SGML: Murata Paper, Abstract

SGML: Murata Paper, Abstract

Alternate abstract: "This paper shows a file format for documents containing both logical and layout structures. First, the logical-layout relationship in structured document models (Interscript, DSSSL, ODA, etc.) is examined. Then, embedding nodes and mold nodes are introduced so as to simplify the logical-layout relationship. Embedding nodes are ones representing logical-layout correspondences; mold nodes are temporary dummy nodes, which will later be replaced by layout nodes. To store a document in a file, it is first augmented with embedding nodes and mold nodes. Then, a data stream is created from this augmented document and stored in the file. This file format has the following advantages. First, it is compact. No leaf nodes are duplicated; tags for embedding nodes and mold nodes do not require much space; the text formatting output can be omitted. Second, the depth-first traversal of logical structures and that of layout structures are efficient and easy to implement. Files are sequentially read only once, and neither logical structures nor layout structures need to be copied into the main memory. Third, the logical-layout correspondences are explicitly represented by tags for embedding nodes. Last, unlike CONCUR of SGML, a large class of WYSIWYG documents, such as documents containing footnotes, can be represented."