{
    "content": [
        {
            "type": "text",
            "text": "# groff_diff (info)\n\n## NAME\n\ngroffdiff - differences between GNU troff and classical troff\n\n## DESCRIPTION\n\nThis  manual page describes the language differences between groff, the\nGNU roff text processing system, and the classical  roff  formatter  of\nthe  freely  available  Unix  7  of  the 1970s, documented in the Troff\nUser's Manual by Ossanna and Kernighan.  This includes  the  roff  lan-\nguage as well as the intermediate output format (troff output).\n\n## Sections\n\n- **NAME**\n- **DESCRIPTION**\n- **GROFF LANGUAGE**\n- **INTERMEDIATE OUTPUT FORMAT**\n- **INCOMPATIBILITIES**\n- **AUTHORS**\n- **SEE ALSO**\n\nUse structuredContent.sections for detailed options, examples, and full documentation.\n"
        }
    ],
    "structuredContent": {
        "command": "groff_diff",
        "section": "",
        "mode": "info",
        "summary": "groffdiff - differences between GNU troff and classical troff",
        "synopsis": null,
        "tldr_summary": null,
        "tldr_examples": [],
        "tldr_source": null,
        "flags": [],
        "examples": [],
        "see_also": [
            {
                "name": "groff",
                "section": "1",
                "url": "https://www.chedong.com/phpMan.php/man/groff/1/json"
            },
            {
                "name": "groff",
                "section": "7",
                "url": "https://www.chedong.com/phpMan.php/man/groff/7/json"
            },
            {
                "name": "roff",
                "section": "7",
                "url": "https://www.chedong.com/phpMan.php/man/roff/7/json"
            }
        ],
        "section_outline": [
            {
                "name": "NAME",
                "lines": 2,
                "subsections": []
            },
            {
                "name": "DESCRIPTION",
                "lines": 9,
                "subsections": []
            },
            {
                "name": "GROFF LANGUAGE",
                "lines": 1437,
                "subsections": []
            },
            {
                "name": "INTERMEDIATE OUTPUT FORMAT",
                "lines": 154,
                "subsections": []
            },
            {
                "name": "INCOMPATIBILITIES",
                "lines": 109,
                "subsections": []
            },
            {
                "name": "AUTHORS",
                "lines": 4,
                "subsections": []
            },
            {
                "name": "SEE ALSO",
                "lines": 25,
                "subsections": []
            }
        ],
        "sections": {
            "NAME": {
                "content": "groffdiff - differences between GNU troff and classical troff\n",
                "subsections": []
            },
            "DESCRIPTION": {
                "content": "This  manual page describes the language differences between groff, the\nGNU roff text processing system, and the classical  roff  formatter  of\nthe  freely  available  Unix  7  of  the 1970s, documented in the Troff\nUser's Manual by Ossanna and Kernighan.  This includes  the  roff  lan-\nguage as well as the intermediate output format (troff output).\n\nSection  \"See Also\" below gives pointers to both the classical roff and\nthe modern groff documentation.\n",
                "subsections": []
            },
            "GROFF LANGUAGE": {
                "content": "In this section, all additional features of groff compared to the clas-\nsical Unix 7 troff are described in detail.\n\nLong names\nThe  names  of number registers, fonts, strings/macros/diversions, spe-\ncial characters (glyphs), and colors can be of any length.   In  escape\nsequences,  additionally to the classical '(xx' construction for a two-\ncharacter glyph name, you can use  '[xxx]'  for  a  name  of  arbitrary\nlength.\n\n\\[xxx] Print the special character (glyph) called xxx.\n\n\\[comp1 comp2 ...]\nPrint  composite glyph consisting of multiple components.  Exam-\nple: '\\[A ho]' is capital letter A  with  ogonek  which  finally\nmaps to glyph name 'u00410328'.  See Groff: The GNU Implementa-\ntion of troff, the groff Texinfo manual, for details  of  how  a\nglyph   name   for   a   composite  glyph  is  constructed,  and\ngroffchar(7) for a list of glyph name components used  in  com-\nposite glyph names.\n\n\\f[xxx]\nSet  font xxx.  Additionally, \\f[] is a new syntax form equal to\n\\fP, i.e., to return to the previous font.\n\n\\*[xxx arg1 arg2 ...]\nInterpolate string xxx, taking arg1, arg2, ..., as arguments.\n\n\\n[xxx]\nInterpolate number register xxx.\n\nFractional point sizes\nA scaled point is equal to 1/sizescale points, where sizescale is spec-\nified  in  the  DESC file (1 by default).  There is a new scale indica-\ntor z that has the effect of multiplying by  sizescale.   Requests  and\nescape  sequences  in  troff interpret arguments that represent a point\nsize as being in units of scaled points, but they  evaluate  each  such\nargument  using  a  default scale indicator of z.  Arguments treated in\nthis way are the argument to the ps request, the third argument to  the\ncs request, the second and fourth arguments to the tkf request, the ar-\ngument to the \\H escape sequence, and those variants of the  \\s  escape\nsequence that take a numeric expression as their argument.\n\nFor  example, suppose sizescale is 1000; then a scaled point is equiva-\nlent to a millipoint; the call .ps 10.25 is  equivalent  to  .ps 10.25z\nand  so  sets  the point size to 10250 scaled points, which is equal to\n10.25 points.\n\nThe number register \\n[.s] returns the point size in points as  decimal\nfraction.  There is also a new number register \\n[.ps] that returns the\npoint size in scaled points.\n\nIt would make no sense to use the z scale indicator in  a  numeric  ex-\npression  whose  default  scale  indicator  was neither u nor z, and so\ntroff disallows this.  Similarly it would make no sense to use a  scal-\ning  indicator  other than z or u in a numeric expression whose default\nscale indicator was z, and so troff disallows this as well.\n\nThere is also new scale indicator s which multiplies by the  number  of\nunits in a scaled point.  So, for example, \\n[.ps]s is equal to 1m.  Be\nsure not to confuse the s and z scale indicators.\n\nNumeric expressions\nSpaces are permitted in a number expression within parentheses.\n\nM indicates a scale of 100ths of an em.  f indicates a scale  of  65536\nunits,  providing fractions for color definitions with the defcolor re-\nquest.  For example, 0.5f = 32768u.\n\ne1>?e2 The maximum of e1 and e2.\n\ne1<?e2 The minimum of e1 and e2.\n\n(c;e)  Evaluate e using c as the default scaling indicator.   If  c  is\nmissing, ignore scaling indicators in the evaluation of e.\n\nNew escape sequences\n\\A'anything'\nThis  expands  to 1 or 0, depending on whether anything is or is\nnot acceptable as the name of a string, macro, diversion, number\nregister, environment, font, or color.  It returns 0 if anything\nis empty.  This is useful if you want to look up user  input  in\nsome sort of associative table.\n\n\\B'anything'\nThis  expands  to 1 or 0, depending on whether anything is or is\nnot a valid numeric expression.  It returns  0  if  anything  is\nempty.\n\n\\C'xxx'\nTypeset  glyph named xxx.  Normally it is more convenient to use\n\\[xxx].  But \\C has the advantage that it is compatible with re-\ncent versions of Unix and is available in compatibility mode.\n\n\\E     This  is equivalent to an escape character, but it is not inter-\npreted in copy mode.  For example, strings to start and end  su-\nperscripting could be defined like this\n\n.ds { \\v'-.3m'\\s'\\En[.s]*6u/10u'\n.ds } \\s0\\v'.3m'\n\nThe  use  of  \\E ensures that these definitions work even if \\*{\ngets interpreted in copy mode (for example, by being used  in  a\nmacro argument).\n\n\\Ff\n\\F(fm\n\\F[fam]\nChange  font family.  This is the same as the fam request.  \\F[]\nswitches back to the previous font family (note that  \\FP  won't\nwork; it selects font family 'P' instead).\n\n\\mx\n\\m(xx\n\\m[xxx]\nSet drawing color.  \\m[] switches back to the previous color.\n\n\\Mx\n\\M(xx\n\\M[xxx]\nSet  background  color for filled objects drawn with the \\D'...'\ncommands.  \\M[] switches back to the previous color.\n\n\\N'n'  Typeset the glyph with index n in the current font.   n  can  be\nany integer.  Most devices only have glyphs with indices between\n0 and 255.  If the current font does not contain  a  glyph  with\nthat  code,  special  fonts are not searched.  The \\N escape se-\nquence can be conveniently used in conjunction with the char re-\nquest, for example\n\n.char \\[phone] \\f(ZD\\N'37'\n\nThe  index  of  each  glyph is given in the fourth column in the\nfont description file after the charset command.  It is possible\nto  include unnamed glyphs in the font description file by using\na name of ---; the \\N escape sequence is the  only  way  to  use\nthese.\n\n\\On\n\\O[n]  Suppress  troff  output.  The escapes \\O2, \\O3, \\O4, and \\O5 are\nintended for internal use by grohtml.\n\n\\O0    Disable any ditroff glyphs from being emitted to the  de-\nvice driver, provided that the escape occurs at the outer\nlevel (see \\O3 and \\O4).\n\n\\O1    Enable output of glyphs, provided that the escape  occurs\nat the outer level.\n\n\\O0  and  \\O1 also reset the registers \\n[opminx], \\n[op-\nminy], \\n[opmaxx], and \\n[opmaxy] to -1.  These four reg-\nisters mark the top left and bottom right hand corners of\na box which encompasses all written glyphs.\n\n\\O2    Provided that the escape occurs at the outer  level,  en-\nable  output  of  glyphs and also write out to stderr the\npage number and four registers  encompassing  the  glyphs\npreviously written since the last call to \\O.\n\n\\O3    Begin  a  nesting  level.  At start-up, troff is at outer\nlevel.  This is really an internal mechanism for  grohtml\nwhile  producing  images.   They are generated by running\nthe troff source through troff to the  PostScript  device\nand ghostscript to produce images in PNG format.  The \\O3\nescape starts a new page if the device is  not  html  (to\nreduce  the  possibility of images crossing a page bound-\nary).\n\n\\O4    End a nesting level.\n\n\\O5[Pfilename]\nThis escape is grohtml specific.  Provided that this  es-\ncape occurs at the outer nesting level, write filename to\nstderr.  The position of the image, P, must be  specified\nand  must be one of l, r, c, or i (left, right, centered,\ninline).  filename is associated with the  production  of\nthe next inline image.\n\n\\R'name +-n'\nThis has the same effect as\n\n.nr name +-n\n\n\\s(nn\n\\s+-(nn\nSet the point size to nn points; nn must be exactly two digits.\n\n\\s[+-n]\n\\s+-[n]\n\\s'+-n'\n\\s+-'n'\nSet the point size to n scaled points; n is a numeric expression\nwith a default scale indicator of z.\n\n\\Vx\n\\V(xx\n\\V[xxx]\nInterpolate the contents of the environment variable xxx, as re-\nturned by getenv(3).  \\V is interpreted in copy mode.\n\n\\Yx\n\\Y(xx\n\\Y[xxx]\nThis  is  approximately  equivalent to \\X'\\*[xxx]'.  However the\ncontents of the string or macro xxx are not interpreted; also it\nis  permitted  for  xxx to have been defined as a macro and thus\ncontain newlines (it is not permitted for the argument to \\X  to\ncontain newlines).  The inclusion of newlines requires an exten-\nsion to the Unix troff output format, and confuses drivers  that\ndo not know about this extension.\n\n\\Z'anything'\nPrint  anything and then restore the horizontal and vertical po-\nsition; anything may not contain tabs or leaders.\n\n\\$0    The name by which the current macro was invoked.   The  als  re-\nquest can make a macro have more than one name.\n\n\\$*    In  a  macro  or  string, the concatenation of all the arguments\nseparated by spaces.\n\n\\$@    In a macro or string, the concatenation  of  all  the  arguments\nwith each surrounded by double quotes, and separated by spaces.\n\n\\$^    In a macro, the representation of all parameters as if they were\nan argument to the ds request.\n\n\\$(nn\n\\$[nnn]\nIn a macro or string, this gives the nn-th or  nnn-th  argument.\nMacros and strings can have an unlimited number of arguments.\n\n\\?anything\\?\nWhen  used in a diversion, this transparently embeds anything in\nthe diversion.  anything is read in copy mode.  When the  diver-\nsion  is reread, anything is interpreted.  anything may not con-\ntain newlines; use \\! if you want to embed newlines in a  diver-\nsion.   The  escape  sequence \\? is also recognized in copy mode\nand turned into a single internal code; it  is  this  code  that\nterminates anything.  Thus\n\n.nr x 1\n.nf\n.di d\n\\?\\\\?\\\\\\\\?\\\\\\\\\\\\\\\\nx\\\\\\\\?\\\\?\\?\n.di\n.nr x 2\n.di e\n.d\n.di\n.nr x 3\n.di f\n.e\n.di\n.nr x 4\n.f\n\nprints 4.\n\n\\/     This  increases  the  width  of  the preceding glyph so that the\nspacing between that glyph and the following glyph is correct if\nthe  following glyph is a roman glyph.  It is a good idea to use\nthis escape sequence whenever an  italic  glyph  is  immediately\nfollowed by a roman glyph without any intervening space.\n\n\\,     This  modifies  the  spacing  of the following glyph so that the\nspacing between that glyph and the preceding glyph is correct if\nthe  preceding glyph is a roman glyph.  It is a good idea to use\nthis escape sequence whenever a roman glyph is immediately  fol-\nlowed by an italic glyph without any intervening space.\n\n\\)     Like  \\&  except  that it behaves like a character declared with\nthe cflags request to be transparent for the purposes of end-of-\nsentence recognition.\n\n\\~     This  produces an unbreakable space that stretches like a normal\ninter-word space when a line is adjusted.\n\n\\:     This causes the insertion of a zero-width break  point.   It  is\nequal to \\% within a word but without insertion of a soft hyphen\nglyph.\n\n\\#     Everything up to and including  the  next  newline  is  ignored.\nThis  is interpreted in copy mode.  It is like \\\" except that \\\"\ndoes not ignore the terminating newline.\n\nNew requests\n.aln xx yy\nCreate an alias xx for number register object named yy.  The new\nname  and  the  old name are exactly equivalent.  If yy is unde-\nfined, a warning of type reg is generated, and  the  request  is\nignored.\n\n.als xx yy\nCreate  an alias xx for request, string, macro, or diversion ob-\nject named yy.  The new name and the old name are exactly equiv-\nalent  (it is similar to a hard rather than a soft link).  If yy\nis undefined, a warning of type mac is generated,  and  the  re-\nquest  is ignored.  The de, am, di, da, ds, and as requests only\ncreate a new object if the  name  of  the  macro,  diversion  or\nstring  is  currently  undefined or if it is defined to be a re-\nquest; normally they modify the value of an existing object.\n\n.am1 xx yy\nSimilar to .am, but compatibility mode is  switched  off  during\nexecution.   To be more precise, a 'compatibility save' token is\ninserted at the beginning of the macro addition, and a 'compati-\nbility  restore'  token  at  the end.  As a consequence, the re-\nquests am, am1, de, and de1 can be intermixed freely  since  the\ncompatibility  save/restore  tokens  only affect the macro parts\ndefined by .am1 and .ds1.\n\n.ami xx yy\nAppend to macro indirectly.  See the dei request below for  more\ninformation.\n\n.ami1 xx yy\nSame  as  the ami request but compatibility mode is switched off\nduring execution.\n\n.as1 xx yy\nSimilar to .as, but compatibility mode is  switched  off  during\nexpansion.   To be more precise, a 'compatibility save' token is\ninserted at the beginning of the string,  and  a  'compatibility\nrestore'  token  at the end.  As a consequence, the requests as,\nas1, ds, and ds1 can be intermixed freely since the  compatibil-\nity  save/restore tokens only affect the (sub)strings defined by\nas1 and ds1.\n\n.asciify xx\nThis request 'unformats' the diversion xx in  such  a  way  that\nASCII and space characters (and some escape sequences) that were\nformatted and diverted into xx are treated like  ordinary  input\ncharacters when xx is reread.  Useful for diversions in conjunc-\ntion with the writem request.  It can be  also  used  for  gross\nhacks; for example, this\n\n.tr @.\n.di x\n@nr n 1\n.br\n.di\n.tr @@\n.asciify x\n.x\n\nsets  register  n to 1.  Note that glyph information (font, font\nsize, etc.) is not preserved; use .unformat instead.\n\n.backtrace\nPrint a backtrace of the input stack on stderr.\n\n.blm xx\nSet the blank line macro to xx.  If there is a blank line macro,\nit  is  invoked  when a blank line is encountered instead of the\nusual troff behaviour.\n\n.box xx\n.boxa xx\nThese requests are similar to the di and da  requests  with  the\nexception  that  a partially filled line does not become part of\nthe diversion (i.e., the diversion  always  starts  with  a  new\nline) but is restored after ending the diversion, discarding the\npartially filled line which possibly comes from the diversion.\n\n.break Break out of a while loop.  See also the while and continue  re-\nquests.  Be sure not to confuse this with the br request.\n\n.brp   This is the same as \\p.\n\n.cflags n c1 c2 ...\nCharacters  c1,  c2, ..., have properties determined by n, which\nis ORed from the following:\n\n1      The character ends sentences  (initially  characters  .?!\nhave this property).\n\n2      Lines  can  be  broken before the character (initially no\ncharacters have this property); a line is not broken at a\ncharacter  with  this  property  unless the characters on\neach side both have non-zero hyphenation codes.  This can\nbe overridden with value 64.\n\n4      Lines  can be broken after the character (initially char-\nacters -\\[hy]\\[em] have this property);  a  line  is  not\nbroken at a character with this property unless the char-\nacters on each side both have non-zero hyphenation codes.\nThis can be overridden with value 64.\n\n8      The  glyph  associated with this character overlaps hori-\nzontally  (initially  characters   \\[ul]\\[rn]\\[ru]\\[radi-\ncalex]\\[sqrtex] have this property).\n\n16     The  glyph associated with this character overlaps verti-\ncally (initially glyph \\[br] has this property).\n\n32     An end-of-sentence character followed by  any  number  of\ncharacters  with this property is treated as the end of a\nsentence if followed by a newline or two spaces; in other\nwords  the  character  is transparent for the purposes of\nend-of-sentence recognition; this is the same as having a\nzero   space   factor   in   TeX   (initially  characters\n\"')]*\\[dg]\\[rq]\\[cq] have this property).\n\n64     Ignore hyphenation code values of the surrounding charac-\nters.   Use this in combination with values 2 and 4 (ini-\ntially no characters have this property).\n\n128    Prohibit a line break before the character, but  allow  a\nline  break after the character.  This works only in com-\nbination with flags 256 and 512 and has no effect  other-\nwise.\n\n256    Prohibit  a  line  break after the character, but allow a\nline break before the character.  This works only in com-\nbination  with flags 128 and 512 and has no effect other-\nwise.\n\n512    Allow line break before or  after  the  character.   This\nworks  only in combination with flags 128 and 256 and has\nno effect otherwise.\n\nContrary to flag values 2 and 4, the flags  128,  256,  and  512\nwork  pairwise.   If,  for example, the left character has value\n512, and the right character 128, no line break  gets  inserted.\nIf  we  use value 6 instead for the left character, a line break\nafter the character can't be suppressed since the  right  neigh-\nbour character doesn't get examined.\n\n.char c string\n[This request can both define characters and glyphs.]\n\nDefine  entity  c  to be string.  To be more precise, define (or\neven override) a groff entity which can be accessed with name  c\non  the  input  side,  and which uses string on the output side.\nEvery time glyph c needs to be printed, string is processed in a\ntemporary environment and the result is wrapped up into a single\nobject.  Compatibility mode is turned off and the escape charac-\nter  is set to \\ while string is being processed.  Any embolden-\ning, constant spacing or track kerning is applied to this object\nrather than to individual glyphs in string.\n\nA  groff  object defined by this request can be used just like a\nnormal glyph provided by the output device.  In particular other\ncharacters  can  be translated to it with the tr request; it can\nbe made the leader glyph by the lc  request;  repeated  patterns\ncan  be  drawn  with  the  glyph  using the \\l and \\L escape se-\nquences; words containing c can be hyphenated correctly, if  the\nhcode request is used to give the object a hyphenation code.\n\nThere  is  a special anti-recursion feature: Use of glyph within\nthe glyph's definition is handled like normal glyphs not defined\nwith char.\n\nA glyph definition can be removed with the rchar request.\n\n.chop xx\nChop  the last element off macro, string, or diversion xx.  This\nis useful for removing the newline from the  end  of  diversions\nthat are to be interpolated as strings.\n\n.class name c1 c2 ...\nAssign name to a set of characters c1, c2, ..., so that they can\nbe referred to from other  requests  easily  (currently  .cflags\nonly).   Character ranges (indicated by an intermediate '-') and\nnested classes are possible also.   This  is  useful  to  assign\nproperties to a large set of characters.\n\n.close stream\nClose  the  stream named stream; stream will no longer be an ac-\nceptable argument to the write request.  See the open request.\n\n.composite glyph1 glyph2\nMap glyph name glyph1 to glyph name glyph2  if  it  is  used  in\n\\[...] with more than one component.\n\n.continue\nFinish  the  current  iteration  of  a while loop.  See also the\nwhile and break requests.\n\n.color n\nIf n is non-zero or missing, enable  colors  (this  is  the  de-\nfault), otherwise disable them.\n\n.cp n  If  n  is non-zero or missing, enable compatibility mode, other-\nwise disable it.  In compatibility mode, long names are not rec-\nognized,  and  the incompatibilities caused by long names do not\narise.\n\n.defcolor xxx scheme colorcomponents\nDefine color xxx.  scheme can be one of  the  following  values:\nrgb  (three components), cmy (three components), cmyk (four com-\nponents), and gray or grey (one  component).   Color  components\ncan be given either as a hexadecimal string or as positive deci-\nmal integers in the range 0-65535.  A  hexadecimal  string  con-\ntains  all color components concatenated; it must start with ei-\nther # or ##.  The former specifies  hex  values  in  the  range\n0-255  (which  are  internally multiplied by 257), the latter in\nthe range 0-65535.   Examples:  #FFC0CB  (pink),  ##ffff0000ffff\n(magenta).   A new scaling indicator f has been introduced which\nmultiplies its value by 65536; this makes it convenient to spec-\nify color components as fractions in the range 0 to 1.  Example:\n\n.defcolor darkgreen rgb 0.1f 0.5f 0.2f\n\nNote  that  f  is the default scaling indicator for the defcolor\nrequest, thus the above statement is equivalent to\n\n.defcolor darkgreen rgb 0.1 0.5 0.2\n\nThe color named default (which is device-specific) can't be  re-\ndefined.  It is possible that the default color for \\M and \\m is\nnot the same.\n\n.de1 xx yy\nSimilar to .de, but compatibility mode is  switched  off  during\nexecution.   On  entry,  the current compatibility mode is saved\nand restored at exit.\n\n.dei xx yy\nDefine macro indirectly.  The following example\n\n.ds xx aa\n.ds yy bb\n.dei xx yy\n\nis equivalent to\n\n.de aa bb\n\n.dei1 xx yy\nSimilar to the dei request but compatibility  mode  is  switched\noff during execution.\n\n.device anything\nThis is (almost) the same as the \\X escape.  anything is read in\ncopy mode; a leading \" is stripped.\n\n.devicem xx\nThis is the same as the \\Y escape (to embed the  contents  of  a\nmacro into the intermediate output preceded with 'x X').\n\n.do xxx\nInterpret .xxx with compatibility mode disabled.  For example,\n\n.do fam T\n\nwould have the same effect as\n\n.fam T\n\nexcept  that  it  would work even if compatibility mode had been\nenabled.  Note that the previous compatibility mode is  restored\nbefore any files sourced by xxx are interpreted.\n\n.ds1 xx yy\nSimilar  to  .ds,  but compatibility mode is switched off during\nexpansion.  To be more precise, a 'compatibility save' token  is\ninserted  at  the  beginning of the string, and a 'compatibility\nrestore' token at the end.\n\n.ecs   Save current escape character.\n\n.ecr   Restore escape character saved with  ecs.   Without  a  previous\ncall to ecs, '\\' will be the new escape character.\n\n.evc xx\nCopy  the contents of environment xx to the current environment.\nNo pushing or popping of environments is done.\n\n.fam xx\nSet the current font family to xx.  The current font  family  is\npart  of the current environment.  If xx is missing, switch back\nto previous font family.  The value at start-up is 'T'.  See the\ndescription of the sty request for more information on font fam-\nilies.\n\n.fchar c string\nDefine fallback character (or glyph) c to be string.  The syntax\nof  this  request is the same as the char request; the only dif-\nference is that a glyph defined with char hides the  glyph  with\nthe  same name in the current font, whereas a glyph defined with\nfchar is checked only if the particular glyph isn't found in the\ncurrent font.  This test happens before checking special fonts.\n\n.fcolor c\nSet  the fill color to c.  If c is missing, switch to the previ-\nous fill color.\n\n.fschar f c string\nDefine fallback character (or glyph) c for font f to be  string.\nThe syntax of this request is the same as the char request (with\nan additional argument to specify the  font);  a  glyph  defined\nwith  fschar  is  searched after the list of fonts declared with\nthe fspecial request but before the list of fonts declared  with\n.special.\n\n.fspecial f s1 s2 ...\nWhen the current font is f, fonts s1, s2, ..., are special, that\nis, they are searched for glyphs not in the current  font.   Any\nfonts  specified in the special request are searched after fonts\nspecified in the fspecial request.  Without argument, reset  the\nlist of global special fonts to be empty.\n\n.ftr f g\nTranslate  font  f to g.  Whenever a font named f is referred to\nin an \\f escape sequence, in the F and S conditional  operators,\nor in the ft, ul, bd, cs, tkf, special, fspecial, fp, or sty re-\nquests, font g is used.  If g is missing, or  equal  to  f  then\nfont f is not translated.\n\n.fzoom f zoom\nSet zoom factor zoom for font f.  zoom must a non-negative inte-\nger multiple of 1/1000th.  If it is missing or is equal to zero,\nit means the same as 1000, namely no magnification.  f must be a\nreal font name, not a style.\n\n.gcolor c\nSet the glyph color to c.  If c is missing, switch to the previ-\nous glyph color.\n\n.hcode c1 code1 c2 code2 ...\nSet the hyphenation code of character c1 to code1 and that of c2\nto code2, and so on.  A hyphenation code must be a single  input\ncharacter  (not  a  special  character)  other than a digit or a\nspace.  Initially each lower-case letter a-z has  a  hyphenation\ncode,  which is itself, and each upper-case letter A-Z has a hy-\nphenation code which is the lower-case version of  itself.   See\nalso the hpf request.\n\n.hla lang\nSet  the  current hyphenation language to lang.  Hyphenation ex-\nceptions specified with the hw request and hyphenation  patterns\nspecified with the hpf request are both associated with the cur-\nrent hyphenation language.  The hla request is  usually  invoked\nby the troffrc file to set up a default language.\n\n.hlm n Set the maximum number of consecutive hyphenated lines to n.  If\nn is negative, there is no maximum.  The default  value  is  -1.\nThis  value  is  associated  with the current environment.  Only\nlines output from an environment count towards the maximum asso-\nciated  with  that  environment.   Hyphens resulting from \\% are\ncounted; explicit hyphens are not.\n\n.hpf file\nRead hyphenation patterns from file; this is searched for in the\nsame  way  that name.tmac is searched for when the -mname option\nis specified.  It should have the same format  as  (simple)  TeX\npatterns files.  More specifically, the following scanning rules\nare implemented.\n\no      A percent sign starts a comment (up to  the  end  of  the\nline) even if preceded by a backslash.\n\no      No support for 'digraphs' like \\$.\n\no      ^^xx  (x  is  0-9 or a-f) and ^^x (character code of x in\nthe range 0-127) are recognized; other use of ^ causes an\nerror.\n\no      No macro expansion.\n\no      hpf  checks  for  the expression \\patterns{...} (possibly\nwith whitespace before and after the braces).  Everything\nbetween  the  braces  is  taken  as hyphenation patterns.\nConsequently, { and } are not allowed in patterns.\n\no      Similarly, \\hyphenation{...} gives a list of  hyphenation\nexceptions.\n\no      \\endinput is recognized also.\n\no      For backwards compatibility, if \\patterns is missing, the\nwhole file is treated as a list of  hyphenation  patterns\n(only  recognizing the % character as the start of a com-\nment).\n\nUse the hpfcode request to map the encoding used in  hyphenation\npatterns  files  to  groff's input encoding.  By default, every-\nthing maps to itself except letters 'A' to 'Z' which map to  'a'\nto 'z'.\n\nThe  set  of hyphenation patterns is associated with the current\nlanguage set by the hla request.  The hpf request is usually in-\nvoked  by  the troffrc file; a second call replaces the old pat-\nterns with the new ones.\n\n.hpfa file\nThe same as hpf except that the hyphenation patterns  from  file\nare  appended to the patterns already loaded in the current lan-\nguage.\n\n.hpfcode a b c d ...\nAfter reading a hyphenation patterns file with the hpf  or  hpfa\nrequest, convert all characters with character code a in the re-\ncently read patterns to character code b, character code c to d,\netc.  Initially, all character codes map to themselves.  The ar-\nguments of hpfcode must be integers in the range 0 to 255.  Note\nthat  it  is  even possible to use character codes which are in-\nvalid in groff otherwise.\n\n.hym n Set the hyphenation margin to n:  when  the  current  adjustment\nmode is not b, the line is not hyphenated if the line is no more\nthan n short.  The default hyphenation margin is 0.  The default\nscaling indicator for this request is m.  The hyphenation margin\nis associated with the current environment.  The current hyphen-\nation margin is available in the \\n[.hym] register.\n\n.hys n Set the hyphenation space to n: When the current adjustment mode\nis b don't hyphenate the line if the line can  be  justified  by\nadding  no  more than n extra space to each word space.  The de-\nfault hyphenation space is 0.  The default scaling indicator for\nthis request is m.  The hyphenation space is associated with the\ncurrent environment.  The current hyphenation space is available\nin the \\n[.hys] register.\n\n.itc n macro\nVariant  of  .it  for  which  a  line interrupted with \\c is not\ncounted as an input line.\n\n.kern n\nIf n is non-zero or missing, enable pairwise kerning,  otherwise\ndisable it.\n\n.length xx string\nCompute  the length of string and return it in the number regis-\nter xx (which is not necessarily defined before).\n\n.linetabs n\nIf n is non-zero or missing, enable  line-tabs  mode,  otherwise\ndisable  it (which is the default).  In line-tabs mode, tab dis-\ntances are computed relative to the (current) output line.  Oth-\nerwise  they are taken relative to the input line.  For example,\nthe following\n\n.ds x a\\t\\c\n.ds y b\\t\\c\n.ds z c\n.ta 1i 3i\n\\*x\n\\*y\n\\*z\n\nyields\n\na         b         c\n\nIn line-tabs mode, the same code gives\n\na         b                   c\n\nLine-tabs mode is associated with the current  environment;  the\nread-only  number register \\n[.linetabs] is set to 1 if in line-\ntabs mode, and 0 otherwise.\n\n.lsm xx\nSet the leading spaces macro to xx.  If there are leading spaces\nin  an  input line, it is invoked instead of the usual troff be-\nhaviour; the leading spaces are removed.  Registers \\n[lsn]  and\n\\n[lss] hold the number of removed leading spaces and the corre-\nsponding horizontal space, respectively.\n\n.mso file\nThe same as the so request except that file is searched  for  in\nthe  same directories as macro files for the -m command-line op-\ntion.  If the file name to be included has  the  form  name.tmac\nand  it  isn't found, mso tries to include tmac.name instead and\nvice versa.  A warning of type file is generated if  file  can't\nbe loaded, and the request is ignored.\n\n.nop anything\nExecute anything.  This is similar to '.if 1'.\n\n.nroff Make  the n built-in condition true and the t built-in condition\nfalse.  This can be reversed using the troff request.\n\n.open stream filename\nOpen filename for writing and associate the stream named  stream\nwith it.  See also the close and write requests.\n\n.opena stream filename\nLike open, but if filename exists, append to it instead of trun-\ncating it.\n\n.output string\nEmit string directly to  the  intermediate  output  (subject  to\ncopy-mode interpretation); this is similar to \\! used at the top\nlevel.  An initial double quote in string is stripped off to al-\nlow initial blanks.\n\n.pev   Print the current environment and each defined environment state\non stderr.\n\n.pnr   Print the names and contents of  all  currently  defined  number\nregisters on stderr.\n\n.psbb filename\nGet  the bounding box of a PostScript image filename.  This file\nmust conform to Adobe's Document  Structuring  Conventions;  the\ncommand  looks for a %%BoundingBox comment to extract the bound-\ning box values.  After a successful call,  the  coordinates  (in\nPostScript  units)  of the lower left and upper right corner can\nbe  found  in  the  registers  \\n[llx],  \\n[lly],  \\n[urx],  and\n\\n[ury],  respectively.   If  some  error has occurred, the four\nregisters are set to zero.\n\n.pso command\nThis behaves like the so request except that  input  comes  from\nthe standard output of command.\n\n.ptr   Print  the names and positions of all traps (not including input\nline traps and diversion traps) on stderr.  Empty slots  in  the\npage  trap list are printed as well, because they can affect the\npriority of subsequently planted traps.\n\n.pvs +-n\nSet the post-vertical line space to n; default  scale  indicator\nis  p.   This value is added to each line after it has been out-\nput.  With no argument, the post-vertical line space is  set  to\nits previous value.\n\nThe total vertical line spacing consists of four components: .vs\nand \\x with a negative value which are applied before  the  line\nis  output,  and .pvs and \\x with a positive value which are ap-\nplied after the line is output.\n\n.rchar c1 c2 ...\nRemove the definitions of glyphs c1, c2, ...   This  undoes  the\neffect of a char request.\n\n.return\nWithin a macro, return immediately.  If called with an argument,\nreturn twice, namely from the current macro and from  the  macro\none level higher.  No effect otherwise.\n\n.rfschar c1 c2 ...\nRemove the font-specific definitions of glyphs c1, c2, ...  This\nundoes the effect of an fschar request.\n\n.rj\n.rj n  Right justify the next n input lines.  Without an argument right\njustify  the  next  input line.  The number of lines to be right\njustified is available in the \\n[.rj] register.  This implicitly\ndoes .ce 0.  The ce request implicitly does .rj 0.\n\n.rnn xx yy\nRename number register xx to yy.\n\n.schar c string\nDefine global fallback character (or glyph) c to be string.  The\nsyntax of this request is the same as the char request; a  glyph\ndefined  with schar is searched after the list of fonts declared\nwith the special request but before the mounted special fonts.\n\n.shc c Set the soft hyphen character to c.  If c is omitted,  the  soft\nhyphen  character  is set to the default \\[hy].  The soft hyphen\ncharacter is the glyph which is inserted when a word is  hyphen-\nated at a line break.  If the soft hyphen character does not ex-\nist in the font of the glyph immediately preceding  a  potential\nbreak point, then the line is not broken at that point.  Neither\ndefinitions (specified with the char request)  nor  translations\n(specified  with the tr request) are considered when finding the\nsoft hyphen character.\n\n.shift n\nIn a macro, shift the arguments by n positions: argument  i  be-\ncomes  argument  i-n;  arguments 1 to n are no longer available.\nIf n is missing, arguments are shifted by 1.  Shifting by  nega-\ntive amounts is currently undefined.\n\n.sizes s1 s2 ... sn [0]\nThis command is similar to the sizes command of a DESC file.  It\nsets the available font sizes for the current font  to  s1,  s2,\n...,  sn  scaled points.  The list of sizes can be terminated by\nan optional 0.  Each si can also be a range of sizes m-n.   Con-\ntrary  to the font file command, the list can't extend over more\nthan a single line.\n\n.special s1 s2 ...\nFonts s1, s2, ..., are special and are searched for  glyphs  not\nin  the current font.  Without arguments, reset the list of spe-\ncial fonts to be empty.\n\n.spreadwarn limit\nMake troff emit a warning if the additional space  inserted  for\neach space between words in an output line is larger or equal to\nlimit.  A negative value is changed to zero; no argument toggles\nthe  warning  on  and  off  without changing limit.  The default\nscaling indicator is m.  At startup, spreadwarn is  deactivated,\nand  limit is set to 3m.  For example, .spreadwarn 0.2m causes a\nwarning if troff must add 0.2m or more for each interword  space\nin  a line.  This request is active only if text is justified to\nboth margins (using .ad b).\n\n.sty n f\nAssociate style f with font position n.  A font position can  be\nassociated either with a font or with a style.  The current font\nis the index of a font position and so is also either a font  or\na  style.  When it is a style, the font that is actually used is\nthe font the name of which is the concatenation of the  name  of\nthe current family and the name of the current style.  For exam-\nple, if the current font is 1 and font position 1 is  associated\nwith  style  R and the current font family is T, then font TR is\nused.  If the current font is not a style, then the current fam-\nily  is ignored.  When the requests cs, bd, tkf, uf, or fspecial\nare applied to a style, then they are  applied  instead  to  the\nmember  of  the current family corresponding to that style.  The\ndefault family can be set with the -f command-line option.   The\nstyles  command  in  the DESC file controls which font positions\n(if any) are initially associated with styles rather than fonts.\n\n.substring xx n1 [n2]\nReplace the string named xx with the substring  defined  by  the\nindices  n1  and  n2.  The first character in the string has in-\ndex 0.  If n2 is omitted,  it  is  taken  to  be  equal  to  the\nstring's length.  If the index value n1 or n2 is negative, it is\ncounted from the end of the string, going  backwards:  The  last\ncharacter  has index -1, the character before the last character\nhas index -2, etc.\n\n.tkf f s1 n1 s2 n2\nEnable track kerning for font f.  When the current font is f the\nwidth  of  every  glyph is increased by an amount between n1 and\nn2; when the current point size is less than or equal to s1  the\nwidth is increased by n1; when it is greater than or equal to s2\nthe width is increased by n2; when the  point  size  is  greater\nthan or equal to s1 and less than or equal to s2 the increase in\nwidth is a linear function of the point size.\n\n.tm1 string\nSimilar to the tm request, string is read in copy mode and writ-\nten on the standard error, but an initial double quote in string\nis stripped off to allow initial blanks.\n\n.tmc string\nSimilar to tm1 but without writing a final newline.\n\n.trf filename\nTransparently output the contents of file filename.   Each  line\nis  output as if preceded by \\!; however, the lines are not sub-\nject to copy-mode interpretation.  If the file does not end with\na newline, then a newline is added.  For example, you can define\na macro x containing the contents of file f, using\n\n.di x\n.trf f\n.di\n\nUnlike with the cf request, the file cannot contain  characters,\nsuch as NUL, that are not valid troff input characters.\n\n.trin abcd\nThis  is  the same as the tr request except that the asciify re-\nquest uses the character code  (if  any)  before  the  character\ntranslation.  Example:\n\n.trin ax\n.di xxx\na\n.br\n.di\n.xxx\n.trin aa\n.asciify xxx\n.xxx\n\nThe result is x a.  Using tr, the result would be x x.\n\n.trnt abcd\nThis  is the same as the tr request except that the translations\ndo not apply to text that is transparently throughput into a di-\nversion with \\!.  For example,\n\n.tr ab\n.di x\n\\!.tm a\n.di\n.x\n\nprints b; if trnt is used instead of tr it prints a.\n\n.troff Make  the  n built-in condition false, and the t built-in condi-\ntion true.  This undoes the effect of the nroff request.\n\n.unformat xx\nThis request 'unformats' the  diversion  xx.   Contrary  to  the\nasciify  request,  which  tries to convert formatted elements of\nthe diversion back to input tokens as much as possible,  .unfor-\nmat  only  handles tabs and spaces between words (usually caused\nby spaces or newlines in the input) specially.  The  former  are\ntreated  as  if  they  were  input  tokens,  and  the latter are\nstretchable again.  Note that the vertical size of lines is  not\npreserved.   Glyph  information  (font,  font size, space width,\netc.) is retained.  Useful in conjunction with the box and  boxa\nrequests.\n\n.vpt n Enable  vertical  position  traps if n is non-zero, disable them\notherwise.  Vertical position traps are traps set by the  wh  or\ndt requests.  Traps set by the it request are not vertical posi-\ntion traps.  The parameter that controls whether vertical  posi-\ntion  traps  are enabled is global.  Initially vertical position\ntraps are enabled.\n\n.warn n\nControl warnings.  n is the sum of the numbers  associated  with\neach  warning that is to be enabled; all other warnings are dis-\nabled.  The number associated with each  warning  is  listed  in\ntroff(1).   For  example,  .warn  0  disables  all warnings, and\n.warn 1 disables all warnings except that about missing  glyphs.\nIf n is not given, all warnings are enabled.\n\n.warnscale si\nSet  the scaling indicator used in warnings to si.  Valid values\nfor si are u, i, c, p, and P.  At startup, it is set to i.\n\n.while c anything\nWhile condition c is true, accept anything as input;  c  can  be\nany condition acceptable to an if request; anything can comprise\nmultiple lines if the first line starts with  \\{  and  the  last\nline ends with \\}.  See also the break and continue requests.\n\n.write stream anything\nWrite  anything  to the stream named stream.  stream must previ-\nously have been the subject of an  open  request.   anything  is\nread in copy mode; a leading \" is stripped.\n\n.writec stream anything\nSimilar to write but without writing a final newline.\n\n.writem stream xx\nWrite the contents of the macro or string xx to the stream named\nstream.  stream must previously have been the subject of an open\nrequest.  xx is read in copy mode.\n\nExtended escape sequences\n\\D'...'\nAll  drawing  commands  of  groff's  intermediate output are ac-\ncepted.  See subsection \"Drawing Commands\" below.\n\nExtended requests\n.cf filename\nWhen used in a diversion, this embeds in the diversion an object\nwhich,  when  reread,  will cause the contents of filename to be\ntransparently copied through to the output.  In Unix troff,  the\ncontents of filename is immediately copied through to the output\nregardless of whether there is a current diversion; this  behav-\niour is so anomalous that it must be considered a bug.\n\n.de xx yy\n.am xx yy\n.ds xx yy\n.as xx yy\nIn  compatibility  mode, these requests behaves similar to .de1,\n.am1, .ds1, and .as1, respectively: A 'compatibility save' token\nis  inserted at the beginning, and a 'compatibility restore' to-\nken at the end, with compatibility mode switched on during  exe-\ncution.\n\n.ev xx If  xx  is  not  a  number, this switches to a named environment\ncalled xx.  The environment should be popped with a matching  ev\nrequest  without  any  arguments,  just as for numbered environ-\nments.  There is no limit on the number of  named  environments;\nthey are created the first time that they are referenced.\n\n.hy n  New  additive values 16 and 32 are available; the former enables\nhyphenation before the last character, the  latter  enables  hy-\nphenation after the first character.\n\n.ss m n\nWhen two arguments are given to the ss request, the second argu-\nment gives the sentence space size.  If the second  argument  is\nnot given, the sentence space size is the same as the word space\nsize.  Like the word space size, the sentence space is in  units\nof one twelfth of the spacewidth parameter for the current font.\nInitially both the word space size and the sentence  space  size\nare  12.  Contrary to Unix troff, GNU troff handles this request\nin nroff mode also; a given value is then rounded  down  to  the\nnearest  multiple of 12.  The sentence space size is used in two\ncircumstances.  If the end of a sentence occurs at the end of  a\nline  in fill mode, then both an inter-word space and a sentence\nspace are added; if two spaces follow the end of a  sentence  in\nthe middle of a line, then the second space is a sentence space.\nNote that the behaviour of Unix troff is exactly that  exhibited\nby  GNU  troff if a second argument is never given to the ss re-\nquest.  In GNU troff, as in Unix troff, you should always follow\na sentence with either a newline or two spaces.\n\n.ta n1 n2 ... nn T r1 r2 ... rn\nSet  tabs  at  positions  n1,  n2,  ..., nn and then set tabs at\nnn+r1, nn+r2, ..., nn+rn and then at  nn+rn+r1,  nn+rn+r2,  ...,\nnn+rn+rn, and so on.  For example,\n\n.ta T .5i\n\nsets tabs every half an inch.\n\nNew number registers\nThe following read-only registers are available:\n\n\\n[.br]\nWithin  a macro call, it is set to 1 if the macro is called with\nthe 'normal' control character ('.' by default), and  set  to  0\notherwise.  This allows the reliable modification of requests.\n\n.als bp*orig bp\n.de bp\n.tm before bp\n.ie \\\\n[.br] .bp*orig\n.el 'bp*orig\n.tm after bp\n..\n\nUsing this register outside of a macro makes no sense (it always\nreturns zero in such cases).\n\n\\n[.C] 1 if compatibility mode is in effect, 0 otherwise.\n\n\\n[.cdp]\nThe depth of the last glyph added to  the  current  environment.\nIt is positive if the glyph extends below the baseline.\n\n\\n[.ce]\nThe  number  of lines remaining to be centered, as set by the ce\nrequest.\n\n\\n[.cht]\nThe height of the last glyph added to the  current  environment.\nIt is positive if the glyph extends above the baseline.\n\n\\n[.color]\n1 if colors are enabled, 0 otherwise.\n\n\\n[.csk]\nThe  skew  of  the  last glyph added to the current environment.\nThe skew of a glyph is how far to the right of the center  of  a\nglyph the center of an accent over that glyph should be placed.\n\n\\n[.ev]\nThe  name  or  number  of  the  current  environment.  This is a\nstring-valued register.\n\n\\n[.fam]\nThe current font family.  This is a string-valued register.\n\n\\n[.fn]\nThe current (internal) real font name.  This is a  string-valued\nregister.   If the current font is a style, the value of \\n[.fn]\nis the proper concatenation of family and style name.\n\n\\n[.fp]\nThe number of the next free font position.\n\n\\n[.g] Always 1.  Macros should use this to determine whether they  are\nrunning under GNU troff.\n\n\\n[.height]\nThe current height of the font as set with \\H.\n\n\\n[.hla]\nThe current hyphenation language as set by the hla request.\n\n\\n[.hlc]\nThe  number  of  immediately  preceding  consecutive  hyphenated\nlines.\n\n\\n[.hlm]\nThe maximum allowed number of consecutive hyphenated  lines,  as\nset by the hlm request.\n\n\\n[.hy]\nThe current hyphenation flags (as set by the hy request).\n\n\\n[.hym]\nThe current hyphenation margin (as set by the hym request).\n\n\\n[.hys]\nThe current hyphenation space (as set by the hys request).\n\n\\n[.in]\nThe indentation that applies to the current output line.\n\n\\n[.int]\nSet  to  a  positive  value  if  last output line is interrupted\n(i.e., if it contains \\c).\n\n\\n[.kern]\n1 if pairwise kerning is enabled, 0 otherwise.\n\n\\n[.lg]\nThe current ligature mode (as set by the lg request).\n\n\\n[.linetabs]\nThe current line-tabs mode (as set by the linetabs request).\n\n\\n[.ll]\nThe line length that applies to the current output line.\n\n\\n[.lt]\nThe title length as set by the lt request.\n\n\\n[.m] The name of the current drawing color.  This is a  string-valued\nregister.\n\n\\n[.M] The name of the current background color.  This is a string-val-\nued register.\n\n\\n[.ne]\nThe amount of space that was needed in the last ne request  that\ncaused  a  trap  to  be  sprung.  Useful in conjunction with the\n\\n[.trunc] register.\n\n\\n[.ns]\n1 if no-space mode is active, 0 otherwise.\n\n\\n[.O] The current output level as set with \\O.\n\n\\n[.P] 1 if the current page is in the output list set with -o.\n\n\\n[.pe]\n1 during a page ejection caused by the bp request, 0 otherwise.\n\n\\n[.pn]\nThe number of the next page, either the value set by  a  pn  re-\nquest, or the number of the current page plus 1.\n\n\\n[.ps]\nThe current point size in scaled points.\n\n\\n[.psr]\nThe last-requested point size in scaled points.\n\n\\n[.pvs]\nThe  current  post-vertical  line  space as set with the pvs re-\nquest.\n\n\\n[.rj]\nThe number of lines to be right-justified as set by the  rj  re-\nquest.\n\n\\n[.slant]\nThe slant of the current font as set with \\S.\n\n\\n[.sr]\nThe  last  requested point size in points as a decimal fraction.\nThis is a string-valued register.\n\n\\n[.ss]\n\\n[.sss]\nThese give the values of the parameters set  by  the  first  and\nsecond arguments of the ss request.\n\n\\n[.sty]\nThe current font style.  This is a string-valued register.\n\n\\n[.tabs]\nA string representation of the current tab settings suitable for\nuse as an argument to the ta request.\n\n\\n[.trunc]\nThe amount of vertical space  truncated  by  the  most  recently\nsprung  vertical position trap, or, if the trap was sprung by an\nne request, minus the amount of vertical motion produced by  the\nne  request.   In other words, at the point a trap is sprung, it\nrepresents the difference of what the  vertical  position  would\nhave been but for the trap, and what the vertical position actu-\nally is.  Useful in conjunction with the \\n[.ne] register.\n\n\\n[.U] Set to 1 if in safer mode and to 0 if in unsafe mode  (as  given\nwith the -U command-line option).\n\n\\n[.vpt]\n1 if vertical position traps are enabled, 0 otherwise.\n\n\\n[.warn]\nThe sum of the numbers associated with each of the currently en-\nabled warnings.  The number  associated  with  each  warning  is\nlisted in troff(1).\n\n\\n[.x] The major version number.  For example, if the version number is\n1.03, then \\n[.x] contains 1.\n\n\\n[.y] The minor version number.  For example, if the version number is\n1.03, then \\n[.y] contains 03.\n\n\\n[.Y] The revision number of groff.\n\n\\n[.zoom]\nThe  zoom  value  of the current font, in multiples of 1/1000th.\nZero if no magnification.\n\n\\n[llx]\n\\n[lly]\n\\n[urx]\n\\n[ury]\nThese four read/write registers are set by the psbb request  and\ncontain the bounding box values (in PostScript units) of a given\nPostScript image.\n\nThe following read/write registers are set by the \\w escape sequence:\n\n\\n[rst]\n\\n[rsb]\nLike the st and sb registers, but take account  of  the  heights\nand depths of glyphs.\n\n\\n[ssc]\nThe  amount  of horizontal space (possibly negative) that should\nbe added to the last glyph before a subscript.\n\n\\n[skw]\nHow far to right of the center of the last glyph in the \\w argu-\nment, the center of an accent from a roman font should be placed\nover that glyph.\n\nOther available read/write number registers are:\n\n\\n[c.] The current input line number.  \\n[.c] is a read-only  alias  to\nthis register.\n\n\\n[hours]\nThe number of hours past midnight.  Initialized at start-up.\n\n\\n[hp] The current horizontal position at input line.\n\n\\n[lsn]\n\\n[lss]\nIf  there  are  leading spaces in an input line, these registers\nhold the number of leading spaces and the corresponding horizon-\ntal space, respectively.\n\n\\n[minutes]\nThe number of minutes after the hour.  Initialized at start-up.\n\n\\n[seconds]\nThe  number  of seconds after the minute.  Initialized at start-\nup.\n\n\\n[systat]\nThe return value of the system() function executed by  the  last\nsy request.\n\n\\n[slimit]\nIf  greater  than  0, the maximum number of objects on the input\nstack.  If less than or equal to 0, there is  no  limit  on  the\nnumber  of objects on the input stack.  With no limit, recursion\ncan continue until virtual memory is exhausted.\n\n\\n[year]\nThe current year.  Note that the traditional troff number regis-\nter \\n[yr] is the current year minus 1900.\n\nMiscellaneous\ntroff  predefines  a single (read/write) string-based register, \\*[.T],\nwhich contains the argument given to the -T command-line option, namely\nthe  current  output  device (for example, latin1 or ascii).  Note that\nthis is not the same as the (read-only) number register \\n[.T] which is\ndefined to be 1 if troff is called with the -T command-line option, and\nzero otherwise.  This behaviour is different from Unix troff.\n\nFonts not listed in the DESC file are automatically mounted on the next\navailable  font  position when they are referenced.  If a font is to be\nmounted explicitly with the fp request on an unused font  position,  it\nshould be mounted on the first unused font position, which can be found\nin the \\n[.fp] register; although troff does not enforce this strictly,\nit  does  not  allow a font to be mounted at a position whose number is\nmuch greater than that of any currently used position.\n\nInterpolating a string does not hide existing macro arguments.  Thus in\na macro, a more efficient way of doing\n\n.xx \\\\$@\n\nis\n\n\\\\*[xx]\\\\\n\nIf  the  font  description  file contains pairwise kerning information,\nglyphs from that font are kerned.  Kerning between two  glyphs  can  be\ninhibited by placing a \\& between them.\n\nIn  a  string comparison in a condition, characters that appear at dif-\nferent input levels to the first delimiter character are not recognized\nas  the  second  or  third delimiters.  This applies also to the tl re-\nquest.  In a \\w escape sequence, a character that appears at a  differ-\nent  input  level to the starting delimiter character is not recognized\nas the closing delimiter character.  The same is true for \\A,  \\b,  \\B,\n\\C,  \\l,  \\L, \\o, \\X, and \\Z.  When decoding a macro or string argument\nthat is delimited by double quotes, a character that appears at a  dif-\nferent  input  level  to the starting delimiter character is not recog-\nnized as the closing delimiter character.  The  implementation  of  \\$@\nensures  that  the  double quotes surrounding an argument appear at the\nsame input level, which is different to the input level of the argument\nitself.   In a long escape name ] is not recognized as a closing delim-\niter except when it occurs at the same input level as  the  opening  [.\nIn compatibility mode, no attention is paid to the input-level.\n\nThere are some new types of condition:\n\n.if rxxx\nTrue if there is a number register named xxx.\n\n.if dxxx\nTrue  if  there  is a string, macro, diversion, or request named\nxxx.\n\n.if mxxx\nTrue if there is a color named xxx.\n\n.if cch\nTrue if there is a character (or glyph) ch available; ch is  ei-\nther  an ASCII character or a glyph (special character) \\N'xxx',\n\\(xx or \\[xxx]; the condition is also true if ch  has  been  de-\nfined by the char request.\n\n.if Ff True  if  font  f exists.  f is handled as if it was opened with\nthe ft request (this is, font translation  and  styles  are  ap-\nplied), without actually mounting it.\n\n.if Ss True  if  style  s has been registered.  Font translation is ap-\nplied.\n\nThe tr request can now map characters onto \\~.\n\nThe space width emitted by the \\| and \\^ escape sequences can  be  con-\ntrolled  on  a per-font basis.  If there is a glyph named \\| or \\^, re-\nspectively (note the leading backslash), defined in  the  current  font\nfile, use this glyph's width instead of the default value.\n\nIt  is now possible to have whitespace between the first and second dot\n(or the name of the ending macro) to end a macro definition.  Example:\n\n.if t \\{\\\n.  de bar\n.    nop Hello, I'm 'bar'.\n.  .\n.\\}\n",
                "subsections": []
            },
            "INTERMEDIATE OUTPUT FORMAT": {
                "content": "This section describes the format output by GNU troff.  The output for-\nmat used by GNU troff is very similar to that used by Unix device-inde-\npendent troff.  Only the differences are documented here.\n\nUnits\nThe argument to the s command is in scaled points (units  of  points/n,\nwhere  n  is  the argument to the sizescale command  in the DESC file).\nThe argument to the x Height command is also in scaled points.\n\nText Commands\nNn     Print glyph with index n (a non-negative integer) of the current\nfont.\n\nIf  the  tcommand line is present in the DESC file, troff uses the fol-\nlowing two commands.\n\ntxxx   xxx is any sequence of characters terminated by  a  space  or  a\nnewline  (to  be  more precise, it is a sequence of glyphs which\nare accessed with the corresponding characters); the first char-\nacter  should  be  printed  at the current position, the current\nhorizontal position should be increased  by  the  width  of  the\nfirst character, and so on for each character.  The width of the\nglyph is that given in the font file, appropriately  scaled  for\nthe  current point size, and rounded so that it is a multiple of\nthe horizontal resolution.  Special characters cannot be printed\nusing this command.\n\nun xxx This  is  same  as the t command except that after printing each\ncharacter, the current horizontal position is increased  by  the\nsum of the width of that character and n.\n\nNote  that  single  characters  can have the eighth bit set, as can the\nnames of fonts and special characters.\n\nThe names of glyphs and fonts  can  be  of  arbitrary  length;  drivers\nshould not assume that they are only two characters long.\n\nWhen  a  glyph  is  to  be printed, that glyph is always in the current\nfont.  Unlike device-independent troff, it is not necessary for drivers\nto search special fonts to find a glyph.\n\nFor color support, some new commands have been added:\n\nmc cyan magenta yellow\nmd\nmg gray\nmk cyan magenta yellow black\nmr red green blue\nSet  the  color  components  of the current drawing color, using\nvarious color schemes.  md resets the drawing color to  the  de-\nfault  value.   The  arguments  are  integers  in the range 0 to\n65536.\n\nThe x device control command has been extended.\n\nx u n  If n is 1, start underlining of spaces.  If n is 0, stop  under-\nlining  of  spaces.   This is needed for the cu request in nroff\nmode and is ignored otherwise.\n\nDrawing Commands\nThe D drawing command has been extended.  These extensions are not used\nby GNU pic if the -n option is given.\n\nDf n\\n Set the shade of gray to be used for filling solid objects to n;\nn must be an integer between 0 and  1000,  where  0  corresponds\nsolid  white and 1000 to solid black, and values in between cor-\nrespond to intermediate shades of gray.  This  applies  only  to\nsolid circles, solid ellipses and solid polygons.  By default, a\nlevel of 1000 is used.  Whatever color a solid  object  has,  it\nshould  completely  obscure  everything  beneath  it.   A  value\ngreater than 1000 or less than 0 can also be  used:  this  means\nfill  with  the  shade  of gray that is currently being used for\nlines and text.  Normally this is black, but  some  drivers  may\nprovide a way of changing this.\n\nThe  corresponding  \\D'f...' command shouldn't be used since its\nargument is always rounded to an integer multiple of  the  hori-\nzontal resolution which can lead to surprising results.\n\nDC d\\n Draw a solid circle with a diameter of d with the leftmost point\nat the current position.\n\nDE dx dy\\n\nDraw a solid ellipse with a horizontal diameter of dx and a ver-\ntical  diameter of dy with the leftmost point at the current po-\nsition.\n\nDp dx1 dy1 dx2 dy2 ... .lf 3501\ndxn dyn\\n Draw a polygon with, for i=1,...,n+1, the i-th  vertex\nat  the  current position +<Sigma>(dxj,dyj).  At the moment, GNU\npic only uses this command to generate triangles and rectangles.\n\nDP dx1 dy1 dx2 dy2 ... .lf 3513\ndxn dyn\\n Like Dp but draw a solid rather than outlined polygon.\n\nDt n\\n Set the current line thickness to n machine  units.   Tradition-\nally Unix troff drivers use a line thickness proportional to the\ncurrent point size; drivers should continue to do this if no  Dt\ncommand has been given, or if a Dt command has been given with a\nnegative value of n.  A zero value of  n  selects  the  smallest\navailable line thickness.\n\nA difficulty arises in how the current position should be changed after\nthe execution of these commands.  This is not of great importance since\nthe code generated by GNU pic does not depend on this.  Given a drawing\ncommand of the form\n\n\\D'c x1 y1 x2 y2 ... .lf 3546 xn yn'\n\nwhere c is not one of c, e, l, a, or ~, Unix troff treats each  of  the\nxi  as a horizontal quantity, and each of the yi as a vertical quantity\nand assumes that the width of the drawn object is <Sigma>xi,  and  that\nthe  height is <Sigma>yi.  (The assumption about the height can be seen\nby examining the st and sb registers after using such a D command in  a\n\\w escape sequence).  This rule also holds for all the original drawing\ncommands with the exception of De.  For the sake of  compatibility  GNU\ntroff also follows this rule, even though it produces an ugly result in\nthe case of the Dt and Df, and, to a lesser extent, DE commands.   Thus\nafter executing a D command of the form\n\nDc x1 y1 x2 y2 ... .lf 3590 xn yn\\n\n\nthe current position should be increased by (<Sigma>xi,<Sigma>yi).\n\nAnother set of extensions is\n\nDFc cyan magenta yellow\\n\nDFd\\n\nDFg gray\\n\nDFk cyan magenta yellow black\\n\nDFr red green blue\\n\nSet  the  color  components  of the filling color similar to the\nm commands above.\n\nThe current position isn't changed by those colour  commands  (contrary\nto Df).\n\nDevice Control Commands\nThere  is  a  continuation convention which permits the argument to the\nx X command to contain newlines: when outputting the  argument  to  the\nx X  command,  GNU  troff follows each newline in the argument with a +\ncharacter (as usual, it terminates the entire argument with a newline);\nthus  if the line after the line containing the x X command starts with\n+, then the newline ending the line containing the x X  command  should\nbe  treated as part of the argument to the x X command, the + should be\nignored, and the part of the line following the  +  should  be  treated\nlike the part of the line following the x X command.\n\nThe first three output commands are guaranteed to be:\n\nx T device\nx res n h v\nx init\n",
                "subsections": []
            },
            "INCOMPATIBILITIES": {
                "content": "In  spite  of  the many extensions, groff has retained compatibility to\nclassical troff to a large degree.  For the cases where the  extensions\nlead  to  collisions, a special compatibility mode with the restricted,\nold functionality was created for groff.\n\nGroff Language\ngroff provides a compatibility mode that allows the processing of  roff\ncode  written  for classical troff or for other implementations of roff\nin a consistent way.\n\nCompatibility mode can be turned on with the  -C  command-line  option,\nand  turned  on or off with the .cp request.  The number register \\n(.C\nis 1 if compatibility mode is on, 0 otherwise.\n\nThis became necessary because the GNU concept  for  long  names  causes\nsome incompatibilities.  Classical troff interprets\n\n.dsabcd\n\nas  defining a string ab with contents cd.  In groff mode, this is con-\nsidered as a call of a macro named dsabcd.\n\nAlso classical troff interprets \\*[ or \\n[ as references to a string or\nnumber  register called [ while groff takes this as the start of a long\nname.\n\nIn compatibility mode, groff interprets these things in the traditional\nway; so long names are not recognized.\n\nOn  the  other hand, groff in GNU native mode does not allow to use the\nsingle-character escapes \\\\ (backslash), \\| (vertical bar), \\^ (caret),\n\\&  (ampersand),  \\{ (opening brace), \\} (closing brace), '\\ ' (space),\n\\' (single quote), \\`  (backquote),  \\-  (minus),  \\  (underline),  \\!\n(bang), \\% (percent), and \\c (character c) in names of strings, macros,\ndiversions, number registers, fonts or environments, whereas  classical\ntroff does.\n\nThe  \\A  escape  sequence  can  be helpful in avoiding these escape se-\nquences in names.\n\nFractional point sizes cause one noteworthy incompatibility.  In  clas-\nsical troff, the ps request ignores scale indicators and so\n\n.ps 10u\n\nsets  the  point  size  to  10 points, whereas in groff native mode the\npoint size is set to 10 scaled points.\n\nIn groff, there is a fundamental difference between  unformatted  input\ncharacters,  and formatted output characters (glyphs).  Everything that\naffects how a glyph is output is stored with the glyph;  once  a  glyph\nhas  been  constructed it is unaffected by any subsequent requests that\nare executed, including the bd, cs, tkf, tr, or fp requests.\n\nNormally glyphs are constructed from input characters at the moment im-\nmediately  before  the  glyph  is  added  to  the  current output line.\nMacros, diversions and strings are all, in fact, the same type  of  ob-\nject; they contain lists of input characters and glyphs in any combina-\ntion.\n\nSpecial characters can be both; before being added to the output,  they\nact as input entities, afterwards they denote glyphs.\n\nA  glyph  does  not  behave like an input character for the purposes of\nmacro processing; it does not inherit any  of  the  special  properties\nthat  the input character from which it was constructed might have had.\nThe following example makes things clearer.\n\n.di x\n\\\\\\\\\n.br\n.di\n.x\n\nWith GNU troff this is printed as \\\\.  So  each  pair  of  input  back-\nslashes '\\\\' is turned into a single output backslash glyph '\\' and the\nresulting output backslashes are not interpreted as  escape  characters\nwhen they are reread.\n\nClassical  troff  would  interpret  them as escape characters when they\nwere reread and would end up printing a single backslash '\\'.\n\nIn GNU, the correct way to get a printable  version  of  the  backslash\ncharacter '\\' is the \\(rs escape sequence, but classical troff does not\nprovide a clean feature for getting  a  non-syntactical  backslash.   A\nclose  method  is the printable version of the current escape character\nusing the \\e escape sequence; this works if the current escape  charac-\nter  is  not  redefined.   It  works in both GNU mode and compatibility\nmode, while dirty tricks like specifying a sequence of  multiple  back-\nslashes do not work reliably; for the different handling in diversions,\nmacro definitions, or text mode quickly leads to a confusion about  the\nnecessary number of backslashes.\n\nTo store an escape sequence in a diversion that is interpreted when the\ndiversion is reread, either the traditional \\! transparent  output  fa-\ncility or the new \\? escape sequence can be used.\n\nIntermediate Output\nThe  groff  intermediate  output format is in a state of evolution.  So\nfar it has some incompatibilities, but it is intended  to  establish  a\nfull  compatibility to the classical troff output format.  Actually the\nfollowing incompatibilities exist:\n\no The positioning after the drawing of the polygons conflicts with  the\nclassical definition.\n\no The  intermediate output cannot be rescaled to other devices as clas-\nsical 'device-independent' troff did.\n",
                "subsections": []
            },
            "AUTHORS": {
                "content": "This document was written by James Clark <jjc@jclark.com> and  modified\nby Werner Lemberg <wl@gnu.org> and Bernd Warken <groff-bernd.warken-72@\nweb.de>.\n",
                "subsections": []
            },
            "SEE ALSO": {
                "content": "Groff: The GNU Implementation of troff, by Trent A. Fisher  and  Werner\nLemberg,  is the primary groff manual.  You can browse it interactively\nwith \"info groff\".\n\ngroff(1)\nA list of all documentation around groff.\n\ngroff(7)\nA description of the groff language, including a short, but com-\nplete  reference  of all predefined requests, registers, and es-\ncapes of plain groff.  From the command line, this is called us-\ning\n\nman 7 groff\n\nroff(7)\nA survey of roff systems, including pointers to further histori-\ncal documentation.\n\n[CSTR #54]\nThe Nroff/Troff User's Manual by J. F. Ossanna of  1976  in  the\nrevision  of  Brian Kernighan of 1992, being the classical troff\ndocumentation <http://cm.bell-labs.com/cm/cs/cstr/54.ps.gz>.\n\ngroff 1.22.4                     23 March 2022                   GROFFDIFF(7)",
                "subsections": []
            }
        }
    }
}