Ann Marks

Translating among processable multi-media document formats using ODA

Introduction listing of the document format features that we supThis paper describes our experiences in translating ported in our translations, _, processable multi-media documents among diverse systems. We performed these translations using the interThe EXPRES Project ! national standard Office Document Architecture (ODA) In June 1986, the National Science Foundation (NSF) solicited proposals for the Experimental Research in : as an intermediate representation. The next two sections provide brief descriptions of the NSF-funded Electronic Submission (EXPRES) project. EXPRES EXPRES project (under which this work was was designed to promote the electronic interchange of performed) and our goals for this phase of the project, multi-media documents among the scientific research Following that is a brief description of the multicommunity. The NSF proposal submittal process was media document systems at Carnegie Mellon University chosen as a controlled sub-environment in which to (CMU) and the University of Michigan (UM), the two pursue the goals of the project, because this process institutions at which this work was performed. This is embodies many of the issues that must be addressed by _EXPRES. In particular, proposals are documents that -_ followed by a brief introduction to ODA and the ODA Tool Kit, a subroutine library that was used in the conusually contain information expressed in several media, struction of the translators, such as text, line drawings and raster images. These sections are followed by a discussion of In September of 1986, the NSF made equal three. general issues relating to the fidelity of document transyear awards to the Information Technology Center at lations. We then describe our approach to document CMU, and to the Center for Information Technology format translation using ODA, the problems we enIntegration at UM. In addition, several vendors have countered and our solutions. We provide some details donated equipment and provided technical support to about the implementation of the translators, our current the EXPRES project: IBM, DEC, Apollo Computer, status and our conclusions on the effort. The last secSun Microsystems and Apple. There are many issues that must be addressed to tion discusses some directions for further work on document interchange and an appendix contains a improve the ability of researchers to interchange multimedia electronic documents. To make the task Permission to copy without fee all or part of this material is granted manageable, we are concentrating on the problem of provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of effective interchange of processable multi-media docuthe publication and its dateappear, and notice is given that copying ments among diverse systems. In particular, we are is by permission of the Association for Computing Machinery. To copy otherwise, or to republish, requires a fee and/or specific ignoring the method by which a document is transferred

Introduction to the Office Document Architecture

The Office Document Architecture (ODA) [30] is designed to aid the representation and interchange of office documents such as memoranda, letters and reports. It provides comprehensive facilities for describing the structure and content of complex multi-media documents.

Experience Implementing ODA

We spent significant resources, in personnel and in time, on the CMU ODA Tool Kit, a portable subroutine library for application programs manipulating ODA documents. The tool kit effort involved specification, design, implementation, testing and distribution.

Initialization and Finalization

To allow proper access to any of the routines, variables or other identifiers exported by the tool kit, the application program must first initialize the tool kit. This is accomplished by calling the routine InitToolKit, described in section 14.1. The programmer is warned that the application program should not attempt to gain access to any of the tool kit facilities until this routine has been called and has returned successfully.

Using the Tool Kits

This chapter provides detailed information enabling a programmer to write application programs that make use of the CMU ODA Tool Kit and Raster Graphics Tool Kit. Section 12.1 explains how to make use of the include files provided by the tool kits. This is, of course, strictly a syntactic issue of putting the proper #include preprocessor statements in the appropriate places. The remaining sections of this chapter describe semantic issues. Section 12.2 describes the manner in which the routines exported by the tool kits should be used: how to call them and how to check for execution errors. Section 12.3 describes the global variables exported by the tool kits, the conditions under which these variables will contain meaningful information and what that information will be. Section 12.4 provides instructions for linking the appropriate tool kit libraries with application programs.

Document Models and Interchange Fidelity

This chapter provides a discussion of the document models and features we considered during the EXPRES project. We describe the structures of a document, the kinds of content a document contains, and the way in which formatting information is provided. We then examine how users edit a document, especially through the use of a style system. After examining each of these aspects of a document model, we consider briefly their impacts on document translation.

Introduction to the CMU ODA Tool Kit

This chapter presents an overview of the architecture of the CMU ODA Tool Kit. We have chosen to structure this discussion around an example application, that of document translation. We have used document translation because we designed the tool kit with this application in mind. This does not imply, however, that this is the only application for which the tool kit is suitable; layout and imaging are other obvious applications. Please note that the model of translation is illustrative only: we are not presenting a general discussion of document translation.

Experience using ODA

This chapter discusses our experiences using ODA as an interchange medium between document processing systems. Following the conventions in chapter 4, the EXPRES participants and their collaborators built a collection of translators between a variety of multi-media formats and ODA. The EXPRES participants, their collaborators and a group from the Technical University of Berlin connected their systems together and interchanged documents at several demonstrations. In this chapter, we discuss each of the participating systems and translators. We then discuss the scenarios presented at the demonstrations, including screen pictures showing the document as it was interchanged among systems. We close the chapter with an evaluation of the translation efforts: how well ODA served our purpose, how well the translators worked and what problems are intrinsic to the translation process.

Tool Kit Data Structures

This chapter describes the user-visible types defined by the Raster Graphics Tool Kit, in the first three sections. The raster formats supported by the tool kit are defined in the remainder of this chapter.

Raster Graphics Tool Kit Routines

This chapter describes the user-visible routines in the Raster Graphics Tool Kit. These routines can be classified broadly into five categories (described in the order in which they are presented): initialization and finalization routines (InitRGToolKit and FinalRGToolKit), creation and destruction routines (MakeRaster, ExtractRaster and DeleteRaster), assignment and access routines (AssignBit, AssignPel, GetBit, GetPel, InsertRaster, RasterPelFormat, RasterHeight and Raste rWidth), input and output routines (ReadMemRaster, ReadRaster, WriteMemRaster and WriteRaster) and miscellaneous routines (HoldRaster, ReleaseRaster and RasterStatus).