These publications could be useful for graphical or textual concrete syntaxes:

• IEEE TRANSACTIONS ON SOFTWARE ENGINEERING, VOL. 35, NO. 5, NOVEMBER-DECEMBER 2009 The “Physics” of Notations: Towards a Scientific Basis for Constructing Visual Notations in Software Engineering

Daniel L. Moody, Member, IEEE

• DSL and their Tools, Markus Voelter
• Introduction to Domain-Specific Languages (DSL tutorial slides)

by Jordi Cabot 

One significant aspect of meta-modelling is that there is a widely accepted meta-modelling language (UML) that aids interoperability. That is, because modellers use the same language, it is easier to build more complex meta-models by composition. For concrete syntax, there is more diversity, but still a tendency to have standard methods to define them for the sake of interoperability.

Standards for concrete syntaxes can be categorized according to their "base format" - strings, XML, ...

String-based:

1. BNF (http://en.wikipedia.org/wiki/Backus%E2%80%93Naur_Form) and its extensions (EBNF, ABNF)

2.van Wijngaarden  (http://en.wikipedia.org/wiki/Van_Wijngaarden_grammar)

XML-based (http://en.wikipedia.org/wiki/XML_schema#XML_schema_languages):

1. XSD

2. Relax NG

3. Schematron

4. DTD

There are other base formats, e.g. JSON, do not yet have accepted standards, but there are efforts to develop JSON schema languages, e.g. http://tools.ietf.org/html/draft-zyp-json-schema-03

For DSLs based in programming languages (e.g. a Ruby or Scala DSL), then the "schema language" is the programming language itself. To my knowledge, there are no "standards" for defining such DSLs, but as these languages become more prevalent and need arises for them to interoperate with each other, such standards may be developed.

I prefer the cement approach.

@Hugh Buckingham ,

What do you mean by the cement approach please?

@Hajara -Hugh's comment is a pun - physically, concrete is a mixture of crushed rock  and cement.

@ Tara _ Thanks

@Hajara - I just noticed I didn't address your final parenthetical about visual syntax. While I don't know the details of the syntax of interest, many such syntaxes can be modelled using markup such as SVG. (http://www.xml.com/pub/a/2001/03/21/svg.html) Since SVG is an XML language, then it may be possible to use a schema language for XML to define a grammar for the subset of SVG which constitutes your concrete syntax model.

Another approach I have seen is to employ a textual XML-based concrete syntax as intermediary, and then use an XSLT stylesheet to transform it to SVG.

Find a syntactic structure to work later in language engineering requires a team of linguists, educators and engineers who study the variables involved in the syntax from childhood to detect parameters and pass them useful for computing.

For example: position of adjectives, names, plural formation; situation of the verb; type of conjugation; length of the sentence; position of pronouns and articles, etc.

@Hosny -  to answer the original question, can you please clarify how would the Z notation be used to formally specify a concrete syntax (textual or visual)?

@Hosny - The Z notation is already being inculcated in defining the variable, parameters and operations of the proposed concrete syntax of the language in question, so i understand the direction you are trying to lead me to (Thank you).

@Tara - Z is a formal specification language that has a defined syntax and semantics in receiving of input, sending of output, manipulation of variables, operations in a system, state changes of a particular system (that may be static, dynamic or discrete), it has a defined concrete syntax that enables it use very easy.

@Hosny - We are looking for a formalized way of defining a concrete syntax (or an informal way that is almost formal). For this language in question we already have a defined abstract syntax  and some semantic domains, we are just trying to formalize the definition of the shapes and connection we want to attach to the abstract syntax.

Thank you

@Stefan - The mapping between syntax is possible if we only have the syntax in question standing independently on their own, then we can build a bridge (i.e mapping in this case) to close the gap between the syntax. The point is there is an abstract syntax, we have an idea of what we want the concrete syntax to look like, how do we formalize these idea (shapes, connections etc.) just as we formalized the abstract syntax through metamodeling.

In some scenarios we can just define shapes (such as  circle, rectangle in correspondence to some classes from the abstract syntax), which we need to formalize, in some scenarios we have collection of classes from the abstract syntax end giving us one shape at the concrete syntax end which also need to be formalized.

Thank you

These other resources might help:

• MID: A MetaCASE Tool For A Better Reuse Of Visual Notations https://www.researchgate.net/publication/262673935_MID_A_MetaCASE_Tool_For_A_Better_Reuse_Of_Visual_Notations
• Concrete Syntax Matters, http://www.infoq.com/presentations/Language-Design
• Design Guidelines for Domain Specific Languages https://www.researchgate.net/publication/265469386_Design_Guidelines_for_Domain_Specific_Languages
• (patent) METHOD AND APPARATUS OF SPECIFYING THE CONCRETE SYNTAX OF GRAPHICAL MODELING LANGUAGES Application number: 20110131546;  http://www.google.com/patents/US20110131546
• Book: DSL Engineering, Designing, Implementing and Using

Domain-Specific Languages, Markus Voelter et al.

Text-based Modeling, Hans Gronniger, Holger Krahn, Bernhard Rumpe, Martin Schindler and Steven Volkel, Institute for Software Systems Engineering, Technische Universitat Braunschweig, Braunschweig, Germany, 2007

http://dslbook.org/

Conference Paper MID: A MetaCASE Tool for a Better Reuse of Visual Notations

Article Design Guidelines for Domain Specific Languages

Article Textbased Modeling

@ Pierre Labreche - Thank you for the links. The last two are new for me, so am going through the documents. I will give you an update if there is any.

Thanks