CA2487312C - Computer-human interface system for compound documents - Google Patents

Abstract

A document-centered user interface architecture for a computer system employs parts as the fundamental building blocks of all documents. All data is stored in the system as a part, which is comprised of contents and an associated editor. The contents and the functionality of the editor are available to the user wherever the part is located, whether in a document, on a desktop or in a folder. Parts function as containers for other parts, thereby facilitating the compilation and editing of multimedia or compound documents.

Description

COMPUTI''~R-Ff CMAN DACE SYSTEM
FOR CONN~"OUND DOCUS
E~1.3~
The present invention is directed to the interface between a computer and the human user of that computer, arid more particularly to a computer-human interface architectwre that provides increased flexibility in the creation and manipulation of compound documents.
Backaronnd of the snv ngtion Curna~tly, many applications for use in personal computers are written as single monolithic programs, each of which is des~ned with a ~ific primary function to be performed. For example, word processing programs are designed to create and edit textuai documents. Although not designed for tent processing, IS many other types of programs also provide for incidental use of text. For example, spreadsheet programs enable text to be inserted in the spreadsheet as descriptors for caws and columns of information. Similarly, graphics applications such as painting and drawing programs provide for the use of text as aabels within the paintings and drawings. While all of these different types of programs provide text editing facilities, they differ in the specific manner in which text is edited. Some of the applications provide a large number of capabilities far this function, while others are quite limited. Consequently, users are required to learn a different way s ,

2 to perform the same task, i.e. edit text, for each of the different types of applications.
It is desirable, therefore, to provide an architecture in which a component of a document, uch as text, can be employed wherever needed while at the same tame be created and edited in a consistent mamier, regardless of the particular environment in which it resides. With such an approach, the sarr~e set of commands can be used to edit tent whether it is present in a word processing document, a spreadsheet, a graphic document or any other document. The user only has to learn about one text exlitor, and can thereby become more comfortable with dhe editing process.
Further along these lines, it is desirable ttt be able to edit the components of a compound document within the document itself. In the context of the present invention, a compound document is one in which two or more different types of programs are used to produce the appearance of a single unified piece of information. For example, it could be a text document which includes a drawing created by a graphics program. These two types of information have a containment relationship to one another. More particularly, the text document functions as a container for the graphic object, such that the graphic object moves with and is affected by the text (container) document.
In the past, a compound document of this type would be created by running two separate applications. A. word processing program was used to generate the textual portion of the document.
A separate graphics application was used to generate the drawing.

3 To incorporate the work products of these two programs into a single document, the drawing would be stored in a particular data type or format, and the contents of this data structure would be placed in the word.processing document, for example by using a cut and paste operation. Often, the merging of these separate pieces of information into a single document was accomplished through a third application program, referred to as a presentation manager.
Once the drawing was embedded in the word processing IO document, its contents became static. If the drawing needed fix be.
changed, it was necessary for the user to call ups the graphics program, retrieve the original drawing, and edit that drawing within that program. The editing took place outside of the compound document, and the user coulid not view the changes in context as I5 they were being made. The revised drawing would tt~n have to be embedded in the word processing document, in plact of the original drawing.
To overcome the need to switch between various applications, it is desirable to provide an appraach which enables 20 the various components of a compound document to be edited within the document itself. Not only does this approach save time and simplify the procedure, it enables the user to see the changes in context as they are being made.
In another aspect of conventional operavng systems for 25 persanal computers, limitations are placid on the manner in which various types of information can be manipulated. For example, documents and other types of files are stored in containers, which

4 are identified as "folders" in the Macintosh operating system provided by Apple Computer, Inc., and as "dirext:ories" or "subdirecwries" in the MS DOS operating system provided by Microsoft Corporation. While folders and directories can contain documents, they cannot themselves be incorporated into a document.
Similarly, applications and documents can be represented as icons in graphical interfaces which employ a desktop metaphor, such as the Macintosh Finder system and the Mie;rosoft Windows environment. Currently, there are limitations in the manner in which icons can be manipulated in these interfaces. For example, an icon can only reside on the desktop or in a window which pertains to a foldex (or subdirectory) . If the window represents something other than a folder or volume, for exa,anple a textual I5 document, an icon cannot be placed in it. In addition, certain lands of icons can only be placed on other specific types of icons. For example, if an icon represents a document, it can only be placed on an icon which represents an application that pertains to that type of document. The document icon cannot be placed on other types of icons, including those which may pertain to other types of applications.
Accordingly, it is an objective of the present invention to provide a system in which a consistent user interface can be provided for a task, regardless of the environment in which the task is to be performed. In addition, it is the obje~dve of the present invention to eliminate the limitations on the manipulation of various types of data, to thereby provide greater flexibility in the creation and editing of all types of compound documents.
u.aaf cr~tpn.pnt of the Invention In pursuit of these objectives, the present invention provides

5 a uniform interface that enables all different Idnds of information-conveying elements to be combined and manipulated within documents in ways that were not previously possible. In the context of the present invention, a document is Considered to be the element of an interface whose purpose is to present information to the user and allow that information to be edited, for example a text document or a graphics document.
The uniformity of the interface is provided through a document-centered architecture which employs a fundamental building block, hereinafter referred to as a "part'" , to accomplish tasks. A part is a self-contained element that is comprised of two primary components, its content and a manipulator for that content.
These two components are always available to the user, regardless of where the part may reside within the system, e.g. as an element of a document, in a folder, or on a desktop. 'The manipulator for the part cart be any type of editor or viewer for the contents of .the part, and it can be selected by the user, if necessary. For example, if the contents of a part is text, the editor which is a component of that part can be any word processor with which the user has familiarity. That word processor can be used to edit the textual contents of the part, wherever they may be embedded. Thus, even if the part is embedded within a nontextual document, e.g. a

6 spreadsheet, the textual contents of the part can still be edited with the user-selected word proc:eisor.
An interfac,~e which exhibits this feature has a significant inherent characteristic: it is simultaneously simpltx and more powerful. The simplicity results from the fact that the user only has to learn one way to perform a particular task, such as editing tead. At the same time, it is more powerful because the text editor can be a fully functional module, which replaces less capable ones curra~tly found in many types of applications.
The two basic components of a part enable it to function as both an object (due to its inherent contents) and as an application (resulting from its inherent manipulator or editor. Due to these characteristics, a part can provide a number of different functions.
For example, every document is a part, and eve~yy part can act as a IS document. To 'illustrate, a text part is a document containing characters, and a graphic part is a document containing Iines and circles. In the present invention, folders, and erren the desktop, cx~nstitute parts, and therefore can contain contents that were previously associated with documents only, such as text. This feature of a part provides a uniformity that does not east in currently available applications, where objects such as folders cannot function as doc~iments.
Another feature of the part is that it furK;tions as a container, i.e. a part is able to contain other parts. Furthermore, a part is not limited to the types of parts that it contains. IL~ther, if a part can contain any type of part, it can contain all types of parts. This feature of the pare renders the present invention an ideal

7 architecture for the creation of compound documents. Since a document can contain parts, it will automatically be able to contain all types of parts repres~ting different types of information, without any modification of the document or the manipulator.
These features result in an interface in which compound docum~ts are assembled by arranging different harts in a manner which produces a desired result. In this sense, the intcrfacx of the present invention is document-centered, rather than being application oriented. More particularly, a eonverntional application pmgram does not have any role in the infeTfaoe of the present invention which is separate from the data 'with which it is bundled to form a part.
Another aspect of the present invention is the manner in which parts are represented to the user. A part can be represented as an icon, which is a small picbue that provides a handle for the part, allowing it to be manipulated as a whole. lFor example, an icon can be dragged on a desldop and betyvoen windows, and can be the desrination for dropping other ions during a drag operation.
Also, the contents of a part can be expanded into the workspace defined by a window on a desktop, similar to conventional applications.
parts can be alternately represented as frames. A frame is an area of a display screen that represents a part. Like an icon, frames provide handles onto parts that allow them to be manipulated as a whole. They can be dragged am the desktop and windows, they can provide the destination for drops during drag operations, and they can be opened into windows and closed back into frames. Unlike an icon, however, a frame allows a part's contents to be seen and edited in place.
A frame is a data structure which functions as an interface between a containing part, e.g. a document, and as embedded part.
The frune consists of a member of attributes, some of which are controlled by the containing part and some by the embedded part.
Through the intermediary of the frame, the manipulators for the containing part and the embedded part can negori~ate the space allotment that is pro»ded to the embedded part within the containing part.
By utilizing parts as the fundamental building blocks for the user interface, the present invention provides capabilities that were not previously available in personal computer applications. Thtse capabilities are described hereinafter witty reference to specific examples of preferned embodiments of the invention.
~i stt~.c Figure 1 is a block diagram of the main components of a computer system;
Figures 2A and ZB are two views of a sateen display, illustrating the dragging of a part from a folder to a document;
Figures 3A and 3B are two views of a screen display, illustrating the activation of fumes;
Figure 4 is an illustration of a selected fr~amme within an active frame;
Figure SA and 5B are diagrams illustrating two examples of frame structures, and Figures SC-5F are flow diagrams for examples of negotiations which take place between an embedded part and a containing part with respect to frame attributes;
Figure 6 is a flow chart of the process for launching an editor;
Figures 7A-71r are a sequence of screen displays illustrating the dragging of a part between a desktop and a document;
Figures 8A and 8B are a flow chart for the process of dragging and dropping a part;
Figures 9A-9C, l0A-IOC and 1IA-11C a~~e respective sequences of screed displays illustrating the copying of material from one part into another part;
Figure 12 is a flow chart of the process for placing material from one part into another pxtts Figure 13 is a screen display illustrating a part having restricted access; and Figure 14 is a screen display illustrating a parts bin.
To facilitate an understanding of the prescnt invention and the advantages offered ther~y, features of the invention are described and illusdated with reference to the use of the invention in a Macintosh~ brand computer supplied by Apple Computer, Inc.
It will be appreciated by those having familiarly with the relevant technology, however, that the principles of the invention are not limited to this particular type of operating environment. Rather, the its which underlie the invention are applicable to any type of computer operating system in which it is desirable to provide a uniform interface for the compilation and editing of docum~ts, particularly compound documents.
A typical computer system, of the type .iii which the present invention can be employed, is illustrated in block diagram form in 5 Figure 1. The structure of the computer itself does not form part of the present invention. It is briefly described here for subsequent understanding of the manner in which the features of the invention cooperate with the structure of the computer. Referring to Figure 1, the system includes a computer 10 having a variety of external 10 peripheral devicxs 12 connected thereto. The computer 10 includes a cxntral processing unit 14, a main memory which is typically implemented in the form of a random acxess memory 16, a static memory that can comprise a read only memory 18, and a permanent storage device, such as a magnetic or optical disk 20.
The CPU 14 communicates with each of these forms of memory through an internal bus 22. The peripheral devices 12 include a data entry device such as a keyboard 24, and a pointing or cursor control device 26 such as a mouse, trackball or the like. A display device 28, such as a CRT monitor or an I,CD peen, provides a visual display of the information that is besng processed within the computer, for example the contents of a document. A hard ovpy of this information cwt be provided through a printer 30, or similar such device. Each of these external peripheral devices commw~icate$ with the CPU 14 by means of one or more , input/output ports 32 on the computer.
The preseait invention is particularly directed to a system which controls the manner in which the CPU 14 handles information that is manipulated by the keyboard 24 and the cursor control device 26, as well as the manner in which that manipulated information is portrayed to the user through the display device 30.
An example of an interface which employs the principles of the present invention is illustrated in Figures 2A andl 2B. These figures illusaabe a desktop 34 which defines a workspace 36. Also included on the desktop is a menu bar 38. Within the workspace 36 are two windows, 40 and 42. The left-hand window 40 is a folder window which contains icons representing various parts. As shown in Figure 2A, three icons 44 pertain to text parts, a fourth icon 45 represent a c part and another icon 46 represents a spreadsheet part. A sixth icon 48 is a stationery icon representing a part that can be used for the creation of presentation slides. In Figure 2A the right hand window 42 contains a text document.
More particularly, the document includes a text part 50. This text part can be created by typing text within the window 42, or by opening in the window a previously-created text part that was represented by an icxan.
Figure 2B illust:af~es the result of an operation in which the graphic icon 45 has beg dragged from the foldear in the left window 40 to the teat document in the right window 42. As is known in interfaces which provide a desktop mexaphor, such as that shown in Figures 2A and 2B, dragging is an opE:ration in which objects can be moved or copied on the desktop and within windows through the actuation of the cursor control device 26. Once the graphic ion is planed within the dxum~t, its contents are displayed in a frame 52. Since the document now contains a graphic element as well as a text element, it is referred to as a compound document:
The text 50 and the graphic element within the frame 52 pertain to separate respective parts which together make up the document. As described previously, a part is a self-contained entity which is comprised of content and a manipulator for that content.
These two components are always available when the part is accessed, regardless of where it is located on the desktop. Thus, when the graphic icon 45 is located in the folder of window 40, as in Figure 2A, it represents a part comprising a graphic element and an editor for that element. When the part is moved to the document in the window 42, its contents go with it and the functionality of an editor for those contents is available to the user.
The significance of this feature is the fact that the contents of a part can always be edited, or otherwise manipulated, wherever they are located. Thus, in the compound documsrnt of 1 figure 2B, the textual content 50 can be edited within the document, for example by using the same word processor that was used to originally create the text. Since the graphic editor is present as a component of the part represented by the frame 52, it enables the graphical contents of the part to be directly edited in place. Thus, unlike prior applications for personal computers, it is not necessary for the uses to open a graphics application, for example in another window on the desktop 34, edit the graphical infbrmation as desired, and then export the edited graphics into the documene in the window 42. Rather, the user can directly edit the contents of the frame 52 within the context of the document in the window 42, without having to implement any intervening steps. The user can focus on document content, rather than an application program, and take advantage of the context provided by the surrounding document.
A part, therefore, is a self-contained object which is autonomous of the underlying system technology. It can be used in any type of software engine or environment, and the user does not have to learn the operation of a particular editor for the part. As long as an editor for that category of part is reprrrsented in the underlying system, the user can talae full advantage of the part's ~ilities. Furthermore, if a text part was created with a particular text editor (word processsor~, the user is not constrained to use only that text editor with the part. If the user has more familiarity with the commands of a different word processor, that processor can be designated as the preferred ediLar for all text parts in the computer system.
parts are not limited to documents and their contents.
Rather, every object in the system comprises a part. Thus, the folder in the window 40 is a part whose aantents are the icons 44-48. Similarly, the desktop 34 is itself a part, as wail as dialog boxes and other user interface objects.
Frames and icons are alternative representations of a part.
users can toggle between the frame and icon representations of a part by means of a suitable keyboard or menu command. If a document is large, e.g. it consists of several pages, only a portion of it is visible in the frame. Therefore, frames; Iike icons, can be opened into windows, to thereby allow all of the contents of large i4 parts to be seen, thraugh scrolling for example, and edited using the functions available with a window, such as magnification.
Similarly, they can be closed from a window back into a frame. A
frame differs from an icon, in that it allows its contents to be edited in place. A frame is also different from a window because a frame, or even multiple frames, can exist inside a window. A
window is a transitory view of objects, which remains open only while a part is being edited or examined. In contrast, a frame is a permanent representation of a part's contents.
A part, whether it is in the form of an icon or a frame, can act as a container for other parts. In the example of Figure 2B, the text 50 is located within a frame whose border is the same as that of the window 42. This larger frame contains the smaller frame 52 for the graphic part. Since the text part is the outermost part in the IS window, it is labelled as the "root part". Generally speaking, the mot part establishes the basic editing behavior fox a window.
A root part can be generated by means of a stationery icon, such as the icon 48 depicted in Figure 2A. In general, a stationery icon represents a predefined part and functions as a metaphor for a pad of paper in the physical world. When it is desired to generate a new document, a person tears a blank sheet of ;paper from the pad and places the contents of the document, e.g. wards, drawings, etc., on it. The stationery icon works in the same manner. When the user desires to create a new teat document, a text stationery icon is opened, for example by double-clicking on it. In response, _ the stationery icon creates (or "tears off") a copy of itself, which is opened into a window or a frame. This copy is ata ordinary part ________.______._ ___ ______._ ______.__.~...~.~~~ _,:~~~:~ ..
~.~y~ø.~.~.~...~n~..~~--_ _____~...~~ ~..~__m._ ______ ' CA 02487312 1994-05-04 zs which can be blank, i.e. it has no contents initially" or it could have certain pre-set information, such as letterhead or a company Logo.
The user then enters additional content, e.g. characaers, into the document.
The stationery itself is not opened as a window or a frame;
it only exists as an icon. Only the tom-off copy is opened. Thus, when a stationery icon is dragged into a document, rather than opening itself into a frame, a copy is "tom off", placed in the document, and that copy is opened into a frame. The stationery icon itself returns to the position it occupied before the move. This prevents the stationery from being accidentally inserted into a document, and possibly lost.
Since a frame can function as a container, it exhibits some of the known properties of a window. For example, when a user selects a part contained within a frame, such as a graphical object, the frame containing that part becomes active. When a part is active, it receives commands and keyboard events, and its menu and other user interface options are displayed. An example of the activation of a frame is illustrated in Figures 3A .and 3B. Figure 3A illustrates a desktop having a window 54 which includes a compound document containing text and a graphic part. In this compound document, the text forms the root part of the document.
In the example of Figure 3A, a portion 56 of the. text has been selected, and is represented by the reverse video display. In this case, the frame which contains the selected part is contiguous with the border of the window itself. The menu bar .'>8 for the desktop includes those commands which are appropriate :for editing the text.

l Figure 3B illustrates an example in which a graphical component, i.e. the triangle 60, has been selected. The frame which contains this graphical component is now active, and its border is indicated by the dotted line 62. Since the active frame is now a graphic part, the commands in the menu bar 58 have changed to those which are appropriate for editing graphical content. In addition, a palette b4 is displayed an tike desktop, to provide those tools which are necessary for editing the graphical contents of the active frame.
In addition to activating frames, a user may also select them. The selection of a frame can be accomplished, for example, by activating a frame so that its border is visible, and then pressing a button on the cursor control device while the pointer of the cursor is positioned over the border of the frame. Figure 4 illustrates a frame 66 that contains both graphical and textual content. The border 68 of the frame is indicated by a dotted line, to denote the fact that it is an active frame. Within this active frame is a frame 70 containing the textual content. This frame has been selected by depressing the button of the cursor control device, while the pointer of the cursor 72 is positioned over the border 73 c~f the frame. To denote the fact that the frame 72 has been selected, the border 73 of the selected frame 72 has a different appearanc~r from that of the active frame 66. For example, the border might be wider as shown in Figure 4. In addition, it can be provided with :resize handles 74, which enable the size and shape of the frame to be changed in a known manner.

In general, whenever a part or some of the contents of a part is selected, the smallest frame containing the selected element becomes the active frame. Thus, if a word within the frame 70 is selected, for example by clicking a cursor control ibutton while the cursor is positioned over the word, the frame 70 changes state from selected to active. At the same time, the border 68 of the frame 66 disappears from view since this frame is neither active nor selected.
Conversely, if the frame 66 is selected by clicking on its border 68, the border 73 disappears, the border 68 changes appearance, and the border of the frame (or window) containing the frame 66 (not shown in Figure 4) is highlighted to show that it is active.
In the examples of Figures 3A, 3B and 4, the frames are illustrated as having a rectangular shape' While this shape is preferable for displaying the contents of most types of parts, the shape of the frame need not be limited to being r~xtangular.
Rather, the frame can have any shape which suits the display of its contents.
The frame is not an element of the part it;~elf. Rather, a frame is a separate data structure which function: as an interface between the container part and the embedded part. Each frame has a number of characteristic attributes, which are denoted as various "shapes", Each of these shapes represents a geometric description of an area that defines certain document context that is shared between the container part and the embedded pa~~t. Some of the shapes are controlled by the container part, whereas others are determined by the embedded part. In general, dhese various shapes can represent any area of the coordinate space of the document context. In this regard, the shapes of a frame can represent disjoint areas as well as contiguous areas.
The various characteristic shapes of a frame: are described with reference to Figures SA and SB. In these figures, the left-s hand illustration depicts a compound document containing text (the root part) and a graphic part, represented by a tri~u~gle. The right-hand illustration shows the shape attributes of the frame for the graphic part.
One of the characteristic attributes of a frame is the Frame Shape. This shape is determined by the container part, and d~ribes to the embedded part the geometric area that it should use to lay itself out. Referring to Figures SA and SB, the Frame Shape is depicted by the outermost solid border 75. The; area defined by this shape might affect the scaling, line breaks, or other similar layout decisions that must be made by the editor for the embedded part. Only the container part can directly change the Frame Shape.
When this shape is changed by the container part, a notification is provided to the embedded part, to enable appropriate changes to be made in the content of the embedded part.
A second characteristic attribute of a frame is its Clip Shape, which is depicted by the shaded border 7f> in Figures SA
and SB. The Clip Shape describes to the embedded part the geometric area in which the contents of the part cyan be drawn. In the example of Figure SA, the Clip Shape is coextensive with the Frame Shape. However, in the example of Figure SB, the lower right-hand portion of the triangular graphic part ivs overlaid by a table. Therefore, the Clip Shape 76 for the graphic part excludes this area which is to be overlaid by the table. The Clip Shape is determined by the container, i.e. the text document which forms the root part of the example illustrated in Figures SA and SB. Only the container can change the Clip Shape, and when it does so it notifies the embedded part of the ~ehange.
Another characteristic attribute of a frame is the Used Shape, depicted by the solid line 77 in Figures SA and 5B. This shape is determined by the embedded part, and describes to the container the part of the geometric area that is actually being used by the embed~d part. This shape is necessarily a subset of the Frame Shape. The container part is not allowed Gd draw inside the Used Shape, but is free to draw in regions outside the Used Shape, even if they are within the area of the Clip Shape and the Frame Shape. The embedded part can change the Used Shape, and when it does so it notifies the container part of the change.
A fourth shape which is a characteristic attribute of a frame is its Active Shape, which is depicted by the shaded line 78 in Figures SA and SB. This shape is determined by the embedded part, and describes to the container part the geometric area in which the embedded part should receive geometrically related events such as cursor control input. Whenever an event occurs within this area, the container part must pass the event to the embedded part, rather than respond to the event itself. However, the container part is free to respond to events which occur in regions outside the Active Shape of the embeddai part.
Another attribute of a frame is its group tag. The frame group tag is an identifier, such as a number, which uniquely identifies a see of related frames in a given container. The container part uses the frame group tag to control the geometric relationship of the various frames in the group.
The frame allows container parts and embedded parts to 5 negotiate document structure with minimal knowledge of one another's internal constraints. Examples of negotiations that can take place between the containing part and the embedded part are illustrated with reference to Figures 5C-5F. In the example of Figure 5C, the Frame Shape is being altered by th~~ container. This 10 procedure may be initiated by a command from the user to change the size of the frame, for example. At Step 501, the containing part determines a new Frame Shape for the embedded frame, and notifies the frame of this new shape. In response thereto, at Step 502, the embedded frame transforms the shape from the containing 15 frame's coordinate space into that of the embedded frame. This new shape is stored as the Frame Shape, and passexl on to the embedded part. At Step 503, the embedded part adjusts itself to the new Frame Shape. For example, a text editor may reset margins for the layout of the text. If the Used and Active Shapes 20 are different from those of the Frame Shape, the embedded part also computes these new shapes. Control is then returned to the containing part, and at Step 504 the 'containing pa~~ changes the Clip Shape of the frame to correspond to the new Frame Shape, and notifies the frame of the new Clip Shape.
Figure 5D illustrates the procedure that tai~:es place when the embedded part requests a new Frame Shape. This request may be brought about as a result of an editing of the embedded part's contents, which necessitates additional space for the contents to be properly laid out. Ai Step 505, a determination is made that the embedded part needs a new Frame Shape, and a rexluest is sent to its display frame for a new shape. The old shape can be stored by the embedded part, in case it will be later needed. At Step 506, the display frame transforms the shape from the embedded frame's coordinate space into that of the containing frame, and passes the request onto the containing part. At Step 507, the containing part determines how to respond to the request. It may grant the request for the desired shape, or determine that a smaller shape is required.
Once the Frame Shape is determined, the Clip Shape for the frame is also modified, and a notification is provided to the frame as to the allowed shape. At Step 508, the display frams: transforms the shape from the containing frame's coordinate shape into that of the embedded frame. The new shape is stored as its Frame Shape, which is passed on to the embedded part. At Step 509, the embedded part changes its display frame's Used aand Active Shapes, to adjust to its new Frame Shape.
Figure SE depicts a procedure in which thf; container part changes the Clip Shape. Referring to Figure SB, for example, the need to change the Clip Shape may result from an action by the user, in which a portion of the area encompassed by the display frame is to be overlaid with the contents of a different part. At Step S I0, the containing part determines a new Clip Shape for its display frame, and notifies the frame of this new shape. At Step 511, the display frame passes the notification on to the embedded part. In response thereto, at Step S I2, the embedded part adjusts itself to the new Clip Shape by changing the manner in which its content is laid out, as appropriate.
Figure 5F depicts a procedure in which the embedded part changes its Used Shape. Again, the need for such a change may arise from an editing of the contents of the part by the user. At Step 513, the embedded part picks a new Used Shsape for its display frame, and notifies the frame of this new shape. At Step 514, the display frame passes the notification on to the containing part. In response thereto, at Step 515, the containing part adjusts itself to the new Used Shape. For ezarnple, it may change the way its content is laid out to flow differently around the new shape.
The frame, therefore, is a data structure which negotiates the required space allocation between an embedded part and a containing part, Through the use of various shape: attributes, the frame enables the containing part and the embedd~ part to utilize the available space in an efficient manner, without requiring that each part have detailed information as to the contents or behavior of the other part. Thus, compound documents can be assembled with a great degroe of flexibility.
The editor, or other manipulator, for a part is analogous to an application program in a conventional computer system. It is a software component which provides the necessary functionality to display a part's contents and, where appropriate, present a user interface for modifying those contents. It may include menus, controls, trool palettes and other interaction techniques. For a part which is fable of containing other parts, its editor takes these contained parts into consideration when displaying its content. For example, a text editor may wrap its part's content; so that it does not appear in the Used Shape of an embedded graphic part.
In order to use the functions provided by a conventional application, the program must be launched or booted. In other words, it must be stared in the system's active memory, e.g. the RAM 16. In a similar manner, the editor for a part must be running in the system's memory or otherwise acc.~;ssible to the CPU. To do so, the editor must be installed on tile system. For example, it can be stored in a folder on a hard disk 20, and called when an associated part is accessed. To conserve. system resources, all installed editors may not be loaded into the system memory at all times. Rather, they can be launchE;d as needed. For example, an editor for a part can be launched when the part is dragged into a document, since this is the time that the editing capabilities are most likely to be needed.
The process for launching an editor, or otlher manipulator, is depicted in the flowchart of Figure 6. Referring thereto, at Step b01 the system looks for a part to he activated. 'The accessing of this step may be in response to an event flag that is generated whenever the mouse button is depressed, or it could be a step in a script. Once the activation of a part is detected, the system identifies the editor associated with that part (Step 602). In this regard, every part has certain properties associatExl with it. One of these is a pointer which points to the editor for that part. Two other properties are the part's category, e.g. whether it is text, graphic; spreadsheet, etc., and its type, which is an indicator of the format for the part's contents. For example, one; text part could be of the "Word Processor A" type, and another's typE: could be "Word Processor B", although both belong to the s~~me category (fit), OfLen, a part's type will be determined by the last editor used on the part's contents. Generally speaking, a category is a set of part types, and determines the set of part editors or viewers that are applicable to a given part. The properties of a part are stored along with its contents. For example, all of a part's properties can be contained in a "Property sheet" that is stored with the part's contents.
Qnce the editor (or part type) has been identified at Step 602, the system determines whether that editor is already running in memory (Step 603). If so, the system returns tra the main routine and awaits further user action. If the editor is not already naming, the system checks whether the editor identified in Step 602 is installed on the system (Step 604). if the editor identified at Step 602 is installed on the system, that editor is launched at Step 605, so that it can be used to edit the part's contents.
It may be the case, however, that the part's editor is not currently installed on the system. In this case, the deasion at Step 604 would be negatide, and in response thereto the system determines whether the user has selected a preferred editor for that category of part (Step 606). More particularly, a user can select a preferred, or default, editor for each different caLcgory of part.
Thus, if the user is most familiar with the function commands of a particular word processor, that processor can be selected as the preferred text editor for all parts in the text category. If a preferred editor has been selected, the system checks whether the preferred editor is installed (Step 607), and if so it is launched at Step 608.
If no preferred editor has been specified, or if the preferred editor does not exist on the system, a check is made at Step 609 5 whether any editor for that category of part is present on the system. If not, an appropriate 'warning is provided to the user (Step 610), and the system then returns to the background routine.
If one or more editors fvr that category of part is available, they are listed and the user is prompted to select one (Step 611). In 10 response to a selection by the user, the selected editor is launched (Step 612) and it is stored as the preferred editor for that category of part (Step 613). The system then returns to the background routine to await further user inert.
The user can be provided with the option oaf changing the 15 . designated editor for a part at any time, for example through an appropriate menu command. By invoking this command, the user causes the pointer that is stored as one of the part's properties to be changed to the location of a different editor installs on the system.
Menu commands can also be provided to enable the user to select 20 preferred editors at different levels, for example a1: the document level (all text parts in a document have the same editor) and at a global (system-wide) level. Whenever the user ac~~esses the menu to select or change an editor for a part, only the set of installed editors which are appropriate for that category of apart are displayed 25 to the user for selection.
If desired, the system can keep a record of the last time that each editor running in memory has been used. If an editor has not been used for a predetermined length of time, the system can automatically delete the editor from memory, to thE:reby conserve system resources. Also, when a document is closed, the editors associated with the document's parts can also be closed, unless they also pertain to other parts which are currently open, for example in another document.
Because of the autonomous nature of parts, the containers for various parts are no longer Limited to specific types. For example, in the past a text document could only contain graphics or other content which was in specific data formats. :Limitations of this type are removed with the architecture of the present invention.
Basically, any part appears as a "black box" to any other pad. A
containing part need not have any information about the internal structure or semantics of other parts contained within it. The containing part merely functians as a wrapper for the other parts contained in it. This type of architecture greatly facilitates the compilation and editing of compound documents.
To further illustrate, a document can be represented by an icon which is located on the desktop itself. Referring to Figure 7A, the illustrated desktop includes a window 79 containing a text document. Also located on the desktop is an icon 80 for a graphic document. Each of the window and the icon reprE:sents a respective part. in the case of the text document located widlin the window, the part is represented in the form of a frame, whose borders are contiguous with that of the window itself. In accordance with the present invention, the part represented by the icon can be directly manipulated by the user and placed within the text document, 2?
without the need for intervening dialog boxes or the like. This operation is described hereinafter with reference to the illustrations of Figures ?A-7E and the flowchart of Figures 8A and 8B.
To place the graphical document represented by the icon 80 S into the body of the text document within the window ?9, the user first positions the mouse pointer over the icon (Step 801) and depresses the mouse button to select the icon (Step 802). In response thereto, the system highlights the display of the icon, to indicate the fact that it is selected (Step 803). In the example of Figure ?A, the highlighting of the icon 80 is indicated by the reverse video representation of the icon's picture and its title.
While holding the mouse button down, the user moves the cursor to the desired location for the graphic part (.step 804). As illustrated in Figure ?B, the selected icon follows the movement of the cursor. The movement of the icon with the cursor is known as a "drag" operation. While the icon is being dragged to the desired location, a ghost outline 81 of the icon remains on the desktop, to indicate its original position. Alternatively, the ghoat outline can move with the cursor, and the full icon representation remain in the original position, as depicted in Step 805.
During the dragging of the icon, the system determines whether the cursor enters a window (Step 806). As the icon 80 is positioned inside the window ?9, the window itself becomes highlighted. If the system detects that the dragged icon enters a 2S frame within the window (Step 80'x, the frame is appropriately highlighted, to indicate that it can receive the dragged object (Step 808). If the icon continues to be dragged so that it goes back _a ..,~,. o, , ~. ~ .. .,~ ..r..._ . ..... _..,.. ~. . ~ , ,,~",& .~,~,xs , w zs outside of the frame, the highlighting on the frame: is removed (Step 809). Conversely, if plural frames are nested inside one another, each successively nested frame becomes the active one as the icon is dragged .inside it. In this manner, the system provides feedback regarding the fiarne that will be active and the location at which the icon will be dropped if the cursor contri~l button is released.
Once the dragged icon is at the desired location, the user releases the mouse button (Step 810). At this point, the system determines the preferred representation for the part, i.e. whether it should appear as an icon or as a frame (Step 811)" In the example of Figures ?A-?E, since the part has been placed in a document, its preferred representation is as a frame. Accordingly, as shown in Figure ?C, the graphic part is displayed as a frame 86, with its upper left corner corresponding to the position of the cursor when the mouse button was releas~l (Step 812). As als~n shown in Figure ?C, the frame is illustrated with a thicker border and resize handles, to indicate the fact that it is the selected part.
The reverse operation can be performed to move the part from the document back onto the desktop. Furthermore, it is possible to copy the part, so that its contents remain within the document while another representation of the part is placed on the desktop. To do so, the user depresses a function key, such as a control or option key an the keyboard, while dragging the frame.
The simultaneous depression of the function key along with the actuation of the mouse button provides an indicati~an to the system that a copy operation is to be performed. Then, while continuing to hold the mouse button down, the user drags the firame to the desktop, as illustrated in Figure 7D. Note that the contents 88 of the graphical part remain within the text document, and a second copy of the part appears in the frame 86. Once the frame is moved to the desired location on the desktop, the mouse button is released, and the system again determines the preferred representation for the part. In this case, since the part is located on the desktop, the preferred representation is as an icon. Thus, an icon representation 80 of the part is displayed on the desktop, as shown in Figure 7E.
10 Since the part is directly manipulated between the desktop and the document in the system of the present invention, the metaphor of the desktop is better preserved. In other words, the user picks up a desired part from the desktop and places it directly in the document, without any intervening operation such as filling 15 out dialog boxes or the like. As a result, the amount of time and number of steps that are required to introduce new content into a document is reduced.
Further details regarding one manner in wlhich the system can implement the dragging and dropping ofobje;cts between a 20 desktop and a document are described in US Patent No. 5,598,524, filed on March 3, 1993 by Robert G. 3ohnston, Jr., Mark L. Stern and David L. Evans, entitled "Method and Apparatus For Improved Manipulation of Data Between an Application Program and the File System on a Computer Controlled Display System".

As described above, each part has a property which is referred to as its preferred representation. This property determines whether parts contained inside a given part aTe normally displayed as icons or as frames, in order to obtain ;a behavior which is traditionally associated ~mth both folder and document windows.
Generally speaiang, all parts are classified as either a desktop part or a document part. A desktop part is one whose purpose is to contain or operate on other parts, such as a folder, printer or mailbox. The preferred representation for a part contained in a desktop part is as an icon. A document part is a part whose purpose is ultimately to be printed on paper or otb~er hard copy, such as a text document or a graphics document. The preferred representation for a part that is contained in a document part is as a frame. Thus, referring to the operation depicted :in Figures ?A-7C, the c part is illustrated as an icon when it is present on the desktop (Figure 7A), but is changed to a frame representation when it is placed within the document (Figure ?C~. Similarly, when the part is moved from a document to the desktop, its representation changes from a frame to an icon (Figures ?C-7E;1. Of course, the user can be provided with the option to change the representation of any part, for example through a keyboard or menu command.
As described previously, a stationery part: is different in the sense that it is only represented as an icon. It ~~n be moved from one desktop part Lo another. Thus, for example, it can be treated vke ar, ordinary part in a folder. However, when an attempt is made to move a stationery iron into a document part, a copy of the stationery part is placed in the document part instead, and that copy appears as a frame.
In addition to moving parts as a whole, the user can select and move some or all of the contents of a part. Every part has its own intrinsic type of contents. For example, text parts contain text characters, graphic parts contain graphic elements, etc. The user can select some of the intrinsic contents of one part, the donor part, and move or copy them into a different part, the destination part.
There are three possible scenarios in which this type of operation can occur: (1) the donor and destination parts both contain the same intrinsic type of data, (2) the donor and destination parts contain different types of data., but the destination part can accommodate the category of data being received :from the donor part, and (3) the donor and destination parts are o~f different data types, and the destination part is not adapted to accommodate the category of material received from the donor part. Each of these three cases is discussed below, with reference to illustrated examples.
In the first scenario, the donor and destina?tion parts are the same type of part. For example, their intrinsic content may be text characters, and they may have been created with the same word processor. Figure 9A illustrates an example in which two text documents 90 and 92 are respectively located wit3ain two windows on a desktop. A portion 94 of the text in the left-hand document 90 has been selected, as indicated by the reverse 'video display.
This selected text is to be copied into the document 92 in the right-hand window. Thus, the user depresses the mouse button, along with a function key, if required, and drags the copy 95 of the selected text 94 into the right-hand document 92, as shown in Figure 9B. Once the cursor enters the second document 92, this document becomes highlighted. Since the Ieft-hand document 90 is now inactive, it is placed behind the document 92, and its title bar is no longer highlighted.
Once the copied text is in the desired location, the user releases the mouse button to place the text in the document, as illustrated in Figure 9C. Since the text that was copied from the IO first document to the second document is of the sanne type as the intrinsic contents of the second document, that text is incorporated into the contents of the second document. In other words, once the selected text is placed in the second document, it n~~ longer has a separate identity of its own. Father, it becomes a .part of the IS intrinsic contents of the destination document. Thos, it can be edited with the same editor that is used for the original contents of the destination document.
In the second scenario described above, the intrinsic contents of the donor and destination parts are of different types, 20 but the destination part can handle the category of information contained in the donor part. lEleferring to Figure LOA, selected text 96 from a document 90 created with a full capability text editor is to be copied into a dialog box 98. In this example, the selected text 96 is stylized, i.e. it has a bold attribute and is in an italicized 25 font. The dialog box, however, is only adapted to handle plain text.
.. __. _ N.;_~ ~ u. ~ .~ .~..:~~..~*..,~r. ~. . ~hA. . _. W ~_.._ __ __._ In a manner similar to the previous example:, a copy of the selected text 96 is dragged from the text document to the dialog box. As the cursor enters the dialog box 98, it moves to the forefront of the display, as illustrated in Figure lOIB. Once the copy of the selected text is located in the appropriate portion of the dialog box, the user releases the mouse button to "drop" it in place.
At this point, the system determines that the dropP~ed text is not of the same type as the intrinsic contents of the dialo;~ box, i.e. it is stylized rather than plain. However, it belongs to a compatible category, in that it is text. Accordingly, the system converts the stylized text into plain text by removing its style properties, and then incorporates it into the dialog box as part of its intrinsic contents, as shown in Figure lOC. Note also that the copied text, which was wrapped over two lines in the original document 90, has been rewrapped onto a single Iine in a dialog box 98 as part of the conversion process, for example by removing a Carriage return in the copy of the selected text.
In the third scenario described above, the intrinsic contents of the donor document are of a different type andl an incompatible category, relative to those of the destination document. Referring to Figure 11A, selected text 94 from a text document 90 is to be copied into a graphic document 100. Again, the user drags a copy 95 of the selected text from the text document 9C1 to the graphic document 100, which is highlighted as soon as tree cursor enters the destination document (Figure 11B). Once the copy 95 of the text is in the appropriate location, the user releases the mouse button to drop the text into the graphic document. In this case, the category of the selected material, i.e. text, is not compatible with the intrinsic contents of the destination document, namely graphic elements. Therefore, the system creates a new part of the same type (text) as the donor document, inserts the copy of the ~1~~
material in it, and embeds the new part in the destination document. With this approach, the new part retains its own identity within the destination document. It has its own intrinsic content, i.e. text characters, and a text editor associated with it.
Since the new material is a separate part, it resides within the graphic document as a frame 102, as illustrated in k~igure 11C.
The frame 102 remains as the selected object, as illustrated by the border around the frame.
As the copy of the selected text was being moved fmm the text document to the graphics document, the menu bar 104 of the desktop retained the commands associated with a text document, as illustrated in Figures 11A and 11B. However, once the copied text was dropped into the graphics document, the commands available on the menu bar 104 switched from those appropriate for a text editor to those for a graphics editor, as illustrated iin Figure I1C.
In addition, a tool bar 106 associated with the graphics editor appeared on the desktop, since the graphics document is naw the active document in which the sel~ted part is located. If, however, the user positions the cursor within the frame 102 and selects a character or a word, the tool bar 106 is removed and the commands in the menu bas 104 return to those appropriate fair editing text.
In summary, when contents are moved from one document to another, the system determines whether the contents can be incorporated into the destination document as part of its intrinsic contents, or must be embedded in the destination document as an autonomous part. The operation of the system to carry out this process is illustrated in the flow chart of Figure 12. After the user 5 has selected an object and dragged it to its intended destination, the system determines the type of part from which the selected objcct originated (Step 1201). For example, the selected object might be plain text, styled text, a graphic element, or the li~:e. ~ne of the pmperties that is stored with each part identifies which type of part IO it is, and each object which is moved or copied frc>m the part is also identified as being of that type. After the part's type is identified, a determination is then made whether this type is the same as that of the destination document (Step 1202). If it is, the selected contents are incorporated or inserted into the destination 15 document, to become part of its intrinsic content (Step 1203). If the selected information is not of the same type a.~ the destination document, the system determines whether they be ong to the same category (Step 1204). For example, if the selected information is styled tent and the destination is plain text, they belong to the same 20 category, although they are different types. If the selected information belongs to the saume category as the destination document, the system convea~tts the selected information into the same type as the destination document (Step 1205). For example, tezt that is generated by one word processor can be converted into 25 the format employed by a different word processt'r, or graphic documents can be converted from one file format: to another. C>nce the conversion is complete, the selected information is incorporated into the contents of the destination document (Step 1203). If the selected information is not of the same category as the destination document, x new part is created with the selected information as its contents (Step 1206), and this new part is embedded in the 5 destination document as a frame (Step 120.
Because of their autonomous nature, parts greatly facilitate the exchange of information between various users of a system. A
document, i.e. a root part, can contain several other parts. A user can retrieve the entire document to view it and edit the individual 10 parts as desired. Alternatively, the respective parts can be individually placed on a desktop that is shared by many users, while access to the document as a whole is protected.
Further along these lines, the individual pares can be protected with limited access rights. For example, .a document 15 containing several different parts may be intended fi5r distribution to a number of different categories of people. Some of the parts may contain confidential information which is not to be revealed to some of the categories of people. In the context of the present invention, these individual parts can be protected with limited 20 access rights, and their contents can be viewed only by those who have an appropriate password, or other form of access. In other words, access restrictability is another characteristic: property of a part.
Referring to Figure I3, a document I08 is illustrated which 25 is comprised of a number of parts, each of which is contained within a frame designated by dashed lines. (?ne of the parts I10 is access protected. For example, this part may conta~,an confidential spreadsheet data. Since the access to this part is limited, its contents are not displayed within the frame. Rather, the frame is opaque, and can include an appropriate symbol, such as a key l I2, to indicatc the fact that a password or the like is required to gain access to it.
The document 108 can be opened by users of the system, to view its general layout and the information in non-confidential parts. However, the data contained within the confidential part I10 can only be seen by those who have an appropriate password for accessing the part- When such a person selects the part, its opaque cover is removed and its contents are displayed.
Although its contents cannot be accessed, the other properties of a restricted part are still available to the users of the system who do not have the password. For example, the location of the part within the document can be changed, and it can be resized by selecting it and moving the resize handles on the border of its frame. Likewise, the limited access part can be copied for placement elsewhere. However, the access protection remains with the copy of the part wherever it is located.
A part with limited access rights can have other parts embedded in it. Consequently, if a user does not possess the appropriate password to.~btain access to the contents of a part, the other parts contained within it will also be restricted from view.
Once a password has 'been accepted, the password can be automatically passed to all of the embedded parts so that they are unlocked arid view able without further user involvement. Of course, however, the embedded parts could have different passwords associated with them, so that even if a viewer can obtain access to one restricted part, access to other parts within it may be further limited.
Thus, the present invention can limit access to the contents of a document on an infra=document level. The access rights are part-specific, and effiective regardless of where the data in the part is located. This approach permits non-privileged parts of a document to be accessed by users without imposing; document-wide access restrictions. In large dacuments having multiple authors, the ability to limit access on a part-by-part basis providles much more flexibility in the viewing of the document and collaboration by the authors.
To facilitate the manipulation of parts and compilation of compound documents, a special type of part, called a "parts bin", IS can be provided. An example of a parts bin is shown in Figure I4.
Referring thereto, a document is open in a window 120, and a parts bin I22 for the document appears in a separate area that is referred to as a windoid. The windoid is associated with the window 120, so that if the document window is brought to the front of the screen, the windoid comes along with it.
The parts bin I22 is essentially a user-definexi palette which functions as a source for other Paris. It contains only other parts and has no intrinsic contents. The parts inside a pa~ris bin have the property of being "frozen". More particularly, wh:rn a user normally depresses the mouse button while the cursor is located within a frame, the smallest element at the location of the cursor is selected. Referring to Figure 14, in the normal case, if the cursor is placed on one cell of a spreadsheet frame 124 and the cursor control button is depressed, the contents of that cell would be selected. The user may have intended to select a diifferent objet, e.g. the entire spreadsheet which contains the call that was actually selected. To do so, the user must move the cursor to another area within the frame, but outside of other parts embedded within the desired part. This requirement is eliminated by fra~zing parts in a parts bin, however. In this case, the user can click the mouse with the cursor located anywhere on a part and select the entire part.
Thus, parts located within parts bins are dealt with as a whole, even if they are represented as frames which contain other embedded parts.
The parts in a parts bin can also be locked so that their contents cannot be modified. The locked property of the parts bin can be denoted by a padlock symbol 126 in the windoid's header.
If a user attempts to drag a part out of a locked parts bin, a copy of the part is automatically made, and the original part: remains in the parts bin, thereby ensuring the integrity of the bin's; contents. To enable users to customize their parts bins, the lock on a parts bin can be turned off, for example by clicking on the padlock symbol 126, and the parts modified.
In a further embodiment of this aspect of the invention, every document can have an associated parts bin. ~~ "document parts bin" enables a standard set of parts to be bundled with a document, so that users of that document will have those parts readily available to them. By activating a command, the user can cause the document parts bin to be displayed in a windoid or a drawer. A document with an associated parts bin Gan have a restrictive property placed on it, so that the only parts which can be placed in the document are those within the bin. A~~ attempt to place other parts into the restricted document is rejected. This 5 property ensures That users of the document work only with a certain set of parts, thereby reducing the training, support and maintenance associated with the document. For example, the parts bin for a data entry form will have only the parts nwcxssary to enter data in the form. As a result, untrained users will not be exposed 10 to unexpected parts, which could otherwise be used improperly.
The restriction on parts also allows an organization to control the contents of documents.
In addition to the user benefits described above, the architecture of the present invention provides benefiits to the 15 software developer. For example, a spreadsheet program is primarily designed to process numbers, and a developer who has expertise in this area can most efficiently spend time on this function. However, a text editor is a necessary part of a spreadsheet application, since the users must be able to edit the text 20 in formulas and values. Having to write a text editor is a distraction for the developer of spreadsheet programs, although necessary with the conventional programs. However, in an arrangement in which a user-selected text editor is ;available for use as a separate part in a spreadsheet; it is no longer wet;essary for the 25 spreadsheet developer to write a text editor. The developer's time can be devoted to the essence of the spreadsheet program, enabling extra features to be implemented, which may not be otherwise available because of time previously devoted to developing other features that were incidental to the number processing function of the spreadsheet.
In addition, when the softvvare developer decides to subsequently add new features or correct existing bugs, the task is made much easier. In particular, the developer need only concentrate on the spreadsheet aspects of the program, and need not be concerned with incidental components, such as a text editor.
It will be appreciated that the present invention is not limited to the specific embodiments which have been described herein to facilitate an understanding of its underlying principles.
For example, text and graphics have been used in the illustrative examples to represent the components of a document. The contents of a part are not limited to these two specific categories, however.
Rather, the contents of a part could be any type of media, including video, sound, animation, and the like. The scope of the invention, therefore, is defined by the claims which are appended hereto, rather than the foregoing description, and all equivalents which are consistent with the meaning of the claims are intended to be embraced therein.

Claims

What is claimed is:

1. A computer-readable storage device containing a computer-human interface which performs the following steps:
storing each component of information that is created by a user as a part which contains intrinsic contents and associated properties;
detecting selection of a part by a user;
determining an editor to be associated with the selected part and automatically launching the determined editor so that its functionality is available to the user whenever the part is selected;
detecting that a user has placed a first pan within the contents of a second part; and creating a data structure which functions as an interface between said first and second parts and defines a geometric area that is shared by said first and second parts and in which the contents of said first part can be edited by its associated editor.