The Threefold Challenge of Integrating Text with Chemical Graphics

Chapter 11

The Threefold Challenge of Integrating Text with Chemical Graphics 1

Robert M. Hanson

Downloaded by UNIV OF PITTSBURGH on May 3, 2016 | http://pubs.acs.org Publication Date: June 15, 1987 | doi: 10.1021/bk-1987-0341.ch011

Integrated Graphics, P.O. Box 401, Northfield, MN 55057

Three major challenges relating to the problem of integrating text with chemical graphics are detailed, and the approaches taken by the author in designing the FLATLAND system to meet these challenges are outlined. The FLATLAND system provides a model from which a new, inclusive definition of "integration" can be made, a definition based on flexibility and involving the entire process of document production. The problem of integrating text and chemical graphics i s readily apparent to any chemist: we get our graphics from many sources, f o r example, from instruments i n the form of spectra and chromatographic traces, from databases i n the form of chemical structures and reactions, from experiments i n the form of synthetic pathways and k i n e t i c s data, and from our own imaginations i n the form of mechanistic insights and hypotheses. We have the need to combine a l l of these various graphics with text to produce a r t i c l e s f o r publication, reports for internal communication, books, and theses. For some time word processors have been capable of dealing with most of the text-writing requirements of the chemist, but only recently have computer-based sj^stems been developed which aid i n the graphics area of the process. Consider the task of putting together a document r e l a t i n g to chemistry, such as a thesis, a report, or an a r t i c l e for publication. What would i t be l i k e i f we could take any table of data, any chart, any picture of a spectrum or molecule, or any chemical scheme or figure and, as we write, integrate that figure or table into our document? What i f we could introduce new raw data, redraw a figure, or update the y i e l d of a reaction i n a scheme without complicated computerized cutting and pasting? What i f v/e could do a l l this and not be limited to using a s p e c i f i c word

'Current address: St. Olaf College, Northfield, MN 55057

0097-6156/87/0341-0120$06.00/0 © 1987 American Chemical Society

Warr; Graphics for Chemical Structures ACS Symposium Series; American Chemical Society: Washington, DC, 1987.


11.

HANSON

Challenge of Integrating Text with Graphics

121

processor? What i f we could integrate into our document raw data from graphics, non-graphics or data base sources already i n service without having to write complicated graphics output interface routines for each? Clearly i f a l l this were possible, document production would be considerably simplified i n many ways. The future for integrated graphics/document processing w i l l depend upon how well program developers respond to three basic challenges. F i r s t i s the challenge of developing a system which can c o l l e c t raw data from a variety of sources and transform those data into graphic form. Second i s the challenge to develop a system f l e x i b l e enough to operate with a variety of word processors, graphic input devices, and printers, not only those available now, but those of the future. Third i s the challenge to develop a system which makes possible the integration of the whole process of document production, allowing rapid access to a l l of the stages of production at any time. We w i l l outline these three challenges i n turn and show how the FLATLAND system addresses each. Challenge I;

Multiple Graphics Sources

Chemical graphics come from a wide variety of sources. Getting this diverse information into one document invariably requires a substantial amount of "cutting and pasting". A major challenge facing those of us interested i n no-cut, no-paste document production i s how to get information, such as data from spectrometers, structures from data bases, graphs from data analyses, and schemes from synthetic work, a l l in a form suitable for insertion into a document. What i s needed i s a general purpose "black box" processor which can convert graphic data from sources such as ORTEP (X-ray crystallographic analysis), MM2 (molecular modeling), MACCS (Molecular Design Limited), and R/S 1 (BBN, Inc.; for data analysis) into the form required by word processors. Three approaches to this challenge w i l l be outlined. One approach to this problem, taken by T a l a r i s , Inc., for use with their laser printers, merges graphics and text f i l e s at the f i n a l printing stage (QDRIVE). One of QDRIVE's p r i n c i p l e strengths i s i t s capability to accept graphics data i n several formats, translating i t on the spot for printing on the T a l a r i s laser p r i n t e r . Thus programs which were o r i g i n a l l y written without device independence i n mind and producing output for Tektronix or Versatec plotters require no changes for t h e i r data to be included i n a document. Apple Computer's Macintosh system uses a r a d i c a l l y different approach. The Apple system emploj^s a standardized device language (Postscript) for a l l of i t s graphics, so any a p p l i c a t i o n automatically creates graphics images i n the proper format for document inclusion. Apple has made i t r e l a t i v e l y easy for the sophisticated user to write applications producing document-ready graphics. The Apple system suffers only i n that i t i s very expensive on a large scale and cannot u t i l i z e graphics data from more t r a d i t i o n a l sources, such as MM2 and R/S 1. The FLATLAND system provides a model, on the other hand, for how one graphics system can, i n p r i n c i p l e , u t i l i z e the raw data from a variety of information sources, including operator input, public


122

GRAPHICS FOR CHEMICAL STRUCTURES

and private structure and scheme l i b r a r i e s , molecular mechanics programs, X-ray crystallographic data, and in-house data bases.


Challenge II;

Device and Word Processor F l e x i b i l i t y

Nev? and more powerful word processors are becoming available each year. Getting graphic data i n a form f l e x i b l e enough to be used by more than one word processor i s an ever-present problem. The problem i s exacerbated by the fact that several laser printer s\7stems e x i s t , or are under development, each of which requires r a d i c a l l y different protocols and graphic languages. To complicate matters, laser printer technology i s not l i k e l y to s t a b i l i z e f o r at least a few years. In order to be compatible with future word processing systems, graphics-producing systems such as FLATLAND must be designed to work with several word processors and generate output suitable for more than one output device. Challenge I I I ;

Process Integration

One of the most frustrating aspects of word processing i s that system designers have generally not considered the p o s s i b i l i t y that t h e i r product might be part of a larger system. Thus, word processors are not designed to allow easy communication with other concurrently running programs. The actual process of combining text and graphics information, however, generally involves going back and forth between writing of text and preparation and modification of graphics. Although the Apple system displays text and graphics together, f a i r l y complicated deletion/reinsertion steps must be carried out when changes are required i n the graphic part of the document. Such complications can be lessened by using switching programs (which allow multiple systems to be quickly accessed). Most word processors do not merge graphics and text u n t i l print time, and thus do not allow for the ready v i s u a l i z a t i o n of graphics with text u n t i l a l l i s on paper. The advantage here i s that alterations i n the graphics can be made e n t i r e l y independently of the text, sc no cutting and pasting i s required. A clear challenge for word processors of the future w i l l be to allow ready v i s u a l i z a t i o n and a l t e r a t i o n of both graphics and text, with even computerized cutting and pasting being unnecessary. The FLATLAND system uses a novel switching technique to allow immediate transfer between text processing and graphics production. The FLATLAND System FLATLAND takes i t s name from the book FLATLAND: A Romance of Many Dimensions written by Edwin A. Abbott i n 1884. In his book, Ahbott depicts his world through the eyes of 19th century V i c t o r i a n s a t i r e , imagining a land of two dimensions. Likewise, the depiction of our three-dimensional world of chemistry i n the two dimensions of the printed page i s the aim of the FLATLAND system, developed by the author for use on a VAX computer operating under the VMS environment. E s s e n t i a l l y , FLATLAND i s a drawing program, capable of producing high quality structures and reaction schemes for output on



11.

HANSON

123


various video terminals and hard copy devices (see Figure 1). FLATLAND, though, i s much more than a two-dimensional drawing program i n that the third dimension i s s t i l l there (see Figure 2). FLATLAND i s unlike other drawing programs i n that information i s stored i n molecular form, that i s , as atoms, groups of atoms, and bonds, rather than as lines and c i r c l e s and l e t t e r s . The data format i s an enhanced version of that used i n molecular modeling, making FLATLAND uniquely suited f o r interaction with data bases and l i b r a r i e s of structural information. Data bases such as those used by the MACCS system might be d i r e c t l y tapped f o r molecular structures needed during the drawing session. FLATLAND also has i t s own structure and structure fragment l i b r a r i e s (see Figure 3 ) . Structures and reaction scheme "templates" can be used over and over while drawing to add whole predrawn molecular or scheme sections to the developing scheme. The l i b r a r i e s , both public (system wide) and private (user owned), are dynamic and can be customized easily and continually. FLATLAND Organization The organization of information i n FLATLAND i s unique (see Figure 4). Documents are seen as a combination of text and graphic scheme "pointers". A pointer i s simply a one-line command indicating to the word processor the space a l l o c a t i o n (in l i n e s ) to be allocated f o r the given scheme along with the name of the picture f i l e (in printer language) to be merged with the text at print time. FLATLAND produces that picture f i l e from a f i l e called the scheme f i l e , which, l i k e a document, includes text along with a set of pointers to other f i l e s (in this case to structure f i l e s ) . Structure f i l e s contain l i s t s of point positions, labels, and connections (atoms, groups, and bonds). Both scheme and structure f i l e s may also contain searchable, nonprintable comments. Though h i e r a r c h i c a l , FLATLAND's design allows f o r substantial f l e x i b i l i t y . Just as schemes can be modified e n t i r e l y independently of the document text, structures can be modified e n t i r e l y independently of the scheme. In addition, a single structure can be used i n many schemes without redrawing or copying, since each scheme merely contains a reference to that structure. Schemes are device independent and can be converted to picture f i l e s f o r output to terminals, p l o t t e r s , or laser printers. Picture f i l e s can then be displayed, plotted, printed, or merged with a document. FLATLAND and Challenge I As mentioned above, a major challenge f o r chemical word processors i s the requirement to deal with graphic information from a variety of sources, p a r t i c u l a r l y on-line services, data bases, and structure/reaction l i b r a r i e s . FLATLAND finds i t s e l f perfectly situated between these sources and the word processor. N f u l l c a p a b i l i t y word processor i s l i k e l y to be developed which can d i r e c t l y tap multiple, specialized graphic-oriented data bases. Likewise, the often used programs SHELLX (for X-ray crystallographic analysis) and MM2 (for molecular modeling) were not designed with graphic manipulation and word processors i n mind. Both are excellent Q



CHP/3A sieves (+)-DET OH

OTs

i

Ti(O Pr) CH C1 Downloaded by UNIV OF PITTSBURGH on May 3, 2016 | http://pubs.acs.org Publication Date: June 15, 1987 | doi: 10.1021/bk-1987-0341.ch011

2

Figure

F i g u r e 2.

ACETND

1.

2

An examnle o f a FLATLAND scheme.

T h r e e - D i m e n s i o n a l FLATLAND s t r u c t u r e s .

.ADAM

.SUGAR F i g u r e 3.

4

.BINAPH

CH3

.TBP

Examples from t h e FLATLAND

library.


11.

HANSON


125


MASS11 Work of K. Barry Sharpless: Scheme I

Document

WWARR

Scheme

0101

0102 >85%

/

0103

WWARR01

yield

\

CHP/3A sieves (+)-DET

OH ΤΪ(0'ΡΓ)

//

Structures

OTs 4

CH2CI2

F i g u r e 4.

FLATLAND o r g a n i z a t i o n

and t h e r e l a t i o n s h i p

FLATLAND schemes and MASS 11 documents.


between

126


at calculating molecular geometries, but neither i s especially good at graphically presenting those geometries in ways that can be e a s i l y integrated into documents. FLATLAND s h i e r a r c h i c a l design allows for structural information from v i r t u a l l y any source to be u t i l i z e d at scheme production time. FLATLAND s design provides for automated run-time access to any number of small, user-written reformatting/searching programs. Thus, while preparing a scheme, one might retrieve a structure from a previous (or concurrent) MM2 modeling session, reading i t in as a FLATLAND structure. That structure could then be independently modified, three-dimensionally rotated, and rescaled anytime p r i o r to or after inclusion i n the scheme. In fact, FLATLAND could be used as a graphics "front end" to manipulate c a l c u l a t i o n a l data three-dimensionally, thus supplying a graphics module to any c a l c u l a t i o n or data base access system. Such use requires no changes to the FLATLAND system whatsoever. A l l that i s required is the writing of a few small, independent data format translation programs. Such programs could be quickly written by an in-house system programmer. 1


1

FLATLAND and Challenge II The Challenge of the ever-changing word processor scene i s also addressed by FLATLAND. Since schemes are not the actual pictures, they can be device- (and thus word-processor-) independent. Schemes are seen as the intermediaries, the organizers of information, rather than the information i t s e l f . Thus, a l l information i s organized i n FLATLAND without any dependence upon destination. The same scheme can be converted to a displayable image, sent to a pen p l o t t e r , or translated into a laser printer f i l e for either direct p r i n t i n g or inclusion i n a document. The key here i s that as new devices become available, new device "drivers" can be added; as new word processors are employed, the schemes can be "repackaged" to remain compatible. FLATLAND and Challenge III FLATLAND i s the f i r s t chemical graphics system ever designed s p e c i f i c a l l y to work concurrently with other systems. Information can be transferred back and forth between systems, and FLATLAND can be instructed to wait for other processes to c a l l for graphics. For example, l e t us say FLATLAND i s being used to create graphics for inclusion i n a SCRIBE (Unilogic, Inc.) text. SCRIBE i s a document production system that r e l i e s on a system editor such as EDT for text entry. SCRIBE i s not a "what-you-see-is-what-you-get" system by any means. Rather i t i s a very smart processor, which can automatically move blocks of text to f i t page boundaries, handle multiple numbering systems (tables, figures, compounds, and references, for example), and insert picture f i l e s from external sources (FLATLAND, for example). A l l SCRIBE needs i n order to insert a picture i s a simple command, for example @pic(size=36 l i n e s , file=WWARR01.LN0), stating the size of a picture i n lines and the VAX filename associated with that picture.



11.

HANSON


127

The novel aspect of FLATLAND i s that both processes, text writing with SCRIBE and scheme production with FLATLAND, can be carried out concurrently on the VAX. When a scheme i s ready, one simply jumps to the text and presses a key. The proper SCRIBE command f o r that scheme i s placed d i r e c t l y into the text where i t i s needed, including the calculated space allotment. No measuring i s necessary. Note that the scheme i s not actually "placed" into the document; only a reference, a "pointer", i s introduced. Using MASS11 (Microsystems Engineering) the procedure i s similar, but i n that case the "embedded" command r e f e r r i n g to the picture, , must be entered manually. FLATLAND n o t i f i e s the user of exactly the format to use, again including the calculated space allotment. In either case, the effect i s an integration of the processes of producing the schemes and writing the text of a document. Summary: Integration and FLATLAND In summary, integration i n the context of FLATLAND takes on a much more encompassing d e f i n i t i o n than has previously been given the term. Integration i s not seen as merely the placing of graphics data with text data on the way to a laser printer. Nor i s integration seen as providing a "what-you-see-is-almost-what-youget" view of a finished document. Rather, integration i s seen as the bringing into a single operation the entire process of putting a document together. Integration i s seen as the c a p a b i l i t y to draw upon graphic information from a wide range of sources at document production time, and to make possible the concurrent production and modification of both text and graphics. It i s hoped that developers of word processors of the future w i l l recognize that no system can "do i t a l l " . The key l i e s i n establishing a d i v i s i o n of labor: i n l e t t i n g modeling programs such as MM2 and data base access systems such as MACCS provide the structures, i n l e t t i n g graphics programs l i k e FLATLAND provide the l i n k , and i n l e t t i n g processors such as MASSll produce the text. Only then w i l l we have complete integration; only then w i l l we have a f l e x i b l e system allowing immediate access to a l l stages of document production.

RECEIVED March 25,1986


The Threefold Challenge of Integrating Text with Chemical Graphics

Recommend Documents