Everyday Information Management
Bob Weiner
Brown University
E-mail: <hyperbole@hub.ucsb.edu> (This is a mail list.)
Edition 3.19.07
August 29, 1995
Copyright © 1991-1995 Free Software Foundation, Inc.
All trademarks referenced herein are trademarks of their respective holders.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This edition of the Hyperbole User Manual is for use with any version 3.19.07 or greater of Hyperbole. Hyperbole is available for free use, distribution, and modification under the terms of version 2 or later of the GNU Public License (GPL). No representations are made about the suitability of this software for any purpose. It is provided "as is" without express or implied warranty.
Hyperbole was designed and written by Bob Weiner. Motorola, Inc. funded and donated this work for free redistribution as part of the InfoDock integrated productivity toolset. Hyperbole and InfoDock can be obtained via anonymous ftp on the Internet from: ‘/anonymous@cs.uiuc.edu:/pub/xemacs/infodock’.
This manual explains user operation and summarizes basic developer facilities of Hyperbole. This major release of Hyperbole concentrates on providing convenient access to information and control over its display. The next major release, version 4, will emphasize flexible views and structure manipulation within bodies of information. (Structure manipulation (outlining) is included within this release but a number of useful functions and additional documentation remain to be written.)
Hyperbole is an open, efficient, programmable information management and hypertext system. It is intended for everyday work on any UNIX platform supported by GNU Emacs. It works well with the versions of Emacs that support multiple X or NEXTSTEP windows: Emacs 19, XEmacs (formerly called Lucid Emacs) and Epoch. Hyperbole allows hypertext buttons to be embedded within unstructured and structured files, mail messages and news articles. It offers intuitive mouse-based control of information display within multiple windows. It also provides point-and-click access to Info manuals, ftp archives, Wide-Area Information Servers (WAIS), and the World-Wide Web (WWW) hypertext system through encapsulations of software that support these protocols.
Hyperbole consists of four parts:
Hyperbole may be used simply for browsing through documents pre-configured with Hyperbole buttons, in which case, one can safely ignore most of the information in this manual. The ‘DEMO’ file included in the Hyperbole distribution demonstrates many of Hyperbole’s standard facilities. It offers a much less technical introduction for Hyperbole users by providing good examples of how buttons may be used and an introduction to the outliner.
So if this manual is too detailed for your taste, you can skip it entirely and just jump right into the demonstration, normally by typing {C-h h d d}, assuming Hyperbole has already been installed at your site. Otherwise, see section Initializing, for Hyperbole installation and initialization information.
Many users, however, will want to do more than browse with Hyperbole, e.g. create their own buttons. The standard Hyperbole button editing user interface is GNU Emacs-based, so a basic familiarity with the Emacs editing model is useful. The material covered in the GNU Emacs tutorial, normally bound to {C-h t} within Emacs, is more than sufficient as background. If some GNU Emacs terms are unfamiliar to you, see Emacs Glossary in the GNU Emacs Manual.
Before we delve into Hyperbole, a number of acknowledgments are in order. Peter Wegner has encouraged the growth in this work. Morris Moore has helped me pursue my own research visions and keeps me striving for excellence. Doug Engelbart has shown me the bigger picture and continues to be an inspiration. His work provides a model from which I am beginning to draw. Kellie Clark and I jointly designed and implemented the Hyperbole outliner.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This chapter describes what Hyperbole is, lists some of its potential applications, and then summarizes the structure of the rest of the manual.
Hyperbole Overview | ||
Manual Overview |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
A Hyperbole user works with buttons; he may create, modify, move or delete buttons. Each button performs a specific action, such as linking to a file or executing a shell command.
There are three categories of Hyperbole buttons:
created by Hyperbole, accessible from within a single document;
created by Hyperbole, accessible anywhere within a user’s network of documents;
created and managed by other programs or embedded within the structure of a document, accessible from within a single document. Hyperbole recognizes implicit buttons by contextual patterns given in their type specifications (explained later).
Explicit Hyperbole buttons may be embedded within any type of text file. Implicit buttons may be recognized anywhere within a text file, depending on the implicit button types that are available. All global buttons are stored in a single location and activated by entering their names, rather than by direct selection, the means used to activate explicit and implicit buttons.
To summarize:
Button Category Active Within Activation Means Managed By ======================================================================== Explicit a single document direct selection Hyperbole Global any document specifying its name Hyperbole Implicit a matching context direct selection other tools ========================================================================
Hyperbole buttons may be clicked upon with a mouse to activate them or to describe their actions. Thus, a user can always check how a button will act before activating it. Buttons may also be activated from a keyboard. (In fact, virtually all Hyperbole operations, including menu usage, may be performed from any standard character terminal interface, so one need not be anchored to a workstation all day). See section Smart Keys.
Hyperbole does not enforce any particular hypertext or information management model, but instead allows you to organize your information in large or small chunks as you see fit. The Hyperbole outliner organizes information hierarchies which may also contain links to external information sources.
Some of Hyperbole’s most important features include:
Typical Hyperbole applications include:
Overlapping link paths provide a variety of views into an information space. A search facility locates buttons in context and permits quick selection.
Cross-references may be embedded within documentation. One can add a point-and-click interface to existing documentation, link code with associated design documents, or jump to the definition of an identifier by selecting its name within code or documentation.
The Hyperbole outliner, See section Outliner Operation, is an effective tool for capturing ideas and then quickly reorganizing them in a meaningful way. Links to related ideas are easy to create, eliminating the need to copy and paste information into a single place.
Tutorials containing buttons can show students how things work while explaining the concepts, e.g. an introduction to local commands. This technique can be much more effective than written documentation alone.
Programs that manage archives from incoming information streams may be supplemented by having them add topic-based buttons that link to the archive holdings. Users can then search and create their own links to archive entries.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Glossary, provides definitions of Hyperbole terms for quick reference, so in some cases terms are not precisely defined within the text. Be sure to reference the glossary if a term is unclear to you. Although you need not have a keen understanding of all of these terms, a quick scan of the Glossary should help throughout Hyperbole use.
Button Concepts, and Hyperbole Views, lay a conceptual foundation needed before the material in Using Hyperbole, can be utilized. See section Smart Key Reference for reference documentation on the innovative, context-sensitive mouse and key handling offered by Hyperbole. See section Questions and Answers for responses to a number of frequent questions.
Developers comfortable with Emacs Lisp will want to continue on through Developing with Hyperbole.
Ongoing Work describes future directions in Hyperbole’s evolution.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This chapter explains the user-level notion of Hyperbole buttons. You should read it before reading Using Hyperbole.
Hyperbole buttons that are stored in files persist across Emacs sessions, so they provide a convenient means of linking from one information source to another.
2.1 Explicit Buttons | ||
2.2 Global Buttons | ||
2.3 Implicit Buttons and Types | ||
2.4 Action Types and Actions | ||
2.5 Button and Type Precedences |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole creates and manages explicit buttons which look like
this <(fake button)>
to a Hyperbole user. They are quickly
recognizable, yet relatively non-distracting as one scans the text in
which they are embedded. The text between the <(
and
)>
delimiters is called the button label. Spacing between
words within a button label is irrelevant to Hyperbole, so button labels
may wrap across several lines without causing a problem.
Hyperbole stores the button data that gives an explicit button its behavior, separately from the button label, in a file named ‘.hypb’ within the same directory as the file in which the button is created. Thus, all files in the same directory share a common button data file. Button data is comprised of individual button attribute values. A user never sees this data in its raw form but may see a formatted version by asking for help on a button.
Explicit buttons may be freely moved about within the buffer in which they are created. (No present support exists for moving buttons between buffers). A single button may also appear multiple times within the same buffer; one simply copies the button label with its delimiters to a new location in such cases.
Each explicit button is assigned an action type which determines the actions that it performs. Link action types connect buttons to particular types of referents. Activation of such buttons then displays the referents.
Hyperbole does not manage referent data; this is left to the applications that generate the data. This means that Hyperbole provides in-place linking and does not require reformatting of data to integrate it with a Hyperbole framework.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Access to explicit buttons depends upon the information on your screen since they are embedded within particular buffers. Sometimes it is useful to activate buttons without regard to the information with which you are presently working. In such instances, you use global buttons, which are simply explicit buttons which may be activated or otherwise operated upon by entering their labels when they are prompted for, rather than selecting the buttons within a buffer.
If you want a permanent link to a file section that you can follow at
any time, you can use a global button. Or what about an Emacs keyboard
macro that you use frequently? Create an exec-kbd-macro
button
with an easy to type name and then you can easily activate it whenever
the need arises.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Implicit buttons are those defined by the natural structure of a document. They are identified by contextual patterns which limit the locations or states in which they can appear. Their behavior is determined by one or more actions which they trigger when activated. An action is derived from either a Hyperbole action type specification, see section Action Types and Actions, or an Emacs Lisp function. Implicit button types may use the same action types that explicit buttons do.
Implicit buttons never have any button data associated with them. They are recognized in context based on predicate matches defined within implicit button types. For example, Hyperbole recognizes file names enclosed in double quotes and can quickly display their associated files in response to simple mouse clicks.
See ‘hibtypes.el’ for complete examples. Standard implicit button types include (in alphabetical order):
annot-bib
Displays annotated bibliography entries referenced internally, delimeters = []. References must be delimited by square brackets, must begin with a word constituent character, and must not be in buffers whose names begin with a ’ ’ or ’*’ character.
completion
Inserts completion at point into minibuffer or other window.
dir-summary
Detects filename buttons in files named "MANIFEST" or "DIR". Displays selected files. Each file name must be at the beginning of the line and must be followed by one or more spaces and then another non-space, non-parenthesis, non-brace character.
doc-id
Displays an index entry for a site-specific document given its id. Ids must be delimited by ’doc-id-start’ and ’doc-id-end’ and must match the function given by ’doc-id-p’. This permits creation of catalogued online libraries. See ‘${hyperb:dir}/hib-doc-id.el’ for more information.
elisp-compiler-msg
Jumps to source code for definition associated with byte-compiler error message. Works when activated anywhere within an error line.
debugger-source
Jumps to source line associated with debugger stack frame or breakpoint lines. This works with gdb, dbx, and xdb. Such lines are recognized in any buffer.
grep-msg
Jumps to line associated with grep or compilation error msgs. Messages are recognized in any buffer.
hyp-address
Turns a Hyperbole e-mail list address into an implicit button which
inserts Hyperbole environment information.
Useful when sending mail to a Hyperbole mail list. See also the
documentation for actypes::hyp-config
.
hyp-source
Turns source location entries in Hyperbole reports into buttons that jump to the associated location.
Info-node
Makes "(file)node" buttons display the associated Info node.
kbd-key
Executes a key sequence delimited by curly braces. Key sequences should be in human readable form, e.g. {C-b}. Forms such as {}, {}, and {^b} will not be recognized.
klink
Follows a link delimited by <> to a koutline cell.
See documentation for actypes::link-to-kotl
for valid link
specifiers.
mail-address
If on an e-mail address in a specific buffer type, mail to that address in another window. Applies to the rolodex match buffer, any buffer attached to a file in rolo-file-list, or any buffer with ‘mail’ or ‘rolo’ (case-insensitive) within its name.
man-apropos
UNIX manual
man pages
man apropos
Makes man apropos entries display associated man pages when selected.
patch-msg
Jumps to source code associated with output from the patch
program. Patch applies diffs to source code.
pathname
Makes a delimited, valid pathname display the path entry.
Also works for delimited and non-delimited ange-ftp and efs pathnames.
See hpath:at-p
for possible delimiters.
See hpath:find
for special file display options.
rfc
Retrieves and displays an Internet rfc referenced at point. Requires ange-ftp or efs when needed for remote retrievals. The following formats are recognized: RFC822, rfc-822, and RFC 822. The hpath:rfc variable specifies the location from which to retrieve RFCs."
rfc-toc
Summarizes contents of an Internet rfc from anywhere within rfc buffer. Each line in summary may be selected to jump to section.
The Hyperbole Smart Keys offer extensive additional context-sensitive point-and-click type behavior beyond these standard implicit button types. See section Smart Keys.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Action types provide action procedures that specify button behavior. The arguments needed by an action type are prompted for at button creation time. When a button is activated, the stored arguments are fed to the action type’s action body to achieve the desired result. Hyperbole handles all of this transparently.
Standard action types include:
annot-bib
Follows internal ref KEY within an annotated bibliography, delimiters=[].
completion
Inserts completion at point into minibuffer or other window. Unless at end of buffer or if completion has already been inserted, then deletes completions window.
eval-elisp
Evaluates a Lisp expression LISP-EXPR.
exec-kbd-macro
Executes KBD-MACRO REPEAT-COUNT times. KBD-MACRO may be a string of editor command characters or a function symbol. Optional REPEAT-COUNT nil means execute once, zero means repeat until error.
exec-shell-cmd
Executes a SHELL-CMD string asynchronously. Optional non-nil second argument INTERNAL-CMD means do not display the shell command line executed. Optional non-nil third argument KILL-PREV means kill last output to shell buffer before executing SHELL-CMD.
exec-window-cmd
Executes an external window-based SHELL-CMD string asynchronously.
hyp-config
Inserts Hyperbole configuration info at end of optional OUT-BUF or current.
hyp-request
Inserts Hyperbole mail list request help into optional OUT-BUF or current.
hyp-source
Displays a buffer or file from a line beginning with ’hbut:source-prefix’.
kbd-key
Executes the function binding for KEY-SEQUENCE, delimited by {}. Returns t if a KEY-SEQUENCE has a binding, else nil.
link-to-buffer-tmp
Displays a BUFFER in another window. Link is generally only good for current Emacs session. Use ’link-to-file’ instead for a permanent link.
link-to-directory
Displays a DIRECTORY in Dired mode in another window.
link-to-doc
Displays online version of a document given by DOC-ID, in other window. If the online version of a document is not found in doc-id-indices, an error is signalled.
link-to-ebut
Performs action given by another button, specified by KEY and KEY-FILE.
link-to-elisp-doc
Displays documentation for FUNC-SYMBOL.
link-to-file
Displays a PATH in another window scrolled to optional POINT. With POINT, buffer is displayed with POINT at the top of the window.
link-to-file-line
Displays a PATH in another window scrolled to LINE-NUM.
link-to-kcell
Displays FILE with kcell given by CELL-REF at the top of the window. CELL-REF may be a kcell’s display label or its permanant idstamp. If FILE is nil, the current buffer is used. If CELL-REF is nil, the first cell in the view is shown.
link-to-kotl
Displays at the top of another window the referent pointed to by LINK. LINK may be of any of the following forms, with or without delimiters:
< pathname [, cell-ref] > < [-!&] pathname > < @cell-ref >
link-to-Info-node
Displays an Info NODE in another window. NODE must be a string of the form ’(file)nodename’.
link-to-mail
Displays mail msg with MAIL-MSG-ID from MAIL-FILE in other window. See documentation for the variable hmail:init-function for information on how to specify a mail reader to use.
link-to-regexp-match
Finds REGEXP’s Nth occurrence in FILE and displays location at window top. Returns t if found, signals an error if not.
link-to-rfc
Retrieves and displays an Internet rfc given by RFC-NUM. RFC-NUM may be a string or an integer. Requires ange-ftp or efs for remote retrievals.
link-to-string-match
Finds STRING’s Nth occurrence in FILE and displays location at window top. Returns t if found, nil if not.
man-show
Displays man page on TOPIC, which may be of the form <command>(<section>
).
rfc-toc
Computes and displays summary of an Internet rfc in BUF-NAME. Assumes point has already been moved to start of region to summarize. Optional OPOINT is point to return to in BUF-NAME after displaying summary.
The use of action types provides a convenient way of specifying button behavior without the need to know how to program. Expert users who are familiar with Emacs Lisp, however, may find that they often want to tailor button actions in a variety of ways not easily captured within a type system. In such cases, hui:ebut-prompt-for-action should be set non-nil. This will cause Hyperbole to prompt for an action to override the button’s action type at each explicit button creation. For those cases where the action type is sufficient, a nil value should be entered for the action. An action may be any Lisp form that may be evaluated.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Explicit buttons always take precedence over implicit buttons. Thus, if a button selection is made which falls within both an explicit and implicit button, only the explicit button will be selected. Explicit button labels are not allowed to overlap; Hyperbole’s behavior in such cases is undefined.
If there is no explicit button at point during a selection request, then each implicit button type predicate is tested in turn until one returns non-nil or all are exhausted. Since two implicit button types may have overlapping domains (those contexts in which their predicates are true), only the first matching type is used. The type predicates are tested in reverse order of definition, i.e. most recently entered types are tested first, so that personal types defined after standard system types take precedence. It is important to keep this order in mind when defining new implicit button types. By making their match predicates as specific as possible, one can minimize any overlapping of implicit button type domains.
Once a type name is defined, its precedence relative to other types remains the same even if you redefine the body of the type, as long as you don’t change its name. This allows incremental modifications to types without having to worry about shifts in type precedence. See section Creating Types, for information on how to develop or modify types.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This chapter explains the role of views in the display of information accessed via Hyperbole. You should read it before reading Using Hyperbole.
This release of Hyperbole does not contain any support for views. A general view model is in development. The descriptions here are meant to familiarize you with the notion of views to be available in future releases.
A Hyperbole view is a perspective on a pool of information. An outline of a document, a document extract, and a decrypted document all represent views of an original document (which may just be a file).
View specifications or view specs, for short, control the application of views to displayed information. An action type which displays a buffer of text, the result buffer, can apply a view spec to the buffer before display. A user can then manipulate the view by interactively altering the view spec associated with the buffer. A ’V’ in an action type’s interactive call string indicates that the view spec of any button which uses the action type should be applied to the result buffer of the action type.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This chapter covers user interaction with Hyperbole. Often a number of overlapping interaction methods are provided to support different work styles and hardware limitations. You need learn only one with which you can become comfortable, in such instances.
Remember that the ‘DEMO’ file included in the Hyperbole distribution demonstrates many of Hyperbole’s standard facilities, See section Preface, for more details.
4.1 Initializing | ||
4.2 Smart Keys | ||
4.3 Window Configurations | ||
4.4 Operating Menus | ||
4.5 Entering Arguments | ||
4.6 Working with Explicit Buttons | ||
4.7 Outliner Operation | ||
4.8 Rolodex Operation |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole must be installed at your site before you can use it. The ‘README’ file that comes with the Hyperbole distribution explains how to install it and how to join the Hyperbole interest mail lists. See the Msg/ menu item in Operating Menus, for a convenient means of joining and mailing to these lists.
If you are installing Hyperbole at your site, please read the ‘README’ file before continuing in this document.
If you want to customize the basic Hyperbole initialization sequence for yourself rather than the users at your site, you should make a personal copy of the ‘hsite.el’ file, modify it the way you want, and then load it. (If you are familiar with Emacs Lisp, see also Hook Variables.)
4.1.1 Internal Viewers | ||
4.1.2 External Viewers | ||
4.1.3 Link Variable Substitution | ||
4.1.4 Configuring Button Colors | ||
4.1.5 Loading Hyperbole |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
When given a file name, Hyperbole will by default display the file for editing within an Emacs buffer. The hpath:display-alist variable can be used to specify file name patterns, such as matching suffixes, which will invoke a special Emacs Lisp function to display any matching files within Emacs. This can be used to format raw data files for convenient display.
Configure the hpath:display-alist variable in ‘hsite.el’. Its value is an association list whose elements are (<file-name-regular-expression> . <function-of-one-arg>) pairs. Any path whose name matches a <file-name-regular-expression> will be displayed by calling the associated <function-of-one-arg> with the file name as the argument.
See section External Viewers, for instructions on associating file names with external, window-system specific viewers.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
If you will be using Hyperbole under a window system, you may want to configure the hpath:find-alist variable in ‘hsite.el’ to support hyperlinks which open files using non-Emacs tools, e.g. a fax reader or a bitmap viewer.
The value of hpath:find-alist is determined when Hyperbole is
initialized based upon the current window system and the version of
Emacs in use. The value is an association list whose elements are
(<file-name-regular-expression> . <viewer-program>) pairs. Any path
whose name matches a <file-name-regular-expression> will be
displayed using the corresponding viewer-program. If a <viewer-program>
entry contains a %s
string, the filename to display will be
substituted at that point within the string. Otherwise, the filename
will be appended to the <viewer-program> entry. See the "x-suffixes"
and "nextstep-suffixes" settings within the definition of
hpath:find-alist as examples.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Another variable to consider modifying in the ‘hsite.el’ file is hpath:variables. This variable consists of a list of Emacs Lisp variable names, each of which may have a pathname or a list of pathnames as a value. Whenever a Hyperbole file or directory link button is created, its pathname is compared against the values in hpath:variables. The first match found, if any, is selected and its associated variable name is substituted into the link pathname, in place of its literal value. When the link is resolved (the button is activated), Hyperbole replaces each variable with the first matching value from this list. (Environment variables are also replaced whenever link paths are resolved.
This permits sharing of links over wide areas, where the variable values may differ between link creator and link activator. The entire process is wholly transparent to the user; it is explained here simply to help you in deciding whether or not to modify the value of hpath:variables.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
When Hyperbole is run under a window system together with Emacs 19,
XEmacs or Epoch, it can automatically highlight any explicit buttons in
a buffer and make them flash when selected. The only configuration
necessary is selection of a color (or style) for button highlighting and
button flashing. See the ‘hui-*-b*.el’ files for lists of
potential colors and the code which supports this behavior. A call to
(hproperty:cycle-but-color)
within a Hyperbole initialization
sequence changes the color used to highlight and flash explicit buttons.
Additionally, under XEmacs, if hproperty:but-emphasize-p is set to
t
in ‘hsite.el’, then whenever the mouse pointer moves over
an explicit button, it will be emphasized in a different color or style.
This emphasis is in addition to any non-mouse-sensitive button
highlighting.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Each time you start a new Emacs session, the Hyperbole code must be loaded for you to use it. The easiest way to achieve this is to add two lines to your ‘~/.emacs’ file, which Emacs reads automatically on startup. The lines should be:
(load "<HYP-DIR>/hversion") (load "<HYP-DIR>/hyperbole")
where the local Hyperbole installation directory is substituted for
<HYP-DIR>
. This establishes a few key bindings and sets up
Hyperbole to automatically load whenever you activate its menu. If you
would rather have the whole Hyperbole system loaded when you start up so
that you can always use the Smart Keys and other facilities, add a third
line after the above two lines.
(require 'hsite)
If you use mouse keys, be sure to add the above lines after any setup of mouse key bindings, to ensure that Hyperbole’s mouse keys are initialized. See section Smart Keys, for further details. If you use any Hyperbole mail or news support, see section Buttons in Mail, be certain to perform all of your personal mail/news initializations before the point at which you load Hyperbole. Otherwise, the mail/news support may not be configured properly. For example, if you use the Emacs add-on Supercite package, its setup should come before Hyperbole initialization.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole provides two special Smart Keys that perform context-sensitive operations, the Action Key and the Assist Key. By default, the Action Key is bound to your shift-middle mouse button (or shift-left on a 2-button mouse) and the Assist Key is bound to your shift-right mouse button, assuming Hyperbole is run under an external window system. (InfoDock users should use the middle mouse button as the Action Key, instead.)
Mouse configuration is automatic for XEmacs and Epoch under the X window system and for GNU Emacs versions 18 and 19 under X, OpenWindows, NEXTSTEP, SunView and Apollo’s Display Manager, assuming your Emacs program has been built with support for any of these window systems.
By default (if hkey-init is left with a setting of t
in ‘hsite.el’), then {M-<RET>} may also be used as the
Action Key and {C-u M-<RET>} may be used as the
Assist Key. In many read-only modes like Dired and Rmail,
{<RTN>} also functions as the Action Key. These key bindings allow
context sensitive operation from any keyboard.
If you prefer other key bindings, simply bind the commands
action-key
and assist-key
to keyboard keys.
hkey-either
may be used instead if you prefer a single
key binding for both commands; a prefix argument then invokes
assist-key
.
You may also bind action-mouse-key
and assist-mouse-key
to mouse keys, as you like.
The Action Key generally selects entities, creates links and activates buttons. The Assist Key generally provides help, such as reporting on a button’s attributes, or serves a complementary function to whatever the Action Key does within a context.
You can get a summary of what the Smart Keys do in all of their different contexts by pressing the Assist Key in the right corner (within the rightmost 3 characters) of a window modeline or by using the Hyperbole Doc/SmartKy menu entry.
The following table is an example of this summary. Much of the browsing power of Hyperbole comes from use of the Smart Keys, so you should spend some time practicing how to use them. This table may appear daunting at first, but as you practice and notice that the Smart Keys do just a few context-sensitive things per editor mode, you will find it easy to just point and click and let Hyperbole do the rest.
For extensive reference documentation on the Smart Keys, see section Smart Key Reference.
Note how the last line in the table explains the default behavior of the Smart Keys. That is what they do when they cannot find a context match at your current location. See the documentation for the variables action-key-default-function and assist-key-default-function for information on how to customize the behavior of the Smart Keys within default contexts.
A prime design criterion of Hyperbole’s user interface is that one should be able to see what an operation will do before using it. The Assist Key provides help for button and menu item actions. Hyperbole also shows the result of directly selecting an argument value, to provide feedback as to whether the right item has been selected.
When you use a mouse and you want to find out what either of the Smart Keys does within a context, depress the one you want to check on and hold it down, then press the other and release as you please. A help buffer will pop up explaining the actions that will be performed in that context, if any. A press of either Smart Key at the end of that help buffer will restore your display to its configuration prior to invoking help.
By default (if hkey-init is left set equal to t
in
‘hsite.el’), then {C-h A} will display this same
context-sensitive help for the Action Key while {C-u C-h
A} will display the help for the Assist Key. Note that
{C-h a} will perform a function unrelated to Hyperbole, so you
must press the shift key when you hit the A character.
When Hyperbole is installed, a key may be bound which allows you
to switch between the Smart Key mouse bindings and your prior ones.
C-h w hmouse-toggle-bindings <RTN> should show you any key
which performs this command. If no key binding has been established or
if you prefer one of your own, simply select a key and bind it
within your ‘~/.emacs’ file. For example, (global-set-key
"\C-ct" 'hmouse-toggle-bindings)
.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole includes the ‘wconfig.el’ package which lets you save and restore window configurations, i.e. the window layout and buffers displayed within an Emacs frame. This is useful to save a particular working context and then to jump back to it at a later time during an Emacs session. It is also useful during demonstrations to pull up many informational artifacts all at once, e.g. all of the windows for a particular subsystem. None of this information is stored between Emacs sessions, so your window configurations will last only through a single session of use.
The wconfig library provides two distinct means of managing window configurations. The first means associates a name with each stored window configuration. The name can then be used to retrieve the window configuration later. The second means uses a ring structure to save window configurations and then allows browsing through the sequence of saved configurations.
The Win/ menu entry on the Hyperbole top-level menu displays a menu of window configuration commands:
WinConfig> AddName DeleteName RestoreName PopRing SaveRing YankRing
Menu Item Command Description ==================================================================== AddName wconfig-add-by-name Name current wconfig DeleteName wconfig-delete-by-name Delete wconfig with name RestoreName wconfig-restore-by-name Restore wconfig by name PopRing wconfig-delete-pop Restore and delete wconfig SaveRing wconfig-ring-save Store wconfig to ring YankRing wconfig-yank-pop Restore next wconfig ====================================================================
Saving and restoring window configurations by name is the easiest method, but it requires that you input the chosen name from the keyboard. The ring commands permit saving and restoring through mouse interaction only, if so desired. The prior section, see section Smart Keys, mentions how to save and restore window configurations with the Smart Keys. Since the ring commands are a bit more complex than their by-name counterparts, the following paragraphs explain them in more detail.
Wconfig creates a ring structure that operates just like the Emacs kill-ring, see Kill Ring in The GNU Emacs Manual, but its elements are window configurations rather than text regions. One can add an element to the ring based upon the current window configuration. After several elements are in the ring, one can walk through all of them in sequence until the desired configuration is restored.
SaveRing executes the wconfig-ring-save
command which
saves the current window configuration to the ring.
YankRing executes the wconfig-yank-pop
command. It restores the
window configuration from the currently pointed to configuration in the
ring. It does not delete this configuration from the ring but it does
move the pointer to the prior ring element. Repeated calls to this
command thus restore successive window configurations until the ring
pointer wraps around. Simply stop when a desired configuration appears.
PopRing calls the wconfig-delete-pop
command.
It is used to restore a previously saved configuration and at the same
time delete it from the ring.
The maximum number of elements the ring can hold is set by the wconfig-ring-max variable whose default is 10. Any saves beyond this value cause deletion of the oldest element in the ring before a new one is added.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Under InfoDock, XEmacs, and Emacs 19, pulldown and popup menus are available to invoke Hyperbole commands, including those from the rolodex and the outliner. These menus operate like any other X window menus. Use the Quit command on the Hyperbole menubar menu to get rid of the menu if you do not need it. Invoking Hyperbole again will add the menu back to the menubar.
This section discusses only the specialized Hyperbole menus that appear in the minibuffer and that work with all Emacs versions. Hyperbole menu items may be selected from either the keyboard or via mouse clicks. When used with the keyboard, they provide rapid command access similar to key bindings.
The top level menu is invoked from a key given in your ‘hsite.el’ file (by default, {C-h h}) or via an Action Key press in a location with no other action defined. The menu will appear in the minibuffer and should look mostly like so:
Hy> Act Butfile/ Doc/ Ebut/ Gbut/ Hist Ibut/ Msg/ Otl/ Rolo/ Win/
The above menu items can be summarized as follows:
Perform the action associated with any button at point or prompt for the name of an explicit button to activate if point is not on one.
Display a local or global file of buttons, providing easy access. ‘HYPB’ for a local button file and ‘~/.hyperb/HYPB’ for your global file. These are good places to start your button creation testing.
All explicit button commands.
Hyperbole documentation quick access. Contains Types/ submenu for documentation on Hyperbole implicit button and action types.
All global button commands. Global buttons are accessed by name rather than by direct selection.
Jumps back to last position in button traversal history.
All implicit button commands.
Hyperbole-specific mail and news messaging support commands. Use this to send mail to the Hyperbole discussion list or to add/modify your entry on a Hyperbole mail list.
Autonumbered, structured outliner and hyper-node manager commands. See section Outliner Operation.
Hierarchical, multi-file rolodex lookup and edit commands. See section Rolodex Operation.
Window configuration management such as adding and restoring window configurations by name. See section Window Configurations.
All menu items are selected via the first character of their names (letter case does not matter) or via a press of the Action Key. "/" at the end of an item name indicates that it brings up a sub-menu. A press of the Assist Key on an item displays help for the item, including the action that it performs.
While a menu is active, to re-activate the top-level Hyperbole menu, you must use {C-t}. This allows you to browse the submenus and then return to the top. You can quit without selecting an item by using {q}. {C-g} aborts whether you are at a menu prompt or any other Hyperbole prompt.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Many Hyperbole commands prompt you for arguments. The standard
Hyperbole user interface has an extensive core of argument types that it
recognizes. Whenever Hyperbole is prompting you for an argument, it
knows the type that it needs and provides some error checking to help
you get it right. More importantly, it allows you to press the Action
Key within an entity that you want to use as an argument and it will grab the
appropriate thing and show it to you at the input prompt within the
minibuffer. If you press the Action Key again at the same point (click
with a mouse) on the same thing again, it accepts the entity as the
argument and moves on. Thus, a double click registers a desired
argument. Double-quoted strings, pathnames, mail messages, Info nodes,
dired listings, buffers, numbers, completion items and so forth are all
recognized at appropriate times. All of the argument types mentioned in
the documentation for the Emacs Lisp (interactive)
function are
recognized. Experiment a little and you will quickly get used to this
direct selection technique.
Wherever possible, standard Emacs completion is offered, see Completion in the Gnu Emacs Manual. Remember to use {?} to see what your possibilities for an argument are. Once you have a list of possible completions on screen, you can double click the Action Key on any one to enter it as the argument.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Explicit buttons provide the building blocks for creating personal or organizational hypertext networks with Hyperbole. This section summarizes the user-level operations available for managing these buttons.
4.6.1 Creation | ||
4.6.2 Renaming | ||
4.6.3 Deletion | ||
4.6.4 Modification | ||
4.6.5 Location | ||
4.6.6 Button Files | ||
4.6.7 Buttons in Mail | ||
4.6.8 Buttons in News |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
4.6.1.1 via Smart Mouse Key Drags | ||
4.6.1.2 via Hyperbole Menus |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
The most efficient way to create an explicit button interactively is to use the mouse Action Key to drag from a button source window to a window showing its link referent. More specifically, you should split your current Emacs frame into two windows: one which contains the point at which you want a button to be inserted and another which shows the point to which you want to link. Depress the mouse Action Key at the point at which the button should be inserted, drag to the other window and release it at the point of the link referent. The process becomes quite simple with a little practice. (See section via Hyperbole Menus, for a more detailed explanation of the explicit button creation process.)
Hyperbole uses the link referent context to determine the type of link to make. If there are a few different types of links which are applicable from the context, you will be prompted with a list of the types. Simply use the Action Key or the first letter of the link type to select one of the type names and to finish the link creation. Hyperbole will then insert explicit button delimiters around the button label and will display a message in the minibuffer indicating both the button name and its action/link type.
If you run Emacs under a window system, you can emulate an Action Key
drag from the keyboard by: hitting {M-o}, the
hkey-operate
command, at the button source location, moving
to the link destination, e.g. with {C-x o}, and then hitting
{M-o} again. This simulates a depress and then release of the
Action Key. {C-u M-o} emulates drags of the Assist Key.
This will not work when Hyperbole is run from a dumb terminal Emacs
session since drag actions are not supported without a window system.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
You can alternatively use the Hyperbole menus to create explicit buttons. First, mark a short region of text in any fashion allowed by GNU Emacs and then select the Hyperbole menu item sequence, Ebut/Create. You will be prompted for the button’s label with the marked region as the default. If you accept the default and enter the rest of the information you are prompted for, the button will be created within the current buffer and Hyperbole will surround the marked region with explicit button delimiters to indicate success.
If you do not mark a region before invoking the button create command, you will be prompted for both a label and a target buffer for the button and the delimited label text will be inserted into the target buffer after a successful button creation.
After Hyperbole has the button label and its target buffer, it will prompt you for an action type for the button. Use the {?} completion help key to see the available types. The type selected determines any following values for which you will be prompted.
If a previous button with the same label exists in the same buffer, Hyperbole will add an instance number to the label when it adds the delimiters so that the name is unique. Thus, you don’t have to worry about accidental button name conflicts. If you want the same button to appear in multiple places within the buffer, just enter the label again and delimit it yourself. Hyperbole will interpret all occurrences of the same delimited label within a buffer as the same button.
If you create link buttons using the Hyperbole menus, the best technique
is to place on screen both the source buffer for the button and the
buffer to which it will link. Mark the region of text to use for your
button label, invoke the button create command from the menu, choose an
action type which begins with link-to-
and then use the direct
selection techniques mentioned in Entering Arguments, to select
the link referent.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Once an explicit button has been created, its label text must be treated specially. Any inter-word spacing within the label may be freely changed, as may happen when a paragraph is refilled. But a special command must be invoked to rename it.
The rename command operates in two different ways. If point is within a button label when it is invoked, it will tell you to edit the button label and then invoke the rename command again. The second invocation will actually rename the button. If instead the command is originally invoked outside of any explicit button, it will prompt for the button label to replace and the label to replace it with and then will perform the rename. All occurrences of the same button in the buffer will be renamed, so you need locate only one occurrence of the button.
The rename command may be invoked from the Hyperbole menu via
Ebut/Rename. A faster method is to use a key bound to the
hui:ebut-rename
command. Your site installation may include such
a key. C-h w hui:ebut-rename <RTN> should show you any key it
is on. If no key binding has been established or if you prefer one
of your own, simply bind it within your ‘~/.emacs’ file. We
recommend the {C-c C-r} key, as in: (global-set-key
"\C-c\C-r" 'hui:ebut-rename)
.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Ebut/Delete works similarly to the Rename command but deletes the selected button. The button’s delimiters are removed to confirm the delete. If the delete command is invoked with a prefix argument, then both the button label and the delimiters are removed as confirmation.
Presently there is no way to recover a deleted button; it must be recreated. Therefore, the hui:ebut-delete-confirm-p variable is true by default, causing Hyperbole to require confirmation before interactively deleting explicit buttons. Set it to nil if you prefer no confirmation.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Ebut/Modify prompts you with each of the elements from the button’s data list and allows you to modify each in turn.
There is a quicker way to modify explicit link buttons. Simply drag with the mouse Action Key from within the button label to a link destination in a different window, just as you would when creating a new button with a mouse drag. Remember that drags may also be emulated from the keyboard. See section Creation.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
The Ebut/Help menu can be used to summarize a single explicit button or all such buttons within a single buffer. The buttons summarized may then be activated directly from the summary.
Ebut/Help/BufferButs summarizes the explicit buttons in the order in which they appear in the buffer. Ebut/Help/CurrentBut summarizes only the button at point. Ebut/Help/OrderedButs summarizes the buttons in alphabetical order. All of these summary commands eliminate duplicate instances of buttons from their help displays.
Ebut/Search prompts for a search pattern and searches across all the locations in which you have previously created explicit buttons. It asks you whether to match to any part of a button label or only complete labels. It then displays a list of button matches with a single line of surrounding context from their sources. Any button in the match list may be activated as usual. An Action Key press on the surrounding context jumps to the associated source line or a press on the filename preceding the matches jumps to the file without selecting a particular line.
There are presently no user-level facilities for globally locating buttons created by others or for searching on particular button attributes.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
It is often convenient to create lists of buttons that can be used as menus to provide centralized access to distributed information pools or for other purposes. These files can serve as useful roadmaps to help efficiently guide a user through both unfamiliar and highly familiar information spaces. Files that are created specifically for this purpose, we call button files.
The Hyperbole menu system provides quick access to two types of these button files: personal and directory-specific, through the ButFile menu. (The variable, hbmap:filename, contains the base name of these standard button files. Its standard value is ‘HYPB’.)
A personal button file may serve as a user’s own roadmap to frequently used resources. Selection of the ButFile/PersonalFile menu item displays this file for editing. The default personal button file is stored within the directory given by the hbmap:dir-user variable whose standard value is ‘~/.hyperb’. The standard Hyperbole configuration also appends all global buttons to the end of this file, one per line, as they are created. So you can edit or annotate them within the file.
A directory-specific button file may exist for each file system directory. Such files are useful for explaining the contents of directories and pointing readers to particular highlights within the directories. Selection of the ButFile/DirFile menu item displays the button file for the current directory; this provides an easy means of updating this file when working on a file within the same directory. If you want to view some other directory-specific button file, simply use the normal Emacs file finding commands.
One might suggest that menu quick access be provided for group-specific and site-specific button files. Instead, link buttons to such things should be placed at the top of your personal button file. This provides a more flexible means of quick access.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole allows the embedding of buttons within electronic mail
messages that are composed in Emacs with the standard (mail)
command, normally bound to {C-x m} or with other Emacs-based
mail composing functions. An enhanced mail reader can then be used
to activate the buttons within messages just like any other buttons.
Hyperbole automatically supports the Rmail, see Rmail in the GNU Emacs Manual, VM, see Introduction in the VM Manual, and MH-e mail readers. Button inclusion and activation within USENET news articles is also supported in the same fashion via the GNUS news reader, see Introduction in the GNUS Manual, if available at your site. (The ‘hmail.el’ file provides a generalized interface that can be used to hook in other mail or news readers if the necessary interface functions are written.)
All explicit buttons to be mailed must be created within the outgoing
message buffer. There is no present support for including text from
other buffers or files which contain explicit buttons, except for the
ability to yank the contents of a message being replied to, together
with all of its buttons, via the (mail-yank-original)
command
bound to {C-c C-y}. From a user’s perspective, buttons are
created in precisely the same way as in any other buffer. They also
appear just like any other buttons to both the message sender and the
reader who uses the Hyperbole enhanced readers. Button operation may be
tested any time before a message is sent. A person who does not use
Hyperbole enhanced mail readers can still send messages with embedded
buttons since mail composing is independent of any mail reader
choice.
Hyperbole buttons embedded within received mail messages act just like any other buttons. The mail does not contain any of the action type definitions used by the buttons, so the receiver must have these or she will receive an error when she activates the buttons. Buttons which appear in message Subject lines are copied to summary buffers whenever such summaries are generated. Thus, they may be activated from either the message or summary buffers.
Nothing bad will happen if a mail message with explicit buttons is sent to a non-Hyperbole user. The user will simply see the text of the message followed by a series of lines of button data at its end. Hyperbole mail users never see this data in its raw form.
In order to alert readers of your mail messages that you can utilize Hyperbole mail buttons, the system automatically inserts a comment into each mail message that you compose to announce this fact. The variable, smail:comment controls this behavior. See its documentation for technical details. By default, it produces a message of the form:
Comments: Hyperbole mail buttons accepted, vX.XX.
where the X’s indicate your Hyperbole version number. You can cut this out of particular messages before you send them. If you don’t want any message at all, add the following to your ‘~/.emacs’ file before the point at which you load Hyperbole.
(setq smail:comment nil)
A final mail-related facility provided by Hyperbole is the ability to
save a pointer to a received mail message by creating an explicit button
with a link-to-mail
action type. When prompted for the mail
message to link to, if you press the Action Key on an Rmail message, the
appropriate parameter will be copied to the argument prompt, as
described in Entering Arguments.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Explicit buttons may be embedded within outgoing USENET news articles and may be activated from news articles that are being read. This support is available for the GNUS news reader. It is enabled by default within ‘hsite.el’ by autoloading the ‘hgnus.el’ file.
All Hyperbole support should work just as it does when reading or sending mail. See section Buttons in Mail. When reading news, buttons which appear in message Subject lines may be activated within the GNUS subject buffer as well as the article buffer. When posting news, the *post-news* buffer is used for outgoing news articles rather than the *mail* buffer.
Remember that the articles you post do not contain the action type definitions used by the buttons, so the receiver must have these or she will receive an error when he activates the buttons. You should also keep in mind that most USENET readers will not be using Hyperbole, so if they receive a news article containing explicit buttons, they will wonder what the button data at the end of the message is. You should therefore limit distribution of such messages. For example, if most people at your site read news with GNUS and use Hyperbole, it would be reasonable to embed buttons in postings to local newsgroups.
In order to alert readers of your postings that you can utilize Hyperbole mail buttons embedded within personal replies, the system automatically inserts the same comment that is included within mail messages to announce this fact. See section Buttons in Mail, for details and an explanation of how to turn this feature off.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
The Hyperbole outliner produces structured, autonumbered documents composed of hierarchies of cells. Each cell has two identifiers, a relative autonumber indicating its present position within the outline and a permanent identifier suitable for use within hyperlink references to the cell.
The outliner only works under GNU Emacs version 19 or higher and XEmacs
version 19.9 or higher. You can tell whether you are running a version
of Emacs which supports the outliner by hitting {C-h h} to
display the Hyperbole menu. If you see an Otl/
entry in the
menu, then the outliner is available. Otherwise, the outliner does not
work with your version of Emacs, so this section of the manual will not
be of interest to you.
The outliner is not yet fully documented within this manual. Full documentation will be available with version 4 of Hyperbole, after the outliner has been through user testing.
In the mean time, a brief explanation of the outliner is offered within
the ‘EXAMPLE.kotl’ file. This is an actual outline file that
explains basic operational details of the outliner. Use the
Otl/Example
menu entry to display this file.
Outliner Keys, for a full summary of the key bindings and commands available in the outliner.
You can create and experiment with outline files by editing a file with
the ‘.kotl’ suffix. ‘.kot’ will also work for
DOS/Windows-impaired users. The outliner menu, See section Outliner Menu,
also contains a Create
item.
4.7.1 Outliner Menu | ||
4.7.2 Outliner History |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
The Otl/ menu entry on the Hyperbole top-level menu provides access to a number of major outliner commands:
Menu Item Command Description ==================================================================== All kotl-mode:show-all Expand all cells Below kotl-mode:hide-sublevels Hide cells deeper than a level Create kfile:find Edit or create an outline Example <sample outliner file> Show self-descriptive example Hide kotl-mode:hide-tree Hide tree with root at point Info <outliner documentation> Show outliner manual section Kill kotl-mode:kill-tree Kill the current tree Link klink:create Create a link to another cell Overview kotl-mode:overview Show first line of each cell Show kotl-mode:show-tree Show tree with root at point Top kotl-mode:top-cells Collapse to top-level cells View kfile:view View an outline read-only ====================================================================
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Much of the Hyperbole outliner design is based upon concepts pioneered in the NLS/AUGMENT system, [Eng84a]. AUGMENT treated buffers as a hierarchical set of nodes, called statements; for example, each paragraph in a document would be treated as a node, with its own address. The system could rapidly change the view of a buffer by collapsing, expanding, clipping, filtering or reordering these nodes. These facilities aided greatly in idea structuring and cross-referencing. Hyperbole version 4 will extend the outliner with AUGMENT-style flexible views. Links will be able to specify not only a referent but also the view style to use when displaying the referent. This view capability will allow fine control over the presentation of information displayed by Hyperbole buttons.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole includes a complete, advanced rolodex system, Wrolo, for convenient management of hierarchical, record-oriented information.
Hyperbole buttons may be included within rolodex records and then manually activated whenever their records are retrieved.
See the description at the top of the ‘wrolo.el’ file for details on programmatic interfacing to the rolodex. The following subsections explain use and basic customization of the rolodex.
4.8.1 Rolo Concepts | ||
4.8.2 Rolo Menu | ||
4.8.3 Rolo Keys | ||
4.8.4 Rolo Settings |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
The rolodex manages and searches rolodex files. A rolodex file consists of an optional header which starts and ends with a line of equal signs (at least three equal signs starting at the beginning of a line), followed by any non-negative number of rolodex records. You must manually add a header to any rolodex file if you want it to have one.
Here is an example of a simple rolodex file.
================================================================== PERSONAL ROLODEX <Last-Name>, <First> <Co/Categ> <Email> W<Work#> F<Fax#> ================================================================== * Smith, John Motorola <js@mot.com> W2001 F1892
We call rolodex records, entries. Entries begin with a delimiter, some number of ‘*’ characters at the beginning of a line. Entries may be arranged in a hierarchy, where child entries begin with one more ‘*’ character than do their parents. Top level entries begin with a single ‘*’.
Beyond this initial delimiter, entries are completely free-form text. It is best to use a "lastname, firstname" format, however, when adding contact entries into a rolodex. Then the rolodex system will automatically keep your entries alphabetized.
Any search done on the rolodex scans the full text of each entry. During a search, the rolodex file header separator lines and anything in between are appended to the buffer of matched entries before any entries are retrieved from the file. Whenever an entry is matched, it and all of its descendant entries are retrieved. If your Emacs version supports textual highlighting, each search match is highlighted for quick, visual location.
For example, a search on "Company" could retrieve the following:
================================================================== COMPANY ROLODEX ================================================================== * Company ** Manager *** Underlings
Thus, searching for Company retrieves all listed employees. Searching for Manager turns up all Underlings.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
The Rolo/ menu entry on the Hyperbole top-level menu provides the user interface to the rolodex. The rolo menu provides access to the following commands:
Menu Item Command Description ==================================================================== Add rolo-add Adds a rolodex entry Display rolo-display-matches Displays last matches again Edit rolo-edit Edits an existing rolodex entry Info Displays Rolodex manual entry Kill rolo-kill Removes an entry from the rolodex Order rolo-sort Sorts all levels in rolodex RegexFind rolo-grep Finds all entries containing a regular expression StringFind rolo-fgrep Finds all entries containing a string WordFind rolo-word Finds all entries containing a string of whole words Yank rolo-yank Inserts first matching rolodex entry at point ====================================================================
A prefix argument used with either of the find commands listed above limits the search to a maximum number of matches given by the argument. The search is terminated whenever that number of matches is found.
For any of the above commands that prompt for a name, you may use the form parent/child to locate a child entry below a parent entry. So for a rolodex which looked like so:
* Company ** Manager *** Underlings
You could edit the Underlings entry by identifying it as Company/Manager/Underlings. Do not use this hierarchical notation in search expressions since the whole rolodex will be searched anyway. Thus, "Underlings" as a search pattern will find an entry containing "Underlings" at any level in a hierarchy, like so:
*** Underlings
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
After a rolodex search is performed, point is left in the rolodex
match buffer, ‘*Rolodex*’, which uses wrolo-mode
to
simplify browsing many rolodex matches. Press {?} when in the
match buffer for a summary of available keys.
If your Emacs version supports textual highlighting, each search match is highlighted for quick, visual location. {<TAB>} moves point forward to successive spans of text which match the search expression. {M-<TAB>} or {r} moves point backward to earlier matches. These keys allow you to quickly find the matching entry of most interest to you if your search expression failed to narrow the matches sufficiently.
Single key outlining commands are also available for browsing matches. If your search matches a large number of entries, use {t} to get a top-level overview of all the entries. Each entry is collapsed so that only its first line shows. Press {s} to show (expand) the entry at point. Use {h} to hide (collapse) the entry again. Press {a} to expand all entries in the buffer.
Many other keys are defined to help you move through matching entries.
Move to the previous entry at the same level as the current entry.
Move to the next entry at the same level as the current entry.
Move to the next entry at any level.
Move to the previous entry at any level.
Move the the previous entry one level up.
Move to the beginning of the buffer.
Move to the end of the buffer.
Scroll backward a windowful.
Scroll forward a windowful.
Once you have found an entry of interest and you want to remove the rolodex match buffer, use {q} to quit. This will restore your current frame to its state prior to the rolodex search.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
If textual highlighting is available in your Emacs on your current display type, the rolodex uses the value of rolo-highlight-face as the face to use to highlight search matches.
The buffers containing the rolodex files are not killed after a search
on the assumption that another search is likely to follow within this
Emacs session. You may wish to change this behavior with the following
setting: (setq rolo-kill-buffers-after-use t)
.
After an entry is killed, the modified rolodex file is automatically
saved. If you would rather always save files yourself, use this
setting: (setq rolo-save-buffers-after-use nil)
.
When adding an entry from within a buffer containing a mail message, the rolodex add function will extract the sender’s name and e-mail address and prompt you with the name as a default. If you accept it, it will enter the name and the email address using the format given by the rolo-email-format variable. See its documentation if you want to change its value.
The files used in any rolodex search are given by the
rolo-file-list variable, whose default value is
("~/.rolodex.otl")
, so that searches initially scan only your
personal rolodex. Any entries added to this list should be file
pathnames. If a file in the list does not exist or is not readable, it
is skipped. Files are searched in the order in which they appear in the
list. In general, you should leave your personal rolodex file as the
first entry in the list, since this is the only file to which the rolo
menu Add command adds entries.
The rolodex entry start delimiter is given by the regular expression variable, rolo-entry-regexp, whose default value is "^\*+".
A rolodex file may begin with an optional header section which is copied to the match display buffer whenever any matches are found during a search. The start and end lines of this header are controlled by the regular expression variable, rolo-hdr-regexp, whose default value is "^===". This allows lines of all equal signs to visually separate matching entries from multiple files retrieved from a single search.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This chapter is only for people who are familiar with Emacs Lisp and wish to extend Hyperbole, customize Hyperbole or develop other systems using Hyperbole as a base.
5.1 Hook Variables | ||
5.2 Creating Types | ||
5.3 Explicit Button Technicalities | ||
5.4 Encapsulating Systems | ||
5.5 Embedding Hyperbole |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole provides a number of hook variables that allow you to adjust its basic operations to meet your own needs, without requiring you to change the code for those operations.
We find it best to always set the value of hook variables either to nil or to a list of function names of no arguments, each of which will be called in sequence when the hook is triggered.
Given the name of a function, a Hyperbole hook variable triggered within
that function has the same name as the function with a -hook
appended. Hyperbole provides the following hook variables:
For customization at Hyperbole initialization time. Use this to load any personal Hyperbole type definitions you might have. It is run after Hyperbole support code is loaded but before Hyperbole session initializations take place.
Run before each Hyperbole button activation. The variable hbut:current contains the button to be activated when this is run.
To add to the Hyperbole explicit button creation process.
To add to the Hyperbole explicit button deletion process.
Executed when an explicit button’s attributes are modified.
Executed prior to loading of standard Hyperbole implicit button types. Used to load site-specific low priority implicit button types since lowest priority ibtypes are loaded first.
Executed after loading of standard Hyperbole implicit button types. Used to load site-specific high priority implicit button types since highest priority ibtypes are loaded last.
Executed when a Hyperbole type (e.g. action type or implicit button type) is added to the environment.
Executed when a type is deleted from the environment.
Executed when a Koutline is created or read in or when kotl-mode is invoked.
Hyperbole also makes use of a number of external Emacs hook variables.
This is called whenever a file is read into a GNU Emacs buffer. Hyperbole uses it to highlight any buttons within files when run under any NEXTSTEP or X window system-based versions of GNU Emacs.
This is called whenever a GNU Emacs buffer is written to a file. Hyperbole uses it to save any modified button data associated with the file’s directory.
Hyperbole mail and news facilities also utilize a number of external hook variables. These hide button data and highlight buttons if possible. See the various support files for details.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
To define or redefine a single Hyperbole type, you may either:
(eval-defun)
(only works in Emacs Lisp mode);
(eval-last-sexp)
(works in most modes).
The functions from the htype
class may be applied to any
Hyperbole types, if needed.
The following subsections explain the specifics of Hyperbole type definitions which are beyond standard practice for Emacs Lisp programming. See the definitions of the standard types in ‘hactypes.el’ and ‘hibtypes.el’ for examples.
5.2.1 Action Types | ||
5.2.2 Implicit Button Types |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
New forms of explicit buttons may be created by adding new action types to a Hyperbole environment. The file, ‘hactypes.el’, provides many examples of working action types.
An action type is created, i.e. loaded into the Hyperbole environment,
with the (defact)
function (which is an alias for
(actype:create)
). The calling signature for this function is
given in its documentation; it is the same as that of (defun)
except that a documentation string is required. (An interactive calling
form is also required if the action type has formal parameters and is to
be used in explicit button definitions. Implicit buttons never use an
action type’s interactive form. It is good practice to include an
interactive form since the type creator cannot know how users may choose
to apply the type.)
An action type’s parameters are used differently than those of a function being called. Its interactive calling form is used when an explicit button is created to prompt for type-specific button attributes. The rest of its body is used when a button with that action type is activated. Then the button attributes together with the action type body are used to form an action that is executed in response to the button activation. The action’s result is returned to the action caller unless it returns nil, in which case t is returned to the caller to ensure that it registers the performance of the action.
An action type body may perform any computation using Emacs Lisp and Hyperbole functions.
The interactive calling form for an action type is of the same form as
that of a regular Emacs Lisp function definition (see the documentation
for the Emacs Lisp (interactive)
form). It may additionally use
Hyperbole command character extensions when the form is given as a
string. Each such extension character must be preceded by a plus
sign, +
, in order to be recognized since such characters may also
have standard interactive form meanings.
The present Hyperbole extension characters are:
Prompts for an existing Info node name and file.
Prompts for an existing kcell identifier, either a full outline level identifier or a permanent idstamp.
Prompts for a mail message date and the file name it resides in. The mail parameters prompted for by this character code are likely to change in the future.
Prompts for a Hyperbole view specification. Not yet available for use.
Arguments are read by the functions in Hyperbole’s hargs
class,
rather than the standard Lisp read
functions, in order to allow
direct selection of arguments via the Action Key.
If an action type create is successful, the symbol that Hyperbole uses
internally to reference the type is returned. Nil
is returned on
failure so that you may test whether or not the operation succeeds.
Once you have defined an action type within your present Hyperbole environment, you can create new explicit buttons which use it. There is no explicit button type beyond its action type, so no further work is necessary.
Call (actype:delete)
to remove an action type from a Hyperbole
environment. It takes a single parameter which should be the same type
symbol used in the type definition call (not the Hyperbole symbol
returned by the call).
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
An implicit button type is created or loaded via the (defib)
function (which is an alias for (ibtype:create)
). The calling
signature for this function is given in its documentation; it is the
same as that of (defun)
, but with a number of constraints. The
parameter list should always be empty since no parameters will be used.
A documentation string is required. The type’s body follows this.
The body of an implicit button type is a predicate which determines
whether or not point is within an implicit button of the type. If not,
the predicate returns nil
. If so, it may optionally setup to
flash the button and then perform one or more actions. A call of the
form: (ibut:label-set label start-pos end-pos)
is used to setup
the button flashing, if desired. This is then typically immediately
followed by an action invocation of the form:
(hact 'actype &rest actype-arguments)
. It is imperative that all
actions (non-predicate code) be invoked through the (hact)
function rather than directly or your ibtypes will not work properly.
(Hyperbole first tests to see if any ibtype matches the current context
before activating any type, so it ensures that (hact)
calls are
disabled during this testing.) Any action types used may be created
before or after the implicit button type definition but obviously should
be defined before any implicit buttons of the given type are activated;
an error will result, otherwise.
If an implicit button type create is successful, the symbol that
Hyperbole uses internally to reference the type is returned. Nil
is returned on failure so that you may test whether or not the operation
succeeds. Implicit button type names and action type names may be the
same without any conflict. In fact, such naming is encouraged when an
implicit button type is the exclusive user of an action type.
Call (ibtype:delete)
to remove an implicit button type from a
Hyperbole environment. It takes a single parameter which should be the
same type symbol used in the type definition call (not the Hyperbole
symbol returned by the call). This will not delete the action type used
by the implicit button; that must be done separately.
By default, a request for help on an implicit button will display the
button’s attributes in the same manner as is done for explicit buttons.
For some implicit button types, other forms of help will be more
appropriate. If an Emacs Lisp function is defined whose name is formed
from the concatenation of the type name followed by :help
, e.g.
my-ibtype:help
, it is used to respond to requests for
help on buttons of that type. Any such function should take a single
argument of an implicit button construct. (This is what
(ibut:at-p)
returns when point is within an implicit button
context.) The button may be queried for its attributes using functions
from the hbut
and hattr
classes. See the ‘hib-kbd.el’
file for an example of a custom help function.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
5.3.1 Button Label Normalization | ||
5.3.2 Operational and Storage Formats | ||
5.3.3 Programmatic Button Creation |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole uses a normalized form of button labels called button keys (or
label keys) for all internal operations. See the documentation for the
function (hbut:label-to-key)
for details of the normalization
process. The normalized form permits Hyperbole to recognize buttons that
are the same but whose labels appear different from one another, due to
text formatting conventions. For example, all of the following would
be recognized as the same button.
<(fake button)> <( fake button)> Pam> <(fake Pam> button)> ;; <(fake ;; button)> /* <( fake */ /* button )> */
The last three examples demonstrate how Hyperbole ignores common fill prefix patterns that happen to fall within the middle of a button label that spans multiple lines. As long as such buttons are selected with point at a location within the label’s first line, the button will be recognized. The variable hbut:fill-prefix-regexps holds the list of fill prefixes recognized when embedded within button labels. All such prefixes are recognized (one per button label), regardless of the setting of the GNU Emacs variable, fill-prefix, so no user intervention is required.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Hyperbole uses a terse format to store explicit buttons and a more meaningful one to show users and to manipulate during editing. The terse format consists solely of button attribute values whereas the edit format includes an attribute name with each attribute value. A button in edit format consists of a Lisp symbol together with its attribute list which holds the attribute names and values. In this way, buttons may be passed along from function to function simply by passing the symbol to which the button is attached. Most functions utilize the pre-defined hbut:current symbol by default to store and retrieve the last encountered button in edit format.
The hbdata
class handles the terse, stored format. The
hbut
, ebut
, and ibut
classes work with the
name/value format. This separation permits the wholesale replacement of
the storage manager with another, with any interface changes hidden from
any Hyperbole client programming.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
A common need when developing with Hyperbole is the ability to create or modify explicit buttons without user interaction. For example, an application might require the addition of an explicit summary button to a file for each new mail message a user reads that contains a set of keywords. The user could then check the summary file and jump to desired messages quickly.
The Hyperbole class ebut
supports programmatic access to explicit
buttons. See it within the ‘hbut.el’ file for full details. The
documentation for (ebut:create)
explains the set of attributes
settings necessary to create an explicit button. For operations over
the whole set of buttons within the visible (non-narrowed) portion of a
buffer, use the (ebut:map)
function.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
A powerful use of implicit button types is to provide a Hyperbole-based interface to external systems. The basic idea is to interpret patterns within the application’s displays as implicit buttons. See the ‘hsys-*’ files for examples of how to do this.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
[NOTE: We have never done this ourselves, though we have done similar things which leads us to infer that the task should not be difficult.]
The standard Emacs-based Hyperbole user interface has purposely been separated from the Hyperbole backend to support the development of alternative interfaces and the embedding of Hyperbole functionality within other system prototypes. The Hyperbole backend functionality that system developers can make use of is called its Application Programming Interface (API). The API may be used to make server-based calls to Hyperbole when Emacs is run as a non-interactive (batch) process, with its input/output streams attached to another process.
The public functions and variables from the following files may be considered the present Hyperbole API:
‘hact.el’, ‘hargs.el’, ‘hbmap.el’, ‘hbut.el’, ‘hhist.el’, ‘hmail.el’, ‘hmoccur.el’, ‘hpath.el’, ‘htz.el’, ‘hypb.el’, ‘set.el’, ‘wconfig.el’, ‘wrolo.el’, and ‘wrolo-logic.el’.
(Note that within a class definition, functions and variables that follow a line of dashes are private.)
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Concepts pertinent to operational usage of Hyperbole are defined here. If some GNU Emacs terms are unfamiliar to you, see Emacs Glossary in the GNU Emacs Manual.
action
An executable behavior associated with a Hyperbole button. A specific class of actions which display entities are called links, such as a link to a file.
Action Key
See Smart Key.
action type
A behavioral specification for use within Hyperbole buttons. Action types usually contain a set of parameters which must be given values for each button with which they are associated. An action type together with a set of values, called arguments, may be considered an action. Actype is a synonym for action type.
activation
Request for a Hyperbole button to perform its action. Ordinarily the user presses a key which selects and activates a button.
ange-ftp
A standard GNU Emacs Lisp package which allows one to use pathnames that are accessible via the Internet File Transfer Protocol (ftp) just like other pathnames, for example when finding a file. The latest version of ange-ftp may always be obtained via anonymous ftp to: ‘/ftp.gnu.ai.mit.edu:ange-ftp/ange-ftp.tar.gz’.
argument
A button-specific value fed to a Hyperbole type specification when the button is activated.
Assist Key
See Smart Key.
attributes
Slot names associated with Hyperbole buttons. An attribute value is associated with each button attribute.
button
A selectable Hyperbole construct which performs an action. A button consists of a set of attributes that includes: a textual label, a category, a type and zero or more arguments. Explicit buttons also have creator, create time, last modifier, and last modifier time attributes.
Buttons provide the user’s gateway to information. The user sees and interacts with button labels, the rest of the button data is managed invisibly by Hyperbole and displayed only in response to user queries.
button activation
See activation.
button attributes
See attributes.
button data
Lists of button attribute values explicitly saved and managed by Hyperbole. One list for each button created by Hyperbole.
button file, local
A per-directory file named ‘HYPB’ that may be used to store any desired buttons and may then be displayed via a menu selection whenever a user is within that directory.
button file, personal
A per-user file named ‘HYPB’ that may be used to store any desired buttons and may then be displayed via a menu selection.
button key
A normalized form of a button label used internally by Hyperbole.
button label
A text string that visually indicates a Hyperbole button location and provides it with a name and unique identifier. Within a buffer, buttons with the same label are considered separate views of the same button and so behave exactly alike. Since button labels are simply text strings, they may be embedded within any text to provide non-linear information or operational access points.
The maximum length of a button label is limited by the variable ebut:max-len.
button selection
The act of designating a Hyperbole button upon which to operate. Use the Action Key to select a button.
category
A high-level, conceptual grouping of Hyperbole buttons into classes. Implicit and explicit groupings represent categories.
class
A group of functions and variables with the same prefix in their names, used to provide an interface to an internal or external Hyperbole abstraction.
context
A programmatic or positional state recognized by Hyperbole. We speak of Smart Key and implicit button contexts. Both are typically defined in terms of surrounding patterns within a buffer, but may be defined by arbitrary Emacs Lisp predicates. (Context may come to have a broader meaning within future versions of Hyperbole.)
environment
See Hyperbole environment.
efs
The much larger successor to ange-ftp. It does the same thing as ange-ftp but works with more types of ftp hosts. See ange-ftp.
explicit button
A button created and managed by Hyperbole. By default, explicit buttons
are delimited like this <(fake button)>
. Direct selection is
used to operate upon an explicit button.
global button
A form of explicit button which is typically accessed by name rather than direct selection. Global buttons are useful when one wants quick access to actions such as jumping to common file locations or for performing sequences of operations. One need not locate them since they are always available by name, with full completion offered. All global buttons are stored in the file given by the variable gbut:file and may be activated as regular explicit buttons by visiting this file. By default, this is the same as the user’s personal button file.
global button file
See button file, personal.
hook variable
A variable that permits customization of an existing function’s
operation without the need to edit the function’s code. See also the
documentation for the function (run-hooks)
.
Hyperbole
A flexible, programmable information management and viewing system built on top of GNU Emacs. It utilizes a button-action model and supports hypertextual linkages. Hyperbole is all things to all people.
Hyperbole environment
A programmatic context within which Hyperbole operates. This includes the set of Hyperbole types defined and the set of Hyperbole code modules loaded. It does not include the set of accessible buttons. Although the entire Emacs environment is available to Hyperbole, we do not speak of this as part of the Hyperbole environment.
hypertext
A text or group of texts which may be explored in a non-linear fashion through associative linkages embedded throughout the text. Instead of simply referring to other pieces of work, hypertext references when followed actually take you to the works themselves.
implicit button
A button recognized contextually by Hyperbole. Such buttons contain no button data. See also implicit button type.
implicit button type
A specification of how to recognize and activate implicit buttons of a specific kind. Implicit button types often utilize structure internal to documents created and managed without Hyperbole assistance, for example, programming documentation. Ibtype is a synonym for implicit button type. See also system encapsulation.
instance number
A colon prefaced number appended to the label of a newly created button when the button’s label duplicates the label of an existing button in the current buffer. This number makes the label unique and so allows any number of buttons with the same base label within a single buffer.
link
A reference from a Hyperbole button to an entity. The referenced entity is sometimes called a node or referent. A specific class of actions which display entities are called links, such as a link to a file.
local button file
See button file, local.
mouse button
mouse key
See Smart Key.
node
See link.
predicate
A boolean (nil = false, non-nil = true) Lisp expression typically evaluated as part of a conditional expression.
referent
See link.
Smart Key
A context-sensitive key used within Hyperbole and beyond. Actually, there are two Smart Keys, the Action Key and the Assist Key. The Action Key, typically bound to the shift-middle mouse key (or shift-left mouse key on a 2-button mouse), activates Hyperbole buttons and scrolls the buffer forward a windowful when pressed at the end of a line. The Assist Key, typically bound to the shift-right mouse key, explains what a Hyperbole button does or scrolls the buffer backward a windowful when pressed at the end of a line.
To see what a Smart Key will do within a particular context, depress and hold the key at the point desired and depress the other Smart Key. A buffer containing a description of its contextual function will then be displayed. You may release the two keys in any order after you have them both depressed. A press of the Assist Key in an unsupported context displays a summary of Smart Key functions in each context, as does the Doc/SmartKy menu item.
source buffer / file
The buffer or file within which a Hyperbole button is embedded.
system encapsulation
Use of Hyperbole to provide an improved or simply consistent user interface to another system. Typically, implicit button types are defined to recognize and activate button-type constructs managed by the other system.
view
A perspective on some information. A view can affect the extent of the information displayed, its format, modes used to operate on it, its display location and so forth.
view spec
A terse (and to the uninitiated, cryptic) string that specifies a particular view of a link referent or some other information.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This appendix supplies complete documentation on Smart Key operation. It is quite extensive and is meant for reference rather than sequential reading. See section Smart Keys, for a description of the Smart Keys. That section also describes how to get context-sensitive Smart Key help, with which you can explore Smart Key operation bit by bit.
Smart Key operations are context-sensitive. Contexts are described herein as conditionals, e.g. when depressed here, if this is true, etc. Each Smart Key context is listed in the order in which it will be checked. The first matching context is always the one applied. Within each context, the actions performed by the Action and Assist Keys are listed.
B.1 Smart Mouse Keys | ||
B.2 Smart Keyboard Keys |
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Smart Key drags and modeline presses can only be used when running under a window system with mouse key support. So keep in mind that the operations in this section apply only if you have mouse support within Hyperbole. The Smart Key operations in, Smart Keyboard Keys, apply to both mouse and keyboard Smart Key usage.
If dragged from a side-by-side window edge or from the immediate left of a vertical scroll bar: ACTION or ASSIST Resizes adjacent window sides to the point of drag release.
If dragged from inside one window to another: ACTION Creates a new link button at the drag start location, linked to the drag end location. If drag start position is within a button, modifies the button to link to drag end location. ASSIST Swaps buffers in the two windows.
If dragged horizontally within a single window while depressed (hmouse-x-drag-sensitivity sets the minimal horizontal movement which registers a drag): ACTION Goes to buffer end if drag was to the right, otherwise goes to beginning. ASSIST Splits window vertically if drag was to the right, otherwise deletes window.
If depressed within a window mode line: ACTION (1) clicked on left edge of a window’s modeline, window’s buffer is buried (placed at bottom of buffer list); (2) clicked on right edge of a window’s modeline, the Info buffer is displayed, or if already displayed and the modeline clicked belongs to a window displaying Info, the Info buffer is hidden; (3) clicked anywhere in the middle of a window’s modeline, the functions listed in ’assist-key-modeline-hook’ are called; (4) dragged vertically from modeline to within a window, the modeline is moved to point of key release, thereby resizing its window and potentially its vertical neighbors. ASSIST (1) clicked on left edge of a window’s modeline, bottom buffer in buffer list is unburied and placed in window; (2) clicked on right edge of a window’s modeline, the summary of Smart Key behavior is displayed, or if already displayed and the modeline clicked belongs to a window displaying the summary, the summary buffer is hidden; (3) clicked anywhere in the middle of a window’s modeline, a popup menu (if available) is displayed; (4) dragged vertically from modeline to within a window, the modeline is moved to point of key release, thereby resizing its window and potentially its vertical neighbors.
If dragged vertically within a single window while depressed (hmouse-y-drag-sensitivity sets the minimal vertical movement which registers a drag): ACTION or ASSIST Splits current window into two side-by-side windows.
If dragged diagonally within a single window while depressed (hmouse-x-diagonal-sensitivity and hmouse-y-diagonal-sensitivity set the minimal diagonal movement which registers a drag): ACTION Save current window configuration onto a ring of window configurations. ASSIST Restores prior window configuration from ring. A prefix argument N specifies the Nth prior configuration from the ring.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
When prompting for a Hyperbole argument, a press in the minibuffer: ACTION Terminates this minibuffer argument. ASSIST Offers completion help for current minibuffer argument.
When reading a Hyperbole menu item or a Hyperbole completion-based argument: ACTION Returns value selected at point if any, else nil. If value is the same as the contents of the minibuffer, it is used as the current minibuffer argument, otherwise, the minibuffer is erased and value is inserted there. ASSIST Displays Hyperbole menu item help when item is selected.
When pressed at the end of a line but not the end of a buffer: ACTION Scrolls up according to value of smart-scroll-proportional. If smart-scroll-proportional is nil (the default) or if point is on the top window line, scrolls up (forward) a windowful. Otherwise, tries to bring current line to top of window. Leaves point at end of line and returns t if scrolled, nil if not. ASSIST Scrolls down according to value of smart-scroll-proportional. If smart-scroll-proportional is nil (the default) or if point is on the bottom window line, scrolls down (backward) a windowful. Otherwise, tries to bring current line to bottom of window. Leaves point at end of line and returns t if scrolled, nil if not.
When pressed on a Hyperbole button: ACTION Activates button. ASSIST Displays help for button, typically a summary of its attributes.
If pressed within a buffer in View major or minor mode: ACTION Scrolls buffer forward a windowful and quits from view mode when at the last line of the buffer. ASSIST Scrolls buffer backward a windowful.
When pressed within a Hyperbole outliner buffer (kotl-mode): ACTION (1) at the end of buffer, uncollapse and unhide all cells in view; (2) within a cell, if its subtree is hidden then show it, otherwise hide it; (3) between cells or within the read-only indentation region to the left of a cell, then move point to prior location and begin creation of a klink to some other outline cell; hit the Action Key twice to select the link referent cell; (4) anywhere else, scroll up a windowful. ASSIST (1) at the end of buffer, collapse all cells and hide all non-level-one cells; (2) on a header line but not at the beginning or end, display properties of each cell in kotl beginning at point; (3) between cells or within the read-only indentation region to the left of a cell, then move point to prior location and prompt to move one tree to a new location in the outline; hit the Action Key twice to select the tree to move and where to move it; (4) anywhere else, scroll down a windowful.
When pressed on a Smart Menu item: ACTION Activates item. ASSIST Displays help for item.
When pressed at the end of a Help buffer: ACTION or ASSIST Restores window configuration prior to help display.
When pressed within an OO-Browser listing window: ACTION (1) in a blank buffer or at the end of a buffer, browser help information is displayed in the viewer window; (2) at the beginning of a (non-single char) class name, the class’ ancestors are listed; (3) at the end of an entry line, scrolls listing up; (4) on the ’...’, following a class name, point is moved to the class descendency expansion; (5) before an element name, the implementor classes of the name are listed; (6) anywhere else on an entry line, the source is displayed for editing. ASSIST (1) in a blank buffer, a selection list of buffer files is displayed; (2) at the beginning of a (non-single char) entry, the class’ descendants are listed; (3) at the end of an entry line, scrolls listing down; (4) on the ’...’, following a class name, point is moved to the class expansion; (5) anywhere else on a class entry line, lists the class’ elements; (6) anywhere else on an element line, lists the element’s implementor classes; (7) on a blank line following all entries, the current listing buffer is exited.
When pressed within an OO-Browser Command Help Menu buffer: ACTION Executes an OO-Browser command whose key binding is at point. ASSIST Displays help for an OO-Browser command whose key binding is at point.
When pressed on an identifier within an OO-Browser source file: ACTION Tries to display identifier definition. ASSIST Not applicable.
When pressed within a C source code file: ACTION Jumps to the definition of selected C construct: (1) on a ’#include’ statement, the include file is displayed; Look for include file in directory lists ’smart-c-cpp-include-dirs’ and ’smart-c-include-dirs’. (2) on a C identifier, the identifier definition is displayed, assuming the identifier is found within an ’etags’ generated tag file in the current directory or any of its ancestor directories. (3) if ’smart-c-use-lib-man’ is non-nil, the C identifier is recognized as a library symbol, and a man page is found for the identifier, then the man page is displayed. ASSIST Jumps to the next tag matching an identifier at point.
When pressed within an assembly source code file: ACTION Jumps to the definition of selected assembly construct: (1) on an include statement, the include file is displayed; Look for include file in directory list ’smart-asm-include-dirs’. (2) on an identifier, the identifier definition is displayed, assuming the identifier is found within an ’etags’ generated tag file in the current directory or any of its ancestor directories. ASSIST Jumps to the next tag matching an identifier at point.
When pressed within a C++ source code file: ACTION Jumps to the definition of selected C+ construct: (1) on a ’#include’ statement, the include file is displayed; Look for include file in directory lists ’smart-c-cpp-include-dirs’ and ’smart-c-include-dirs’. (2) on a C++ identifier, the identifier definition is displayed, assuming the identifier is found within an ’etags’ generated tag file in the current directory or any of its ancestor directories. (3) if ’smart-c-use-lib-man’ is non-nil, the C++ identifier is recognized as a library symbol, and a man page is found for the identifier, then the man page is displayed. ASSIST Jumps to the next tag matching an identifier at point.
When pressed within a Objective-C source code file: ACTION Jumps to the definition of selected C+ construct: (1) on a ’#include’ statement, the include file is displayed; Look for include file in directory lists ’smart-c-cpp-include-dirs’ and ’smart-c-include-dirs’. (2) on an Objective-C identifier, the identifier definition is displayed, assuming the identifier is found within an ’etags’ generated tag file in the current directory or any of its ancestor directories. (3) if ’smart-c-use-lib-man’ is non-nil, the Objective-C identifier is recognized as a library symbol, and a man page is found for the identifier, then the man page is displayed. ASSIST Jumps to the next tag matching an identifier at point.
When pressed on a Lisp symbol within a Lisp code buffer: ACTION Jumps to the definition of any selected Lisp construct. If on an Emacs Lisp require, load, or autoload clause and ’find-library’ from load-library package by Hallvard Furuseth <hallvard@ifi.uio.no> has been loaded, jumps to library source, if possible. ASSIST Jumps to the next tag matching an identifier at point or if using the "wtags" package and identifier is an Emacs Lisp symbol, then displays documentation for the symbol.
When the OO-Browser has been loaded and the press is within a C++ buffer: ACTION or ASSIST Jumps to the definition of selected C++ construct via OO-Browser support. (1) on a ’#include’ statement, the include file is displayed; Look for include file in directory lists ’smart-c-cpp-include-dirs’ and ’smart-c-include-dirs’. (2) within a method declaration, its definition is displayed; (3) on a class name, the class definition is shown.
When the OO-Browser has been loaded and the press is within a Objective-C buffer: ACTION or ASSIST Jumps to the definition of selected Objective-C construct via OO-Browser support. (1) on a ’#include’ statement, the include file is displayed; Look for include file in directory lists ’smart-c-cpp-include-dirs’ and ’smart-c-include-dirs’. (2) within a method declaration, its definition is displayed; (3) on a class name, the class definition is shown.
When pressed within an occur-mode or moccur-mode buffer: ACTION or ASSIST Jumps to the source buffer and line of the current occurrence.
When pressed within a calendar-mode buffer: ACTION (1) at the end of the buffer, the calendar is scrolled forward 3 months; (2) to the left of any dates on a calendar line, the calendar is scrolled backward 3 months; (3) on a date, the diary entries for the date, if any, are displayed. ASSIST (1) at the end of the buffer, the calendar is scrolled backward 3 months; (2) to the left of any dates on a calendar line, the calendar is scrolled forward 3 months; (3) anywhere else, all dates with marking diary entries are marked in the calendar window.
When pressed within a man page apropos buffer: ACTION (1) on a UNIX man apropos entry, the man page for that entry is displayed in another window; (2) on or after the last line, the buffer in the other window is scrolled up a windowful. ASSIST (1) on a UNIX man apropos entry, the man page for that entry is displayed in another window; (2) on or after the last line, the buffer in the other window is scrolled down a windowful.
If Smart Menu package has been loaded and ’hkey-always-display-menu’ is non-nil: ACTION or ASSIST Pops up a window with a Smart Menu of commands. Menu displayed is selected by (smart-menu-choose-menu).
If pressed within an outline-mode buffer or when ’selective-display’ is non-nil: ACTION Collapses, expands, and moves outline entries. (1) after an outline heading has been cut via the Action Key, then paste the cut heading at point; (2) at the end of buffer, show all buffer text (3) at the beginning of a heading line, cut the headings subtree from the buffer; (4) on a header line but not at the beginning or end, if headings subtree is hidden then show it, otherwise hide it; (5) anywhere else, scroll up a windowful. ASSIST (1) after an outline heading has been cut via the Action Key, allow multiple pastes throughout the buffer (last paste should be done with the Action Key, not the Assist Key); (2) at the end of buffer, hide all bodies in buffer; (3) at the beginning of a heading line, cut the current heading (sans subtree) from the buffer; (4) on a header line but not at the beginning or end, if heading body is hidden then show it, otherwise hide it; (5) anywhere else, scroll down a windowful.
If pressed within an Info manual node: ACTION (1) the first line of an Info Menu Entry or Cross Reference, the desired node is found; (2) the Up,Next,or Previous entries of a Node Header (first line), the desired node is found; (3) the File entry of a Node Header (first line), the ’Top’ node within that file is found; (4) at the end of the current node, the Next node is found (this will descend subtrees if the function ’Info-global-next’ is bound); (5) anywhere else (e.g. at the end of a line), the current node entry is scrolled up a windowful. ASSIST (1) the first line of an Info Menu Entry or Cross Reference, the desired node is found; (2) the Up,Next,or Previous entries of a Node Header (first line), the last node in the history list is found; (3) the File entry of a Node Header (first line), the ’DIR’ root-level node is found; (4) at the end of the current node, the Previous node is found (this will return from subtrees if the function ’Info-global-prev is bound); (5) anywhere else (e.g. at the end of a line), the current node entry is scrolled down a windowful.
If pressed within a Hyperbole-supported mail reader, ’hmail:reader’, or mail summary mode, ’hmail:lister’, buffer at: ACTION (1) a msg buffer, within the first line or at the end of a message, the next undeleted message is displayed; (2) a msg buffer within the first line of an Info cross reference, the reference is followed; (3) anywhere else in a msg buffer, the window is scrolled up one windowful; (4) a msg summary buffer on a header entry, the message corresponding to the header is displayed in the msg window; (5) a msg summary buffer, on or after the last line, the messages marked for deletion are expunged. ASSIST (1) a msg buffer, within the first line or at the end of a message, the previous undeleted message is displayed; (2) a msg buffer within the first line of an Info cross reference, the reference is followed; (3) anywhere else in a msg buffer, the window is scrolled down one windowful; (4) a msg summary buffer on a header entry, the message corresponding to the header is marked as deleted; (5) a msg summary buffer, on or after the last line, all messages are marked undeleted.
If pressed within a GNUS listing of newsgroups buffer at: ACTION (1) a GNUS-GROUP line, that newsgroup is read; (2) to the left of any GNUS-GROUP line, on any of the whitespace, the current group is unsubscribed or resubscribed; (3) at the end of the GNUS-GROUP buffer, after all lines, checks for new news. ASSIST (1) a GNUS-GROUP line, that newsgroup is read; (2) to the left of any GNUS-GROUP line, on any of the whitespace, the user is prompted for a group name to subscribe or unsubscribe to; (3) at the end of the GNUS-GROUP buffer, after all lines, quits from the newsreader.
If pressed within a GNUS newsreader subject listing buffer at: ACTION (1) a GNUS-SUBJECT line, that article is read, marked deleted, and scrolled forward; (2) at the end of the GNUS-SUBJECT buffer, the next undeleted article is read or the next group is entered. ASSIST (1) a GNUS-SUBJECT line, that article is read and scrolled backward; (2) at the end of the GNUS-SUBJECT buffer, the subject is exited, the user is returned to group mode.
If pressed within a GNUS newsreader article buffer at: ACTION (1) the first line or end of an article, the next unread message is displayed; (2) the first line of an Info cross reference, the reference is followed; (3) anywhere else, the window is scrolled up a windowful. ASSIST (1) the first line or end of an article, the previous message is displayed; (2) the first line of an Info cross reference, the reference is followed; (3) anywhere else, the window is scrolled down a windowful.
If pressed within a listing of buffers (Buffer-menu-mode): ACTION (1) on the first column of an entry, the selected buffer is marked for display; (2) on the second column of an entry, the selected buffer is marked to be saved; (3) anywhere else within an entry line, all saves and deletes are done, and selected buffers are displayed, including the one just clicked on (if in the OO-Browser, only the selected buffer is displayed); (4) on or after the last line in the buffer, all saves and deletes are done. ASSIST (1) on the first or second column of an entry, the selected buffer is unmarked for display and for saving or deletion; (2) anywhere else within an entry line, the selected buffer is marked for deletion; (3) on or after the last line in the buffer, all display, save, and delete marks on all entries are undone.
If pressed within a dired-mode buffer: ACTION (1) within an entry line, the selected file/directory is displayed for editing in the other window; (2) on or after the last line in the buffer, if any deletes are to be performed, they are executed after user verification, otherwise, this dired invocation is quit. ASSIST (1) on a ’~’ character, all backup files in the directory are marked for deletion; (2) on a ’#’ character, all auto-save files in the directory are marked for deletion; (3) anywhere else within an entry line, the current entry is marked for deletion; (4) on or after the last line in the buffer, all delete marks on all entries are undone.
If pressed within a tar-mode buffer: ACTION (1) within an entry line, the selected file/directory is displayed for editing in the other window; (2) on or after the last line in the buffer, if any deletes are to be performed, they are executed after user verification, otherwise, this tar file browser is quit. ASSIST (1) on an entry line, the current entry is marked for deletion; (2) on or after the last line in the buffer, all delete marks on all entries are undone.
If pressed on a cross reference within a man page entry section labeled NAME, SEE ALSO, or PACKAGES USED, or within a man page C routine specification (see ’smart-man-c-routine-ref’) and man page buffer has either an attached file or else a man-path local variable containing its pathname: ACTION or ASSIST Displays man page or source code for cross reference.
If pressed within an entry in the wrolo match display buffer: ACTION or ASSIST The entry is edited in the other window.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This appendix summarizes the specialized key bindings available when editing an outline with Hyperbole. Each key is shown together with its command binding and the documentation for that command. Normal emacs editing keys are modified to account for the structure within outlines. An outliner command which overloads an Emacs command named cmd would be named kotl-mode:cmd.
kfile:write {C-x C-w}
Write current outline to FILE.
klink:create {C-c l}
Insert at point an implicit link to REFERENCE. REFERENCE should be a cell-ref or a list of (filename cell-ref). See documentation for ’kcell:ref-to-id’ for valid cell-ref formats.
kotl-mode:add-cell {<LFD>}
Add a cell following current cell at optional RELATIVE-LEVEL with CONTENTS string. Optional prefix arg RELATIVE-LEVEL means add as sibling if nil or >= 0, as child if equal to universal argument, C-u, and as sibling of current cell’s parent, otherwise. If added as sibling of current level, RELATIVE-LEVEL is used as a repeat count for the number of cells to add.
Return last newly added cell.
kotl-mode:add-child {C-c a}
Add a new cell to current kview as first child of current cell.
kotl-mode:add-parent {C-c p}
Add a new cell to current kview as sibling of current cell’s parent.
kotl-mode:back-to-indentation {M-m}
Move point to the first non-read-only non-whitespace character on this line.
kotl-mode:backward-cell {C-c C-b}
Move to prefix ARGth prior cell (same level) within current view. Return number of cells left to move.
kotl-mode:backward-char {C-b}
Move point backward ARG (or 1) characters and return point.
kotl-mode:backward-kill-word {M-DEL}
Kill up to prefix ARG words preceding point within a single cell.
kotl-mode:backward-sentence {M-a}
Move point backward ARG (or 1) sentences and return point.
kotl-mode:backward-word {M-b}
Move point backward ARG (or 1) words and return point.
kotl-mode:beginning-of-buffer {M-<}
Move point to beginning of buffer and return point.
kotl-mode:beginning-of-cell {C-c ,}
Move point to beginning of current or ARGth - 1 prior cell and return point.
kotl-mode:beginning-of-line {C-a}
Move point to beginning of current or ARGth - 1 line and return point.
kotl-mode:beginning-of-tree {C-c ^}
Move point to the level 1 root of the current cell’s tree. Leave point at the start of the cell.
kotl-mode:center-line {M-s}
Center the line point is on, within the width specified by fill-column. This means adjusting the indentation so that it equals the distance between the end of the text and fill-column.
kotl-mode:center-paragraph {M-S}
Center each nonblank line in the paragraph at or after point.
See center-line
for more info.
kotl-mode:copy-after {C-c c}
Copy tree rooted at FROM-CELL-REF to follow tree rooted at TO-CELL-REF. If prefix arg CHILD-P is non-nil, make FROM-CELL-REF the first child of TO-CELL-REF, otherwise make it the sibling following TO-CELL-REF.
Leave point at the start of the root cell of the new tree.
kotl-mode:copy-before {C-c C-c}
Copy tree rooted at FROM-CELL-REF to precede tree rooted at TO-CELL-REF. If prefix arg PARENT-P is non-nil, make FROM-CELL-REF the first child of TO-CELL-REF’s parent, otherwise make it the preceding sibling of TO-CELL-REF.
Leave point at the start of the root cell of the new tree.
kotl-mode:copy-to-buffer {C-c M-c}
Copy outline tree rooted at CELL-REF to a non-koutline BUFFER. Use 0 to copy the whole outline buffer.
kotl-mode:copy-to-register {C-x x}
Copy into REGISTER the region START to END. With optional prefix arg DELETE-FLAG, delete region.
kotl-mode:delete-backward-char {DEL}
Delete up to the preceding prefix ARG characters. Return number of characters deleted. Optional KILL-FLAG non-nil means save in kill ring instead of deleting. Does not delete across cell boundaries.
kotl-mode:delete-blank-lines {C-x C-o}
On blank line within a cell, delete all surrounding blank lines, leaving just one. On isolated blank line, delete that one. On nonblank line, delete all blank lines that follow it.
If nothing but whitespace follows point until the end of a cell, delete all whitespace at the end of the cell.
kotl-mode:delete-char {C-d}
Delete up to prefix ARG characters following point. Return number of characters deleted. Optional KILL-FLAG non-nil means save in kill ring instead of deleting. Does not delete across cell boundaries.
kotl-mode:delete-indentation {M-^}
Join this line to previous and fix up whitespace at join. If there is a fill prefix, delete it from the beginning of this line. With argument, join this line to following line.
kotl-mode:demote-tree {TAB}
Move current kotl a maximum of prefix ARG levels lower in current view. Each cell is refilled iff its no-fill attribute is nil and kotl-mode:refill-flag is non-nil. With prefix ARG = 0, cells are demoted up to one level and kotl-mode:refill-flag is treated as true.
kotl-mode:down-level {C-c d}
Move down prefix ARG levels lower within current tree.
kotl-mode:end-of-buffer {M->}
Move point to end of buffer and return point.
kotl-mode:end-of-cell {C-c .}
Move point to end of current or ARGth - 1 succeeding cell and return point.
kotl-mode:end-of-line {C-e}
Move point to end of current or ARGth - 1 line and return point.
kotl-mode:end-of-tree {C-c $}
Move point to the last cell in tree rooted at the current cell. Leave point at the start of the cell.
kotl-mode:exchange-cells {C-c e}
Exchange CELL-REF-1 with CELL-REF-2 in current view. Don’t move point.
kotl-mode:fill-cell {C-c M-j}
Fill current cell within current view if it does not have the ’no-fill attribute. With optional JUSTIFY, justify cell as well. IGNORE-COLLAPSED-P is used when caller has already expanded cell, indicating it is not collapsed.
kotl-mode:fill-paragraph {C-x f}
Fill current paragraph within cell. With optional JUSTIFY, justify paragraph as well. Ignore any non-nil no-fill attribute attached to the cell.
kotl-mode:fill-tree {C-M-j}
Refill each cell within the tree whose root is at point.
kotl-mode:first-sibling {C-c <}
Move point to the first sibling of the present cell. Leave point at the start of the cell or at its present position if it is already within the first sibling cell.
kotl-mode:fkey-backward-char {left}
Move point backward ARG (or 1) characters and return point.
kotl-mode:fkey-forward-char {right}
Move point forward ARG (or 1) characters and return point.
kotl-mode:fkey-next-line {down}
Move point to ARGth next line and return point.
kotl-mode:fkey-previous-line {up}
Move point to ARGth previous line and return point.
kotl-mode:forward-cell {C-c C-f}
Move to prefix ARGth following cell (same level) within current view. Return number of cells left to move.
kotl-mode:forward-char {C-f}
Move point forward ARG (or 1) characters and return point.
kotl-mode:forward-para {M-n}
Move to prefix ARGth next cell (any level) within current view.
kotl-mode:forward-paragraph {M-]}
Move to prefix ARGth next cell (any level) within current view.
kotl-mode:forward-sentence {M-e}
Move point forward ARG (or 1) sentences and return point.
kotl-mode:forward-word {M-f}
Move point forward ARG (or 1) words and return point.
kotl-mode:goto-cell {C-c g}
Move point to start of cell given by CELL-REF. (See ’kcell:ref-to-id’.) Return point iff CELL-REF is found within current view. With a prefix argument, CELL-REF is assigned the argument value for use as an idstamp.
Optional second arg, ERROR-P, non-nil means signal an error if CELL-REF is not found within current view. Will signal same error if called interactively when CELL-REF is not found.
kotl-mode:hide-sublevels {C-X $}
Hide all cells in outline at levels deeper than LEVELS-TO-KEEP (a number). Shows any hidden cells within LEVELS-TO-KEEP. 1 is the first level.
kotl-mode:hide-subtree {C-M-h}
Hide subtree, ignoring root, at optional CELL-REF (defaults to cell at point).
kotl-mode:hide-tree {C-c BS}
Collapse kotl rooted at optional CELL-REF (defaults to cell at point).
kotl-mode:insert-register {C-c r i}
Insert contents of register REGISTER at point in current cell. REGISTER is a character naming the register to insert. Normally puts point before and mark after the inserted text. If optional second arg is non-nil, puts mark before and point after. Interactively, second arg is non-nil if prefix arg is supplied.
kotl-mode:just-one-space {M-\}
Delete all spaces and tabs around point and leave one space.
kotl-mode:kcell-help {C-c h}
Display a temporary buffer with CELL-REF’s properties. CELL-REF defaults to current cell. Optional prefix arg CELLS-FLAG selects the cells to print: If = 1, print CELL-REF’s cell only; If > 1, print CELL-REF’s visible kotl (the tree rooted at CELL-REF); If < 1, print all visible cells in current view (CELL-REF is not used).
See also the documentation for kotl-mode:properties
.
kotl-mode:kill-contents {C-c k}
Kill contents of cell from point to cell end. With prefix ARG, kill entire cell contents.
kotl-mode:kill-line {C-k}
Kill ARG lines from point.
kotl-mode:kill-region {C-w}
Kill region between START and END within a single kcell. With optional COPY-P equal to ’t, copy region to kill ring but does not kill it. With COPY-P any other non-nil value, return region as a string without affecting kill ring.
If the buffer is read-only and COPY-P is nil, the region will not be deleted but it will be copied to the kill ring and then an error will be signaled.
kotl-mode:kill-ring-save {M-w}
Copy region between START and END within a single kcell to kill ring.
kotl-mode:kill-sentence {M-k}
Kill up to prefix ARG (or 1) sentences following point within a single cell.
kotl-mode:kill-tree {C-c C-k}
Kill ARG following trees starting with tree rooted at point. If ARG is not a non-positive number, nothing is done.
kotl-mode:kill-word {M-d}
Kill up to prefix ARG words following point within a single cell.
kotl-mode:last-sibling {C-c >}
Move point to the last sibling of the present cell. Leave point at the start of the cell or at its present position if it is already within the last sibling cell.
kotl-mode:mail-tree {C-c @}
Mail outline tree rooted at CELL-REF. Use "0" for whole outline buffer.
kotl-mode:move-after {C-c m}
Move tree rooted at FROM-CELL-REF to follow tree rooted at TO-CELL-REF. If prefix arg CHILD-P is non-nil, make FROM-CELL-REF the first child of TO-CELL-REF, otherwise make it the sibling following TO-CELL-REF. With optional COPY-P, copies tree rather than moving it.
Leave point at original location but return the tree’s new start point.
kotl-mode:move-before {C-c RET}
Move tree rooted at FROM-CELL-REF to precede tree rooted at TO-CELL-REF. If prefix arg PARENT-P is non-nil, make FROM-CELL-REF the first child of TO-CELL-REF’s parent, otherwise make it the preceding sibling of TO-CELL-REF. With optional COPY-P, copies tree rather than moving it.
Leave point at original location but return the tree’s new start point.
kotl-mode:newline {RET}
Insert a newline. With ARG, insert ARG newlines. In Auto Fill mode, if no numeric arg, break the preceding line if it is too long.
kotl-mode:next-cell {C-c C-n}
Move to prefix ARGth next cell (any level) within current view.
kotl-mode:next-line {C-n}
Move point to ARGth next line and return point.
kotl-mode:open-line {C-o}
Insert a newline and leave point before it. With arg N, insert N newlines.
kotl-mode:overview {C-c C-o}
Show only the first line of each cell in the current outline.
kotl-mode:previous-cell {C-c C-p}
Move to prefix ARGth previous cell (any level) within current view.
kotl-mode:previous-line {C-p}
Move point to ARGth previous line and return point.
kotl-mode:promote-tree {M-TAB}
Move current kotl a maximum of prefix ARG levels higher in current view. Each cell is refilled iff its no-fill attribute is nil and kotl-mode:refill-flag is non-nil. With prefix ARG = 0, cells are promoted up to one level and kotl-mode:refill-flag is treated as true.
kotl-mode:scroll-down {M-v}
Scroll text of current window downward ARG lines; or a windowful if no ARG.
kotl-mode:scroll-up {C-v}
Scroll text of current window upward ARG lines; or a windowful if no ARG.
kotl-mode:set-fill-prefix {C-x l}
Sets fill prefix to line up to point. With prefix arg TURN-OFF or at begin of line, turns fill prefix off.
kotl-mode:show-all {C-c C-a}
Show (expand) all cells in current view.
kotl-mode:show-subtree
Show subtree, ignoring root, at optional CELL-REF (defaults to cell at point).
kotl-mode:show-tree {C-c C-s}
Display fully expanded kotl rooted at CELL-REF.
kotl-mode:split-cell {C-c s}
Split cell into two cells and move to new cell. Cell contents after point become part of newly created cell. Default is to create new cell as sibling of current cell. With optional universal ARG, C-u, new cell is added as child of current cell.
kotl-mode:top-cells {C-c C-t}
Collapse all level 1 cells in view, thereby hiding any sublevels.
kotl-mode:transpose-cells {C-c t}
Exchange current and previous visible cells, leaving point after both. If no previous cell, exchange current with next cell. With prefix ARG, take current cell and move it past ARG cells. With prefix ARG = 0, interchange the cell that contains point with the cell that contains mark.
kotl-mode:transpose-chars {C-t}
Interchange characters around point, moving forward one character. With prefix ARG, take character before point and drag it forward past ARG other characters (backward if ARG negative). If no prefix ARG and at end of line, the previous two characters are exchanged.
kotl-mode:transpose-lines {C-x C-t}
Exchange current line and previous line, leaving point after both. If no previous line, exchange current with next line. With prefix ARG, take previous line and move it past ARG lines. With prefix ARG = 0, interchange the line that contains point with the line that contains mark.
kotl-mode:transpose-words {M-t}
Interchange words around point, leaving point after both words. With prefix ARG, take word before or around point and drag it forward past ARG other words (backward if ARG negative). If ARG is zero, the words around or after point and around or after mark are interchanged.
kotl-mode:up-level {C-c u}
Move up prefix ARG levels higher in current outline view.
kotl-mode:yank {C-y}
Reinsert the last stretch of killed text. More precisely, reinsert the stretch of killed text most recently killed OR yanked. Put point at end, and set mark at beginning. With just C-u as argument, same but put point at beginning (and mark at end). With argument N, reinsert the Nth most recently killed stretch of killed text. See also the command M-x kotl-mode:yank-pop.
kotl-mode:yank-pop {M-y}
Replace just-yanked stretch of killed text with a different stretch.
This command is allowed only immediately after a yank
or a
yank-pop
. At such a time, the region contains a stretch of
reinserted previously-killed text. yank-pop
deletes that text
and inserts in its place a different stretch of killed text.
With no argument, the previous kill is inserted. With argument N, insert the Nth previous kill. If N is negative, this is a more recent kill.
The sequence of kills wraps around, so that after the oldest one comes the newest one.
kotl-mode:zap-to-char {M-z}
Kill up to and including prefix ARG’th occurrence of CHAR. Goes backward if ARG is negative; error if CHAR not found.
kview:set-label-type {C-c C-l}
Change kview’s label display type to NEW-TYPE, updating all displayed labels. See documentation for variable, kview:default-label-type, for valid values of NEW-TYPE.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
See the ‘README’ file in the Hyperbole distribution directory for complete details on Hyperbole mailing lists and how to subscribe.
If you find any errors in Hyperbole’s operation or documentation, feel free to report them to the Hyperbole discussion list: hyperbole@hub.ucsb.edu. Be sure to use the Msg/Compose-Hypb-Mail menu item whenever you send a message to the mail list since it will insert important system version information for you.
If you use Hyperbole mail or news support, see section Buttons in Mail, a click with your Action Key on the Hyperbole mail list address will insert a description of your Hyperbole configuration information into your outgoing message, so that you do not have to type it. This is useful when composing a reply for the Hyperbole mail list. Otherwise, be sure to include your Emacs, Hyperbole and window system versions in your message. Your Hyperbole version number can be found in the top-level Hyperbole menu.
Please use your subject line to state the position that your message takes on the topic that it addresses, e.g. send "Subject: Basic bug in top-level Hyperbole menu." rather than "Subject: Hyperbole bug". This simple rule makes all e-mail communication much easier.
If you have suggestions on how to improve Hyperbole, send them to the same address. Here are some issues you might address:
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Since the Smart Mouse Keys are set up for use under five different Emacs
configurations, there is no easy way to provide user level
customization. Therefore, to change the bindings you must edit the
hmouse-setup
and hmouse-get-bindings
functions in the
‘hmouse-key.el’ file.
The hkey-alist and hmouse-alist variable settings in ‘hui-mouse.el’ and ‘hui-window.el’ must be altered if you want to change what the Smart Keys do in particular contexts. You should then update the Smart Key summary documentation in the file, ‘hypb-mouse.txt’, and potentially the same summary in this manual.
What if someone sends a mail message with a button for which I do not have the action type? Or a button whose link referent I can’t access?
You receive an error that an action type is not defined or a link referent is not accessible/readable if you try to use the button. This is hardly different than trying to get through a locked door without a key; you try the doorknob, find that it is locked, and then realize that you need to take a different approach or else give up.
Like all communication, people need to coordinate, which usually requires an iterative process. If you get a mail message with a button for which you don’t have the action type, you mail the sender and request it.
Rather than typing the name for each, it is quicker to jump to the global button file and edit the buttons there as you would any explicit buttons. By default, the ButFile/PersonalFile menu item takes you to the file where global buttons are saved.
When you think of a hyper-space that you depend on every day, you don’t want to have a single point of failure make you incapable of doing work. With Hyperbole, if some directories become unavailable for a particular time (e.g. the filesystems on which they reside are dismounted) you can still work elsewhere with minimal effect. We believe this to be a compelling factor to leave the design with external button data storage.
This design also permits the potential addition of buttons to read-only media.
Any category of button can make use of an action type. Some action types are useful as behavior definitions for a variety of button categories, so all action types are defined separately to give them independence from those types which apply them.
For implicit button types that require a lot of code, it is useful to add a module that includes the implicit button type definition, its action type definition and supporting code.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
This appendix is included for a number of reasons:
Note that due to a lack of volunteers to further develop Hyperbole, much of this work may not be done. So if you want to see these features, encourage qualified people to volunteer.
There is as yet no means of transferring buttons among buffers. We realize this is a critical need. Users should be able to manipulate text with embedded buttons in ordinary ways. Some of the Emacs Lisp functions that must be augmented to support this behavior include:
(append-next-kill) (copy-region-as-kill) (delete-region) ;; a built-in function (get-register) (insert-buffer-substring) ;; a built-in function (insert-file-contents) (kill-append) (kill-region) (rotate-yank-pointer) (set-register) (view-register) (yank) (yank-pop) (undo)
Trails are an extension to the basic history mechanism presently offered by Hyperbole. Trails will allow a user to capture, edit and store a specific sequence and set of views of information for later replay by other users. Conditional branching may also be supported.
The current design choice of storing buttons external to the source file was made under the assumption that people should be able to look at files that contain Hyperbole buttons with any standard editor or tool and not be bothered by the ugly button data (since they won’t be able to utilize the buttons anyway, they don’t need to see or have access to them).
In many contexts, embedding the button data within the source files may be a better choice, so a provision which would allow selection of either configuration may be added. Here are some of the PROs and CONs of both design choices:
POSITIVE NEGATIVE Bdata in src file Documents can stand alone. All edit operators have Normal file operations apply. to account for file structure and hide Simplifies creation and internal components. facility expansion for structured and multi-media files. Bdata external to src file Files can be displayed and Currently, bdata for printed exactly as they look. whole directory is No special display formatting locked when any bdata is necessary. entry is locked. Button-based searches and database-type lookup operations need only search one file per directory.
This will allow one to create buttons more flexibly. For example, button attributes could be given in any order. Entry of long code sequences, quick note taking and cross-referencing would also be made easier.
From the early stages of Hyperbole design, collaborative work environments have been considered. A simple facility has demonstrated broadcast of button activations to a number of workstations on a local area network, so that one user can lead others around an information space, as during an online design review. (This facility is not yet ready for release.) We shall do some work in specific collaborative mechanisms, but also expect that others who concentrate in collaborative work will provide more extensive capabilities.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Akscyn, R. M., D. L. McCracken and E. A. Yoder. KMS: A Distributed Hypermedia System for Managing Knowledge in Organizations. Communications of the ACM, Vol. 31, No. 7, July 1988, pp. 820-835.
Brown, P. J. Turning Ideas into Products: The Guide System. Proceedings of Hypertext ’87, November 13-15, 1987, Chapel Hill, NC. ACM: NY, NY, pp. 33-40.
Conklin, Jeff. Hypertext: An Introduction and Survey. IEEE Computer, Vol. 20, No. 9, September 1987, pp. 17-41.
Engelbart, D., and W. English. A research center for augmenting human intellect. Proceedings of the Fall Joint Computer Conference, 33, 1, AFIPS Press: Montvale, NJ, 1968, pp. 395-410.
Engelbart, D. C. Authorship Provisions in Augment. Proceedings of the 1984 COMPCON Conference (COMPCON ’84 Digest), February 27-March 1, 1984, San Francisco, CA. IEEE Computer Society Press, Spring, 1984. 465-472. (OAD,2250,)
Engelbart, D. C. Collaboration Support Provisions in Augment. Proceedings of the AFIPS Office Automation Conference (OAC ’84 Digest), February, 1984, Los Angeles, CA, 1984. 51-58. (OAD,2221,)
Foss, C. L. Effective Browsing in Hypertext Systems. Proceedings of the Conference on User-Oriented Content-Based Text and Image Handling (RIAO 88), March 21-24, MIT, Cambridge MA. Centre de Hautes Etudes Internationales d’Informatique Documentaire, 1988, pp. 82-98.
Garrett, N., K. E. Smith and N. Meyrowitz. Intermedia: Issues, Strategies, and Tactics in the Design of a Hypermedia Document System. Computer-Supported Cooperative Work (CSCW ’86) Proceedings, December 3-5, Austin, TX, 1986, pp. 163-174.
Halasz, F. G., T. P. Moran and R. H. Trigg. NoteCards in a Nutshell. Proceedings of the CHI and GI ’87 Conference on Human Factors in Computing Systems, Toronto, J. M. Carroll and P. P. Tanner, (editors), ACM: NY, NY, April 1987, pp. 45-52.
Harvey, G. Understanding HyperCard. Alameda, CA: SYBEX, Inc., 1988.
Kaplan, S., A. M. Carroll, C. Love and D. M. LaLiberte. Epoch 4.0 Manual. Department of Computer Science, University of Illinois, Urbana-Champaign, March 1992.
Kaplan, S. J., M. D. Kapor, E. J. Belove, R. A. Landsman, and T. R. Drake. AGENDA: A personal Information Manager. Communications of the ACM, No. 33, July 1990, pp. 105-116.
Nelson, T. H. Computer Lib/Dream Machines. MicroSoft Press, Redmond, WA, 1987.
Nelson, T. H. Literary Machines, Edition 87.1. Available from the Distributors, 702 South Michigan, South Bend, IN 46618, 1987.
Norman, D. A. and S. W. Draper, editors. User Centered System Design. Lawrence Erlbaum Associates: Hillsdale, New Jersey, 1986.
Shneiderman, B. The future of interactive systems and the emergence of direct manipulation. Behavior and Information Technology, Vol. 1, 1982, pp. 237-256.
Stallman, R. GNU Emacs Manual. Free Software Foundation, Cambridge: MA, March 1987.
Trigg, R., L. Suchman, and F. Halasz. Supporting collaboration in NoteCards. Proceedings of the CSCW ’86 Conference, Austin, TX, December 1986, pp. 147-153.
Trigg, R. H., T. P. Moran and F. G. Halasz. Adaptability and Tailorability in NoteCards. Proceedings of INTERACT ’87, Stuttgart, West Germany, September 1987.
Weiner, B. PIEmail: A Personalized Information Environment Mail Tool. Department of Computer Science Masters Project, Brown University: Providence, RI, May 10, 1992.
Yankelovich, N., B. J. Haan, N. Meyrowitz and S. M. Drucker. Intermedia: The Concept and the Construction of a Seamless Information Environment. IEEE Computer, Vol. 21, No. 1, January 1988, pp. 81-96.
Yoder, E. A., R. M. Akscyn and D. L. McCracken. Collaboration in KMS, A Shared Hypermedia System. Proceedings of the 1989 ACM Conference on Human Factors in Computer Systems (CHI ’89), April 30-May 4, 1989, Austin, TX, ACM: NY,NY, 1989, pp. 37-42.
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Jump to: | ,
.
<
>
A B C D F H M N P Q R S T U |
---|
Jump to: | ,
.
<
>
A B C D F H M N P Q R S T U |
---|
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Jump to: | A C D E F G H I K M R S W |
---|
Jump to: | A C D E F G H I K M R S W |
---|
[ << ] | [ < ] | [ Up ] | [ > ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Jump to: | A B C D E F G H I K L M N O P R S T U V W X |
---|
Jump to: | A B C D E F G H I K L M N O P R S T U V W X |
---|
[Top] | [Contents] | [Index] | [ ? ] |
[Top] | [Contents] | [Index] | [ ? ] |
[Top] | [Contents] | [Index] | [ ? ] |
This document was generated on December 6, 2024 using texi2html 5.0.
The buttons in the navigation panels have the following meaning:
Button | Name | Go to | From 1.2.3 go to |
---|---|---|---|
[ << ] | FastBack | Beginning of this chapter or previous chapter | 1 |
[ < ] | Back | Previous section in reading order | 1.2.2 |
[ Up ] | Up | Up section | 1.2 |
[ > ] | Forward | Next section in reading order | 1.2.4 |
[ >> ] | FastForward | Next chapter | 2 |
[Top] | Top | Cover (top) of document | |
[Contents] | Contents | Table of contents | |
[Index] | Index | Index | |
[ ? ] | About | About (help) |
where the Example assumes that the current position is at Subsubsection One-Two-Three of a document of the following structure:
This document was generated on December 6, 2024 using texi2html 5.0.