freebsd-nq/contrib/groff/doc/groff-7

This is groff, produced by makeinfo version 4.3d from ./groff.texinfo.

This manual documents GNU `troff' version 1.19.

   Copyright (C) 1994-2000, 2001, 2002, 2003 Free Software Foundation,
Inc.

     Permission is granted to copy, distribute and/or modify this
     document under the terms of the GNU Free Documentation License,
     Version 1.1 or any later version published by the Free Software
     Foundation; with no Invariant Sections, with the Front-Cover texts
     being `A GNU Manual," and with the Back-Cover Texts as in (a)
     below.  A copy of the license is included in the section entitled
     `GNU Free Documentation License."

     (a) The FSF's Back-Cover Text is: `You have freedom to copy and
     modify this GNU Manual, like GNU software.  Copies published by
     the Free Software Foundation raise funds for GNU development."
   
INFO-DIR-SECTION Typesetting
START-INFO-DIR-ENTRY
* Groff: (groff).               The GNU troff document formatting system.
END-INFO-DIR-ENTRY


File: groff,  Node: Page Location Traps,  Next: Diversion Traps,  Prev: Traps,  Up: Traps

Page Location Traps
-------------------

   "Page location traps" perform an action when `gtroff' reaches or
passes a certain vertical location on the page.  Page location traps
have a variety of purposes, including:

   * setting headers and footers

   * setting body text in multiple columns

   * setting footnotes

 - Request: .vpt flag
 - Register: \n[.vpt]
     Enable vertical position traps if FLAG is non-zero, or disables
     them otherwise.  Vertical position traps are traps set by the `wh'
     or `dt' requests.  Traps set by the `it' request are not vertical
     position traps.  The parameter that controls whether vertical
     position traps are enabled is global.  Initially vertical position
     traps are enabled.  The current setting of this is available in the
     `.vpt' read-only number register.

     Note that a page can't be ejected if `vpt' is set to zero.

 - Request: .wh dist [macro]
     Set a page location trap.  Non-negative values for DIST set the
     trap relative to the top of the page; negative values set the trap
     relative to the bottom of the page.  Default scaling indicator is
     `v'.

     MACRO is the name of the macro to execute when the trap is sprung.
     If MACRO is missing, remove the first trap (if any) at DIST.

     The following is a simple example of how many macro packages set
     headers and footers.


          .de hd                \" Page header
          '  sp .5i
          .  tl 'Title''date'
          '  sp .3i
          ..
          .
          .de fo                \" Page footer
          '  sp 1v
          .  tl ''%''
          '  bp
          ..
          .
          .wh 0   hd            \" trap at top of the page
          .wh -1i fo            \" trap one inch from bottom

     A trap at or below the bottom of the page is ignored; it can be
     made active by either moving it up or increasing the page length
     so that the trap is on the page.

     It is possible to have more than one trap at the same location; to
     do so, the traps must be defined at different locations, then
     moved together with the `ch' request; otherwise the second trap
     would replace the first one.  Earlier defined traps hide later
     defined traps if moved to the same position (the many empty lines
     caused by the `bp' request are omitted in the following example):


          .de a
          .  nop a
          ..
          .de b
          .  nop b
          ..
          .de c
          .  nop c
          ..
          .
          .wh 1i a
          .wh 2i b
          .wh 3i c
          .bp
              => a b c


          .ch b 1i
          .ch c 1i
          .bp
              => a


          .ch a 0.5i
          .bp
              => a b


 - Register: \n[.t]
     A read-only number register holding the distance to the next trap.

     If there are no traps between the current position and the bottom
     of the page, it contains the distance to the page bottom.  In a
     diversion, the distance to the page bottom is infinite (the
     returned value is the biggest integer which can be represented in
     `groff') if there are no diversion traps.

 - Request: .ch macro [dist]
     Change the location of a trap.  The first argument is the name of
     the macro to be invoked at the trap, and the second argument is
     the new location for the trap (note that the parameters are
     specified in opposite order as in the `wh' request).  This is
     useful for building up footnotes in a diversion to allow more
     space at the bottom of the page for them.

     Default scaling indicator for DIST is `v'.  If DIST is missing,
     the trap is removed.


 - Register: \n[.ne]
     The read-only number register `.ne' contains the amount of space
     that was needed in the last `ne' request that caused a trap to be
     sprung.  Useful in conjunction with the `.trunc' register.  *Note
     Page Control::, for more information.

     Since the `.ne' register is only set by traps it doesn't make much
     sense to use it outside of trap macros.

 - Register: \n[.trunc]
     A read-only register containing the amount of vertical space
     truncated by the most recently sprung vertical position trap, or,
     if the trap was sprung by an `ne' request, minus the amount of
     vertical motion produced by the `ne' request.  In other words, at
     the point a trap is sprung, it represents the difference of what
     the vertical position would have been but for the trap, and what
     the vertical position actually is.

     Since the `.trunc' register is only set by traps and it doesn't
     make much sense to use it outside of trap macros.

 - Register: \n[.pe]
     A read-only register which is set to 1 while a page is ejected with
     the `bp' request (or by the end of input).

     Outside of traps this register is always zero.  In the following
     example, only the second call to `x' is caused by `bp'.


          .de x
          \&.pe=\\n[.pe]
          .br
          ..
          .wh 1v x
          .wh 4v x
          A line.
          .br
          Another line.
          .br
              => A line.
                        .pe=0
                        Another line.
          
                        .pe=1


   An important fact to consider while designing macros is that
diversions and traps do not interact normally.  For example, if a trap
invokes a header macro (while outputting a diversion) which tries to
change the font on the current page, the effect will not be visible
before the diversion has completely been printed (except for input
protected with `\!' or `\?') since the data in the diversion is already
formatted.  In most cases, this is not the expected behaviour.


File: groff,  Node: Diversion Traps,  Next: Input Line Traps,  Prev: Page Location Traps,  Up: Traps

Diversion Traps
---------------

 - Request: .dt dist macro
     Set a trap _within_ a diversion.  DIST is the location of the trap
     (identical to the `wh' request; default scaling indicator is `v')
     and MACRO is the name of the macro to be invoked.  The number
     register `.t' still works within diversions.  *Note Diversions::,
     for more information.


File: groff,  Node: Input Line Traps,  Next: Blank Line Traps,  Prev: Diversion Traps,  Up: Traps

Input Line Traps
----------------

 - Request: .it n macro
 - Request: .itc n macro
     Set an input line trap.  N is the number of lines of input which
     may be read before springing the trap, MACRO is the macro to be
     invoked.  Request lines are not counted as input lines.

     For example, one possible use is to have a macro which prints the
     next N lines in a bold font.


          .de B
          .  it \\$1 B-end
          .  ft B
          ..
          .
          .de B-end
          .  ft R
          ..

     The `itc' request is identical except that an interrupted text
     line (ending with `\c') is not counted as a separate line.

     Both requests are associated with the current environment (*note
     Environments::); switching to another environment disables the
     current input trap, and going back reactivates it, restoring the
     number of already processed lines.


File: groff,  Node: Blank Line Traps,  Next: End-of-input Traps,  Prev: Input Line Traps,  Up: Traps

Blank Line Traps
----------------

 - Request: .blm macro
     Set a blank line trap.  `gtroff' executes MACRO when it encounters
     a blank line in the input file.


File: groff,  Node: End-of-input Traps,  Prev: Blank Line Traps,  Up: Traps

End-of-input Traps
------------------

 - Request: .em macro
     Set a trap at the end of input.  MACRO is executed after the last
     line of the input file has been processed.

     For example, if the document had to have a section at the bottom
     of the last page for someone to approve it, the `em' request could
     be used.


          .de approval
          .  ne 5v
          .  sp |(\\n[.t] - 6v)
          .  in +4i
          .  lc _
          .  br
          Approved:\t\a
          .  sp
          Date:\t\t\a
          ..
          .
          .em approval



File: groff,  Node: Diversions,  Next: Environments,  Prev: Traps,  Up: gtroff Reference

Diversions
==========

   In `gtroff' it is possible to "divert" text into a named storage
area.  Due to the similarity to defining macros it is sometimes said to
be stored in a macro.  This is used for saving text for output at a
later time, which is useful for keeping blocks of text on the same
page, footnotes, tables of contents, and indices.

   For orthogonality it is said that `gtroff' is in the "top-level
diversion" if no diversion is active (i.e., the data is diverted to the
output device).

 - Request: .di macro
 - Request: .da macro
     Begin a diversion.  Like the `de' request, it takes an argument of
     a macro name to divert subsequent text into.  The `da' macro
     appends to an existing diversion.

     `di' or `da' without an argument ends the diversion.

 - Request: .box macro
 - Request: .boxa macro
     Begin (or appends to) a diversion like the `di' and `da' requests.
     The difference is that `box' and `boxa' do not include a
     partially-filled line in the diversion.

     Compare this:


          Before the box.
          .box xxx
          In the box.
          .br
          .box
          After the box.
          .br
              => Before the box.  After the box.
          .xxx
              => In the box.

     with this:


          Before the diversion.
          .di yyy
          In the diversion.
          .br
          .di
          After the diversion.
          .br
              => After the diversion.
          .yyy
              => Before the diversion.  In the diversion.

     `box' or `boxa' without an argument ends the diversion.

 - Register: \n[.z]
 - Register: \n[.d]
     Diversions may be nested.  The read-only number register `.z'
     contains the name of the current diversion (this is a string-valued
     register).  The read-only number register `.d' contains the current
     vertical place in the diversion.  If not in a diversion it is the
     same as register `nl'.

 - Register: \n[.h]
     The "high-water mark" on the current page.  It corresponds to the
     text baseline of the lowest line on the page.  This is a read-only
     register.


          .tm .h==\n[.h], nl==\n[nl]
              => .h==0, nl==-1
          This is a test.
          .br
          .sp 2
          .tm .h==\n[.h], nl==\n[nl]
              => .h==40, nl==120

     As can be seen in the previous example, empty lines are not
     considered in the return value of the `.h' register.

 - Register: \n[dn]
 - Register: \n[dl]
     After completing a diversion, the read-write number registers `dn'
     and `dl' contain the vertical and horizontal size of the diversion.
     Note that only the just processed lines are counted: For the
     computation of `dn' and `dl', the requests `da' and `boxa' are
     handled as if `di' and `box' had been used - lines which have been
     already stored in a macro are not taken into account.


          .\" Center text both horizontally & vertically
          .
          .\" Enclose macro definitions in .eo and .ec
          .\" to avoid the doubling of the backslash
          .eo
          .\" macro .(c starts centering mode
          .de (c
          .  br
          .  ev (c
          .  evc 0
          .  in 0
          .  nf
          .  di @c
          ..


          .\" macro .)c terminates centering mode
          .de )c
          .  br
          .  ev
          .  di
          .  nr @s (((\n[.t]u - \n[dn]u) / 2u) - 1v)
          .  sp \n[@s]u
          .  ce 1000
          .  @c
          .  ce 0
          .  sp \n[@s]u
          .  br
          .  fi
          .  rr @s
          .  rm @s
          .  rm @c
          ..
          .\" End of macro definitions, restore escape mechanism
          .ec


 - Escape: \!
 - Escape: \?anything\?
     Prevent requests, macros, and escapes from being interpreted when
     read into a diversion.  Both escapes take the given text and
     "transparently" embed it into the diversion.  This is useful for
     macros which shouldn't be invoked until the diverted text is
     actually output.

     The `\!' escape transparently embeds text up to and including the
     end of the line.  The `\?' escape transparently embeds text until
     the next occurrence of the `\?' escape.  Example:


          \?ANYTHING\?

     ANYTHING may not contain newlines; use `\!'  to embed newlines in
     a diversion.  The escape sequence `\?' is also recognized in copy
     mode and turned into a single internal code; it is this code that
     terminates ANYTHING.  Thus the following example prints 4.


          .nr x 1
          .nf
          .di d
          \?\\?\\\\?\\\\\\\\nx\\\\?\\?\?
          .di
          .nr x 2
          .di e
          .d
          .di
          .nr x 3
          .di f
          .e
          .di
          .nr x 4
          .f

     Both escapes read the data in copy mode.

     If `\!' is used in the top-level diversion, its argument is
     directly embedded into the `gtroff' intermediate output.  This can
     be used for example to control a postprocessor which processes the
     data before it is sent to the device driver.

     The `\?' escape used in the top-level diversion produces no output
     at all; its argument is simply ignored.

 - Request: .output string
     Emit STRING directly to the `gtroff' intermediate output (subject
     to copy-mode interpretation); this is similar to `\!' used at the
     top level.  An initial double quote in STRING is stripped off to
     allow initial blanks.

     This request can't be used before the first page has started - if
     you get an error, simply insert `.br' before the `output' request.

     Without argument, `output' is ignored.

     Use with caution!  It is normally only needed for mark-up used by a
     postprocessor which does something with the output before sending
     it to the output device, filtering out STRING again.

 - Request: .asciify div
     "Unformat" the diversion specified by DIV in such a way that ASCII
     characters, characters translated with the `trin' request, space
     characters, and some escape sequences that were formatted and
     diverted are treated like ordinary input characters when the
     diversion is reread.  It can be also used for gross hacks; for
     example, the following sets register `n' to 1.


          .tr @.
          .di x
          @nr n 1
          .br
          .di
          .tr @@
          .asciify x
          .x

     *Note Copy-in Mode::.

 - Request: .unformat div
     Like `asciify', unformat the specified diversion.  However,
     `unformat' only unformats spaces and tabs between words.
     Unformatted tabs are treated as input tokens, and spaces are
     stretchable again.

     The vertical size of lines is not preserved; glyph information
     (font, font size, space width, etc.) is retained.


File: groff,  Node: Environments,  Next: Suppressing output,  Prev: Diversions,  Up: gtroff Reference

Environments
============

   It happens frequently that some text should be printed in a certain
format regardless of what may be in effect at the time, for example, in
a trap invoked macro to print headers and footers.  To solve this
`gtroff' processes text in "environments".  An environment contains
most of the parameters that control text processing.  It is possible to
switch amongst these environments; by default `gtroff' processes text
in environment 0.  The following is the information kept in an
environment.

   * font parameters (size, family, style, glyph height and slant, space
     and sentence space size)

   * page parameters (line length, title length, vertical spacing, line
     spacing, indentation, line numbering, centering, right-justifying,
     underlining, hyphenation data)

   * fill and adjust mode

   * tab stops, tab and leader characters, escape character, no-break
     and hyphen indicators, margin character data

   * partially collected lines

   * input traps

   * drawing and fill colours

   These environments may be given arbitrary names (see *Note
Identifiers::, for more info).  Old versions of `troff' only had
environments named `0', `1', and `2'.

 - Request: .ev [env]
 - Register: \n[.ev]
     Switch to another environment.  The argument ENV is the name of
     the environment to switch to.  With no argument, `gtroff' switches
     back to the previous environment.  There is no limit on the number
     of named environments; they are created the first time that they
     are referenced.  The `.ev' read-only register contains the name or
     number of the current environment.  This is a string-valued
     register.

     Note that a call to `ev' (with argument) pushes the previously
     active environment onto a stack.  If, say, environments `foo',
     `bar', and `zap' are called (in that order), the first `ev'
     request without parameter switches back to environment `bar'
     (which is popped off the stack), and a second call switches back
     to environment `foo'.

     Here is an example:


          .ev footnote-env
          .fam N
          .ps 6
          .vs 8
          .ll -.5i
          .ev
          
          ...
          
          .ev footnote-env
          \(dg Note the large, friendly letters.
          .ev


 - Request: .evc env
     Copy the environment ENV into the current environment.

     The following environment data is not copied:

        * Partially filled lines.

        * The status whether the previous line was interrupted.

        * The number of lines still to center, or to right-justify, or
          to underline (with or without underlined spaces); they are
          set to zero.

        * The status whether a temporary indent is active.

        * Input traps and its associated data.

        * Line numbering mode is disabled; it can be reactivated with
          `.nm +0'.

        * The number of consecutive hyphenated lines (set to zero).

 - Register: \n[.w]
 - Register: \n[.cht]
 - Register: \n[.cdp]
 - Register: \n[.csk]
     The `\n[.w]' register contains the width of the last glyph added
     to the current environment.

     The `\n[.cht]' register contains the height of the last glyph
     added to the current environment.

     The `\n[.cdp]' register contains the depth of the last glyph added
     to the current environment.  It is positive for glyphs extending
     below the baseline.

     The `\n[.csk]' register contains the "skew" (how far to the right
     of the glyph's center that `gtroff' should place an accent) of the
     last glyph added to the current environment.

 - Register: \n[.n]
     The `\n[.n]' register contains the length of the previous output
     line in the current environment.


File: groff,  Node: Suppressing output,  Next: Colors,  Prev: Environments,  Up: gtroff Reference

Suppressing output
==================

 - Escape: \Onum
     Disable or enable output depending on the value of NUM:

    `\O0'
          Disable any glyphs from being emitted to the device driver,
          provided that the escape occurs at the outer level (see
          `\O[3]' and `\O[4]').  Motion is not suppressed so
          effectively `\O[0]' means _pen up_.

    `\O1'
          Enable output of glyphs, provided that the escape occurs at
          the outer level.

     `\O0' and `\O1' also reset the four registers `opminx', `opminy',
     `opmaxx', and `opmaxy' to -1.  *Note Register Index::.  These four
     registers mark the top left and bottom right hand corners of a box
     which encompasses all written glyphs.

     For example the input text:


          Hello \O[0]world \O[1]this is a test.

     produces the following output:


          Hello       this is a test.

    `\O2'
          Provided that the escape occurs at the outer level, enable
          output of glyphs and also write out to `stderr' the page
          number and four registers encompassing the glyphs previously
          written since the last call to `\O'.

    `\O3'
          Begin a nesting level.  At start-up, `gtroff' is at outer
          level.

    `\O4'
          End a nesting level.

    `\O[5PFILENAME]'
          This escape is `grohtml' specific.  Provided that this escape
          occurs at the outer nesting level write the `filename' to
          `stderr'.  The position of the image, P, must be specified
          and must be one of `l', `r', `c', or `i' (left, right,
          centered, inline).  FILENAME will be associated with the
          production of the next inline image.


File: groff,  Node: Colors,  Next: I/O,  Prev: Suppressing output,  Up: gtroff Reference

Colors
======

 - Request: .color [n]
 - Register: \n[.color]
     If N is missing or non-zero, activate colors (this is the default);
     otherwise, turn it off.

     The read-only number register `.color' is 1 if colors are active,
     0 otherwise.

     Internally, `color' sets a global flag; it does not produce a
     token.  Similar to the `cp' request, you should use it at the
     beginning of your document to control color output.

     Colors can be also turned off with the `-c' command line option.

 - Request: .defcolor ident scheme color_components
     Define color with name IDENT.  SCHEME can be one of  the following
     values: `rgb' (three components), `cmy' (three components), `cmyk'
     (four components), and `gray' or `grey' (one component).

     Color components can be given either as a hexadecimal string or as
     positive decimal integers in the range 0-65535.  A hexadecimal
     string contains all color components concatenated.  It must start
     with either `#' or `##'; the former specifies hex values in the
     range 0-255 (which are internally multiplied by 257), the latter
     in the range 0-65535.  Examples: `#FFC0CB' (pink), `##ffff0000ffff'
     (magenta).  The default color name value is device-specific
     (usually black).  It is possible that the default color for `\m'
     and `\M' is not identical.

     A new scaling indicator `f' has been introduced which multiplies
     its value by 65536; this makes it convenient to specify color
     components as fractions in the range 0 to 1 (1f equals 65536u).
     Example:


          .defcolor darkgreen rgb 0.1f 0.5f 0.2f

     Note that `f' is the default scaling indicator for the `defcolor'
     request, thus the above statement is equivalent to


          .defcolor darkgreen rgb 0.1 0.5 0.2


 - Escape: \mc
 - Escape: \m(co
 - Escape: \m[color]
     Set drawing color.  The following example shows how to turn the
     next four words red.


          \m[red]these are in red\m[] and these words are in black.

     The escape `\m[]' returns to the previous color.

     The drawing color is associated with the current environment
     (*note Environments::).

     Note that `\m' doesn't produce an input token in `gtroff'.  As a
     consequence, it can be used in requests like `mc' (which expects a
     single character as an argument) to change the color on the fly:


          .mc \m[red]x\m[]


 - Escape: \Mc
 - Escape: \M(co
 - Escape: \M[color]
     Set background color for filled objects drawn with the `\D'...''
     commands.

     A red ellipse can be created with the following code:


          \M[red]\h'0.5i'\D'E 2i 1i'\M[]

     The escape `\M[]' returns to the previous fill color.

     The fill color is associated with the current environment (*note
     Environments::).

     Note that `\M' doesn't produce an input token in `gtroff'.


File: groff,  Node: I/O,  Next: Postprocessor Access,  Prev: Colors,  Up: gtroff Reference

I/O
===

   `gtroff' has several requests for including files:

 - Request: .so file
     Read in the specified FILE and includes it in place of the `so'
     request.  This is quite useful for large documents, e.g. keeping
     each chapter in a separate file.  *Note gsoelim::, for more
     information.

     Since `gtroff' replaces the `so' request with the contents of
     `file', it makes a difference whether the data is terminated with
     a newline or not: Assuming that file `xxx' contains the word `foo'
     without a final newline, this


          This is
          .so xxx
          bar

     yields `This is foobar'.

 - Request: .pso command
     Read the standard output from the specified COMMAND and includes
     it in place of the `pso' request.

     This request causes an error if used in safer mode (which is the
     default).  Use `groff''s or `troff''s `-U' option to activate
     unsafe mode.

     The comment regarding a final newline for the `so' request is valid
     for `pso' also.

 - Request: .mso file
     Identical to the `so' request except that `gtroff' searches for
     the specified FILE in the same directories as macro files for the
     the `-m' command line option.  If the file name to be included has
     the form `NAME.tmac' and it isn't found, `mso' tries to include
     `tmac.NAME' and vice versa.

 - Request: .trf file
 - Request: .cf file
     Transparently output the contents of FILE.  Each line is output as
     if it were preceded by `\!'; however, the lines are not subject to
     copy mode interpretation.  If the file does not end with a newline,
     then a newline is added (`trf' only).  For example, to define a
     macro `x' containing the contents of file `f', use


          .di x
          .trf f
          .di

     Both `trf' and `cf', when used in a diversion, embeds an object in
     the diversion which, when reread, causes the contents of FILE to
     be transparently copied through to the output.  In UNIX `troff',
     the contents of FILE is immediately copied through to the output
     regardless of whether there is a current diversion; this behaviour
     is so anomalous that it must be considered a bug.

     While `cf' copies the contents of FILE completely unprocessed,
     `trf' disallows characters such as NUL that are not valid `gtroff'
     input characters (*note Identifiers::).

     Both requests cause a line break.

 - Request: .nx [file]
     Force `gtroff' to continue processing of the file specified as an
     argument.  If no argument is given, immediately jump to the end of
     file.

 - Request: .rd [prompt [arg1 arg2 ...]]
     Read from standard input, and include what is read as though it
     were part of the input file.  Text is read until a blank line is
     encountered.

     If standard input is a TTY input device (keyboard), write PROMPT
     to standard error, followed by a colon (or send BEL for a beep if
     no argument is given).

     Arguments after PROMPT are available for the input.  For example,
     the line


          .rd data foo bar

     with the input `This is \$2.' prints


          This is bar.


   Using the `nx' and `rd' requests, it is easy to set up form letters.
The form letter template is constructed like this, putting the
following lines into a file called `repeat.let':


     .ce
     \*(td
     .sp 2
     .nf
     .rd
     .sp
     .rd
     .fi
     Body of letter.
     .bp
     .nx repeat.let

When this is run, a file containing the following lines should be
redirected in.  Note that requests included in this file are executed
as though they were part of the form letter.  The last block of input
is the `ex' request which tells `groff' to stop processing.  If this
was not there, `groff' would not know when to stop.


     Trent A. Fisher
     708 NW 19th Av., #202
     Portland, OR  97209
     
     Dear Trent,
     
     Len Adollar
     4315 Sierra Vista
     San Diego, CA  92103
     
     Dear Mr. Adollar,
     
     .ex

 - Request: .pi pipe
     Pipe the output of `gtroff' to the shell command(s) specified by
     PIPE.  This request must occur before `gtroff' has a chance to
     print anything.

     `pi' causes an error if used in safer mode (which is the default).
     Use `groff''s or `troff''s `-U' option to activate unsafe mode.

     Multiple calls to `pi' are allowed, acting as a chain.  For
     example,


          .pi foo
          .pi bar
          ...

     is the same as `.pi foo | bar'.

     Note that the intermediate output format of `gtroff' is piped to
     the specified commands.  Consequently, calling `groff' without the
     `-Z' option normally causes a fatal error.

 - Request: .sy cmds
 - Register: \n[systat]
     Execute the shell command(s) specified by CMDS.  The output is not
     saved anyplace, so it is up to the user to do so.

     This request causes an error if used in safer mode (which is the
     default).  Use `groff''s or `troff''s `-U' option to activate
     unsafe mode.

     For example, the following code fragment introduces the current
     time into a document:


          .sy perl -e 'printf ".nr H %d\\n.nr M %d\\n.nr S %d\\n",\
                       (localtime(time))[2,1,0]' > /tmp/x\n[$$]
          .so /tmp/x\n[$$]
          .sy rm /tmp/x\n[$$]
          \nH:\nM:\nS

     Note that this works by having the `perl' script (run by `sy')
     print out the `nr' requests which set the number registers `H',
     `M', and `S', and then reads those commands in with the `so'
     request.

     For most practical purposes, the number registers `seconds',
     `minutes', and `hours' which are initialized at start-up of
     `gtroff' should be sufficient.  Use the `af' request to get a
     formatted output:


          .af hours 00
          .af minutes 00
          .af seconds 00
          \n[hours]:\n[minutes]:\n[seconds]

     The `systat' read-write number register contains the return value
     of the `system()' function executed by the last `sy' request.

 - Request: .open stream file
 - Request: .opena stream file
     Open the specified FILE for writing and associates the specified
     STREAM with it.

     The `opena' request is like `open', but if the file exists, append
     to it instead of truncating it.

     Both `open' and `opena' cause an error if used in safer mode
     (which is the default).  Use `groff''s or `troff''s `-U' option to
     activate unsafe mode.

 - Request: .write stream data
 - Request: .writec stream data
     Write to the file associated with the specified STREAM.  The
     stream must previously have been the subject of an open request.
     The remainder of the line is interpreted as the `ds' request reads
     its second argument: A leading `"' is stripped, and it is read in
     copy-in mode.

     The `writec' request is like `write', but only `write' appends a
     newline to the data.

 - Request: .writem stream xx
     Write the contents of the macro or string XX to the file
     associated with the specified STREAM.

     XX is read in copy mode, i.e., already formatted elements are
     ignored.  Consequently, diversions must be unformatted with the
     `asciify' request before calling `writem'.  Usually, this means a
     loss of information.

 - Request: .close stream
     Close the specified STREAM; the stream is no longer an acceptable
     argument to the `write' request.

     Here a simple macro to write an index entry.


          .open idx test.idx
          .
          .de IX
          .  write idx \\n[%] \\$*
          ..
          .
          .IX test entry
          .
          .close idx


 - Escape: \Ve
 - Escape: \V(ev
 - Escape: \V[env]
     Interpolate the contents of the specified environment variable ENV
     (one-character name E, two-character name EV) as returned by the
     function `getenv'.  `\V' is interpreted in copy-in mode.


File: groff,  Node: Postprocessor Access,  Next: Miscellaneous,  Prev: I/O,  Up: gtroff Reference

Postprocessor Access
====================

   There are two escapes which give information directly to the
postprocessor.  This is particularly useful for embedding POSTSCRIPT
into the final document.

 - Escape: \X'xxx'
     Embeds its argument into the `gtroff' output preceded with `x X'.

     The escapes `\&', `\)', `\%', and `\:' are ignored within `\X',
     `\ ' and `\~' are converted to single space characters.  All other
     escapes (except `\\' which produces a backslash) cause an error.

     If the `use_charnames_in_special' keyword is set in the `DESC'
     file, special characters no longer cause an error; the name XX is
     represented as `\(XX)' in the `x X' output command.  Additionally,
     the backslash is represented as `\\'.

     `use_charnames_in_special' is currently used by `grohtml' only.

 - Escape: \Yn
 - Escape: \Y(nm
 - Escape: \Y[name]
     This is approximately equivalent to `\X'\*[NAME]'' (one-character
     name N, two-character name NM).  However, the contents of the
     string or macro NAME are not interpreted; also it is permitted for
     NAME to have been defined as a macro and thus contain newlines (it
     is not permitted for the argument to `\X' to contain newlines).
     The inclusion of newlines requires an extension to the UNIX `troff'
     output format, and confuses drivers that do not know about this
     extension (*note Device Control Commands::).

   *Note Output Devices::.


File: groff,  Node: Miscellaneous,  Next: Gtroff Internals,  Prev: Postprocessor Access,  Up: gtroff Reference

Miscellaneous
=============

   This section documents parts of `gtroff' which cannot (yet) be
categorized elsewhere in this manual.

 - Request: .nm [start [inc [space [indent]]]]
     Print line numbers.  START is the line number of the _next_ output
     line.  INC indicates which line numbers are printed.  For example,
     the value 5 means to emit only line numbers which are multiples
     of 5; this defaults to 1.  SPACE is the space to be left between
     the number and the text; this defaults to one digit space.  The
     fourth argument is the indentation of the line numbers, defaulting
     to zero.  Both SPACE and INDENT are given as multiples of digit
     spaces; they can be negative also.  Without any arguments, line
     numbers are turned off.

     `gtroff' reserves three digit spaces for the line number (which is
     printed right-justified) plus the amount given by INDENT; the
     output lines are concatenated to the line numbers, separated by
     SPACE, and _without_ reducing the line length.  Depending on the
     value of the horizontal page offset (as set with the `po'
     request), line numbers which are longer than the reserved space
     stick out to the left, or the whole line is moved to the right.

     Parameters corresponding to missing arguments are not changed; any
     non-digit argument (to be more precise, any argument starting with
     a character valid as a delimiter for identifiers) is also treated
     as missing.

     If line numbering has been disabled with a call to `nm' without an
     argument, it can be reactivated with `.nm +0', using the
     previously active line numbering parameters.

     The parameters of `nm' are associated with the current environment
     (*note Environments::).  The current output line number is
     available in the number register `ln'.


          .po 1m
          .ll 2i
          This test shows how line numbering works with groff.
          .nm 999
          This test shows how line numbering works with groff.
          .br
          .nm xxx 3 2
          .ll -\w'0'u
          This test shows how line numbering works with groff.
          .nn 2
          This test shows how line numbering works with groff.

     And here the result:


           This  test shows how
           line numbering works
           999 with   groff.   This
          1000 test shows how  line
          1001 numbering works with
          1002 groff.
                This test shows how
                line      numbering
           works  with  groff.
           This test shows how
          1005  line      numbering
                works with groff.


 - Request: .nn [skip]
     Temporarily turn off line numbering.  The argument is the number
     of lines not to be numbered; this defaults to 1.

 - Request: .mc glyph [dist]
     Print a "margin character" to the right of the text.(1) (*note
     Miscellaneous-Footnote-1::)  The first argument is the glyph to be
     printed.  The second argument is the distance away from the right
     margin.  If missing, the previously set value is used; default is
     10pt).  For text lines that are too long (that is, longer than the
     text length plus DIST), the margin character is directly appended
     to the lines.

     With no arguments the margin character is turned off.  If this
     occurs before a break, no margin character is printed.

     For empty lines and lines produced by the `tl' request no margin
     character is emitted.

     The margin character is associated with the current environment
     (*note Environments::).

     This is quite useful for indicating text that has changed, and, in
     fact, there are programs available for doing this (they are called
     `nrchbar' and `changebar' and can be found in any
     `comp.sources.unix' archive.


          .ll 3i
          .mc |
          This paragraph is highlighted with a margin
          character.
          .sp
          Note that vertical space isn't marked.
          .br
          \&
          .br
          But we can fake it with `\&'.

     Result:


          This  paragraph is highlighted |
          with a margin character.       |
          
          Note that vertical space isn't |
          marked.                        |
                                         |
          But we can fake it with `\&'.  |


 - Request: .psbb filename
 - Register: \n[llx]
 - Register: \n[lly]
 - Register: \n[urx]
 - Register: \n[ury]
     Retrieve the bounding box of the PostScript image found in
     FILENAME.  The file must conform to Adobe's "Document Structuring
     Conventions" (DSC); the command searches for a `%%BoundingBox'
     comment and extracts the bounding box values into the number
     registers `llx', `lly', `urx', and `ury'.  If an error occurs (for
     example, `psbb' cannot find the `%%BoundingBox' comment), it sets
     the four number registers to zero.


File: groff,  Node: Miscellaneous-Footnotes,  Up: Miscellaneous

   (1) "Margin character" is a misnomer since it is an output glyph.


File: groff,  Node: Gtroff Internals,  Next: Debugging,  Prev: Miscellaneous,  Up: gtroff Reference

`gtroff' Internals
==================

   `gtroff' processes input in three steps.  One or more input
characters are converted to an "input token".(1) (*note Gtroff
Internals-Footnote-1::)  Then, one or more input tokens are converted
to an "output node".  Finally, output nodes are converted to the
intermediate output language understood by all output devices.

   Actually, before step one happens, `gtroff' converts certain escape
sequences into reserved input characters (not accessible by the user);
such reserved characters are used for other internal processing also -
this is the very reason why not all characters are valid input.  *Note
Identifiers::, for more on this topic.

   For example, the input string `fi\[:u]' is converted into a
character token `f', a character token `i', and a special token `:u'
(representing u umlaut).  Later on, the character tokens `f' and `i'
are merged to a single output node representing the ligature glyph `fi'
(provided the current font has a glyph for this ligature); the same
happens with `:u'.  All output glyph nodes are `processed' which means
that they are invariably associated with a given font, font size,
advance width, etc.  During the formatting process, `gtroff' itself
adds various nodes to control the data flow.

   Macros, diversions, and strings collect elements in two chained
lists: a list of input tokens which have been passed unprocessed, and a
list of output nodes.  Consider the following the diversion.


     .di xxx
     a
     \!b
     c
     .br
     .di

It contains these elements.

node list            token list   element number
line start node      --           1
glyph node `a'       --           2
word space node      --           3
--                   `b'          4
--                   `\n'         5
glyph node `c'       --           6
vertical size node   --           7
vertical size node   --           8
--                   `\n'         9

Elements 1, 7, and 8 are inserted by `gtroff'; the latter two (which
are always present) specify the vertical extent of the last line,
possibly modified by `\x'.  The `br' request finishes the current
partial line, inserting a newline input token which is subsequently
converted to a space when the diversion is reread.  Note that the word
space node has a fixed width which isn't stretchable anymore.  To
convert horizontal space nodes back to input tokens, use the `unformat'
request.

   Macros only contain elements in the token list (and the node list is
empty); diversions and strings can contain elements in both lists.

   Note that the `chop' request simply reduces the number of elements
in a macro, string, or diversion by one.  Exceptions are "compatibility
save" and "compatibility ignore" input tokens which are ignored.  The
`substring' request also ignores those input tokens.

   Some requests like `tr' or `cflags' work on glyph identifiers only;
this means that the associated glyph can be changed without destroying
this association.  This can be very helpful for substituting glyphs.
In the following example, we assume that glyph `foo' isn't available by
default, so we provide a substitution using the `fchar' request and map
it to input character `x'.


     .fchar \[foo] foo
     .tr x \[foo]

Now let us assume that we install an additional special font `bar'
which has glyph `foo'.


     .special bar
     .rchar \[foo]

Since glyphs defined with `fchar' are searched before glyphs in special
fonts, we must call `rchar' to remove the definition of the fallback
glyph.  Anyway, the translation is still active; `x' now maps to the
real glyph `foo'.


File: groff,  Node: Gtroff Internals-Footnotes,  Up: Gtroff Internals

   (1) Except the escapes `\f', `\F', `\H', `\m', `\M', `\R', `\s', and
`\S' which are processed immediately if not in copy-in mode.


File: groff,  Node: Debugging,  Next: Implementation Differences,  Prev: Gtroff Internals,  Up: gtroff Reference

Debugging
=========

   `gtroff' is not easy to debug, but there are some useful features
and strategies for debugging.

 - Request: .lf line [filename]
     Change the line number and optionally the file name `gtroff' shall
     use for error and warning messages.  LINE is the input line number
     of the _next_ line.

     Without argument, the request is ignored.

     This is a debugging aid for documents which are split into many
     files, then put together with `soelim' and other preprocessors.
     Usually, it isn't invoked manually.

     Note that other `troff' implementations (including the original
     AT&T version) handle `lf' differently.  For them, LINE changes the
     line number of the _current_ line.

 - Request: .tm string
 - Request: .tm1 string
 - Request: .tmc string
     Send STRING to the standard error output; this is very useful for
     printing debugging messages among other things.

     STRING is read in copy mode.

     The `tm' request ignores leading spaces of STRING; `tm1' handles
     its argument similar to the `ds' request: a leading double quote
     in STRING is stripped to allow initial blanks.

     The `tmc' request is similar to `tm1' but does not append a
     newline (as is done in `tm' and `tm1').

 - Request: .ab [string]
     Similar to the `tm' request, except that it causes `gtroff' to
     stop processing.  With no argument it prints `User Abort.' to
     standard error.

 - Request: .ex
     The `ex' request also causes `gtroff' to stop processing; see also
     *Note I/O::.

   When doing something involved it is useful to leave the debugging
statements in the code and have them turned on by a command line flag.


     .if \n(DB .tm debugging output

To activate these statements say


     groff -rDB=1 file

   If it is known in advance that there will be many errors and no
useful output, `gtroff' can be forced to suppress formatted output with
the `-z' flag.

 - Request: .pm
     Print the entire symbol table on `stderr'.  Names of all defined
     macros, strings, and diversions are print together with their size
     in bytes.  Since `gtroff' sometimes adds nodes by itself, the
     returned size can be larger than expected.

     This request differs from UNIX `troff': `gtroff' reports the sizes
     of diversions, ignores an additional argument to print only the
     total of the sizes, and the size isn't returned in blocks of 128
     characters.

 - Request: .pnr
     Print the names and contents of all currently defined number
     registers on `stderr'.

 - Request: .ptr
     Print the names and positions of all traps (not including input
     line traps and diversion traps) on `stderr'.  Empty slots in the
     page trap list are printed as well, because they can affect the
     priority of subsequently planted traps.

 - Request: .fl
     Instruct `gtroff' to flush its output immediately.  The intent is
     for interactive use, but this behaviour is currently not
     implemented in `gtroff'.  Contrary to UNIX `troff', TTY output is
     sent to a device driver also (`grotty'), making it non-trivial to
     communicate interactively.

     This request causes a line break.

 - Request: .backtrace
     Print a backtrace of the input stack to the standard error stream.

     Consider the following in file `test':


          .de xxx
          .  backtrace
          ..
          .de yyy
          .  xxx
          ..
          .
          .yyy

     On execution, `gtroff' prints the following:


          test:2: backtrace: macro `xxx'
          test:5: backtrace: macro `yyy'
          test:8: backtrace: file `test'

     The option `-b' of `gtroff' internally calls a variant of this
     request on each error and warning.

 - Register: \n[slimit]
     Use the `slimit' number register to set the maximum number of
     objects on the input stack.  If `slimit' is less than or equal
     to 0, there is no limit set.  With no limit, a buggy recursive
     macro can exhaust virtual memory.

     The default value is 1000; this is a compile-time constant.

 - Request: .warnscale si
     Set the scaling indicator used in warnings to SI.  Valid values for
     SI are `u', `i', `c', `p', and `P'.  At startup, it is set to `i'.

 - Request: .spreadwarn [limit]
     Make `gtroff' emit a warning if the additional space inserted for
     each space between words in an output line is larger or equal to
     LIMIT.  A negative value is changed to zero; no argument toggles
     the warning on and off without changing LIMIT.  The default scaling
     indicator is `m'.  At startup, `spreadwarn' is deactivated, and
     LIMIT is set to 3m.

     For example,


          .spreadwarn 0.2m

     will cause a warning if `gtroff' must add 0.2m or more for each
     interword space in a line.

     This request is active only if text is justified to both margins
     (using `.ad b').

   `gtroff' has command line options for printing out more warnings
(`-w') and for printing backtraces (`-b') when a warning or an error
occurs.  The most verbose level of warnings is `-ww'.

 - Request: .warn [flags]
 - Register: \n[.warn]
     Control the level of warnings checked for.  The FLAGS are the sum
     of the numbers associated with each warning that is to be enabled;
     all other warnings are disabled.  The number associated with each
     warning is listed below.  For example, `.warn 0' disables all
     warnings, and `.warn 1' disables all warnings except that about
     missing glyphs.  If no argument is given, all warnings are enabled.

     The read-only number register `.warn' contains the current warning
     level.

* Menu:

* Warnings::