You can use the RTF parser library to parse an RTF document. Like an XML document, an RTF document consists of markup and data content. The OmniMark RTF parser library maps the markup structures of RTF to OmniMarks markup rules according to the rules detailed below. When you use the RTF parser, you launch it just like an external XML parser, using do markup-parse and you access the structure and data content of the RTF file just as you would with an XML file -- by writing markup rules.

The following program illustrates the operation of the RTF parser by implementing a crude RTF to XML converter:

  import "omrtf.xmd" unprefixed
  process
    do markup-parse rtf
     scan file #args[1]
       output "%c"
    done

  element #implied
     output "<%q"
     repeat over attributes as a
        output " " || key of attribute a || '="%v(a)"'
     again
     output ">%c</%q>%n"

  translate "<"
     output "&lt;"
  translate ">"
     output "&gt;"
  translate "&"
     output "&amp;"

How RTF structures are mapped to markup rules

The only markup rules rules fired by the RTF parser are:

• element
• data-content
• translate

RTF commands are translated into XML-like elements, as follows:

• When an RTF command designated by the RTF 1.7 Specification as a "destination" command (see list below) appears at the start of an RTF group (i.e. immediately following "{"), it starts an element, with the RTF command as its element name, and its content is the content of that RTF group. If the RTF command has a numeric value specified, this value is provided as the value of the element's "value" attribute. If the RTF group in question is marked "ignorable" (i.e. using the "\*" designator), an attribute is provided named "ignorable", with a value of "yes". Both the "value" and "ignorable" attributes are of type "CDATA". The element type is "ANY".
• When an RTF command isn't specified as a destination, but does appear at the start of a group with an "ignorable" designator, then it's treated as if it were a destination, in the above manner.
• An RTF command not specified or treated as a destination is an "empty tag" element (as if its tag were ended with "/>"), with its value, if any, provided by the "value" attribute. The attribute types are as above, and the element type is "EMPTY".
• The "\u" (UNICODE character) command is treated specially. It's an "empty tag" element, with the UNICODE character number as its "value" attribute. In addition, it has an "alt" attribute, which contains the "alternative" text provided immediately following the "\u" command. The "alt" attribute is of type "CDATA".
• An RTF group not recognized as grouping the content of an RTF destination is provided as an element, with the name "group_", with the content of the group as its content. The only possible attribute for this element is the "ignorable" one described above, if the "{" is followed by "\*". The "_" in the name of this element and following ones is used so that the command doesn't conflict, or potentially conflict, with any RTF command.
• An RTF command whose name is a special character rather than a name, such as "\_" or "\~", is provided as an element with the name "special_", and with the "name" as the value of the "value" attribute.
• An RTF comand that consists of "\" followed by a new line, which is an alternative to the "\par" command, is provided as a "par_" element.
• Line ends, which are intended to be ignored by RTF readers, are not returned by the RTF parser.
• All other data is returned as data content, including that encoded in the RTF document as "binary" (with the "\bin" command). Hexadecimal data that uses the "\'xx" RTF command is returned as data content. However, there are RTF commands whose "content" is implicitly hexadecimal. In the latter case, the hexadecimal data is made available "as is" -- the RTF parser has no special knowledge of these commands.

RTF commands considered destinations

The OMRTF library is based on version 1.7 of the RTF spec , according to which, the RTF command names that are "destinations" are the following:

     aftncn
     aftnsep
     aftnsepc
     annotation
     atnauthor
     atndate
     atnicn
     atnid
     atnparent
     atnref
     atntime
     atrfend
     atrfstart
     author
     background
     bkmkend
     bkmkstart
     buptim
     category
     colortbl
     comment
     company
     creatim
     datafield
     do
     doccomm
     docvar
     dptxbxtext
     falt
     fchars
     ffdeftext
     ffentrymcr
     ffexitmcr
     ffformat
     ffhelptext
     ffl
     ffname
     ffstattext
     field
     file
     filetbl
     fldinst
     fldrslt
     fldtype
     fname
     fontemb
     fontfile
     fonttbl
     footer
     footerf
     footerl
     footerr
     footnote
     formfield
     ftncn
     ftnsep
     ftnsepc
     g
     generator
     gridtbl
     header
     headerf
     headerl
     headerr
     htmltag
     info
     keycode
     keywords
     lchars
     levelnumbers
     lfolevel
     list
     listlevel
     listname
     listoverride
     listoverridetable
     listpicture
     listtable
     listtext
     manager
     mhtmltag
     nesttableprops
     nextfile
     nonesttables
     objalias
     objclass
     objdata
     object
     objname
     objsect
     objtime
     oldcprops
     oldpprops
     oldsprops
     oldtprops
     operator
     panose
     pgp
     pgptbl
     picprop
     pict
     pn
     pnseclvl
     pntext
     pntxta
     pntxtb
     printim
     private
     pwd
     pxe
     result
     revtbl
     revtim
     rsidtbl
     rtf
     rxe
     shp
     shpinst
     shppict
     stylesheet
     subject
     tc
     template
     title
     txe
     ud
     upr
     urtf
     userprops
     xe

Functions    omrtf-version    rtf

Platforms    HP/UX    IBM OS/2    Linux (Intel)    MS Windows 98/ME    MS Windows NT/2000/XP    Sun Solaris

[ INDEX ] [ CONCEPTS ] [ TASKS ] [ SYNTAX ] [ LIBRARIES ] [ LEGACYLIBRARIES ] [ ERRORS ]