IMPORTANT NOTICE

CML is being developed in close conjunction with the XML-working_group of the W3C (Feb 1997). TecML/CML is intended to be completely XML-compliant but there will be a few small differences as both languages develop. Therefore you should be prepared to make changes if required, especially if the XML project provides something that TecML/CML does not. The following Elements may become obsolete:

• RELATION and Internal addressing scheme. It is highly likely that TecML will adopt the TEI addressing scheme (or whatever minor variant emerges from XML. However it should be straightforward to reprocess any TecML documents with XADDR elements. (Whether XML will use a GI or attributes for addressing is unclear).
  RELATION: Link between 2 or more information components. This will be managed by the XML Multilink tool

• REL, REV and METHODS. I never understood these, no one uses them and they will disappear (they will remain in the HTML subset for compatibility, but will not be processed).
• HREF and NAME on non-HTML elements. HREF was never used and will disappear.NAME is used, but may be superseded by XML-link.

Other effects of XML

CML documents may not be strictly compatible with XML until it is clear what the minimum header is. My software (JUMBO) ignores PIs, but others may not. (CML does not require PIs at present). If in doubt, convert XML/CML documents to ESIS which JUMBO can read

CML was developed as a valid SGML application and so it is legitimate to use entities, marked sections and so on. If in doubt, use sgmls to produce an ESIS stream and read that into JUMBO (which doesn't recognise these).

XML is also developing an SGMLOpen Catalog approach. CML will embrace this and it is likely that chemical components can be catalog'ed. JUMBO does not (yet) support catalogs, but will use NXP to do so when that is ready.

DTDs required for CML

• CML.DTD

  This is a minimal DTD which contains the three DTDs listed below. It serves to:
  • Define the overal structure of a CML document (which can be very flexible).
  • Define a few parameter entities (since this is the most convenient place).
  • Give some overall DTD documentation.
  The DTD can be viewed as:
  • A raw text file.
  • Top-level documentation from dtd2html
  • Complete tree of ELEMENTS produced by dtd2html

• HTML20.DTD

  This is the draft standard for HTML (V2.0). I am very grateful to Dan Connolly (one of the authors) for permission to use this. I shall include a pointer to the documentation, but it's assumed that most people who want to write CML will know the basics of HTML. (There are some slight changes - OMITTAG is not allowed in keeping with XML and FORMS, etc are IGNOREd).

  The DTD can be viewed as:

  • A raw text file.
  • Complete tree of ELEMENTS produced by dtd2html

• TecML.DTD

  Technical MarkupLanguage is an application-independent DTD which can describe a large variety of technical information. It includes HTML 2.0 (within the XHTML element). It can 'inherit' some elements and attributes from other DTDs which use it.

  The DTD can be viewed as:

  • A raw text file.
  • Description of TecML and its contents produced by dtd2html

• MOL.DTD

  MOL.DTD is a DTD for describing single molecules and related information. It includes various elements from TecML and can use its containers (e.g. X.LIST).

  The DTD can be viewed as:

  • A raw text file.
  • Description of MOL and its contents produced by dtd2html

Quick Reference

• A description of the tags in CML (without attributes).