@style(spacing 1 lines, margins .7 inches, indent 8 chars, font Times10) @Define(goodstuff, fill n, font Helvetica10) @Define(col3, columns 3, fill n) @style(indent 8 chars, spacing 1 lines, BindingMargin .5 inches) @comment(This manual is normally generated using Mark of the Unicorn's Final Word II product. Alternativily, Borland's SPRINT (derived from Final Word II) should also build it properly. Both of these formatters are based on the SCRIBE language originally used on the DECSystem-20, and will form the basis of a shareware text formatter I hope, but not promise, to complete some day.) @pagefooting() @big(@b[MicroEMACS]) Full Screen Text Editor Reference Manual @i(- Preliminary Version) Version 3.12 April 22, 1993 (C)Copyright 1988 - 1993 by Daniel M. Lawrence Reference Manual (C)opyright 1988 - 1993 by Brian Straight and Daniel M. Lawrence All Rights Reserved @i[(C)Copyright 1988 - 1993 by Daniel M. Lawrence MicroEMACS 3.12 can be copied and distributed freely for any non-commercial purposes. Commercial users may use MicroEMACS 3.12 inhouse. Shareware distributors may redistribute MicroEMACS 3.12 for media costs only. MicroEMACS 3.12 can only be incorporated into commercial software or resold with the permission of the current author.] @newpage @flushleft(@big[Introduction]) MicroEMACS is a tool for creating and changing documents, programs, and other text files. It is both relatively easy for the novice to use, but also very powerful in the hands of an expert. MicroEMACS can be extensively customized for the needs of the individual user. MicroEMACS allows several files to be edited at the same time. The screen can be split into different windows and screens, and text may be moved freely from one window on any screen to the next. Depending on the type of file being edited, MicroEMACS can change how it behaves to make editing simple. Editing standard text files, program files and word processing documents are all possible at the same time. There are extensive capabilities to make word processing and editing easier. These include commands for string searching and replacing, paragraph reformatting and deleting, automatic word wrapping, word move and deletes, easy case controlling, and automatic word counts. For complex and repetitive editing tasks editing macroes can be written. These macroes allow the user a great degree of flexibility in determining how MicroEMACS behaves. Also, any and all the commands can be used by any keystroke by changing, or rebinding, what commands various keys invoke. Special features are also available to perform a diverse set of operations such as file encryption, automatic backup file generation, entabbing and detabbing lines, executing operating system commands and filtering of text through other programs (like SORT to allow sorting text). @flushleft(@big[History]) EMACS was originally a text editor written by Richard Stallman at MIT in the early 1970s for Digital Equipment computers. Various versions, rewrites and clones have made an appearance since. This version of MicroEMACS is derived from code written by Dave G. Conroy in 1985. Later modifications were performed by Steve Wilhite and George Jones. In December of 1985 Daniel Lawrence picked up the then current source (version 2.0) and made extensive modifications and additions to it over the course of the next eight years. Updates and support for the current version are still available. Commercial support and usage and resale licences are also available. The current program author can be contacted by writing to: @begin(verbatim) USMAIL: Daniel Lawrence 617 New York St Lafayette, IN 47901 UUCP: pur-ee!mdbs!dan ARPA: mdbs!dan@@ee.ecn.purdue.edu Support is provided through: The Programmer's Room Opus 201/10 300/1200/2400 and 9600 (Hayes V series only) (317) 742-5533 no parity 8 databits no stop bits @end(verbatim) @newpage @flushleft(@big[Credits]) Many people have been involved in creating this software and we wish to credit some of them here. Dave Conroy, of course, wrote the very first version of MicroEMACS, and it is a credit to his clean coding that so much work was able to be done to expand it. John Gamble is responsible for writing the MAGIC mode search routines, and for maintaining all the search code. Jeff Lomicka wrote the appendix on DEC VMS and has supplied a lot of code to support VMS and the ATARI 1040ST versions. Curtis Smith wrote the original VMS code and help support the Commodore AMIGA. Also Lance Jones has done a lot of work on the AMIGA code. Professor Suresh Konda at Carnegie Mellon University has put a lot of effort into writing complex macroes and finding all the bugs in the macro language before anyone else does. A special thanks to Dana Hoggatt who has provided an almost daily sounding board for ideas, algorythms and code. He is responsible for the encryption code directly and has prodded me into adding many features with simple but poignant questions (Dan? How do we move the upper left corner of the screen? . . . which forced me to write the text windowing system). Pierre Perrot dealt with my restrictive path to a generalized windowing version, and produced an excellent version for MicroSoft Windows. He continues to assist with this environment, forcing me to keep pace with him, making all the version more interesting. As to people sending source code and text translations over computer networks like USENET and ARPA net, there are simply more than can be listed here. [The comments in the edit history in the history.c file mention each and the piece they contributed]. All these people should be thanked for the hard work they have put into MicroEMACS. @closing(Daniel M. Lawrence) @newpage @string(ChapterTitle="") @set(page = 0) @pageheading(odd, left "@title[Chapter]", right "MicroEMACS Reference Manual") @pageheading(even, left "MicroEMACS Reference Manual", right "@title[Chapter]") @pagefooting(odd, left="@value(page)") @pagefooting(even, right="@value(page)") @chapter(Installation) MicroEMACS is a programmer's text editor which is very powerfull, customizable, and exists for a large number of different types of computer systems. It is particularly usefull for people who work on a lot of different computers and want to have a familiar and powerful editor which works identically no matter what computer they are using. But before using MicroEMACS, you must INSTALL it on your computer system. Since each computer is different, there is usually a different way to install MicroEMACS for each type of computer. @section(MSDOS - IBM-PCs) @section(UNIX) @section(ATARI ST) @chapter(Basic Concepts) The current version of MicroEMACS is 3.12 (Third major re-write, twelveth public release), and for the rest of this document, we shall simply refer to this version as "EMACS". Any modifications for later versions will be in the file README on the MicroEMACS distribution disk. @section(Keys and the Keyboard) Many times throughout this manual we will be talking about @index(special keys) commands and the keys on the keyboard needed to use them. There are a number of "special" keys which can be used and are listed here: @begin(description) @\NewLine which is also called RETURN, ENTER, or , this key is used to @index(newline) end different commands. ^@\The control key can be used before any alphabetic character and some symbols. For example, ^C means to hold down the key and type @index(control key) the C key at the same time. ^X@\The CONTROL-X key is used at the beginning of many different @index(control-x) commands. META or M-@\This is a special EMACS key used to begin many commands. @index(meta key)This key is pressed and then released before typing the next character. On most systems, this is the key, but it can be changed. (consult appendix E to learn what key is used for META on your computer). @end(description) Whenever a command is described, the manual will list the actual keystrokes needed to execute it in @b(boldface) using the above conventions, and also the name of the command in @i(italics). @section(Getting Started) In order to use EMACS, you must call it up from your system or computer's command prompt. On UNIX and MSDOS machines, just type "emacs" from the command prompt and follow it with the or key (we will refer to this key as for "new-line" for the remainder of this manual). On the Macintosh, the Amiga, the ATARI ST, and under OS/2 and other icon based operating systems, double click on the uEMACS icon. Shortly after this, a screen similar to the one below should appear. @section(Parts and Pieces) The screen is divided into a number of areas or @b. On some systems the top window contains a function list of unshifted and @index(windows) shifted function keys. We will discuss these keys later. @index(mode line) Below them is an EMACS @b which, as we will see, informs you of the present mode of operation of the editor--for example "(WRAP)" if you set EMACS to wrap at the end of each line. @index(text window) Under the mode line is the @b where text appears and is manipulated. Since each window has its own mode line, below the text window is it's mode line. The last line of the screen is the @b(command line) where EMACS takes commands and reports on what it is doing. @begin(verbatim) =============================================================================== f1 search-> f2 <-search | MicroEMACS: Text Editor f3 hunt-> f4 <-hunt | f5 fkeys f6 help | Available function key Pages include: f7 nxt wind f8 pg[ ] | WORD BOX EMACS PASCAL C cObal Lisp f9 save f10 exit | [use the f8 key to load Pages] =============================================================================== MicroEMACS 3.12 () Function Keys =============================================================================== =============================================================================== ---- MicroEMACS 3.12 () -- Main ----------------------------------------------- =============================================================================== Fig 1: EMACS screen on an IBM-PC @end(verbatim) @section(Entering Text) Entering text in EMACS is simple. Type the following sentence fragment: @quotation @flushleft(The text is displayed at the top of the text window. Now type:) @quotation Notice that some of your text has dissapeared off the left side of the screen. Don't panic--your text is safe!!! You've just discovered that EMACS doesn't "wrap" text to the next line like most word processors unless you hit . But since EMACS is used for both word processing, and text editing, it has a bit of a dual personality. You can change @index(modes) the way it works by setting various @b(modes). In this case, you need to set @b(WRAP) mode, using the @i(add-mode) @index(add-mode) command, by typing @b(^XM). The command line at the base of the screen will prompt you for the mode you wish to add. Type @b followed by the key and any text you now enter will be wrapped. However, the command doesn't wrap text already entered. To get rid of the long line, press and hold down the key until the line is gone. Now type in the words you deleted, watch how EMACS goes down to the next line at the right time. @i{(In some versions of EMACS, @b is a default mode in which case you don't have to worry about the instructions relating to adding this mode.)} Now let's type a longer insert. Hit a couple of times to tab down from the text you just entered. Now type the following paragraphs. Press twice to indicate a paragraph break. @quotation @section Now let's practice moving around in this text. To move the cursor back to the word "Winding," enter @b @i(previous-word) @index(previous-word). This command moves the cursor backwards by one word at a time. Note you have to press the key combination every time the cursor steps back by one word. Continuously pressing META and toggling B produces an error message. To move forward to the word "stairs" enter @b @i(next-word)@index(next-word), which moves the cursor forward by one word at a time. Notice that EMACS commands are usually mnemonic--F for forward, B for backward, for example. To move the cursor up one line, enter @b<^P> @i(previous-line) @index(previous-line), down one line @b<^N> @i(next-line) @index(next-line). Practice this movement by moving the cursor to the word "terrifying" in the second line. The cursor may also be moved forward or backward in smaller increments. To move forward by one character, enter @b<^F> @i(forward-character) @index(forward-character), to move backward, @b<^B> @i(backward-character) @index(backward-character). EMACS also allows you to specify a number which is normally used to tell a command to execute many times. To repeat most commands, press META and then the number before you enter the command. Thus, the command META 5 ^F (@b) will move the cursor forward by five characters. Try moving around in the text by using these commands. For extra practice, see how close you can come to the word "small" in the first paragraph by giving an argument to the commands listed here. Two other simple cursor commands that are useful to help us move around in the text are @b @i(next-paragraph) @index(next-paragraph) which moves the cursor to the second paragraph, and @b @i(previous-paragraph) @index(previous-paragraph) which moves it back to the previous paragraph. The cursor may also be moved rapidly from one end of the line to the other. Move the cursor to the word "few" in the second line. Press @b<^A> @i(beginning-of-line) @index(beginning-of-line). Notice the cursor moves to the word "events" at the beginning of the line. Pressing @b<^E> @i(end-of-line) @index(end-of-line) moves the cursor to the end of the line. Finally, the cursor may be moved from any point in the file to the end or beginning of the file. Entering @b{M->} @i(end-of-file) @index(end-of-file) moves the cursor to the end of the buffer, @b{M-<} @i(beginning-of-file) @index(beginning-of-file) to the first character of the file. @i(On the IBM-PC, the ATARI ST and many other machines, the cursor keys @index(cursor keys) can also be used to move the cursor.) Practice moving the cursor in the text until you are comfortable with the commands we've explored in this chapter. @section(Saving your text) When you've finished practicing cursor movement, save your file. Your @index(buffer) file currently resides in a @b. The buffer is a temporary storage area for your text, and is lost when the computer is turned off. You can save the buffer to a file by entering @b<^X^S> @i(save-file) @index(save-file). Notice that EMACS informs you that your file has no name and will not let you save it. To save your buffer to a file with a different name than it's current one (which is empty), press @b<^X^W> @i(write-file) @index(write-file). EMACS will prompt you for the filename you wish to write. Enter the name @b and press return. On a micro, the drive light will come on, and EMACS will inform you it is writing the file. When it finishes, it will inform you of the number of lines it has written to the disk. Congratulations!! You've just saved your first EMACS file! @newpage @heading(Chapter @value(chapter) Summary) In chapter @value(chapter), you learned how to enter text, how to use wrap mode, how to move the cursor, and to save a buffer. The following is a table of the commands covered in this chapter and their corresponding key bindings: @begin{verbatim} @u(Key Binding Keystroke Effect) abort-command @b<^G> aborts current command add-mode @b<^XM> allows addition of EMACS mode such as @b(WRAP) backward-character @b<^B> moves cursor left one character beginning-of-file @b{M-<} moves cursor to beginning of file beginning-of-line @b<^A> moves cursor to beginning of line end-of-file @b{M->} moves cursor to end of file end-of-line @b<^E> moves cursor to end of line forward-character @b<^F> moves cursor right one character next-line @b<^N> moves cursor to next line next-paragraph @b moves cursor to next paragraph next-word @b moves cursor forward one word previous-line @b<^P> moves cursor backward by one line previous-paragraph @b moves cursor to previous paragraph previous-word @b moves cursor backward by one word save-file @b<^X^S> saves current buffer to a file write-file @b<^X^W> save current buffer under a new name @end(verbatim) @newpage @chapter(Basic Editing--Simple Insertions and Deletions) @section In the first chapter, you learned how to create and save a file in EMACS. Let's do some more editing on this file. Call up emacs by typing in the following command. @b @i(On icon oriented systems, double click on the uEMACS icon, usually a file dialog box of some sort will appear. Choose @b(FANG.TXT) from the appropriate folder.) Shortly after you invoke EMACS, the text should appear on the screen ready for you to edit. The text you are looking at currently resides in a @b. A buffer is a temporary area of computer memory which is @index(buffer) the primary unit internal to EMACS -- this is the place where EMACS goes to work. The mode line at the bottom of the screen lists the buffer name, @b and the name of the file with which this buffer is associated, @b The computer talks to you through the use of its @b(screen). This @index(screen) screen usually has an area of 24 lines each of 80 characters across. You can use EMACS to subdivide the screen into several separate work areas, or @b(windows), each of which can be @index(window) 'looking into' different files or sections of text. Using windows, you can work on several related texts at one time, copying and moving blocks of text between windows with ease. To keep track of what you are editing, each window is identified by a @b(mode line) on the @index(mode line) @index(buffer) last line of the window which lists the name of the @b(buffer) which it is looking into, the file from which the text was read, and how the text is being edited. An EMACS @b tells EMACS how to deal with user input. As we have already seen, the mode 'WRAP' controls how EMACS deals with long lines (lines with over 79 characters) while the user is typing them in. The 'VIEW' mode, allows you to read a file without modifying it. Modes are associated with buffers and not with files; hence, a mode needs to be explicitly set or removed every time you edit a file. A new file read into a buffer with a previously specified mode will be edited under this mode. If you use specific modes frequently, EMACS allows you to set the modes which are used by all new buffers, called @b modes. @section Your previously-saved text should look like this: @quotation Let's assume you want to add a sentence in the second paragraph after the word "base." Move the cursor until it is on the "W" of "Winding". Now type the following: @quotation If the line fails to wrap and you end up with a '$' sign in the right margin, just enter @b{M-Q} @i(fill-paragraph) @index(fill-paragraph) to reformat the paragraph. This new command attempts to fill out a paragraph. Long lines are divided up, and words are shuffled around to make the paragraph look nicer. Notice that all visible EMACS characters are self-inserting -- all you had to do was type the characters to insert and the existing text made space for it. With a few exceptions discussed later, all non-printing characters (such as control or escape sequences) are commands. To insert spaces, simply use the space bar. Now move to the first line of the file and type @b{^O} @i(open-line) @index(open-line) (Oh, not zero). You've just learned how to insert a blank line in your text. @section EMACS offers a number of deletion options. For example, move the cursor until it's under the period at the end of the insertion you just did. Press the backspace key. Notice the "n" on "lantern" disappeared. The backspace implemented on EMACS is called a @b backspace--it removes text immediately before the current cursor position from the buffer. Now type @b<^H> @i(delete-previous-character) @index(delete-previous-character). Notice that the cursor moves back and obliterates the "r"--either command will backspace the cursor. Type in the two letters you erased to restore your text and move the cursor to the beginning of the buffer @b{M->} @i(beginning-of-file) @index(beginning-of-file). Move the cursor down one line to the beginning of the first paragraph. To delete the forward character, type @b{^D} @i(delete-next-character) @index(delete-next-character). The "F" of "Fang" disappears. Continue to type @b{^D} until the whole word is erased EMACS also permits the deletion of larger elements of text. Move the cursor to the word "center" in the first line of text. Pressing @b{M-} @i(delete-previous-word) @index(delete-previous-word) kills the word immediately before the cursor. @b{M-^H} has the same effect. Notice that the commands are very similar to the control commands you used to delete individual letters. As a general rule in EMACS, control sequences affect small areas of text, META sequences larger areas. The word forward of the cursor position can therefore be deleted by typing @b{M-D} @i(delete-next-word) @index(delete-next-word). Now let's take out the remainder of the first line by typing @b{^K} @i(kill-to-end-of-line) @index(kill-to-end-of-line). You now have a blank line at the top of your screen. Typing @b{^K} again or @b{^X^O} @i(delete-blank-lines) @index(delete-blank-lines) deletes the blank line and flushes the second line to the top of the text. Now exit EMACS by typing @b{^X^C} @i(exit-emacs) @index(exit-emacs). Notice EMACS reminds you that you have not saved your buffer. Ignore the warning and exit. This way you can exit EMACS without saving any of the changes you just made. @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter), you learned about the basic 'building blocks' of an EMACS text file--buffers, windows, and files. @begin{verbatim} @u(Key binding Keystroke Effect) delete-previous-character @b{^H} deletes character immediately before the current cursor position delete-next-character @b{^D} deletes character immediately after current cursor position delete-previous-word @b{M-^H} deletes word immediately before current cursor position delete-next-word @b{M-D} deletes word immediately after current cursor position kill-to-end-of-line @b<^K> deletes from current cursor position to end of line insert-space @b<^C> inserts a space to right of cursor open-line @b{^O} inserts blank line delete-blank-lines @b{^X^O} removes blank line exit-emacs @b{^X^C} exits emacs @end(verbatim) @chapter(Using Regions) @section(Defining and Deleting a Region) At this point its time to familiarize ourselves with two more EMACS terms--the @b and the @b. The point is located directly @index(point) @index(mark) behind the current cursor position. The mark (as we shall see shortly) is user defined. These two elements together are called the current @b(region) and limit the @b of text on which EMACS performs many of its editing functions. Let's begin by entering some new text. Don't forget to add @b(wrap) mode if its not set on this buffer. Start EMACS and open a file called @b{PUBLISH.TXT}. Type in the following text: @quotation{One of the largest growth areas in personal computing is electronic publishing. There are packages available for practically every machine from elegantly simple programs for the humble Commodore 64 to sophisticated professional packages for PC and Macintosh computers. Electronic publishing is as revolutionary in its way as the Gutenburg press. Whereas the printing press allowed the mass production and distribution of the written word, electronic publishing puts the means of production in the hands of nearly every individual. From the class magazine to the corporate report, electronic publishing is changing the way we produce and disseminate information. Personal publishing greatly increases the utility of practically every computer. Thousands of people who joined the computer revolution of this decade only to hide their machines unused in closets have discovered a new use for them as dedicated publishing workstations.} Now let's do some editing. The last paragraph seems a little out of place. To see what the document looks like without it we can cut it from the text by moving the cursor to the beginning of the paragraph. Enter @b(M-) @i(set-mark) @index(set-mark). EMACS will respond with "[Mark set]". Now move the cursor to the end of the paragraph. You have just defined a region of text. To remove this text from the screen, type @b<^W> @i(kill-region) @index(kill-region). The paragraph disappears from the screen. On further consideration, however, perhaps the paragraph we cut wasn't so bad after all. The problem may have been one of placement. If we could tack it on to the end of the first paragraph it might work quite well to support and strengthen the argument. Move the cursor to the end of the first paragraph and enter @b<^Y> @i(yank) @index(yank). Your text should now look like this: @quotation{One of the largest growth areas in personal computing is electronic publishing. There are packages available for practically every machine from elegantly simple programs for the humble Commodore 64 to sophisticated professional packages for PC and Macintosh computers. Personal publishing greatly increases the utility of practically every computer. Thousands of people who joined the computer revolution of this decade only to hide their machines unused in closets have discovered a new use for them as dedicated publishing workstations. Electronic publishing is as revolutionary in its way as the Gutenburg press. Whereas the printing press allowed the mass production and distribution of the written word, electronic publishing puts the means of production in the hands of nearly every individual. From the class magazine to the corporate report, electronic publishing is changing the way we produce and disseminate information.} @section(Yanking a Region) The text you cut initially didn't simply just disappear, it was cut into a buffer that retains the 'killed' text appropriately called the @b. @b<^Y> "yanks" the text back from this buffer into the current buffer. If you have a long line (indicated, remember, by the "$" sign), simply hit @b{M-Q} to reformat the paragraph. There are other uses to which the kill buffer can be put. Using the @index(kill buffer) method we've already learned, define the last paragraph as a region. Now type @b @i(copy-region) @index(copy-region). Nothing seems to have happened; the cursor stays blinking at the point. But things have changed, even though you may not be able to see any alteration. To see what has happened to the contents of the kill buffer, move the cursor down a couple of lines and "yank" the contents of the kill buffer back with @b<^Y>. Notice the last paragraph is now repeated. The region you defined is "tacked on" to the end of your file because @b @b a region to the kill buffer while leaving the original text in your working buffer. Some caution is needed however, because the contents of the kill buffer are updated when you delete any regions, lines or words. If you are moving large quantities of text, complete the operation before you do any more deletions or you could find that the text you want to move has been replaced by the most recent deletion. Remember--a buffer is a temporary area of computer memory that is lost when the machine is powered down or switched off. In order to make your changes permanent, they must be saved to a file before you leave EMACS. Let's delete the section of text we just added and save the file to disk. @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter), you learned how to achieve longer insertions and deletions. The EMACS terms @b and @b were introduced and you learned how to manipulate text with the kill buffer. @begin{verbatim} @begin(group) @u(Key Binding Keystroke Effect) set-mark @b{M-} Marks the beginning of a region delete-region @b{^W} Deletes region between point and mark and places it in KILL buffer copy-region @b{M-W} Copies text between point and mark into KILL buffer yank-text @b{^Y} Inserts a copy of the KILL buffer into current buffer at point @end(group) @end(verbatim) @chapter(Search and Replace) @section Load EMACS and bring in the file you just saved. Your file should look like the one below. @quotation{One of the largest growth areas in personal computing is electronic publishing. There are packages available for practically every machine from elegantly simple programs for the humble Commodore 64 to sophisticated professional packages for PC and Macintosh computers. Personal publishing greatly increases the utility of practically every computer. Thousands of people who joined the computer revolution of this decade only to hide their machines unused in closets have discovered a new use for them as dedicated publishing workstations. Electronic publishing is as revolutionary in its way as the Gutenburg press. Whereas the printing press allowed the mass production and distribution of the written word, electronic publishing puts the means of production in the hands of nearly every individual. From the class magazine to the corporate report, electronic publishing is changing the way we produce and disseminate information.} Let's use EMACS to search for the word "revolutionary" in the second paragraph. Because EMACS searches from the current cursor position toward the end of buffers, and we intend to search forward, move the cursor to the beginning of the text. Enter @b<^S> @i(search-forward) @index(search-forward). Note that the command line now reads "Search [] :" EMACS is prompting you to enter the @b -- the text you want to find. Enter the word @b and hit the @b key. The cursor moves to the end of the word "revolutionary." Notice that you must enter the key to start the search. If you @index() simply press the command line responds with "". Although this may seem infuriating to users who are used to pressing the return key to execute any command, EMACS' use of to begin searches allows it to pinpoint text with great accuracy. After every line wrap or carriage return, EMACS 'sees' a new line character (). If you need to search for a word at the end of a line, you can specify this word uniquely in EMACS. In our sample text for example, the word "and" occurs a number of times, but only once at the end of a line. To search for this particular occurrence of the word, move the cursor to the beginning of the buffer and type @b(^S). Notice that EMACS stores the last specified @index(default string) search string as the @b string. If you press @b{} now, EMACS will search for the default string, in this case, "revolutionary." To change this string so we can search for our specified "and" simply enter the word @b{and} followed by @b{}. The command line now shows: "search [and]:" Press @b{} and the cursor moves to "and" at the end of the second last line. @section If the mode EXACT is active in the current buffer, EMACS searches on a case sensitive basis. Thus, for example you could search for @b{Publishing} as distinct from @b{publishing}. @section Backward searching is very similar to forward searching except that it is implemented in the reverse direction. To implement a reverse search, type @b{^R} @i(search-reverse) @index(search-reverse). Because EMACS makes no distinction between forward and backward stored search strings, the last search item you entered appears as the default string. Try searching back for any word that lies between the cursor and the beginning of the buffer. Notice that when the item is found, the point moves to the beginning of the found string (i.e., the cursor appears under the first letter of the search item). Practice searching for other words in your text. @section Searching and replacing is a powerful and quick way of making changes to your text. Our sample text is about electronic publishing, but the correct term is 'desktop' publishing. To make the necessary changes we need to replace all occurrences of the word "electronic" with "desktop." First, move the cursor to the top of the current buffer with the @b(M-<) command. Then type @b[M-R] @i(replace-string) @index(replace-string). The command line responds: "Replace []:" where the square brackets enclose the default string. Type the word @b and hit @b{}. The command line responds: "with []" type @b{desktop}. EMACS replaces all instances of the original word with your revision. Of course, you will have to capitalize the first letter of "desktop" where it occurs at the beginning of a sentence. You have just completed an @b. In this operation, EMACS replaces every instance of the found string with the replacement string. @section You may also replace text on a case by case basis. The @b{M-^R} @i(query-replace-string) @index(query-replace-string) command causes EMACS to pause at each instance of the found string. For example, assume we want to replace some instances of the word "desktop" with the word "personal." Go back to the beginning of the current buffer and enter the @b(M-^R) @i(query-replace) @index(query-replace) command. The procedure is very similar to that which you followed in the unconditional search/replace option. When the search begins however, you will notice that EMACS pauses at each instance of "publishing" and asks whether you wish to replace it with the replacement string. You have a number of options available for response: @begin(verbatim) @u( Response Effect) Y(es) Make the current replacement and skip to the next occurrence of the search string N(o) Do not make this replacement but continue ! Do the rest of the replacements with no more queries U(ndo) Undo just the last replacement and query for it again (This can only go back ONE time) ^G Abort the replacement command (This action does not undo previously-authorized replacements . Same effect as ^G, but cursor returns to the point at which the replacement command was given ? This lists help for the query replacement command @end(verbatim) Practice searching and searching and replacing until you feel comfortable with the commands and their effects. @begin(group) @heading(Chapter @value(chapter) Summary) In this chapter, you learned how to search for specified strings of text in EMACS. The chapter also dealt with searching for and replacing elements within a buffer. @begin(verbatim) @u(Key Binding Keystroke Effect) search-forward @b{^S} Searches from point to end of buffer. Point is moved from current location to the end of the found string search-backward @b{^R} Searches from point to beginning of buffer. Point is moved from current location to beginning of found string replace @b{M-R} Replace ALL occurrences of search string with specified (null) string from point to the end of the current buffer query-replace @b{M-^R} As above, but pause at each found string and query for action @end(verbatim) @end(group) @chapter(Windows) @section@index{Windows, Creating} We have already met windows in an earlier chapter. In this chapter, we will explore one of EMACS' more powerful features -- text manipulation through multiple windowing. Windows offer you a powerful and easy way to edit text. By manipulating a number of windows and buffers on the screen simultaneously, you can perform complete edits and revisions on the computer screen while having your draft text or original data available for reference in another window. You will recall that windows are areas of buffer text that you can see @index(windows) on the screen. Because EMACS can support several screen windows simultaneously you can use them to look into different places in the same buffer. You can also use them to look at text in different buffers. In effect, you can edit several files at the same time. Let's invoke EMACS and pull back our file on desktop publishing by typing @quotation When the text appears, type the @b{^X2} @i(split-current-window) @index{split-current-window} command. The window splits into two windows. The window where the cursor resides is called the @b window -- in this case the bottom window. Notice that each window has a text area and a mode line. The @b(command line) @index{command line} is however, common to all windows on the screen. The two windows on your screen are virtually mirror images of each other because the new window is opened into the same buffer as the one you are in when you issue the @i{open-window command} @index{open-window}. All commands issued to EMACS are executed on the current buffer in the current window. To move the cursor to the upper window (i.e., to make that window the current window, type @b{^XP} @i(previous-window) @index{previous-window}. Notice the cursor moves to the upper or @b window. Entering @b{^XO} @i(next-window) moves to the @b{next} window. Practice moving between windows. You will notice that you can also move into the Function Key menu by entering these commands. Now move to the upper window. Let's open a new file. On the EMACS disk is a tutorial file. Let's call it into the upper window by typing: @quotation<^X^F> and press return. Enter the filename @b. In a short time, the tutorial file will appear in the window. We now have two windows on the screen, each looking into different buffers. We have just used the @b(^X^F) @i(find-file) @index{find-file} command to find a file and bring it into our current window. You can scroll any window up and down with the cursor keys, or with the commands we've learned so far. However, because the area of visible text in each window is relatively small, you can scroll the current window a line at a time. Type @b{^X^N} @i(move-window-down)@index{move-window-down} The current window scrolls down by one line -- the top line of text scrolls out of view, and the bottom line moves towards the top of the screen. You can imagine, if you like, the whole window slowly moving down to the end of the buffer in increments of one line. The command @b{^X^P} @i(move-window-up)@index{move-window-up} scrolls the window in the opposite direction. As we have seen, EMACS editing commands are executed in the current window, but the program does support a useful feature that allows you to scroll the @b window. @b @i(scroll-next-up) @index{scroll-next-up} scrolls the next window up, @b{M-^V} @i(scroll-next-down)@index{scroll-next-down} scrolls it downward. From the tutorial window, practice scrolling the window with the desktop publishing text in it up and down. When you're finished, exit EMACS without saving any changes in your files. Experiment with splitting the windows on your screen. Open windows into different buffers and experiment with any other files you may have. Try editing the text in each window, but don't forget to save any changes you want to keep -- you still have to save each buffer separately. @section(Deleting Windows)@index{Windows, Deleting} Windows allow you to perform complex editing tasks with ease. However, they become an inconvenience when your screen is cluttered with open windows you have finished using. The simplest solution is to delete unneeded windows. The command @b{^X0} @i{delete-window} will delete the window you are currently working in and move you to the next window. If you have a number of windows open, you can delete all but the current window by entering @b{^X1} @i{delete-other-windows}. @section(Resizing Windows)@index{Windows, Resizing} During complex editing tasks, you will probably find it convenient to have a number of windows on the screen simultaneously. However this situation may present inconveniences because the more windows you have on the screen the smaller they are; in some cases, a window may show only a couple of lines of text. To increase the flexibility and utility of the window environment, EMACS allows you to resize the window you are working in (called, as you will recall, the @b window) to a convenient size for easier editing, and then shrink it when you no longer need it to be so large. Let's try an example. Load in any EMACS text file and split the current window into two. Now type @b{^X^(Shift-6)}, @i{grow-window}@index{grow-window}. Your current window should be the lower one on the screen. Notice that it increases in size upwards by one line. If you are in the upper window, it increases in size in a downward direction. The command @b{^X^Z}, @i{shrink-window}@index{shrink-window} correspondingly decreases window size by one line at a time. EMACS also allows you to resize a window more precisely by entering a numeric argument specifying the size of the window in lines. To resize the window this way, press the META key and enter a numeric argument (remember to keep it smaller than the number of lines on your screen display) then press @b{^XW} @i{resize-window}@index{resize-window}. The current window will be enlarged or shrunk to the number of lines specified in the numeric argument. For example entering: @quotation{@b[M-8 ^XW]}will resize the current window to 8 lines. @section(Repositioning within a Window) The cursor may be centered within a window by entering @b{M-! or M-^L} @i{redraw-display} @index{redraw-display}. This command is especially useful in allowing you to quickly locate the cursor if you are moving frequently from window to window. You can also use this command to move the line containing the cursor to any position within the current window. This is done by using a numeric argument before the command. Type @b(M- M-^L) where is the number of the line within the window that you wish the current line to be displayed. The @b{^L} @i{clear-and-redraw} @index{clear-and-redraw} command is useful for 'cleaning up' a 'messy' screen that can result of using EMACS on a mainframe system and being interrupted by a system message. @newpage @u In Chapter @value(chapter) you learned how to manipulate windows and the editing flexibility they offer. @begin(verbatim) @u(Key Binding Keystroke Effect) open-window @b{^X2} Splits current window into two windows if space available close-windows @b{^X1} Closes all windows except current window next-window @b{^XO}[oh] Moves point into next (i.e. downward) window previous-window @b{^XP} Moves point to previous (i.e. upward) window move-window-down @b{^X^N} Scrolls current window down one line move-window-up @b{^X^P} Scrolls current window up one line redraw-display @b{M !} or Window is moved so line with point @b{M ^L} (with cursor) is at center of window grow-window @b{M-X ^} Current window is enlarged by one line and nearest window is shrunk by one line shrink-window @b{^X^Z} Current window is shrunk by one line and nearest window is enlarged by one line clear-and-redraw @b{^L} Screen is blanked and redrawn. Keeps screen updates in sync with your commands scroll-next-up @b{M-^Z} Scrolls next window up by one line scroll-next-down @b{M-^V} Scrolls next window down by one line delete-window @b{^X0} Deletes current window delete-other-windows @b{^X1} Deletes all but current window resize-window @b{^X^W} Resizes window to a given numeric argument @end(verbatim) @chapter(Using a Mouse) On computers equipped with a mouse@index(mouse), the mouse can usually be used to make editing easier. If your computer has a mouse, let's try using it. Start MicroEMACS by typing: emacs publish.txt This brings EMACS up and allows it to edit the file from the last chapter. If the function key window is visible on the screen, press the F5 key to cause it to disappear. Now use the @b(^X2) @i(split-current-window) command to split the screen into two windows. Next use the @b(^X^F) @i(find-file) command to read in the @b(fang.txt) file. Now your screen should have two windows looking into two different files. Grab the mouse and move it around. On the screen an arrow, or block of color appears. This is called the mouse cursor @index(mouse cursor) and can be positioned on any character on the screen. On some computers, positioning the mouse cursor in the extreme upper right or left corner may bring down menus which allow you to access that computers utilities, sometimes called @b(Desk Accessories) @index(desk accessories). @section(Moving around with the mouse) Using the mouse button (or the left button if the mouse has more than one), position the mouse over some character in the current window. Click the mouse button once. The @b(point) will move to where the mouse cursor is. If you place the mouse cursor past the end of a line, the point will move to the end of that line. Move the mouse cursor into the other window and click on one of the characters there. MicroEMACS will automatically make this window the current window (notice that the mode line changes) and position the point to the mouse cursor. This makes it very easy to use the mouse to switch to a different window quickly. @section(Dragging around) Besides just using the mouse to move around on the screen, you can use the same button to move text. Move the mouse cursor to a character in one of the windows, and click down... but don't let the button up yet! The point will move to where the mouse cursor is. Now move the mouse cursor up or down on the screen, and release the button. The point will again move to where the mouse cursor is, but this time it will bring the text under it along for the ride. This is called @b(dragging)@index(dragging), and is how you can make the text appear just where you want it to. If you try to drag text out of the current window, EMACS will ignore your attempt and leave the point where you first clicked down. @index(vertical scrolling) Now, click down on a word in one of the windows, and drag it directly to the left. Release the button and watch as the entire window slides, or @b(scrolls) @index(horizontal scrolling) to the left. The missing text has not been deleted, it is simply not visible, off the left hand side of the screen. Notice the mode line has changed and now looks like: @flushleft(==== MicroEMACS 3.12 [<12] () == fang.txt == File: fang.txt ==============) The number insided the brackets [] shows that the screen is now scrolled 12 characters from the left margin. Now grab the same text again, and drag it to the right, pulling the rest of the text back into the current window. The [<] field will disappear, meaning that the window is no longer scrolled to the left. This feature is very useful for looking at wide charts and tables. Remember, MicroEMACS will only scroll the text in the current window sideways if you drag it straight to the side, otherwise it will drag the text vertically. Now, place the mouse cursor over a character on the upper mode line, click down, move the mouse cursor up or down a few lines and let go of the button. The mode line moves to where you dragged it, changing the size of the windows above and below it. If you try to make a window with less than one line, EMACS will not let you. Dragging the mode lines can make it very fast and easy for you to rearrange the windows as you would like. If you have a number of different windows visible on the screen, positioning the mouse over the mode line of one window and clicking the right mouse button will cause that window to be deleted. @section(Cut and Paste) If your mouse has two buttons, then you can use the right button to do some other things as well. Earlier, we learned how to define a @b(region)@index(region) by using the @b(M-) @i(set-mark) command. Now, position the mouse over at the beginning of a region you would like to copy. Next click and hold down the right mouse button. Notice that the point jumps to the mouse cursor and EMACS reports "[Mark Set]". Holding the button down move the mouse to the end of the text you wish to copy and release the mouse button. Emacs reports "[Region Copied]" to let you know it has copied the region into the KILL buffer. This has done the same job as the @b(M-W) @i(copy-region) command. If you now click the right mouse button, without moving the mouse, the region you defined dissapear, being @b(cut)@index(cut) from the current buffer. This works just like the @b(^W) @i(kill-region) command. If you move the mouse away from where you cut the text, and click the right mouse button down and up without moving the mouse, the text in the KILL buffer gets inserted, or pasted@index(paste) into the current buffer at the point. @section(Screens) MicroEMACS can use more than one screen@index(screen) at once. Each screen is a collection of @i(windows) along with a mode line. These screens usually fill the terminal or computer screen on text based systems, but can also be held in different @b(windows) on graphically based systems like MicroSoft Windows, OS/2, the Macintosh Finder and X-Windows. Don't be confused by the two different uses of the term "window". Inside EMACS style editors, a @i(window) lets you view part of a buffer. Under graphical operating systems, a @b(window) holds a "virtual terminal", allowing you to manipulate more than one job, editing session or program at once. Within MicroEMACS, these operating system @b(window)s are called screens. All these screens are displayed on your current desktop@index(desktop). @section(Resizing a Screen) You can change the size of a screen. Move the mouse to the last position of the command line. Press the left mouse button down. Holding it, move the mouse to the place you want the new lower right corner. Release the mouse. The desktop redraws, with your newly resized screen. MicroEMACS will ignore size changes that can not be done, like attempting to pull the lower left corner above the upper right corner of the current screen. @section(Moving a Screen) To change where on the desktop a screen is placed, move the mouse to the upper right corner of the screen, press the left mouse button down, move the mouse and release it where you want the screen displayed. Again, MicroEMACS will ignore placements that can not be done. @section(Creating a Screen) Creating a new screen is just like moving a screen, but using the right button. Move to the upper right of an existing screen, press the right mouse button down, and move the mouse, releasing the button where the new screen should appear. A new screen will have a single @b(window), containing the contents of the current window in the copied screen, and will have that @b(window)'s colors. The new screen will have the copied screen's size. @section(Switching to a Screen) This is simple. Any mouse command can be done in any screen by placing the mouse on a visible part of the screen and clicking. The last screen the mouse is used on comes to front and is the current screen. Also, the @b(A-C) @i(cycle-screens)@index(cycle-screens) command brings the rearmost screen to front. @section(Deleting a Screen) Place the mouse on the command line of the screen you want to delete. Click the right mouse button, the screen will disapear. If you delete the only remaining screen on the desktop, MicroEMACS will exit. @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter), you learned how to use the mouse to move the point, switch windows, drag text, and resize windows. You also learned how to use the right mouse button in order to copy and delete regions and yank them back at other places. And lastly, you learned how to control multiple screens with the mouse. @begin{verbatim} @begin(group) @u(Action Mouse Directions) Move Cursor position mouse cursor over desired location click down and up with left button Drag Text position mouse cursor over desired text click left button down move to new screen location for text release mouse button Resize Windows position mouse cursor over mode line to move click left button down move to new location for mode line release mouse button Delete Window position mouse cursor over mode line of window to delete click right mouse button Activate Screen Move mouse over existing screen click left button down and up Resize Screen position mouse cursor over last character on message line click left button down move to new lower right corner of screen release mouse button Copy Region position mouse at beginning of region click right button down move to end of region release mouse button Cut Region position mouse at beginning of region click right button down move to end of region release mouse button click right button down and up Paste Region position mouse at place to paste click right button down and up Create Screen position mouse at upper left corner of existing screen click right button down move to position of new screen release mouse button Resize Screen position mouse at lower right corner of screen click left button down move to new lower left corner release mouse button Move Screen position mouse at upper right corner of screen click left button down move to new screen position release mouse button Delete Screen position to command line of existing screen click right button down release mouse button @end(group) @end(verbatim) @chapter(Buffers) @index(buffer)We have already learned a number of things about buffers. As you will recall, they are the major internal entities in EMACS -- the place where editing commands are executed. They are characterized by their @b, their @b, and by the file with which they are associated. Each buffer also "remembers" its @b(mark) and @b(point). This convenient feature allows you to go to other buffers and return to the original location in the "current" buffer. Advanced users of EMACS frequently have a number of buffers in the computer's memory simultaneously. In the last chapter, for example, you opened at least two buffers -- one into the text you were editing, and the other into the EMACS on-line tutorial. If you deal with complex text files -- say, sectioned chapters of a book, you may have five or six buffers in the computer's memory. You could select different buffers by simply calling up the file with @b{^X^F} @i(find-file) @index(find-file), and let EMACS open or reopen the buffer. However, EMACS offers fast and sophisticated buffering techniques that you will find easy to master and much more convenient to use. Let's begin by opening three buffers. You can open any three you choose, for example call the following files into memory: @b(fang.txt), @b(publish.txt), and @b(emacs.tut) in the order listed here. When you've finished this process, you'll be looking at a screen showing the EMACS tutorial. Let's assume that you want to move to the fang.txt buffer. Enter: @b{^XX} @i(next-buffer) @index(next-buffer) This command moves you to the @u buffer. Because EMACS cycles through the buffer list, which is alphabetized, you will now be in the @b(fang.txt) buffer. Using @b(^XX) again places you in the @b(publish.txt) buffer. @i(If you are on a machine that supports function keys, using @b[^XX] again places you in the @b(Function Keys) buffer). Using @b(^XX) one last time cycles you back to the beginning of the list. If you have a large number of buffers to deal with, this cycling process may be slow and inconvenient. The command @b{^XB} @i(select-buffer) @index(select-buffer) allows you to specify the buffer you wish to be switched to. When the command is entered, EMACS prompts, "Use buffer:". Simply enter the buffer name (NOT the file name), and that buffer will then become the current buffer. If you type in part of the file name and press the space bar, EMACS will attempt to complete the name from the list of current buffers. If it succeeds, it will print the rest of the name and you can hit to switch to that buffer. If EMACS beeps the bell, there is no such buffer, and you may continue editing the name on the command line. Multiple buffer manipulation and editing is a complex activity, and you will probably find it very inconvenient to re-save each buffer as you modify it. The command @b{^X^B} @i(list-buffers) @index(list-buffers) creates a new window that gives details about all the buffers currently known to EMACS. Buffers that have been modified are identified by the "buffer changed" indicator (an asterisk in the second column). You can thus quickly and easily identify buffers that need to be saved to files before you exit EMACS. The buffer window also provides other information -- buffer specific modes, buffer size, and buffer name are also listed. To close this window, simply type the close-windows command, @b{^X1}. To delete any buffer, type @b{^XK} @i(delete-buffer) @index(delete-buffer). EMACS prompts you "Kill buffer:". Enter the buffer name you want to delete. As this is destructive command, EMACS will ask for confirmation if the buffer was changed and not saved. Answer Y(es) or N(o). As usual @b{^G} cancels the command. @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter) you learned how to manipulate buffers. @begin{verbatim} @u(Key Binding Keystroke Effect) next-buffer @b(^X^X) Switch to the next buffer in the buffer list select-buffer @b(^XB) Switch to a particular buffer list-buffers @b(^X^B) List all buffers delete-buffer @b(^XK) Delete a particular buffer if it is off-screen @end(verbatim) @chapter(Modes) EMACS allows you to change the way it works in order to customized it to the style of editing you are using. It does this by providing a number of different @b(modes) @index(modes). These modes can effect either a single buffer, or any new buffer that is created. To add a mode to the current buffer, type @b(^XM) @i(add-mode) @index(add-mode). EMACS will then prompt you for the name of a mode to add. When you type in a legal mode name, and type a , EMACS will add the mode name to the list of current mode names in the mode line of the current buffer. To remove an existing mode, typing the @b(^X^M) @i(delete-mode) @index(delete-mode) will cause EMACS to prompt you for the name of a mode to delete from the current buffer. This will remove that mode from the mode list on the current mode line. Global modes are the modes which are inherited by any new buffers which are created. For example, if you wish to always do string searching with character case being significant, you would want global mode EXACT to be set so that any new files read in inherent the EXACT mode. Global modes are set with the @b(M-M) @i(add-global-mode) @index(add-global-mode) command, and unset with the @b(M-^M) @i(delete-global-mode) @index(delete-global-mode) command. Also, the current global modes are displayed in the first line of a @b(^X^B) @i(list-buffers) @index(list-buffers) command. On machines which are capable of displaying colors, @index(color) the mode commands can also set the background and foreground character colors. Using @i(add-mode) or @i(delete-mode) with a lowercase color will set the background color in the current window. An uppercase color will set the foreground color in the current window. Colors that EMACS knows about are: white, cyan, magenta, yellow, blue, red, green, and black. If the computer you are running on does not have eight colors, EMACS will attempt to make some intelligent guess at what color to use when you ask for one which is not there. @section(ASAVE mode)@index(ASAVE mode) Automatic Save mode tells EMACS to automatically write out the current buffer to its associated file on a regular basis. Normally this will be every 256 characters typed into the file. The environment variable $ACOUNT counts down to the next auto-save, and $ASAVE is the value used to reset $ACOUNT after a save occurs. @section(CMODE mode)@index(CMODE mode) CMODE is useful to C programmers. When CMODE is active, EMACS will try to assist the user in a number of ways. This mode is set automatically with files that have a .c or .h extension. The key will normally attempt to return the user to the next line at the same level of indentation as the last non blank line, unless the current line ends with a open brace ({) in which case the new line will be further indented by one tab position. A close brace (}) will search for the corresponding open brace and line up with it. A pound sign (#) with only leading white space will delete all the white space preceding itself. This will always bring preprocessor directives flush to the left margin. Whenever any close fence is typed, IE )]>}, if the matching open fence is on screen in the current window, the cursor will briefly flash to it, and then back. This makes balancing expressions, and matching blocks much easier. @section(CRYPT mode)@index(CRYPT mode) When a buffer is in CRYPT mode, @index(encryption) it is encrypted whenever it is written to a file, and decrypted when it is read from the file. The encryption key can be specified on the command line with the -k switch, or with the @b(M-E) @i(set-encryption-key) @index(set-encryption-key) command. If you attempt to read or write a buffer in crypt mode and now key has not been set, EMACS will execute @i(set-encryption-key) automatically, prompting you for the needed key. Whenever EMACS prompts you for a key, it will not echo the key to your screen as you type it (IE make SURE you get it right when you set it originally). The encryption algorithm used changes all characters into normal printing characters, thus the resulting file is suitable for sending via electronic mail. All version of MicroEMACS should be able decrypt the resulting file regardless of what machine encrypted it. Also available with EMACS is the stand alone program, MicroCRYPT, which can en/decrypt the files produced by CRYPT mode in EMACS. @section(EXACT mode)@index(EXACT mode) All string searches and replacements will take upper/lower case into account. Normally the case of a string during a search or replace is not taken into account. @section(MAGIC mode)@index(MAGIC mode) In the MAGIC mode certain characters gain special meanings when used in a search pattern. Collectively they are know as @index(regular expressions) regular expressions, and a limited number of them are supported in MicroEmacs. They grant greater flexibility when using the search command. They have no affect on the incremental search command. The symbols that have special meaning in MAGIC mode are ^, $, ., &, ?, *, +, [ (and ], used with it), and \. The characters ^ and $ fix the search pattern to the beginning and end of line, respectively. The ^ character must appear at the beginning of the search string, and the $ must appear at the end, otherwise they lose their meaning and are treated just like any other character. For example, in MAGIC mode, searching for the pattern "t$" would put the cursor at the end of any line that ended with the letter 't'. Note that this is different than searching for "t", that is, 't' followed by a newline character. The character $ (and ^, for that matter) matches a position, not a character, so the cursor remains at the end of the line. But a newline is a character that must be matched like any other character, which means that the cursor is placed just after it - on the beginning of the next line. The character . has a very simple meaning -- it matches any single character, except the newline. Thus a search for "bad.er" could match "badger", "badder" (slang), or up to the 'r' of "bad error". The character [ indicates the beginning of a character class. It is similar to the 'any' character ., but you get to choose which characters you want to match. The character class is ended with the character ]. So, while a search for "ba.e" will match "bane", "bade", "bale", "bate", et cetera, you can limit it to matching "babe" and "bake" by searching for "ba[bk]e". Only one of the characters inside the [ and ] will match a character. If in fact you want to match any character except those in the character class, you can put a ^ as the first character. It must be the first character of the class, or else it has no special meaning. So, a search for [^aeiou] will match any character except a vowel, but a search for [aeiou^] will match any vowel or a ^. If you have many characters in order, that you want to put in the character class, you may use a dash (-) as a range character. So, [a-z] will match any letter (or any lower case letter if EXACT mode is on), and [0-9a-f] will match any digit or any letter 'a' through 'f', which happen to be the characters for hexadecimal numbers. If the dash is at the beginning or end of a character class, it is taken to be just a dash. The ? character indicates that the preceding character is optional. The character may or may not appear in the matched string. For example, a search for "bea?st" would match both "beast" and "best". If there is no preceding charcter for ? to modify, it is treated as a normal question mark character. The * character is known as closure, and means that zero or more of the preceding character will match. If there is no preceding character, * has no special meaning and is treated as a normal asterisk. The closure symbol will also have no special meaning if it is preceded by the beginning of line symbol ^, since it represents a position, not a character. The notion of zero or more characters is important. If, for example, your cursor was on the line @quotation(This line is missing two vowels.) and a search was made for "a*", the cursor would not move, because it is guaranteed to match no letter 'a' , which satisfies the search conditions. If you wanted to search for one or more of the letter 'a', you could search for "aa*", which would match the letter a, then zero or more of them. A better way, however, is to use the + character. The + character behaves in every respect like the * character, with the exception that its minimum match range is one, not zero. Thus the pattern "a+" is identical to "aa*". Under older versions of MicroEMACS, the closure symbols would not operate on newlines. The current versions no longer have this restriction. The \ is the escape character. With the exception of groups, which are explained below, the \ is used at those times when you want to be in MAGIC mode, but also want a regular expression character to be just a character. It turns off the special meaning of the character. So a search for "it\." will search for a line with "it.", and not "it" followed by any other character. Or, a search for "TEST\*+" would match the word TEST followed by one or more asterisks. The escape character will also let you put ^, -, or ] inside a character class with no special side effects. The character pair \( represent the start of a group in a search string. A group is ended by the character pair \). All characters matched within the \( and \) are part of a numbered group, and may be referenced with the &GROUP function, or with a \ followed by the group number in the replacement string of @i(replace-string) @index(replace-string) or the @i(query-replace-string) @index(query-replace-string) commands. For example, a search for "INDEX\([0-9]+\)", to be replaced by "getind(\1)" would change @quotation(indptr := INDEX42) to @quotation(indptr := getind(42)). There may be up to nine groups. Groups may be nested. The character & (ampersand) is a replacement character, and represents all the characters which were matched by the search string. When used in the @b[M-R] @i(replace-string) @index(replace-string) or the @b[M-^R] @i(query-replace-string) @index(query-replace-string) commands, the & will be substituted for the search string. @section(OVER mode)@index(OVER mode) OVER mode stands for overwrite mode. When in this mode, when characters are typed, instead of simply inserting them into the file, EMACS will attempt to overwrite an existing character past the point. This is very useful for adjusting tables and diagrams. @section(WRAP mode)@index(WRAP mode) Wrap mode is used when typing in continuous text. Whenever the cursor is past the currently set fill column @index(fill column) (72 by default) and the user types a space or a , the last word of the line is brought down to the beginning of the next line. Using this, one just types a continuous stream of words and EMACS automatically inserts s at appropriate places. @center(NOTE to programmers:) @quotation{The EMACS variable $wraphook contains the name of the function which executes when EMACS detects it is time to wrap. This is set to the function @i(wrap-word) @index(wrap-word) by default, but can be changed to activate different functions and macroes at wrap time.} @section(VIEW mode)@index(VIEW mode) VIEW mode disables all commands which can change the current buffer. EMACS will display an error message and ring the bell every time you attempt to change a buffer in VIEW mode. @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter) you learned about modes and their effects. @begin{verbatim} @u(Key Binding Keystroke Effect) add-mode @b(^XM) Add a mode to the current buffer delete-mode @b(^X^M) Delete a mode from the current buffer add-global-mode @b(M-M) Add a global mode to the current buffer delete-global-mode @b(M-^M) Delete a global mode from the current buffer @end(verbatim) @chapter(Files) A file is simply a collection of related data. In EMACS we are dealing with text files -- named collections of text residing on a disk (or some other storage medium). You will recall that the major entities EMACS deals with are buffers. Disk-based versions of files are only active in EMACS when you are reading into or writing out of buffers. As we have already seen, buffers and physical files are linked by associated file names. For example, the buffer "ch7.txt" which is associated with the physical disk file "ch7.txt." You will notice that the file is usually specified by the drive name or (in the case of a hard drive) a path. Thus you can specify full file names in EMACS, e.g. disk:\directories\filename.extension If you do not specify a disk and directories, the default disk and the current directory is used. IMPORTANT -- If you do not explicitly save your buffer to a file, all your edits will be lost when you leave EMACS (although EMACS will prompt you when you are about to lose edits by exiting). In addition, EMACS does not protect your disk-based files from overwriting when it saves files. Thus when you instruct EMACS to save a file to disk, it will create a file if the specified file doesn't exist, or it will overwrite the previously saved version of the file thus replacing it. Your old version is gone forever. If you are at all unsure about your edits, or if (for any reason) you wish to keep previous versions of a file, you can change the name of the associated file with the command @b{^XN} @i(change-file-name)@index(change-file-name). When this file is saved to disk, EMACS will create a new physical file under the new name. The earlier disk file will be preserved. For example, let's load the file @b{fang.txt} into EMACS. Now, type @b{^XN}. The EMACS command line prompts "Name:". Enter a new name for the file -- say @b(new.txt) and press . The file will be saved under the new filename, and your disk directory will show both @b(fang.txt) and @b(new.txt). An alternative method is to write the file directly to disk under a new filename. Let's pull our "publish.txt" file into EMACS. To write this file under another filename, type @b{^X^W} @i(write-file)@index(writefile). EMACS will prompt you "write file:". Enter an alternate filename -- @b{desktop.txt}. Your file will be saved as the physical file "desktop.txt". Note that in the examples above, although you have changed the names of the related files, the buffer names remain the same. However, when you pull the physical file back into EMACS, you will find that the buffer name now relates to the filename. For example -- You are working with a buffer "fang.txt" with the related file "fang.txt". You change the name of the file to "new.txt". EMACS now shows you working with the buffer "fang.txt" and the related file "new.txt". Now pull the file "new.txt" into EMACS. Notice that the buffer name has now changed to "new.txt". If for any reason a conflict of buffer names occurs,(if you have files of the same name on different drives for example) EMACS will prompt you "use buffer:". Enter an alternative buffer name if you need to. For a list of file related commands (including some we`ve already seen), see the summary page. @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter) you learned some of the more advanced concepts of file naming and manipulation. The relationship between files and buffers was discussed in some detail. @begin(verbatim) @u(Key Binding Keystroke Effect) save-file @b{^X^S} Saves contents of current buffer with associated filename on default disk/ directory (if not specified) write-file @b{^X^W} Current buffer contents will be saved under specified name change-file-name @b{^XN} The associated filename is changed (or associated if not previously specified) as specified find-file @b{^X^F} Reads specified file into buffer and switches you to that buffer, or switches to buffer in which the file has previously been read read-file @b{^X^R} Reads file into buffer thus overwriting buffer contents. If file has already been read into another buffer, you will be switched to it view-file @b{^X^V} The same as read-file except the buffer is automatically put into VIEW mode thus preventing any changes from being made @end{verbatim} @chapter(Screen Formatting) @section As we learned in the introduction, EMACS is not a word processor, but an editor. Some simple formatting options are available however, although in most cases they will not affect the appearance of the finished text @index(wrapping text) when it is run through the formatter. We have already encountered WRAP mode which wraps lines longer than a certain length (default is 75 characters). You will recall that WRAP is enabled by entering @b{^XM} and responding to the command line prompt with @b{wrap}. You can also set your own wrap margin with the command @b{^XF} @i(set-fill-column) @index(set-fill-column). Notice EMACS responds "[Fill column is 1]." Now try typing some text. You'll notice some very strange things happening -- your text wraps at every word!! This effect occurs because the set wrap margin command must be preceded by a numeric argument or EMACS sets it to the first column. Thus any text you type that extends past the first column will wrap at the most convenient line break. To reset the wrap column to 72 characters, press the @b{} key and enter 72. EMACS will respond "Arg: 72". Now press @b<^XF>. EMACS will respond "[Fill column is 72]". Your text will again wrap at the margin you've been using up to this point. @section After an intensive editing session, you may find that you have paragraphs containing lines of differing lengths. Although this disparity will not affect the formatted text, aesthetic and technical concerns may make it desirable to have consistent paragraph blocks on the screen. If you are in WRAP mode, you can reformat a paragraph with the command @b{M-Q} @i(fill-paragraph) @index(fill-paragraph). This command 'fills' the current paragraph reformatting it so all the lines are filled and wrap logically. @section There may be occasions when you find it necessary to change the case of the text you've entered. EMACS allows you to change the case of even large amounts of text with ease. Let's try and convert a few of the office traditionalists to the joy of word processing. Type in the following text: @quotation{Throw away your typewriter and learn to use a word processor. Word processing is relatively easy to learn and will increase your productivity enormously. Enter the Computer Age and find out just how much fun it can be!!} Let's give it a little more impact by capitalizing the first four words. The first step is to define the region of text just as you would if you were doing an extensive deletion. Set the mark at the beginning of the paragraph with @b{M-} @i(set-mark) and move the cursor to the space beyond "typewriter." Now enter @b{^X^U} @i(case-region-upper). Your text should now look like this: @quotation{THROW AWAY YOUR TYPEWRITER and learn to use a word processor. Word processing is relatively easy to learn and will increase your productivity enormously. Enter the Computer Age and find out just how much fun it can be!!} If you want to change the text back to lower case, type @b{^X^L} @i(case-region-lower) @index(case-region-lower). You can also capitalize individual words. To capitalize the word "fun", position the cursor in front of the word and type @b{M-U} @i(case-word-upper) @index(case-word-upper). The word is now capitalized. To change it ck to lower case, move the cursor back to the beginning of the word and type @b{M-L} @i(case-word-lower) @index(case-word-lower). You may also capitalize individual letters in EMACS. The command @b{M-C} @i(case-word-capitalize) @index(case-word-capitalize) capitalizes the first letter after the point. This command would normally be issued with the cursor positioned in front of the first letter of the word you wish to capitalize. If you issue it in the middle of a word, you can end up with some strAnge looking text. @section Unless your formatter is instructed to take screen text literally (as MicroSCRIBE does in the 'verbatim' environment for example), tabs in EMACS generally affect screen formatting only. When EMACS is first started, it sets the default tab to every eighth column. As long as you stay with default, every time you press the tab key a tab character, @b(^I) is inserted. This character, like other control characters, is invisible -- but it makes a subtle and significant difference to your file and editing. For example, in default mode, press the tab key and then type the word @b{Test}. "Test" appears at the eighth column. Move your cursor to the beginning of the word and delete the backward character. The word doesn't move back just one character, but flushes to the left margin. The reason for this behavior is easily explained. In tab default, EMACS inserts a 'real' tab character when you press the tab key. This character is inserted at the default position, but NO SPACES are inserted between the tab character and the margin (or previous tab character). As you will recall, EMACS only recognizes characters (such as spaces or letters) and thus when the tab character is removed, the text beyond the tab is flushed back to the margin or previous tab mark. This situation changes if you alter the default configuration. The default value may be changed by entering a numeric argument before pressing the tab key. As we saw earlier, pressing the @b{META} key and entering a number allows you to specify how EMACS performs a given action. In this case, let's specify an argument of 10 and hit the tab key. Now hit the tab key again and type @b{Test}. Notice the word now appears at the tenth column. Now move to the beginning of the word and delete the backward character. "Test" moves back by one character. EMACS behaves differently in these circumstances because the @b(^I) @index(tab handling) @i(handle-tab) @index(handle-tab) function deals with tabbing in two distinct ways. In default conditions, or if the numeric argument of zero is used, @i(handle-tab) inserts a true tab character. If, however, a non-zero numeric argument is specified, @i(handle-tab) inserts the correct number of spaces needed to position the cursor at the next specified tab position. It does NOT insert the single tab character and hence any editing functions should take account of the number of spaces between tabbed columns. The distance which a true tab character moves the cursor can be modified by changing the value of the $hardtab environment variable. Initially set to 8, this will determine how far each tab stop is placed from the previous one. (Use the ^XA @i(set)@index(set) command to set the value of an environment variable). Many times you would like to take text which has been created using the tab character and change it to use just spaces. The command @b(^X^D) @i(detab-region) @index(detab-region) changes any tabs in the currently selected region into the right number of spaces so the text does not change. This is very useful for times when the file must be printed or transferred to a machine which does not understand tabs. Also, the inverse command, @b(^X^E) @i(entab-region) @index(entab-region) changes multiple spaces to tabs where possible. This is a good way to shrink the size of large documents, especially with data tables. Both of these commands can take a numeric argument which will be interpreted as the number of lines to en/detab. Another function, related to those above is provided for by the @b(^X^T) @i(trim-region)@index(trim-region) when invoked will delete any trailing white space in the selected region. A preceding numeric argument will do this for that number of lines. @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter) introduced some of the formatting features of EMACS. Text-wrap, paragraph reformatting, and tabs were discussed in some detail. The commands in the following table were covered in the chapter. @begin{verbatim} @u(Key Binding Keystroke Effect) add-mode/WRAP @b{^XM}[WRAP] Add wrap mode to current buffer delete-mode/WRAP @b{^X^M}[WRAP] Remove wrap mode from current buffer set-fill-column @b{^XF} Set fill column to given numeric argument fill-paragraph @b{M-Q} Logically reformats the current paragraph case-word-upper @b{M-U} Text from point to end of the current word is changed to uppercase case-word-lower @b{M-L} Text from point to end of the current word is changed to lowercase case-word-capitalize @b{M-C} First word (or letter) after the point is capitalized case-region-upper @b{^X^U} The current region is uppercased case-region-lower @b{^X^L} The current region is lowercased handle-tab @b{^I} Tab interval is set to the given numeric argument entab-region @b(^X^E) Changes multiple spaces to tabs characters where possible detab-region @b(^X^D) Changes tab characters to the appropriate number of spaces trim-region @b(^X^T) Trims white space from the end of the lines in the current region @end{verbatim} @chapter(Access to the Outside World) EMACS has the ability to interface to other programs and the environment of the computer outside of itself. It does this through a series of commands that allow it to talk to the computer's @b(command processor) @index(command processor) or @b(shell) @index(shell). Just what this is varies between different computers. Under MSDOS or PCDOS this is the @b(command.com) @index(command.com) command processor. Under UNIX it is the @b(csh) @index(cshell) shell. On the Atari ST is can be the Mark Williams @b(MSH) or the Beckmeyer shell. In each case, it is the part of the computer's operating system that is responsible for determining what programs are executed, and when. The @b(^X!) @i(shell-command) @index(shell-command) command prompts the user for a command line to send out to the shell to execute. This can be very useful for doing file listings and changing the current directory or folder. EMACS gives control to the shell, which executed the command, and then types @b([END]) and waits for the user to type a character before redrawing the screen and resuming editing. If the @i(shell-command) command is used from within the macro language, there is no pause. @b(^X@@) @i(pipe-command) @index(pipe-command) command allows EMACS to execute a shell command, and if the particular computer allows it, send the results into a buffer which is automatically displayed on the screen. The resulting buffer, called "command" can be manipulated just like any other editing buffer. Text can be copied out of it or rearranged as needed. This buffer is originally created in @b(VIEW) mode, so remember to @b(^X^Mview) in order to change it. Many computers provide tools which will allow you to @b(filter) @index(filter) text, making some modifications to it along the way. A very common tool is the @b(SORT) program which accepts a file, sorts it, and prints the result out. The EMACS command, @b(^X#) @i(filter-buffer)@index(filter-buffer) sends the current buffer through such a filter. Therefore, if you wished to sort the current buffer on a system which supplied a sort filter, you would type @b(^X#sort). You can also create your own filters by writing programs and utilities which read text from the keyboard and display the results. EMACS will use any of these which would normally be available from the current shell. If you would like to execute another program directly, without the overhead of an intervening shell, you can use the @b(^X$) @i(execute-program) @index(execute-program) command. It will prompt you for an external program and its arguments and attempt to execute it. Like when EMACS looks for command files, EMACS will look first in the HOME directory, then down the execute PATH, and finally in the current directory for the named program. On some systems, it will automatically tack the proper extension on the file name to indicate it is a program. On some systems that don't support this function, @b(^X$) will be equivalent to @b(^X!) @i(shell-command). Sometimes, you would like to get back to the shell and execute other commands, without losing the current contents of EMACS. The @b(^XC) @i(i-shell) @index(i-shell) command shells out of EMACS, leaving EMACS in the computer and executing another command shell. Most systems would allow you to return to EMACS with the "exit" command. @i(On some systems, mainly advanced versions of UNIX, you can direct EMACS to "go into the background" with the @b(^XD) suspend-emacs @index(suspend-emacs) command. This places EMACS in the background returning you to the original command shell. EMACS can then be returned to at any time with the "fg" foreground command.) @newpage @heading(Chapter @value(chapter) Summary) In Chapter @value(chapter) introduced different ways to access the computers shell or command processor from within EMACS. The commands in the following table were covered in the chapter. @begin{verbatim} @u(Key Binding Keystroke Effect) execute-program @b(^X$) Execute an external program directly filter-command @b(^X#) Send the current buffer through a shell filter i-shell @b(^XC) Escape to a new shell pipe-command @b(^X@@) Send the results of an external shell command to a buffer shell-command @b(^X!) Execute one shell command suspend-emacs @b(^XD) Place EMACS in the background (some UNIX systems only) @end{verbatim} @chapter(Keyboard Macroes) In many applications, you may need to repeat a series of characters or commands frequently. For example, a paper may require the frequent repetition of a complex formula or a long name. You may also have a series of EMACS commands that you invoke frequently. Keyboard macroes offer a convenient method of recording and repeating these commands. Imagine, for example, you are writing a scholarly paper on @i{Asplenium platyneuron}, the spleenwort fern. Even the dedicated botanist would probably find it a task bordering on the agonizing to type @i{Asplenium platyneuron} frequently throughout the paper. An alternative method is 'record' the name in a keyboard macro. Try it yourself. The command @b{^X(} @i(begin-macro) @index(begin-macro) starts recording the all the keystrokes and commands you input. After you've typed it, enter @b{Asplenium platyneuron}. To stop recording, type @b{^X)} @i(end-macro) @index(end-macro). EMACS has stored all the keystrokes between the two commands. To repeat the name you've stored, just enter @b{^XE} @i(execute-macro) @index(execute-macro), and the name "Asplenium platyneuron" appears. You can repeat this action as often as you want, and of course as with any EMACS command, you may precede it with a numerical argument to repeat it many times. Because EMACS records keystrokes, you may freely intermix commands and text. Unfortunately, you can only store one macro at a time. Thus, if you begin to record another macro, the previously defined macro is lost. Be careful to ensure that you've finished with one macro before defining another. If you have a series of commands that you would like to 'record' for future use, use the procedure facilities detailed in chapter @ref(macroes). @newpage @heading(Chapter @value(chapter) Summary) Chapter @value(chapter) covered keyboard macroes. You learned how to record keystrokes and how to repeat the stored sequence. @begin{verbatim} @u(Key Binding Keystroke Effect) start-macro @b{^X(} Starts recording all keyboard input end-macro @b{^X)} Stops recording keystrokes for macro execute-macro @b{^XE} Entire sequence of recorded keystrokes is replayed @end{verbatim} @chapter(MicroEMACS Procedures) @tag(macroes) Procedures, or macroes, are programs that are used to customize the editor and to perform complicated editing tasks. They may be stored in files or buffers and may be executed using an appropriate command, or bound to a particular keystroke. Portions of the standard start-up file are implemented via procedures, as well as the built in help system. The @b(M-^E) @i(run) @index(run) command causes named procedures to be executed. The @i(execute-file) @index(execute-file) command allows you to execute a procedure stored in a disk file, and the @i(execute-buffer) @index(execute-buffer) command allows you to execute a procedure stored in a buffer. Procedures are stored for easy execution by executing files that contain the store-procedure command. In a command file, the @i(store-procedure) @index(store-procedure) command takes a string argument which is the name of a procedure to store. These procedures than can be executed with the @b(M-^E) @i(run) @index(run) command. Also, giving the name of a stored procedure within another procedure will executed that named procedure as if it had been called up with the @i(run) command. Some fairly length examples of MicroEMACS procedures can be seen by examining the standard files that come with EMACS. The @b(emacs.rc) @index(emacs.rc) file (called @b[.emacsrc]@index[.emacsrc]) under UNIX) is the MicroEMACS command file which is executed when EMACS is normally run. It contains a number of different stored procedures along with the lines to setup and display the Function key window @index(function key window) and to call up other procedures and command files using function keys. There are many different aspects to the language within MicroEMACS. Editor commands are the various commands that manipulate text, buffers, windows, et cetera, within the editor. Directives are commands which control what lines get executed within a macro. Also there are various types of variables. Environmental variables both control and report on different aspects of the editor. User variables hold string values which may be changed and inspected. Buffer variables allow text to be placed into variables. Interactive variable allow the program to prompt the user for information. Functions can be used to manipulate all these variables. @section(Constants) All constants and variable contents in EMACS are stored as strings of characters. Numbers are stored digit by digit as characters. This allows EMACS to be "typeless", not having different variables types be legal in different contexts. This has the disadvantage of forcing the user to be more careful about the context of the statements variables are placed in, but in turn gives them more flexibility in where they can place variables. Needless to say, this also allows EMACS's expression evaluator to be both concise and quick. Wherever statements need to have arguments, it is legal to place constants. A constant is a double quote character, followed by a string of characters, and terminated by another double quote character. To represent various special characters within a constant, the tilde (~) @index(tilde, special use) character is used. The character following the tilde is interpreted according to the following table: @begin(verbatim) @u(Sequence Result) ~n EMACS newline character (breaks lines) ~r ^M carriage return ~l ^J linefeed ~~ ~ tilde ~b ^H backspace ~f ^L formfeed ~t ^I tab ~" " quote @end(verbatim) Any character not in the table which follows a tilde will be passed unmodified. This action is similar to the @b(^Q) @i(quote-character) command available from the keyboard. EMACS may use different characters for line terminators on different computers. The ~n combination will always get the proper line terminating sequence for the current system. The double quotes around constants are not needed if the constant contains no internal white space and it also does not happen to meet the rules for any other EMACS commands, directives, variables, or functions. This is reasonable useful for numeric constants. @section(Variables) Variables in MicroEMACS procedures can be used to return values within expressions, as repeat counts to editing commands, or as text to be inserted into buffers and messages. The value of these variables is set using the set @b(^XA) command. For example, to set the current fill column to 64 characters, the following macro line would be used: set $fillcol 64 or to have the contents of @b(%name) inserted at the point in the current buffer, the command to use would be: insert-string %name @subsection(Environmental Variables) "What good is a quote if you can't change it?" These variables are used to change different aspects of the way the editor works. Also they will return the current settings if used as part of an expression. All environmental variable names begin with a dollar sign ($) and are in lower case. @begin(description) @index($acount) $acount@\The countdown of inserted characters until the next save-file. @index($asave) $asave@\The number of inserted characters between automatic file-saves in ASAVE mode. @index($bufhook) $bufhook@\The function named in this variable is run when a buffer is entered. It can be used to implement modes which are specific to a paricular file or file type. @index($cbflags) $cbflags@\Current buffer attribute flags (See appendix G for details). @index($cbufname) $cbufname@\Name of the current buffer. @index($cfname) $cfname@\File name of the current buffer. @index($cmdhook) $cmdhook@\Name of function to run before accepting a command. This is by default set to @i(nop). @index($cmode) $cmode@\Integer containing the mode of the current buffer. (See Appendix F for values). @index($curchar) $curchar@\Ascii value of the character currently at the point. @index($curcol) $curcol@\Current column of point in current buffer. @index($curline) $curline@\Current line of point in current buffer. @index($curwidth) $curwidth@\Number of columns used currently. @index($curwind) $curwind@\Current window number. @index($cwline) $cwline@\Current display line in current window. @index($debug) $debug@\Flag to trigger macro debugging. @index($deskcolor) $deskcolor@\Color to use for current desktop, default to BLACK. @index($diagflag) $diagflag@\If set to TRUE, diagonal dragging of text and mode lines is enabled. If FALSE, text and modelines can only be dragged horizontally or vertically at one time. @index($discmd) $discmd@\Controls the echoing of command prompts. Default is TRUE. @index($disinp) $disinp@\Controls the echoing of input at the command prompts. Default is TRUE. @index($disphigh) $disphigh@\If set to TRUE, high-bit characters (single byte characters that are greater than 127 in value) will be displayed in a pseudo-control format. The characters "^!" will lead off the sequence, followed by the character stripped of its high bit. Default is FALSE. @index($exbhook) $exbhook@\This variable holds the name of a function or macro which is run whenever you are switching out of a buffer. @index($fcol) $fcol@\The current line position being displayed in the first column of the current window. @index($fillcol) $fillcol@\Current fill column. @index($flicker) $flicker@\Flicker Flag set to TRUE if IBM CGA set to FALSE for most others. @index($fmtlead) $fmtlead@\lists all formatter command leadin characters. Lines beginning with these characters will be considered the beginning of paragraphs. @index($gflags) $gflags@\Global flags controlling some EMACS internal functions (See appendix G for details). @index($gmode) $gmode@\Global mode flags. (See Appendix F for values). @index($hardtab) $hardtab@\Number of spaces between hard tab stops. Normally 8, this can be used to change indentation only within the editor.@index(tabs) @index($hjump) $hjump@\The number in here tells EMACS how many columns to scroll the screen horizontally when a horizontal scroll is required. @index($hscroll) $hscroll@\This flag determines if EMACS will scroll the entire current window horizontally, or just the current line. The default value, TRUE, results in the entire current window being shifted left and right when the cursor goes off the edge of the screen. @index($kill) $kill@\This contains the first 127 characters currently in the kill buffer and can be used to set the contents of the kill buffer. @index($language) $language@\[READ ONLY]Contains the name of the language which the current EMACS's message will display. (Currently EMACS is available in English, French, Spanish, Latin, Portuguese, Dutch, German, and Pig Latin). @index($lastkey) $lastkey@\[READ ONLY]Last keyboard character typed. @index($lastmesg) $lastmesg@\[READ ONLY]Contains the text of the last message which emacs wrote on the command line. @index($line) $line@\The current line in the current buffer can be retrieved and set with this environment variable. @index($lterm) $lterm@\Character(s) to write as a line terminator when writing a file to disk. Default is null, which causes a '\n' character to be written. Not all operating systems support this. @index($lwidth) $lwidth@\[READ ONLY]Returns the number of characters in the current line. @index($match) $match@\[READ ONLY]Last string matched in a search. @index($modeflag) $modeflag@\Determines if mode lines are currently displayed. @index($msflag) $msflag@\If TRUE, the mouse (if present) is active. If FALSE, no mouse cursor is displayed, and no mouse actions are taken.@index(mouse) @index($numwind) $numwind@\The number of windows displayed. @index($oldcrypt) $oldcrypt@\Use the old method of encryption (which had a bug in it). Default is FALSE. This variable was dropped from EMACS with version 3.12. @index($orgrow) $orgrow@\The desktop row position of current screen. @index($orgcol) $orgcol@\The desktop column position of current screen. @index($pagelen) $pagelen@\The number of screen lines used currently. @index($palette) $palette@\A string used to control the palette register settings on graphics versions. The usual form consists of groups of three octal digits setting the red, green, and blue levels.@index(color pallette) @index($paralead) $paralead@\A string containing all paragraph start characters. @index($pending) $pending@\[READ ONLY]A flag used to determine if there are user keystrokes waiting to be processed. @index($popflag) $popflag@\Use pop-up windows. Default is TRUE. @index($posflag) $posflag@\Display the line and column position on the modeline. Default is FALSE. @index($progname) $progname@\[READ ONLY]Always contains the string "MicroEMACS" for standard MicroEMACS. Could be something else if EMACS is incorporated as part of someone else's program. @index($ram) $ram@\The amount of remaining memory if MicroEMACS was compiled with RAMSIZE set. A debugging tool. @index($readhook) $readhook@\This variable holds the name of a function to execute whenever a file is read into EMACS. Normally, using the standard @b(emacs.rc) file, this is bound to a function which places EMACS into CMODE if the extension of the file read is .c or .h. @index($region) $region@\Contains the string of the current region. It will truncate at the stringsize limit, 255. @index($replace) $replace@\The current replace pattern used in replace commands. @index($rval) $rval@\This contains the return value from the last subprocess which was invoked from EMACS. @index($scrname) $scrname@\The current screen name. @index($search) $search@\The current search pattern used in search and replace commands. @index($searchpnt) $searchpnt@\Set the placement of the of the cursor on a successful search match. $searchpnt = 0 (the default), causes the cursor to be placed at the end of the matched text on forward searches, and at the beginning of the text on reverse searches. $searchpnt = 1 causes the cursor to be placed at the the beginning of the matched text regardless of the search direction, while $searchpnt = 2 causes the cursor to be placed at the end. @index($seed) $seed@\Integer seed of the random number generator. @index($softtab) $softtab@\Number of spaces inserted by EMACS when the handle-tab command (which is normally bound to the TAB key) is invoked.@index(tabs) @index($sres) $sres@\Current screen resolution (CGA, MONO, EGA or VGA on the IBM-PC driver. LOW, MEDIUM, HIGH or DENSE on the Atari ST1040, NORMAL on most others).@index(screen resolution) @index($ssave) $ssave@\A variable which flags EMACS's method of saving files. If set to TRUE, EMACS will write all files out to a temporary file, delete the original, then rename the temporary to the old file name. The default value of this is TRUE. @index($sscroll) $sscroll@\When set to TRUE, EMACS will smoothly scroll windows one line at a time when cursoring off the ends of the current window. Default is FALSE. @index($status) $status@\[READ ONLY]Status of the success of the last command (TRUE or FALSE). This is usually used with !force to check on the success of a search, or a file operation. @index($sterm) $sterm@\This is the character used to terminate search string inputs. The default for this is the last key bound to @i(meta-prefix). @index($target) $target@\Current target for line moves (setting this fools EMACS into believing the last command was a line move). @index($time) $time@\[READ ONLY]Contains a string corresponding to the current date and time. Usually this is in a form similar to "Mon May 09 10:10:58 1988". Not all operating systems will support this. @index($timeflag) $timeflag@\Flag to determine if the time of day is displayed on the modeline. Default is FALSE. The time is updated only after a keystroke. @index($tpause) $tpause@\Controls the length of the pause to display a matched fence when the current buffer is in CMODE and a close fence has been typed. @index($version) $version@\[READ ONLY]Contains the current MicroEMACS version number. @index($wchars) $wchars@\When set, MicroEMACS uses the characters listed in it to determine if it is in a word or not. If it is not set (the default), the characters it uses are the upper and lower case letters, and the underscore. @index($wline) $wline@\Number of display lines in current window. @index($wraphook) $wraphook@\This variable contains the name of an EMACS function which is executed when a buffer is in WRAP mode and it is time to wrap. By default this is bound to @i(wrap-word). @index($writehook) $writehook@\This variable contains the name of an EMACS function or macro which is invoked whenever EMACS attempts to write a file out to disk. This is executed before the file is written, allowing you to process a file on the way out. @index($xpos) $xpos@\The column the mouse was at the last mouse button press. @index($yankflag) $yankflag@\Controls the placement of the cursor after a yank command or an insert. When $yankflag is FALSE (the default), the cursor is placed at the end of the yanked or inserted text. When it is TRUE, the cursor remains at the start of the text. @index($ypos) $ypos@\The line which the mouse was on during the last mouse button press. @end(description) @subsection(User variables) User variables allow you to store strings and manipulate them. These strings can be pieces of text, numbers (in text form), or the logical values @b(TRUE) and @b(FALSE). These variables can be combined, tested, inserted into buffers, and otherwise used to control the way your macroes execute. At the moment, up to 512 user variables may be in use in one editing session. All users variable names must begin with a percent sign (%) and may contain any printing characters. Only the first 10 characters are significant (IE differences beyond the tenth character are ignored). Most operators will truncate strings to a length of 128 characters. @subsection(Buffer Variables) Buffer variables are special in that they can only be queried and cannot be set. What buffer variables are is a way to take text from a buffer and place it in a variable. For example, if I have a buffer by the name of RIGEL2, and it contains the text: @begin(verbatim) @begin(group) Richmond Lafayette <*>Bloomington (where <*> is the current point) Indianapolis Gary =* MicroEMACS 3.12 (WRAP) == rigel2 == File: /data/rigel2.txt ===== @end(group) @end(verbatim) and within a command I reference #rigel2, like: insert-string #rigel2 MicroEMACS would start at the current point in the RIGEL2 buffer and grab all the text up to the end of that line and pass that back. Then it would advance the point to the beginning of the next line. Thus, after our last command executes, the string "Bloomington" gets inserted into the current buffer, and the buffer RIGEL2 now looks like this: @begin(verbatim) @begin(group) Richmond Lafayette Bloomington <*>Indianapolis (where <*> is the current point) Gary =* MicroEMACS 3.12 (WRAP) == rigel2 == File: /data/rigel2.txt ===== @end(group) @end(verbatim) as you have probably noticed, a buffer variable consists of the buffer name, preceded by a pound sign (#). @subsection(Interactive variables) Interactive variables are actually a method to prompt the user for a string. This is done by using an at sign (@@) followed either with a quoted string, or a variable containing a string. The string is the placed on the bottom line, and the editor waits for the user to type in a string. Then the string typed in by the users is returned as the value of the interactive variable. For example: @begin(verbatim) set %quest "What file? " find-file @@%quest @end(verbatim) will ask the user for a file name, and then attempt to find it. Note also that complex expressions can be built up with these operators, such as: @verbatim(set %default "file1" @@&cat &cat "File to decode[" %default "]: ") which prompts the user with the string: File to decode[file1]: @section(Functions) Functions can be used to act on variables in various ways. Functions can have one, two, or three arguments. These arguments will always be placed after the function on the current command line. For example, if we wanted to increase the current fill column by two, using emacs's set (^XA) command, we would write: @begin(group) @begin(verbatim) set $fillcol &add $fillcol 2 \ \ \ \ \____second operand \ \ \ \_________first operand \ \ \_______________function to execute \ \_____________________variable to set \___________________________set (^XA) command @end(verbatim) @end(group) Function names always begin with the ampersand (&) character, and are only significant to the first three characters after the ampersand. Functions will normal expect one of three types of arguments, and will automatically convert types when needed. Different argument types include: @begin(description) @\an ascii string of digits which is interpreted as a numeric value. Any string which does not start with a digit or a minus sign (-) will be considered zero. @\An arbitrary string of characters. At the moment, strings are limited to 128 characters in length. @\A logical value consisting of the string "TRUE" or "FALSE". Numeric strings will also evaluate to "FALSE" if they are equal to zero, and "TRUE" if they are non-zero. Arbitrary text strings will have the value of "FALSE". @end(description) A list of the currently available functions follows. Functions are always used in lower case, the uppercase letters in the function table are the short form of the function (IE &div for ÷). @begin(verbatim) Numeric Functions: (returns ) &ADD Add two numbers &SUB Subtract the second number from the first &TIMes Multiply two numbers &DIVide Divide the first number by the second giving an integer result &MOD Return the reminder of dividing the first number by the second &NEGate Multiply the arg by -1 &LENgth Returns length of string &SINdex Finds the position of within . Returns zero if not found. &ASCii Return the ascii code of the first character in &RND Returns a random integer between 1 and &ABS Returns the absolute value of &BANd Bitwise AND function &BOR Bitwise OR function &BXOr Bitwise XOR function &BNOt Bitwise NOT function String manipulation functions: (returns ) &CAT Concatenate the two strings to form one &LEFt return the leftmost characters from &RIGht return the rightmost characters from &MID Starting from position in , return characters. &REVerse return a string with reversed-ordered characters &UPPer Uppercase &LOWer Lowercase &CHR return a string with the character represented by ascii code >C returns a string of characters containing a EMACS command input from the user >K return a string containing a single keystroke from the user &ENV If the operating system is capable, this returns the environment string associated with &BIND return the function name bound to the keystroke &XLATE &FINd Find the named file along the path and return its full file specification or an empty string if none exists &TRIM Trim the trailing whitespace from a string Logical Testing functions: (returns ) &NOT Return the opposite logical value &AND Returns TRUE if BOTH logical arguments are TRUE &OR Returns TRUE if either argument is TRUE &EQUal If and are numerically equal, return TRUE &LESs If is less than , return TRUE. &GREater If is greater than , return TRUE. &SEQual If the two strings are the same, return TRUE. &SLEss If is less alphabetically than , return TRUE. &SGReater If is alphabetically greater than or equal to , return TRUE. &EXIst Does the named file exist? &ISNum Is the given argument a legitimate number? Special Functions: &GROup Return group as set by a MAGIC mode search. &SUPper Translate the first char in to the first char in when uppercasing. &SLOwer Translate the first char in to the first char in when lowercasing. &INDirect Evaluate as a variable. @end(verbatim) This last function deserves more explanation. The &IND function evaluates its argument, takes the resulting string, and then uses it as a variable name. For example, given the following code sequence: @begin(verbatim) ; set up reference table set %one "elephant" set %two "giraffe" set %three "donkey" set %index "%two" insert-string &ind %index @end(verbatim) the string "giraffe" would have been inserted at the point in the current buffer. This indirection can be safely nested up to about 10 levels. @section(Directives) Directives are commands which only operate within an executing procedure, IE they do not make sense as a single command. As such, they cannot be called up singly or bound to keystroke. Used within command files, they control what lines are executed and in what order. Directives always start with the exclamation mark (!) character and must be the first non-white space placed on a line. Directives executed interactively (via the execute-command-line command) will be ignored. @subsection(!ENDM Directive) This directive is used to terminate a procedure or macro being stored. For example, if a file is being executed contains the text: @begin(verbatim) ; Read in a file in view mode, and make the window red store-procedure get-red-viewed-file find-file @@"File to view: " add-mode "view" add-mode "red" !endm print "[Consult procedure has been loaded]" @end(verbatim) only the lines between the store-macro command and the !ENDM directive are stored in procedure get-red-viewd-file. Both named procedures and numbered macroes (via the @i(store-macro) command) should be terminated with this directive. @subsection(!FORCE Directive) When MicroEMACS executes a procedure, if any command fails, the procedure is terminated at that point. If a line is preceded by a !FORCE directive, execution continues whether the command succeeds or not. For example: @begin(verbatim) ; Merge the top two windows save-window ;remember what window we are at 1 next-window ;go to the top window delete-window ;merge it with the second window !force restore-window ;This will continue regardless add-mode "red" @end(verbatim) Often this is used together with the $status environment variable to test if a command succeeded. For example: @begin(verbatim) set %seekstring @"String to Find: " !force search-forward %seekstring !if &seq $status TRUE print "Your string is Found" !else print "No such STRING!" !endif @end(verbatim) @subsection(!IF, !ELSE, and !ENDIF Directives) This directive allows statements only to be executed if a condition specified in the directive is met. Every line following the !IF directive, until the first !ELSE or !ENDIF directive, is only executed if the expression following the !IF directive evaluates to a TRUE value. For example, the following commands creates the portion of a text file automatically. (yes believe me, this will be easier to understand then that last explanation....) @begin(verbatim) !if &sequal %curplace "timespace vortex" insert-string "First, rematerialize~n" !endif !if &sequal %planet "earth" ;If we have landed on earth... !if &sequal %time "late 20th century" ;and we are then write-message "Contact U.N.I.T." !else insert-string "Investigate the situation....~n" insert-string "(SAY 'stay here Sara')~n" !endif !else set %conditions @@"Atmosphere conditions outside? " !if &sequal %conditions "safe" insert-string &cat "Go outside......" "~n" insert-string "lock the door~n" !else insert-string "Dematerialize..try somewhen else" newline !endif !endif @end(verbatim) @subsection(!GOTO Directive) Flow can be controlled within a MicroEMACS procedure using the !GOTO directive. It takes as an argument a label. A label consists of a line starting with an asterisk (*) and then an alphanumeric label. Only labels in the currently executing procedure can be jumped to, and trying to jump to a non-existing label terminates execution of a procedure. For example: @begin(verbatim) ;Create a block of DATA statements for a BASIC program insert-string "1000 DATA " set %linenum 1000 *nxtin update-screen ;make sure we see the changes set %data @@"Next number: " !if &equal %data 0 !goto finish !endif !if &greater $curcol 60 2 delete-previous-character newline set %linenum &add %linenum 10 insert-string &cat %linenum " DATA " !endif insert-string &cat %data ", " !goto nxtin *finish 2 delete-previous-character newline @end(verbatim) @subsection(!WHILE and !ENDWHILE Directives) This directive allows you to set up repetitive tasks easily and efficiently. If a group of statements need to be executed while a certain condition is true, enclose them with a while loop. For example, @begin(verbatim) !while &less $curcol 70 insert-string &cat &cat "[" #stuff "]" !endwhile @end(verbatim) places items from buffer "item" in the current line until the cursor is at or past column 70. While loops may be nested and can contain and be the targets of !GOTOs with no ill effects. Using a while loop to enclose a repeated task will run much faster than the corresponding construct using !IFs. @subsection(!BREAK Directive) This lets you abort out of the most executing currently inner while loop, regardless of the condition. It is often used to abort processing for error conditions. For example: @begin(verbatim) ; Read in files and substitute "begining" with "beginning" set %filename #list !while ¬ &seq %filename "" !force find-file %filename !if &seq $status FALSE write-message "[File read error]" !break !endif beginning-of-file replace-string "begining" "beginning" save-file set %filename #list !endwhile @end(verbatim) This while loop will process files until the list is exhausted or there is an error while reading a file. @subsection(!RETURN Directive) The !RETURN Directive causes the current procedure to exit, either returning to the caller (if any) or to interactive mode. For example: @begin(verbatim) ; Check the monitor type and set %mtyp !if &sres "CGA" set %mtyp 1 !return !else set %mtyp 2 !endif insert-string "You are on a MONOCHROME machine!~n" @end(verbatim) @chapter(Debugging MicroEMACS Procedures) @index(debugging)@index($debug)When developing new procedures, it is very convenient to be able to trace their execution to discover errors. The $debug environment variable enables procedure debugging. While this variable is TRUE, emacs will stop at each line it intends to execute and allow you to view it, and issue a number of different commands to help determine how the procedure is executing. For example, we will step through the procedure which toggles the function key window off. The first thing to do, is to set $debug, using the @b(^XA) @i(set) command. Type ^XA and emacs will prompt you on the command line with "Variable to set: ". Type in "$debug" and press the enter key. Emacs will then ask "Value: ". Type in "TRUE" (in capital letters) and press the enter key. While macro debugging is enabled (as it is now) emacs will report each time a variable is assigned a value, by displaying the variable and its value on the command line. Right now, @flushleft{((($debug <- TRUE)))} appears on the command line to tell you that $debug now has been assigned the value of TRUE. Press the space bar to continue. Now, lets try to debug a macro. Press function key 5 which normally toggles the function key window. The first thing that appears is: @flushleft{<<<[Macro 01]:!if %rcfkeys>>>} At this point, emacs is waiting for a command. It is prepared to see if the user variable %rcfkeys is TRUE, and execute some lines if they are. Suppose we want to see the value of this variable, type the letter "e" to evaluate an expression. Emacs will prompt with "EXP: ". Type "%rcfkeys" followed by the enter key. Emacs should then respond with "TRUE" to indicate that the function key window is currently on screen. Press the space bar to allow the !if directive to execute. Emacs will decide that it is TRUE, and then display the next command to execute. @flushleft{<<<[Macro 01]:!goto rcfoff>>>} Notice emacs tells us what procedure we are currently executing (in this case, the macro bound to execute-macro-1). Press the space bar again to execute the !goto directive. @flushleft{<<<[Macro 01]:save-window>>>} Emacs is saving the position of the current window so that it can attempt to return to it after it has brought up the function key window. [...THIS CHAPTER IS NOT FINISHED...] @chapter(Key Bindings, What they are and why) One of the features which makes MicroEMACS very adaptable is its ability to use different keystrokes to execute different commands. The process of changing the particular command a key invokes is called @i(rebinding)@index(rebinding). This allows us to make the editor look like other popular editors and programs. Each command in MicroEMACS has a name which is used for binding purposes. For example, the command to move the cursor down one page is called @i(next-line) and is normally bound to the ^N key. If you decided that you also wanted to use the ^D key to move the cursor down one line, you would use the M-K @i(bind-to-key)@index(bind-to-key) command. EMACS would respond with ": bind-to-key " on the command line and allow you to type in a command name. Then type in the name of the command you want to change, in this case @i(next-line), followed by the key. EMACS will then wait for you to type in the keys you want to activate the named function. Type a single ^D. From now on, typing ^D will cause EMACS to move down one line, rather than its original function of deleting characters. To find out the name of a command, consult the list of valid EMACS commands in Appendix B. Also, you can use the ^X? @i(describe-key)@index(describe-key) command to look up the name of a command. Type ^X? and then the key to use that command, and EMACS will show you the name of the command. After you have experimented with changing your key bindings, you may decide that you want to change some bindings permanently. To have EMACS rebind keys to your pleasure each time you start EMACS, you can add statements to the end of your startup file (@b(emacs.rc) or @b(.emacsrc) depending on the system). For example, bind-to-key next-line ^D Notice, that control D character in the startup file is represented visibly as an uparrow key followed by a capital D. To know how to represent any keys you want to bind, use the @i(describe-key) command on the key, and use the sequence that is displayed. bind-to-key split-current-window FN1 This example would make function key 1 activate the command that splits the current window in two. EMACS will let you define a large number of keys, but will report "Binding table FULL!" when it runs out of space to bind keys. Normally EMACS will allow up to 512 key bindings (including approx. 300 originally bound keys). If you want to get a current listing of all the commands and the keys bound to them, use the @i(describe-bindings) @index(describe-bindings) command. Notice, that this command is not bound to any keys! There are some @index(key bindings, declined) key bindings that cannot be made without special precautions. Alternative bindings for ^X, META, ^G, and ^U (which bind respectively to ctlx-prefix, meta-prefix, abort-command, and universal-argument) must be made before re-binding ^X, META, ^G, or ^U. The reason is to protect the innocent user from losing the prefix and other commands inadvertantly. @appendix(MicroEMACS Command Line Switches and Startup Files) @index(startup files) When EMACS first executes, it always searches for a file, called @b(.emacsrc) @i(under most UNIX systems) or @b(emacs.rc) @i(on most other systems) @index(emacs.rc) @index(.emacsrc) which it will execute as EMACS macroes before it reads in the named source files. This file normally contains EMACS macroes to bind the function keys to useful functions and load various useful macroes. The contents of this file will probably vary from system to system and can be modified by the user as desired. When searching for this file, EMACS looks for it in this order. First, it attempts to find a definition for "@b(HOME)" in the environment. It will look in that directory first. Then it searches all the directories listed in the "@b(PATH)" environment variable. Then it looks through a list of predefined standard directories which vary from system to system. Finally, failing all of these, it looks in the current directory. This is also the same method EMACS uses to look up any files to execute, and to find it's help file @b(EMACS.HLP). @index(PATH environment variable) @index(HOME environment variable) @index(Help File) @index(command line) @index(switches) On computers that call up EMACS via a command line process, such as MSDOS and UNIX, there are different things that can be added to the command line to control the way EMACS operates. These can be s