[Mirrored from: http://www.ile.com/opentag/otspecs.htm, Feb-28-1997]

OpenTag is not a finalized format, it is just a proposed base the members of the consortium will modify and build upon, taking in account the various experience the makers of existing extracted text formats have accumulated.

This document describes the current OpenTag format element by element. You can find more information on how to support the format in the OpenTag Implementation Notes document. You can also refer to the EBNF grammar and download other useful reference material, such as the SGML DTD for OpenTag, from the Resources page. If you need any additional information or for any questions that you may have, please send an e-mail to opentag@ile.com.

1 - Markup

OpenTag is XML/SGML compliant. The markup rules of an OpenTag file follow the SGML/XML rules. You can find the XML working draft specifications wd-xml-961114 at http://www.textuality.com/sgml-erb/WD-xml.html and http://www.w3.org/pub/WWW/TR/.

The notion of a paragraph in an OpenTag file does not necessarily corresponds to the same definition of a paragraph in the original file. For example, an OpenTag paragraph generated from RTF can originate from a \row element and is not restricted to originate only from the \par element.

2 - Character Entities

OpenTag uses character entities to specify the characters that do not exist in the default code set of the file. A character entity is composed of the prefix character '&', followed by the SGML name of the character, or its Unicode value (in hexadecimal, with a leading '#' and a trailing ';').

Example:

<P id=1/>
Lowercase "a grave" can be coded &agrave; or &#00e0;

3 - File Header

The header is composed of the XML prologue that sets the default values for the code set and the locale of the file.

<?XML 
  version='1.0'
  type='OpenTag1'
  encoding='ISO8859-1'
  locale='FR-FR'
  original='MIF4'
?>

4 - Body

The body of an OpenTag file or a data stream coded in OpenTag is enclosed by the <OPENTAG> element.

The following table lists all the formatting elements defined in OpenTag:

B
<B id=1>...</B>

I
<I id=1>...</I>

U
<U id=1>...</U>

D
<D id=1>...</D>

G
<G id=1>...</G>

P
<P id=1/>
<P id=1 s="Body"/>
<P id=1 cs=EUC-JA/>
<P id=1 lc=JA-JP cs=EUC-JA/>
<P id=1 seg=23/>

PB
<PB id=1/>
<PB id=1 seg=45/>

X
<X id=1/>

CELL
<CELL id=1/>

CB
<CB id=1/>

OB
<OB id=1/>

OPENTAG
<OPENTAG>...</OPENTAG>

FN
<FN id=1/>

IX
<IX id=1/>

RF
<RF id=1/>

FND
<FND id=1 seg=1>...</FND>

IXD
<IXD id=1>...</IXD>

RFD
<RFD id=1 seg=1>...</RFD>

OC
<OC cs=EUC-JA>...</OC>

CS
<CS cs=EUC-JA>...</CS>

PR
<PR id=1>...</PR>

CT
<CT cond="doc" id=1>...</CT>

LVL
<LVL  id=1 seg=1>...</LVL>

SO
<SO id=1 seg=1>...</SO>

The following table lists all the property elements defined in OpenTag. Property elements are used to process information. They are not used for merging back the text in its original form.

S
<S id=1/>

TERM
<TERM>...</TERM>