MRM parser and domain
You can use the MRM domain to parse and write a wide range of message formats.
The MRM domain can be used to parse and write a wide variety of message formats. It is primarily intended for non-XML message formats, but it can also parse and write XML. For guidance on when to consider using the MRM parser, instead of one of the XML parsers, to parse XML, see Which XML parser should you use?
- Support for messages from applications that are written in C, COBOL, PL/I and other languages, by using the Custom Wire Format (CWF) physical format. This support includes the ability to create a message model directly from a C header file or COBOL copybook.
- Support for text messages, perhaps with field content that is identified by tags, separated by specific delimiters, or both, by using the Tagged Delimited String (TDS) physical format. This includes industry standards such as CSV, HL7, SWIFT, EDIFACT, and X12.
- Support for XML messages, including those that use XML namespaces, by using the XML physical format.
IBM® Integration Bus uses the MRM parser to read and write messages that belong to the MRM domain. When reading a message, the MRM parser constructs a message tree from a bit stream. When writing a message, the MRM parser creates a bit stream from a message tree. The MRM parser is always model-driven, and it is guided by a message dictionary that describes the shape of the message tree (the logical model) and the physical layout of the bytes or characters in the bit stream (the physical format). A message dictionary is created automatically from the content of a message set when it is added to the BAR file. Therefore, when you create a message set for use with the MRM domain, you must define both the logical model and the appropriate physical format information.
- For a binary message, the parser reads a set sequence of bytes according to information in the CWF physical format, and translates them into the fields and values in the message tree.
- For a text message, the parser uses a key piece of TDS physical format information called Data Element Separation to decide how to parse each portion of the message bit stream. This informs the parser whether the message uses delimiters, tags, fixed length elements, patterns, and so on. The parser then reads the data according to information in the TDS physical format, and translates it into the fields and values in the message tree.
- For an XML message, the parser reads the XML markup language (element tags and attributes), guided by information in the XML physical format, and translates them into the fields and values in the message tree.
Because the MRM parser is model-driven, it can perform validation of messages against the model that is defined in the deployed dictionary. The level of validation that is performed by the MRM parser is similar to that defined by XML Schema 1.0, but is not fully compliant. If you use XML messages, and you want fully compliant XML Schema 1.0 validation, use the XMLNSC domain.
The MRM parser is an on-demand parser. See Parsing on demand.
If you want to use the MRM domain to parse a particular message:
- Create a new message set with an appropriate CWF, TDS, or XML physical format; or locate an existing message set.
- Ensure that the message set has its Default message domain set to MRM, or that the MRM check box under Supported message domains is selected to indicate that the message set supports the MRM domain.
- Create a message definition file in the message set to represent your message, ensuring that both logical and physical format information is provided. If you have an existing C, COBOL, XML Schema, or DTD description of your message, you can import the description by using a wizard.
- Add the message set to a BAR file which will generate a message dictionary for use by the MRM parser, and deploy the BAR file to the integration node.
- Select MRM as Message Domain on the appropriate node in your message flow.
- Additionally set values for the Message model, Message, and Physical format properties on the node. The Message property specifies the name of the message in the message definition file.
Some predefined message models are supplied with the IBM Integration Toolkit and can be imported by using the New Message Definition File From IBM supplied Message wizard. The CSV, ALE IDoc, and File IDoc models are specifically for use with the MRM domain. See Message Sets: IBM supplied messages that you can import.
IBM supplies predefined message sets for industry standard formats SWIFT, X12, EDIFACT, and FIX. For more information, contact Dublin Adapters at firstname.lastname@example.org.