Presentation-oriented edit operations require that an edited presentation can be parsed to yield an updated document structure. However, conventional parsing techniques cannot readily be applied, since presentations in Proxima are not restricted to text but may also contain graphical elements. For example, an exponential may be presented as 32. Although this graphical presentation may not be directly edited at the presentation level, its components may. Hence, instead of simply parsing the changed representation, we have to take into account the existing structure.

This paper explains the scanning and parsing process for presentations that are a possibly nested combination of text and graphical elements. For textual parts of the presentation a Haskell combinator parser needs to be provided. The parser for graphical parts, on the other hand, is constructed by Proxima, based on information in the presentation. White space in the presentation can be handled automatically, if desired.