELF.request &= NEW HTTPRequest
SELF.request.headers &= NEW HeaderQueue
SELF.request.queryParams &= NEW HeaderQueue
SELF.request.pathParams &= NEW HeaderQueue
SELF.parsed = FALSE
HTTPRequestParser.DESTRUCT PROCEDURE()
CODE
IF NOT SELF.request &= NULL
IF NOT SELF.request.headers &= NULL
DISPOSE(SELF.request.headers)
END
IF NOT SELF.request.queryParams &= NULL
DISPOSE(SELF.request.queryParams)
END
IF NOT SELF.request.pathParams &= NULL
DISPOSE(SELF.request.pathParams)
END
DISPOSE(SELF.request)
END
HTTPRequestParser.ParseRequest PROCEDURE(STRING pRawRequest)
st StringTheory
firstLine StringTheory
headerLine StringTheory
parts StringTheory
i LONG
colonPos LONG
queryPos LONG
bodyStart LONG
CODE
st.SetValue(pRawRequest)
st.Split('<13><10>')
IF st.Records() < 1
RETURN HTTP_BAD_REQUEST
END
! Parse request line (GET /path?query HTTP/1.1)
firstLine.SetValue(st.GetLine(1))
parts.SetValue(firstLine.GetValue())
parts.Split(' ')
IF parts.Records() < 3
RETURN HTTP_BAD_REQUEST
END
SELF.request.method = UPPER(CLIP(parts.GetLine(1)))
SELF.request.url = CLIP(parts.GetLine(2))
SELF.request.httpVersion = CLIP(parts.GetLine(3))
Be easier to use nettalk
for a large project with a Dictionary and where you want to deploy the Clarion IDE + NetTalk that is correct… and you using Javascript and a browser as your front end…and your a Clarion developer on windows Yes. this isnt that…